draS^ METEOROLOGY THE MACMILLAN COMPANY NEW YORK BOSTON CHICAGO DALLAS SAN FRANCISCO MACMILLAN & CO., LIMITED LONDON BOMBAY CALCUTTA MELBOURNE THE MACMILLAN CO. OF CANADA, LTD. TORONTO METEOROLOGY A TEXT-BOOK ON THE WEATHER, THE CAUSES OP ITS CHANGES, AND WEATHER FORECASTING FOE THE STUDENT AND GENEEAL EEADEE BY WILLIS ISBISTER MILHAM, PH.D. FIELD MEMORIAL PROFESSOR OF ASTRONOMY IN WILLIAMS COLLEGE Neto gorfc THE ,MACMILLAN COMPANY 1912 All rights reserved COPYRIGHT, 1912, BY THE MACMILLAN COMPANY. Set up and electrotyped. Published March, 1912. NortoooD J. 8. Cashing Co. Berwick & Smith Co. Norwood, Mass., U.S.A. PREFACE THIS book owes its existence to a course on meteorology which has been given by the author in Williams College for the last eight years. This course is a Junior -and Senior elective course with three exercises a week during a half year. A syllabus, covering both the text-book used and the added material, was prepared for the course. This was at first mimeographed, then revised and printed. Later it was again revised and reprinted. This book follows the order of topics in this last syllabus very closely, and is thus essentially a resume of the material which has been gathered for the course. This book is essentially a text-book. For this reason, the marginal comments at the sides of the pages, the questions, topics for investiga- tion, and practical exercises have been added. A syllabus of each chapter has been placed at its beginning, and the book has been divided into numbered sections, each treating a definite topic. The book is also intended for the general reader of scientific tastes, and it is hoped that these earmarks of a text-book will not be found objectionable by him. It can hardly be called an elementary treatise, but it starts at the begin- ning and no previous knowledge of meteorology itself is anywhere assumed. It is assumed, however, that the reader is familiar with the great general facts of science. References have been added at the end of each chapter. These include pamphlets and articles in the periodical literature as well as books. These are the first things which a student would naturally look up in order to gain further information. In appendix IX an attempt has been made to summarize the literature of meteorology. Here the books are arranged in alphabetical order without regard to age or value. Both the metric and English system of units and the Fahrenheit and centigrade thermometer scales have been used in the book. It seems unnecessary in quoting facts and data from many sources to change every- thing to conform to one set of units. In appendices I, II, and III, the English and metric systems of units and conversion tables have been added. The facts of meteorology have now become so general and accepted that there can be but little that is new in such a book. The originality must lie in the arrangement, use, and perhaps interpretation v 236401 v i PREFACE of these facts. Whenever a distinctive idea has been introduced by some investigator, credit is always given in the text when this idea is quoted. Such credit has never been intentionally omitted by the author. Although this book has assumed a considerable size, it can lay no claim to compTeteness. No single volume can be a treatise containing all known facts to date and all the explanations which have been offered. Four aspects or applications of meteorology have been entirely omitted. These are: (1) Mathematical Meteorology. (2) Meteorology applied to living things ; including phenology and the influence of climate on man. (3) Meteorology and medicine ; including climate and disease. (4) A History of Meteorology ; including a biography of the men who have contributed much to its development. To each of these subjects a long chapter could be devoted, and they could easily be expanded so as to become large books. Such topics as " Meteor- ological Apparatus," " The Daily, Annual, and Irregular Variation in the Various Meteorological Elements," "The Isothermal Layer," etc., could be easily treated at such length as to become a large book. In fact, each chapter in this book could be expanded into a sizable volume. This book, then, makes no attempt at completeness, but it does attempt to give a fairly full presentation of the present state of the science, and also to point the way for further acquisition of information on the part of him who desires it. In the preparation of this book, the author is particularly indebted to the United States Weather Bureau and to Professor Willis L. Moore, its chief. Every opportunity was given to make use of what is probably the largest meteorological library in the world; tables of data to illustrate various points were supplied ; and free permission to reproduce and quote much that has appeared in government publications was given. The author is under great obligations to many persons who have helped him in various ways: to Professor William J. Humphreys, Professor of Meteorological Physics, United States Weather Bureau, who has read the entire manuscript and made many helpful and valuable suggestions ; to Edward H. Bowie of the forecast division, Preston C. Day and Mait- land C. Bennett of the climatological division, C. Fitzhugh Talman in charge of the library, and Cleveland Abbe, Jr., of the library division, for assistance and suggestions during the final revision of this book while in Washington ; to Professor Cleveland Abbe, who has read a portion of the manuscript and whose kindly interest is an inspiration to any one who is teaching meteorology or doing research in that subject ; to George PREFACE vii T. Todd. local forecaster in charge of the Albany station, and Herbert E. Vail, the first assistant. These last gentlemen have read the entire manuscript, made many suggestions, and guarded it against minor mis- takes in connection with the routine work of the United States Weather Bureau. They have also responded with unfailing cheerfulness and promptness to the many calls for data in connection with the Albany station. W. I. M. WILLIAMS COLLEGE, WILLIAMS-TOWN, MASS., July, 1911. CONTENTS PART I CHAPTER PAGB I. INTRODUCTION THE ATMOSPHERE ........ 1 II. THE HEATING AND COOLING OF THE ATMOSPHERE ..... 28 III. THE OBSERVATION AND DISTRIBUTION OF TEMPERATURE ... 60 IV. THE PRESSURE AND CIRCULATION OF THE ATMOSPHERE .... 112 A. The Observation and Distribution of Pressure 114 B. The Observation and Distribution of the Winds . . . .136 C. The Convectional Theory and its Comparison with the Observed Facts .156 D. A General Classification of the Winds 164 V. THE MOISTURE IN THE ATMOSPHERE ....... 189 ^. The Water Vapor of the Atmosphere 191 X Dew, Frost, Fog 210 "$!, Clouds 218 D. Precipitation . . .239 VI. THE SECONDARY CIRCULATION OF THE ATMOSPHERE . . . . 264 A. Tropical Cyclones . 266 B. Extratropical Cyclones and Anticyclones 283 C. Thundershowers 320 D. Tornadoes .335 E. Waterspouts and Whirlwinds 342 F. Cyclonic and Local Winds 343 VII. WEATHER BUREAUS AND THEIR WORK ....... 353 VIII. WEATHER PREDICTIONS .......... 378 PAET II IX. CLIMATE 426 X. FLOODS AND RIVER STAGES 442 XL ATMOSPHERIC ELECTRICITY 454 XII. ATMOSPHERIC OPTICS 483 XIII. ATMOSPHERIC ACOUSTICS . . . 496 ILLUSTRATIONS FIGUBE 1. Composition of the Atmosphere at Various Heights 10 2. Aitken's Dust Counter 11 3. The Cause of Twilight 19 4. Typical Undisturbed Daily Variation in Temperature 23 5. Daily Variation disturbed by a Thundershower . . . . . .23 ^ 6. Energy received by Perpendicular and Oblique Incidence .... 32 ^7. The Revolution of the Earth around the Sun .33 8. The Presentation of the Earth to the Sun on June 21 and December 21 . 33 9. Variation in the Insolation with Latitude at Five Different Dates ... 34 10. Annual Variation in the Insolation received at Three Different Latitudes . 35 11. Variation in the Insolation with Latitude and Time 36 12. Showing the Contrast in the Thickness of Air passed through by Vertical and Oblique Rays * 38 13. Insolation on Mont Ventoux 40 14. Insolation at Montpellier ........... 40 15. Diagram Illustrating Convection .45 16. Diagram Illustrating Mirage 46 17. Balloon Equipped for Meteorological Observations .... facing 48 18. Kite Equipped for Meteorological Observations .... facing 49 19. Graph Illustrating Vertical Temperature Gradient . . . . . .50 20. Temperature Gradients 51 21. The Original Thermometer of Sanctorius 6T" 22. Thermometers of Various Forms . . . . . x . . . .64 23. A Thermometer in the Making ......... 64 24. The Error of Parallax in Reading a Thermometer 65 25. The Thermometer Shelter of the U. S. Weather Bureau . . . .67 26. The French Thermometer Shelter . . 68 27. The English Thermometer Shelter 68 28. The Russian Thermometer Shelter % . .69 29. The Sling Thermometer 69 30. Assmann's Ventilated Thermometer 70 31. The Draper Thermograph 71 32. The Richard Freres Thermograph . 72 33. Weather Bureau Maximum and Minimum Thermometers .... 73 34. Six Maximum and Minimum Thermometer ....... 74 35. Black Bulb Thermometer 75 36. Graphical Representation of Station Normals of Temperature at Albany, N. Y. 80 xi xii ILLUSTRATIONS 37. Thermo-isopleths at Berlin 81 38. Thermograph Record showing Typical Daily Variation of Temperature at Albany, N.Y., Oct. 14-18, 1908 84 39. Diagram illustrating the Energy received and given off by the Earth during a Day 85 40. Annual Variation in Temperature at Five Different Places . . .86 41. Isanomalous Temperature Lines for January .98 42. Isanomalous Temperature Lines for July 99 43. Isothermal Lines for Spain and Portugal for January and July ... 99 44. Annual Range of Temperature . 100 45. Highest Temperatures ever observed in the United States . 102 46. Lowest Temperatures ever observed in the United States .... 103 47. The Variability of Temperature for January in the United States . . . 104 48. Torricelli's Experiment . . . 115 49. Standard Barometer . . . . . 116 50. Cross-section of the Cistern of a Barometer ..117 51. The Meniscus 118 52. An Aneroid Barometer 120 53. The Internal Construction of an Aneroid Barometer 120 54. The Richard Freres Barograph 121 55. The Mouth-barometer 122 56. Barograph Record Showing the Diurnal Variation in Pressure . . . 125 67. The Diurnal Variation in Pressure at Sitka, New York, St. Louis, San Fran- cisco, New Orleans, and Mexico City 126 58. Diurnal Barometric Wave at 10 A. M 127 59. Diurnal Barometric Wave at 4 P.M. . . . . . . . . 127 60. The Distribution of Pressure in a Vertical Section along a Meridian . . 134 61. Areas of High and Low Pressure . . . . . . . . . 135 62. The Electrically Recording Wind Vane of the Weather Bureau . . .139 63. A Simple Deflection Anemometer 141 64. A Simple Pressure Anemometer 142 65. Lind's Pressure Anemometer .......... 142 66. Robinson's Cup Anemometer ...... .... 143 67. Pocket Anemometer 144 68. Wind Rose for January, 1909, at Syracuse, N.Y .147 69. Wind Rose for January. 1909, at Albany, N.Y 148 70. The Daily Variation in Wind Velocity at New York, St. Louis, and San Francisco for January and July ........ 149 71. Diurnal Variation in Wind Direction and Velocity at the Top of the Eiffel Tower during June, July, and August 151 72. The Annual Variation in Wind Velocity at Philadelphia, Chicago, Phoenix, and San Francisco 152 73. Diagram illustrating Convection in a Long Tank of Fluid . . . .156 74. The Meridional Section before Convection Started 157 76. The Meridional Section after Convection was Permanently Established . 158 76. A Meridional View of the Air Circulation ... . 158 ILLUSTRATIONS xiii FIGURE . PAGE 77. The Relation of Wind Direction to Pressure Gradients . . . .159 78. Air Motion about " Highs " and ' Lows " on a Non-rotating Earth . . 160 79. The Meridional Section of the Isobaric Surfaces in a Convectional Circula- tion on a Rotating Earth .......... 162 80. Air Motion about " Highs" and " Lows" on a Rotating Earth . . . 163 81. Surface Distribution of the Planetary Winds ...... 166 82. The Air Currents in the Outer Layer of the Atmosphere .... 166 83. The Air Circulation in the Intermediate Layer ...... 167 84. The North-south Component of the Circulation of the Atmosphere . . 167 85. The Sub-equatorial and Sub-tropical Wind Belts 171 86. Pressure and Wind Distribution over India during January . . . 175 87. Pressure and Wind Distribution over India during July .... 175 88. Pressure and Wind Direction in Spain and Portugal during January . . 176 89. Pressure and Wind Direction in Spain and Portugal during July . . . 176 90. Wind Direction in Australia during January and July 176 91. The Effect of Land and Sea Breezes on the Prevailing West Wind on the Shores of Long Island . 178 92. Cross-section of a Valley showing the Isobaric Surfaces .... 180 93. Piche Evaporimeter 193 94. The Hair Hygrometer 199 95. The Psychrometer . . 201 96. The Whirled Psychrometer . . .201 97. The Recording Hygrometer 203 98. The Annual Variation in Absolute Humidity at Washington, San Francisco, New Orleans, St. Louis, and Bismarck . . . . . . . 206 99. Relative Humidity at St. Louis, Mo., Sept. 16-19, 1908 . . . .207 100. The Annual Variation in Relative Humidity at New Orleans, Washington, Bismarck, and Phoenix .......... 208 101. A Nephoscope 226 102. The Burnt Paper Sunshine Recorder 227 103. The Photographic Sunshine Recorder 228 104. The U. S. Weather Bureau Sunshine Recorder 228 105. Snowflakes facing 241 106. The Effects of an Ice Storm in New England .... facing 242 107. The U. S. Weather Bureau Rain Gauge . 244 108. The Annual Variation in the Amount of Precipitation at Twelve Stations in the United States , . . . .249 109. The Annual Variation in the Amount of Precipitation at Eight Foreign Stations 250 110. The Distribution of the Meteorological Elements about a Tropical Cyclone . 268 111. Temperature and Moisture Changes in a Tropical Cyclone .... 269 112. The Path of the Hurricane of Sept. 1-12, 1900 272 113. Locating the Center of a Tropical Cyclone . . . . . . . 273 114. The Five Regions of Occurrence of Tropical Cyclones . . . . 274 115. The Direction of Rotation, Dangerous Half, and Path of Tropical Cyclones 275 116. The Hurricanes of the West Indies during September, from 1878 to 1900 . 276 xiv ILLUSTRATIONS FIGITKB PAQB 117. The Application of Ferrel's Law to the Air coming towards a Tropical Cyclone 280 118. The Distribution of the Meteorological Elements about an Extratropical Cyclone 284 119. Distribution of the Meteorological Elements about a Low with a Typical Wind Shift Line 288 120. The Structure of an Extratropical Cyclone at Various Levels . . . 289 121. The Distribution of the Meteorological Elements about a Typical Anticyclone or High 295 122. The Motion of the Cirrus Clouds about an Area of High Pressure . . 297 123. The Tracks of a Number of Lows in the Northern Hemisphere . . . 300 124. Storm Tracks for Europe . . . . . . . . . .301 125. The Bigelow System of Storm Tracks across the United States ... . 302 126. The Russell System of Tracks across the United States . . . .302 127. The Van Cleef System of Storm Tracks across the United States . . . 303 128. The Van Cleef System of Tracks for Highs across the United States . . 306 129. Diagram Illustrating the Determination of the Wind Direction and Velocity at Any Point near a Passing Low . . . . . . . .317 130. The Changes in the Meteorological Elements during a Hot Summer Day with a Typical Thundershower in the Afternoon . . 323 131. The Cross-section of a Typical Thundershower .... . 324 132. The Typical Form of a Thundershower . 328 133. The Typical Path of a Thundershower ...... . 329 134. 329 135. Two Views of the Same Tornado at Goddard, Kansas . facing 337 136. A Tornado at Oklahoma City, May 12, 1896 facing 337 137. Damage Caused by a Tornado at Rochester, Minn., Aug. 21, 1883 facing 337 138. Wreckage of Anchor Hall, St. Louis, May 27, 1896 facing 337 139. The Distribution of All Recorded Tornadoes from 1794 to 1881 . , 339 140. The Cottage City Waterspout, Aug. 19, 1896, 1:02 P.M. facing 343 141. Diagram Illustrating the Formation of the Foehn Wind . 346 142. The Climatological Districts and Forecast Sections of the Weather Bureau . 357 143. The U. S. Weather Bureau Station at Washington, D.C. Frontispiece 144. International Storm Warnings ....... . 374 145. Diagram Illustrating the Area for which to Predict a Cold Wave . 402 146. The Price Current Meter . 445 147. Two Simple Electroscopes ........ . 458 148. The Equipotential Surfaces over an Irregular Surface . 459 149. A Lightning Flash .......... facing 467 150. An Oak Struck by Lightning facing 473 151. A Black Walnut Struck by Lightning ...... facing 473 152. A Church Spire Struck by Lightning 474 153. The Effect of Refraction ........ . 484 154. Halos and Related Phenomena ....... . 488 155. The Formation of the Rainbow . 489 156. Cloud Shadows ; " the Sun drawing Water " . facing 490 157. Selective Scattering bv a Turbid Medium 491 CHARTS CHART I. Isothermal Lines for the Year. II. Ocean Currents. III. Isothermal Lines for the Year for the United States. IV. Isothermal Lines for January. V. Isothermal Lines for July. VI. Isothermal Lines for July for the United States. VII. Isothermal Lines for January for the United States. VIII. Isotherms for the North Polar Regions for January. IX. Isotherms for the North Polar Regions for July. X. Isobaric Lines of the World for the Year. XI. Isobaric Lines of the World for January. XII. Isobaric Lines of the World for July. XIII. Air Circulation of the Atlantic Ocean for January and February. XIV. Air Circulation of the Atlantic Ocean for July and August. XV. Normal Relative Humidity of the United States for January. XVI. Normal Relative Humidity of the United States for July. XVII. Normal Date of the First Killing Frost of the Autumn. XVIII. Normal Date of the Last Killing Frost of the Spring. XIX. Sunshine of the United States for January. XX. Sunshine of the United States for July. XXI. Cloudiness of the World for the Year. XXII. Normal Annual Precipitation for the World. XXIII. Normal Annual Precipitation for the United States. XXIV. Normal Precipitation for the United States for January, February, and March. XXV. Normal Precipitation for the United States for July, August, and September. XXVI. Normal Annual Snowfall for the United States. XXVII. Weather Map, 8 A.M., Sept. 8, 1900, showing Galveston Hurricane. XXVIII. Weather Map, 8 A.M., Dec. 30, 1907, showing a Typical Extratropical Cyclone or Low. XXIX. Weather Map, 8 A.M., March 3, 1904, showing a Typical Extratropical Cyclone or Low with a Pronounced Wind Shift Line. XXX. Weather Map, 8 A.M., April 23, 1906, showing a Typical Anticyclone or High. XXXI-XXXIV. Four Weather Maps Illustrating Distribution of the Meteorological Elements, Tracks followed, and Velocity of Motion for Highs and Lows (Jan. 30, 31, Feb. 1, 2, 1908). 1 The fifty charts are placed together at the end of the book. XV XVI CHARTS CHART XXXV. Weather Map, 8 A.M., Sept. 2, 1904, showing the Presence of Many Thundershowers in the Southern Quadrants of the Low. XXXVI. Weather Map, 3 P.M., March 11, 1884, showing a Low which gave rise to Tornadoes. XXXVII. Weather Map, 8 A.M., April 25, 1906, showing a Low which gave rise to a Tornado. XXXVIII-XL. Three Weather Maps to illustrate Forecasting (Dec. 8, 9, 10, 1907). XLI-XLII. Two Weather Maps to illustrate the Founding of a Low from a V-shaped Depression (Dec. 27, 28, 1904). XLIII-XLVIII. Six Weather Maps illustrating Forecasting by Similarity (Jan. 9, 10, 14, 15, 27, 28, 1908). XLIX. Weather Map illustrating Cold Wave Prediction (Jan. 6, 1909). L. Weather Map of the Northern Hemisphere for Jan. 28, 1910. METEOROLOGY METEOROLOGY CHAPTER I INTRODUCTION THE ATMOSPHERE INTRODUCTION The science of meteorology, i. Outline history of meteorology, 2, 3. Utility, 4-6. Relation of meteorology to the other natural sciences, 7. THE ATMOSPHERE The atmosphere and its properties, 8. Composition of the atmosphere, 9-14. Offices and activities of the atmosphere, 15, 16. Atmospheres of other heavenly bodies, 17. Evolution of the atmosphere, 18, 19. Future of the atmosphere, 20. THE PRESSURE AND HEIGHT OP THE ATMOSPHERE Gravity and its effects, 21. Geosphere, hydrosphere, atmosphere, 22, 23. Height of the atmosphere, 24. THE METEOROLOGICAL ELEMENTS The meteorological elements, 25. Weather and climate, 26. Periodic and irregular variation and normal value, 27, 28. Graphical representation, 29. THE THREE METHODS OF INVESTIGATION, 30. THE PLAN OF THE BOOK, 31. INTRODUCTION i. The science of meteorology. The natural sciences deal with the phenomena of the world of nature about us. As examples of these varied phenomena one might mention the fall of a Definition stone, the rusting of iron, the growth of a plant, the change of a natural in the phases of the moon, the formation of a cloud, or the erosion of a valley by a stream. These are all occurrences in or phe- nomena of this world of nature, and it is thus the province of some one of the natural sciences to treat fully each one of these phenomena. 1 2 METEOROLOGY Since these phenomena are so numerous, complex, and varied, they are divided among several natural sciences. Physics and chemistry are the two fundamental natural "sciences because they Enumera- tion of the treat matter and energy, the two components of the mate- natural r j a j worid^ i n the abstract. They are also fundamental be- cause so many of their facts, laws, principles, and methods are usexLjn the other sciences. Biology with its many subdivisions includes all those phenomena where life is involved. The remaining phenomena, those of inanimate matter, are divided up between astron- omy, meteorology, and geology ; astronomy treating the heavenly bodies beyond the earth, meteorology the earth's atmosphere or envelope of gas, and geology the earth itself. (^Meteorology is thus one of the natural sciences, since it has a group of phenomena to investigate. It treats of the condition of the atmosphere, Definition of ^s changes of condition, and the causes of these changes, meteorol- Its duty is to arrange the facts in an orderly way so that the relation of cause and effect can be traced and generaliza- tions can be formed. Meteorology is often defined briefly as the study of atmospheric phenomena. Since so many of the laws and principles of physics find their application in the atmosphere on a stupendous scale, meteorology is also often defined as the physics of the atmosphere. 2. Outline history of meteorology. -- The annual change from summer to winter and back again from winter to summer, with all the Meteorol- attendant changes in vegetation, the daily change from the ogy an old heat of the day to the cool of the night, the falling of rain and snow, the coming of a thunder shower, all such things must have profoundly occupied the mind of man from remote ages. The antiquity of many weather proverbs and of much of our weather lore also shows that meteorological observations and generalizations were among the first acts of an intelligent race. 2 The word meteorology comes from the Greek. Socrates is spoken On 'n of the ^ m one ^ P^o's dialogues as " a wise man, both a thinker word mete- on supra-terrestrial things and an investigator of all things oroiogy. beneath the earth." 3 The word for supra-terrestrial fur- 1 The abbreviation used by the U.S. Weather Bureau for meteorological is met'l, and this form will be used in this book whenever the word is abbreviated. J In the British Museum in London there are Babylonian clay tablets dating from about 4000 B.C. which contain weather proverbs. One, for example, reads : "When a ring surrounds the sun, then will rain fall." See HELLMANN, "The Dawn of Meteorology," Quarterly Journal of the Royal Met. Society, October, 1908, vol. XXXIV, No. 148, p. 227. 3 Sw/cpdtTTjs, &s &VTIP, "rdre /ueT^wpa pojTi<7T7js, ical T&. virb 7775 Apologia Socratis, cap. II. INTRODUCTION 3 nishes the root for the word meteorology. 1 About fifty years later (350 B.C.) Aristotle wrote the first treatise 2 on meteorology, consist- ing of four books or parts, containing a large amount of information of very mixed value. Not only were the things considered to-day under the term meteorology treated in this book, but the appearance of the stars^ comets, meteors, earthquakes, northern lights, the composition of matter, and other things as well. 3. Four periods may be recognized in the history of meteorology. The first lasts from antiquity until about 1600 A.D. The observations were crude, nearly all made without the help of instruments, The four and they were often inaccurate and much influenced by periods in superstition and imagination. 'The explanations were often ^e history fantastic and supernatural. ^* ogy an d The coming of the second period was brought about by theiTch&T ~ ,, . ,. e . :,. .. , acteristics. the invention of instruments for making observations. The- two most important instruments in meteorology were invented at about this time, the thermometer in 1590 by Galileo and Sanctorius of Padua, and the barometer in 1643 by Torricelli ; and in 1653 Ferdinand II, Grand Duke of Tuscany, established stations throughout northern Italy for making careful observations of meteorological phenomena. The chief characteristics of this period are a larger number of observations and a great increase in accuracy. The third period begins a little before 1800 and runs until about 1850. Its great characteristic is the attempt to give logical explana- tions for the various phenomena which Were now being observed with ever increasing accuracy. The fourth period is the modern period, and it began about 1850. ' >r,tly after this the various governments began establishing weather bureaus, interest in meteorology increased markedly, and great advances began to be made. The three great characteristics of this period are : ceasehss activity in gaining information, the utmost accuracy in ob- serving all possil e phenomena, and the rigid testing of all explanations and hypotheses. 4. Utility- The utility of meteorology may be noted in two en- tirely different directions. First, there is the financial saving caused by the thflely forecast : -->g of tho^e weather conditions which Two Hnes of do damage to commerce, agriculture, products in transit, usefulness. . ~ ^upra-tenv- .:. \67ot = description or treatise. Thus meteorology was ... rial thing 2 T><" - v;is TO. u,(.~f' * - 4 METEOROLOGY and businesses of many sorts; and, secondly, there is the educational advantage in the study of meteorology. The storms on the coast and also those on the lakes and rivers do immense damage to shipping and commerce and often cause an appalling loss of life. The late saving by f rosts of spring and the early frosts of autumn work havoc with the farm produce which may be exposed. Sudden changes in temperature from hot to cold or from cold to hot cause the complete loss of produce or merchandise which is being trans- j ported by boat or train from one point to another. Traffic on railroad and street car lines may be delayed or even stopped entirely by a heavy, unexpected fall of snow. The timely warning of the approach of these damage-causing weather conditions allows every precaution to be taken and has resulted in a tremendous financial saving. The total cost in maintaining the Weather Bureau now amounts to about a million and a half dollars, while the most conservative estimates place the saving to this country brought about by timely forecasts at many times the cost of maintaining the bureau. Mark W. Harrington, a former chief of the United States Weather Bureau, says in the preface of his book, About the Weather : l " The Harrington's more than twelve hundred thousand dollars expended every opinion. year by the Government may perhaps be considered an ex- orbitant price to pay for learning what weather we are likely to have for the coming twenty-four hours, but the truth is that no public investment is so immediately and so immensely profitable as that applied to the main- tenance of the Weather Bureau. " Not only are cyclones in the West, but the floods of streams and rivers, especially those of the Mississippi Valley, are foretold, ' and* in- calculable saving of life and property/ as Mr. J. E. Prindle says in ? port on the subject, 'results from their warnings. Before the daysof the bureau/ he proceeds, ' the West India hurricanes came unannounced, and sometimes two thousand lives were lost in a single storm, warnings of the Weather Bureau three such storms have parsed in succession without the loss of a single life, and the propen /ed in one storm would support the service for two years, in the winter of 1895-1896, by forecasting six very g hundred and fifty vessels, valued at seventeen million dollars carrying eighteen hundred persons, were held safe in port ' warnings/ ' 1 Reprinted from HABKI^OTON'S About the Weather, copyright, 1899., by D. Appi. and Company. INTRODUCTION 5 5. There is probably no subject which so fully occupies our attention and is of so much importance to us as the weather. And yet there is probably no subject about which the ordinary person knows The ^^^ less and where ignorance and superstition are more universal bility of an than in connection with the weather and the causes of its ed j^ ated public* changes. Meteorology has long since dispelled the mystery connected with weather and weather changes. It is thus the duty of | our schools and colleges to produce an educated public which can dis- ' tinguish between truth and error, fact and superstition. It is also a real pleasure to know the causes of the changes and to appreciate the mechan- ism back of the ever shifting panorama which constitutes our weather. There are many facts to be arranged in an orderly way so that the relation of cause and effect can be traced. Many observations have been made so that generalizations and laws can be derived. Mete- orology thus has a well-developed descriptive side which offers tionai value (training in exactness of statement and description and in ofmeteor- logical reasoning. It also has a mathematical side which is being developed more and more at the present time. This develops extreme exactness in thought and expression along abstract lines. Ap- paratus must be tested and improved, and there are many observations to be made. It thus has a laboratory side which gives increased familiar- ity with the details of a subject, sense training, and skill in manipulation. Meteorology thus offers to the student training along three different lines, and these are the three kinds of training offered by any one of the natural sciences. 6. In discussing the utility of meteorology attention should be called to how intimately the weather enters into every aspect of human life. Our plans are daily made and unmade on account of the weather. Our very moods depend upon the weather. As influences mentioned above, untold damage to property and even loss of of th life may be caused by storms, frosts, hail, lightning, and sudden changes in temperature. The student of criminology tells us that certain crimes are more prevalent at one season of the year than at an- other or with one particular type of weather. Theft, for example, is more common during the winter, and the cause is not far to seek. During the winter there is less employment than during the summer, and incomes are smaller. In addition the cold makes food, clothing, and shelter more imperative. The result is a decided seasonal variation in this form of crime. The physician tells us that certain diseases show a well-marked seasonal variation and are more prevalent witbt one kind of weather than 6 METEOROLOGY another. Meteorological statistics are being made use of more and more by lawyers not only in damage cases but also in criminal cases. A single illustration will suffice. A certain burglary case turned largely on the certain identification of a person seen to come from a building during the early hours of the morning. The observer was in a near-by building across the street. It was put in evidence by the defense that on the morn- ing in question, according to the observations of the Weather Bureau, a dense fog hung over the city so that positive identification at the dis- tance in question would have been impossible. 7. Relation of meteorology to the other natural sciences. The relation of meteorology to physics is a close one. Many of its terms and Relation of ^ ac ^ s are used, and so many of its laws and principles find their meteorology application on a large scale in the atmosphere that mete- to physics. oro i O gy i s sometimes defined as the physics of the atmosphere. A course on elementary physics at least should precede a study of mete- orology. Such terms as the following must be used from time to time,| and but little space can be given to defining or illustrating them : mass, t volume, density, velocity, acceleration, rotation, revolution, force, in- ertia, centrifugal force, gravitation, gravity, weight, pressure; atom, molecule, ether, solid, liquid, gas; sound; heat, temperature, expan- sion, specific heat, latent heat, conduction; light, reflection, trans- mission, absorption, radiant energy; magnetism, electricity. But two groups of facts are borrowed from astronomy. They are, first, the facts concerning the position of the earth with reference t9 the sun Relation of at different times of year, and, secondly, the facts concerning tTt'hTnatu- tne rotation of the earth on its axis. These will be presented rai sciences, at the appropriate place. (See section 36.) Meteorology touches chemistry in but one place, namely, in the discus- sion of the composition of the atmosphere. Almost no facts or principles are taken from biology or geology. THE ATMOSPHERE 8. The atmosphere and its properties. The atmosphere * may be defined as the envelope of gas surrounding the earth. It is an odorless, Definition colorless, tasteless gas and when at rest one might almost oftheatmos- doubt its substantiality. When it is in motion, however, ph . ere> in the form of wind, and hinders walking and even overturns trees and houses, there can be no doubt of its existence. 1 From the Greek : dr/*6s= vapor or gas ; and ly constant. (1) The air is very mobile. It is being the con- constantly mixed and transported great distances by the wind. (2) Gases diffuse easily, so that any irregularity would quickly eliminate itself even if there were no wind. INTRODUCTION 9 1 1 . Although the composition of the air is so uniform all over the earth's surface and to any height at which man can live, this is far 'from the case when great heights above the earth's surface are considered. Com og . Since air is only a mechanical mixture of its gaseous coin- tion at great ponents, each behaves as if the others were not present.^ This hei s hts - . means that the heavier gases will be held closer to the earth's surface, and the lighter components will predominate at great heights^ This is, of course, based on the assumption that the atmosphere is not mixed by the wind, but that each gas is free to distribute itself in accordance with its density. Accordingly, from a knowledge of the composition of air at the earth's surface and the temperature at different elevations, the percentage composition at any height can be computed. The following table com- puted by Humphreys l gives the percentage composition at various heights. In Fig. 1 these results are shown graphically. HEIGHT IN KILOMETERS N Ar C0 2 H HELIUM 5 77.89 20.95. 0.94 0.03, O.OL- o.oo 15 79.56 19.66 0.7l 0.02 o.oe 0.00* 30 84.48 15.10 0.22 0.00 0.20 0.00 50 86.16 10.01 0.08 0.00 3.72 0.03 80 22.70 1.38 0.00 0.00 75.47 0.45 100 1.63 0.07 0.00 0.00 97.84 0.46 150 0.00 0.00 0.00 0.00 99.73 0.27 It will be seen that at the height of 150 Jdlometers, or less than 100 miles, the atmosphere should be composed almost entirely of hydrogen with a little helium. This is in good agreement with the observed fact that the spectrum of shooting stars' shows prominently the hydrogen and helium lines (section 24). The outside of the atmosphere of the sun is also composed largely of hydrogen and helium. 12. The minor constituents of the atmosphere, often spoken of as impurities, are water vapor, nitric acid, sulfuric acid, ozone, organic and inorganic particles, and minute traces of several other The minor things. Of these the water vapor and the 4 particles and per- ? tile*!^* 8 haps ozone are the only ones which deserve attention. mosphere. The amount of water vapor in the atmosphere is always small, as it never exceeds 4 per cent and the amount is constantly changing with every change in the weather. Yet it is one of the most water inmortaTvt components, for without it both plant and va P r - 1 M.Mint Weather Bulletin, vol. II, p. 66, 1909. 10 METEOROLOGY animal life would be an impossibility. The discussion of its various forms, such as dew, frost, fog, cloud, rain, hail, and snow must form a large portion of every book on meteorology. 405 10 20 30 ^40 60 80 70 80 90 100 78 VOLUME PER CENT. FIG. 1. Composition of the Atmosphere at Various Heights. (After HUMPHREYS in Mount Weather Bulletin, Vol. II.) \ 13. The organic particles include bacteria and the^BQres of plants, and these minute organisms are scattered throughout t'he atmosphere. It is Organic estimated that even on high mountains and over the oceans particles. there is at least one in every cubic meter of air. In the streets of our cities the number probably runs up to 3000 per cubic meter and in crowded houses and hospital wards it probably reaches at least 80,000. INTRODUCTION 11 The Inorganic particles are usually spoken of as dust. These dust particles are much more numerous than the organic particles and are for the most part entirely invisible to the naked eye. A few of the giant members of the family may be seen when carried up by the wind from the earth's surface or when a beam of sunlight falls through a small opening into a darkened room. At the present time the number of these particles in a given volume can be determined by means of Aitken's dust counter. This in- jy^^ig genious instrument, as illustrated in Fig. 2, consists of a. dust pump P by means of which the air in the box B may be- cc suddenly rarefied. This is provided with a graduated part G at the lower end so that a known quantity of air may be withdrawn from B if desired. The box B is about a centimeter thick with other dimensions in proportion. At the bottom is a glass plate divided by lines usually into square millimeters and illuminated by the mirror M. At the top is a lens L for viewing and magnifying the spaces on the glass plate. At the sides are pieces of filter paper saturated with water. There are two stop cocks at C and C f . The working principle of the apparatus is this : Whenever the air in B is suddenly rarefied, it becomes colder and can no longer hold all the water vapor in it. (See section 183.) This water vapor collects on the dust particles and forms a fog which slowly settles and collects on the glass plate. The number of these small water drops can then be counted, and. this is the number of dust particles which were present. The method of conducting this experiment is at once ap- parent. By repeated rarefactions all dust must be removed from the box B. A known quantity of the air to be investigated is then introduced. The instrument is shaken so as to mix the air FIG. 2. Aitken's Dust Counter. with the dust-free air already in the box and saturate it with moisture. The next rarefaction will coat the introduced dust particles with water and cause them to collect on the glass plate, where they may be counted. determinations have been made with this or similar instruments 12 METEOROLOGY in all parts of the world and at various heights on mountains and under The amount many conditions. The following table gives the results ob- ofdustun- tained by Fridlander for the various oceans and at various der various . . , ,, -r^. , conditions, heights on the Bieshorn : DCST PARTICLES DCST PARTICLES PER CUBIC ELEVATION PER CUBIC CENTIMETER CENTIMETER Atlantic Ocean 2,053 6,700 950 Pacific Ocean . . 613 8,200 480 Indian Ocean 512 8,400 513 10,665 406 11,000 257 13,200 219 13,600 157 In a dusty city 100,000 dust particles per cubic centimeter is by no means large v and it has been found that a single puff of cigarette smoke contains about 4,000,000,000 particles. There are four chief sources of the dust of the atmosphere. (1) It is blown up from the earth by the wind. (2) It is injected into the atmos- Sources of phere by volcanoes while in eruption. During the volcanic dust. explosion in Krakatoa between Sumatra and Java in 1883 it is estimated that dust and steam were thrown into the air to a height of nearly twenty miles, and the presence of this dust could be detected in sunset colors all over the world for more than three years. (3) Shoot- ing stars put an immense quantity of dust into the upper atmosphere as a result of their combustion and disintegration. (4) Ocean spray when evaporated adds dust to the atmosphere, especially fine particles of salt. Atmospheric dust plays an important part in at least four ways. (1) It is one of the chief causes of haze. (2) It probably serves as cen- Effects of ters of condensation for all fog particles and raindrops. It dust. was once thought that condensation was impossible without it. (3) It is the cause of the sunrise and sunset colors and perhaps of the blue color of the sky. (4) It is the cause of twilight. These various effects of the dust will be fully considered later. 14. Ozone is a peculiar form of oxygen, in that its molecule is composed of three atoms of oxygen linked together, while the ordinary oxygen mole- Properties of cule consists of two. This third atom has a tendency to ozone. leave the molecule, and to this is due the oxidizing power of ozone and its value as a sanitary agent. Ozone is usually con- INTRODUCTION 13 sidered a powerful disinfectant and is supposed to be particularly useful where organic matter is decomposing. It is even stated by some that the bracing effect of certain climates and of certain types of weather is due to a slightly larger amount of ozone present in the air. In this last instance, however, a lower temperature, the absence of moisture, and an increase in the amount of electricity in the air probably play a far more important part in determining one's feelings than any change in the amount of ozone. The discovery of ozone is usually attributed to Schonbein in 1848, al- though a few investigators probably detected its peculiar odor earlier. It is this peculiar odor which gives it its name. 1 The The quan _ amount present in the atmosphere is extremely small, usually tity of about one part in a million. The amount shows a decided daily and annual variation and irregular fluctuations which are closely correlated with the type of weather. There is much more during the winder than during the summer. The quantity of ozone present in the air is determined by means of its oxidizing power on certain chemical compounds. Potassium iodide is ordinarily used, and a known quantity of this substance is added to a paste formed by dissolving a known quantity of starch in water. This is then spread on pieces of paper and exposed to the air a known length of time. From the depth of the blue color which results from the decom- position of the potassium iodide, the amount of ozone is estimated. The determinations are none too accurate, even when a standard paste and a standard color scale are used, as other things affect the decomposition of the potassium iodide slightly. Ozone is formed in the laboratory by allowing electricity to discharge through oxygen or air, and here the peculiar odor can be Formation readily detected. In nature it may be formed by electrical of ozone - discharges, or by the action of ultra-violet light on oxygen, or possibly in connection with the evaporation of water. 15. Offices and activities of the atmosphere and its components. The atmosphere as a whole has the following offices and activities: (1) It disseminates bacteria and the spores of plants. It also carries the seeds of some plants, particularly those provided with Activitiesof down like the thistle, long distances. (2) It makes flight theatmos- possible in the case qf birds and insects and even some ani- phere as a mals. (3) It furnishes power to sailing vessels and windmills. (4) It transports moisture, thus making possible animal and plant life 1 From the Greek 6fr = I smell. 14 METEOROLOGY The func- tions of the theatmos- on large portions of the earth's surface. (5) It produces sana mounus or sand dunes and is the cause of " weathering." (6) It produces waves on bodies of water. (7) It makes sound possible. 1 6. The various components of the atmosphere all have their own in- dividual offices and functions. Nitrogen is the inert component. When taken into the lungs of ani- mals, it appears to have no physiological effects. Plants also are unable to use it. It serves simply to dilute the oxygen, which is the active ele- ment in the atmosphere. Nitrogen does not readily form compounds, and to this may be due the fact that it makes up such a large part of the atmosphere and is found to such a slight extent in the compounds in the earth's crust. ^Oxygen is the active, energizing component. All animals derive their energy and power to do work from the oxygen. It is taken into the lungs, makes its way into the blood, and joins with the tissues of the body, liberating energy. As a result of this process, large amounts of CO 2 are added to the atmosphere by animals. Oxygen is also consumed whenever combustion takes place. A small amount of oxygen is also lost to the earth's atmos- phere by forming compounds with certain substances in the earth's crust. Oxygen is supplied to the atmosphere chiefly by green plants. A small amount comes from volcanoes and other vents in the earth. Carbon dioxide is also an important component of the earth's atmos- phere, for without it plant life would be impossible. The sap which comes up from the roots consists largely of water with certain organic and in- organic substances in solution. The green cells in the leaves in the pres- ence of sunshine have the power of taking the CO 2 from the air and com- bining it with the sap, thus building the complex molecules which make up its own tissues, at the same time liberating oxygen. CO 2 is supplied to the atmosphere from many sources. It comes from volcanoes and vents in the earth's crust. The water of- the ocean contains a large amount of it. Meteors when consumed in the atmosphere add a small quantity. A large quantity is added as the result of combustion, but the largest quantity is put into the atmosphere by animal life and the slow decay of vegetation. It is estimated that nearly a billion tons of coal are now burned annually. This alone would put into the atmosphere nearly four billion tons of CO 2 each year. Slightly more CO 2 is found ove the oceans than over the land, more in the southern hemisphere than i the northern, more over cities than in the country, and more at night tha during the day TVp rpqsnns for this are apparent. INTRODUCTION Oxygen and carbon dioxide are thus held in a state oi equilibrium by -is of plant and animal life. The animal consumes O and liberates CO 2 , while the plant consumes CO 2 and liberates O. Should the quan- tity of COo increase, plant life would become more luxuriant and animal life would be hindered. Should the quantity of oxygen increase, animal life would become more active and exhilarated and plant life would be stunted. In each case the tendency would be towards a restoration of equilibrium. Argon, hydrogen, and the rare gases of the atmosphere, like nitrogen^ have no individual functions. The sources and effects of the secondary components of the atmosphere have already been considered. 17. Atmosphere of other heavenly bodies. There are many in- direct methods of determining the amount of atmosphere possessed by the various heavenly bodies which make up our solar system. The giant planet Jupiter, with an equatorial diam- p heres of " eter of 90,190 miles and a temperature certainly above that the various of boiling water, has an immense atmosphere which is con- bodies. y stantly filled with clouds. Saturn, Uranus, and Neptune seem to be in about the same condition. Venus has an atmosphere somewhat less bulky than that of the earth. Mars, an older, colder planet with a diameter of 4352 miles, has only one twelfth as dense an atmosphere as the earth, while the moon with a diameter of 2163 miles has practically no atmosphere at all. Thus great diversity is found in the amount of atmosphere. The determining factors seem to be tempera- ture, size, and perhaps age. For a full discussion of this topic the reader must be referred to books on astronomy. 1 8. Evolution of the atmosphere. If during ' the early history of the earth it was molten to any extent s , or even if the surface tem- peratures were high, the water of .the oceans and the Early his- volatile mineral substances of the earth itself must have tory of the been in the atmosphere as vapor or gas. The earth must a then have possessed an immense atmosphere lade^i with vapor and clouds. As the earth cooled, the water and other substances would be precipitated upon the earth, thus reducing the atmosphere to its present bulk. * If. however, as some prefer to believe, the earth grew verjRlowly by accretion, each small mass of matter as it was joined to the earth bring- ,ng its quantum of gas, it may be that the surface temperatures were never au<) : "' ;1 "~" ^We screw as gradually in bulk as the earth 16 METEOROLOGY 19. Whatever may have been the initial bulk or condition of the at- mosphere, there are several reasons for believing that great changes in composition have taken place during geological times. The composition luxuriant plant growth during the carboniferous age when during geo- the g rea t coal beds were laid down is usually explained by assuming a much larger amount of CO 2 in the atmosphere In fact some go so far as to state that there was probably extremely little oxygen in the atmosphere at that time, as it had all been used in forming compounds such as water and the oxides of the earth's crust. They attribute our present supply of oxygen entirely to the very luxuriant plant life of that time. Great changes have also taken place in the tem- peratures of the earth. Glaciation has extended down the Mississippi Valley as far as Kansas, and traces of luxuriant vegetation have been found in northern countries where it is at present impossible. One of the explanations sometimes given is to attribute it to changes in the composition of the atmosphere. So great is the influence of composi- tion on temperature that the statement is sometimes made that if the amount of C0 2 in the atmosphere were multiplied by four, the vegeta- tion of Florida would be found in Greenland. 20. Future of the atmosphere. At present a small amount of gas is lost to the atmosphere by escape into space beyond the gravitational control of the earth. A small amount, particularly oxygen, * s a ^ so ^ os ^ m ^ e f rma tin of compounds which become constant in part of the ocean or the solid earth. The atmosphere gains composition. a smal l amount of gas from 'meteors and from volcanoes and other vents in the earth. The amount of and CO 2 in the atmosphere is kept constant by the balance between plant and animal life. It thus seems that the future will see no such changes in bulk or composition as have been witnessed in the past. A very gradual diminution in the amount of the atmosphere is, however, to be expected. The sun, however, may eventually grow cold. The only two explana- tions of the maintenance of the sun's heat which have withstood mod- ern investigation are that the heat is maintained by slow The future . . , disappear- contraction or by the presence of radio-active material, anceofthe However the outpour of heat may be maintained, it must eventually come to an end. As the sun grows cold, so must the earth. A glance at the accompanying table of the boiling points of the components of the atmosphere shows the inevitable results. INTRODUCTION 17 Water, f 100 C. When the temperature reaches - 78 C. the CO 2 CO 2 , 78, will come out of the atmosphere, and plant and ani- O, 183 mal life must cease. As the temperature falls lower, Ar, 187 the various components will come out in order as N, 194 their boiling points are reached, until finally a cold, H, 253 dark, airless world will be revolving about a dying sun. And a strange world it will indeed be, for the sky will be black in- stead of blue, the stars will be visible in the daytime, shadows will be entirely black, sound will be impossible, and the constant bombardment by meteors will make life in the open more dangerous than on a modern battlefield. THE PRESSURE AND HEIGHT OF THE ATMOSPHERE 21. Gravity and its effects. All objects near the earth's surface are pulled downward by what is called the force of gravity. Gravity in mag- nitude and direction is in reality the resultant of three forces ; the attraction of the earth as a whole, the attraction of sur- definition, 5 rounding objects, and the centrifugal force due to the earth's magnitude, rotation. It is ordinarily considered that the direction of the j^ n force of gravity is towards the center of the earth and that its magnitude is constant all over the earth. These statements are, how- ever, only approximately true. It is only at the equator and poles of the earth that the direction of gravity is directly towards the earth's center. At other places it is nearly but not exactly towards it. The magnitude also varies slightly with latitude and elevation. Two effects of this ever present force deserve attention. (1) All masses, since they are acted on by gravity, will have weight and exert a downward pressure on whatever supports them. (2) Since a fluid does not resist a Effects of change of form, it will set itself with its lower surface con- & &vli * forming to the shape of the containing vessel and with its upper surface at right angles to the direction of gravity. 22. Geosphere, hydrosphere, atmosphere. The great mass of the earth is solid at least in the outer crust and has a very uneven surface. It is covered in part by the oceans, which under the influence Q eosphere }f gravity conform with their under surfaces to the irregular- hydro- ties of the solid earth and have an upper free surface at right ingles to gravity. Both solid earth and fluid oceans are sur- ounded by an envelope of gas. These three are often spoken of as the posphere, the hydrosphere, and the atmosphere. 1 .= gas ; aipa = sphere. 18 METEOROLOGY The surface of the hydrosphere is considered a level surface, and all elevations are reckoned from it. When the dimensions of the earth are stated, it is always the dimensions of the hydrosphere that are given. Its form is not that of a perfect sphere, but that of a sphere flattened at the poles and known as an oblate spheroid. The equatorial diameter of the hydrosphere exceeds the polar by 27 miles. The following are a few numerical facts concerning the geosphere, hydrosphere, and atmosphere. Polar diameter of hydrosphere 7899.580 miles Equatorial diameter of hydrosphere 7926.592 miles Area of oceans I of earth's surface Mass of oceans ? ^ of earth Mass of atmosphere vvkinsTS f earth 23. Since the atmosphere is substantial, that is, possesses mass, and is acted upon by gravity, it must have weight and exert a downward pres- Pressure of sure - The pressure of the atmosphere is simply the weight of theatmos- the column of air above the point in question. With eleva- tion above the earth's surface the pressure must thus grow less. The average pressure of the atmosphere at the surface of the hydro- sphere is 14.7 pounds per square inch. It is usually measured, however, not in pounds per square inch, but by the length of the balancing or equivalent mercury column, as will be fully discussed in Chapter IV. 24. Height of the atmosphere. Since a gas tends to expand and occupy all the space open to it, no theoretical limit can be set to the Notheoreti- height of the atmosphere. We cannot picture the free sur- cai limit. f ace o f a g as> an( j m ust think of the earth's atmosphere as growing gradually thinnc** and thinner with elevation until it merges with that mere trace of gas which fills interplanetary space. The question of the height of the atmosphere may be put, however, in a more practical form. To what height above the earth's surface does a sufficient quanti- ty of air extend to give us any indication whatever of its presence ? This is sometimes called the sensible height of the atmosphere, and there are several methods of detecting the presence of air at considerable hei;. The five ^ Twilight is caused by the reflection of sunlight i ways of de- the dust and perhaps moisture particles of the upper air termining after the sun has gone below the horizon of the place in bie height question. Diffraction may also play a part. The cause of I the twilight is illustrated in Fig. 3. By noting the duration of twilight and knowing the dimensions of the earth, the height of the air producing the t.,lll te -ht may be "etermined* It has been found INTRODUCTION 19 AIR PRODUCING TWILIGHT FIG. 3. The Cause of Twilight. that a sufficient quantity of air for deflecting an appreciable amount of sunlight extends to a height of 63 kilometers. (2) The ordinary height of clouds is from a few meters to perhaps 15 kilometers, but on certain rare occasions, at night, par- ticularly in high latitudes in midsummer, faint luminous clouds have been observed as high as 83 kilometers l above the earth's surface. (3) The Aurora Borealis, or northern lights, is supposed to be due to the electrical discharges in the rarefied gases of the upper atmosphere. 2 The height of the aurora has been determined and is found in some cases to vary from 60 to 200 kilometers. 1 (4) Meteors are masses of matter from pinhead size up which are flying haphazard through space and often enter the earth's atmosphere with velocities from 12 to 50 miles per second. The heat caused by the resistance of the air raises them to incandescence and they become visible as shooting stars. The height at which these become visible has often been determined, and the larger values vary from 240 to 300 kilometers. 1 Thus there is sufficient quantity of air above this height, even, to make a meteor incandescent. (5) From observations of eclipses of the moon it is also possible to determine the extent of the atmosphere. This method also gives elevations as great as 300 kilometers. As a general conclusion, then, it may be stated that a sufficient quantity )f air extends to a height of 300, kilometers to give us an indication of its )resence. It should be noted in this connection that the highest moun- tain does not rise above 10 kilometers. The greatest height attained by a manned balloon is not over 11 kilometers (10.3 kilometers or 6J miles by Dr. Berson and Professor Siiring in 1901), and unmanned balloons and kites have not gone above 29 and 8 kilometers respectively. 1 These heights are determined trigonometrically, by means of surveying instruments. rather than the particular values of the meteorological ele- ments. The term is only used in connection with larger areas and longer periods of time. Thus one should speak of the weather on December 25, 1910, in New York City, but of the winter climate of New England. 27. Periodic and irregular variation and normal values. The numeri- Periodic ca ^ vames f the meteorological elements are by no nu'tins co^- and irregu- stant, but are always t undergoing change or variatio on ' are two kinds of change or variation, periodic r Whenever in the course of a variation the initial va 1 NTRODUCTION 21 after the lapse of approximately equal intervals of time, the variation is said to be periodic. If the changes are irregular or haphazard as regards amount or time of occurrence, the variation is said to be irregular. Now the meteorological elements are undergoing both kinds of variation simultaneously, and sometimes two or more periodic variations are present at the same time. For example, it usually grows warmer during the morning and early afternoon and then cooler during the rest of the after- noon and night. This is aperiodic variation in temperature. A sudden thunder shower may, however, lower the temperature twenty degrees or more during a few minutes. This is an irregular variation. Sometimes the irregular variations are of such magnitude as to cloak or render almost imperceptible the periodic variations. (See Figs. 4 and 5.) There are no examples in the realm of meteorology of a pure undisturbed periodic variation except for a short time. Examples of this must be taken from mechanics or physics. The amount of snow which falls during the win- ter, the highest or lowest temperature which occurs on a definite date for successive years, the number of thunder showers during a year, are all examples of an irregular variation only. In all the changes of tempera- ture and pressure, and in fact of all the meteorological elements from moment to moment, we have examples of periodic and irregular variation combined. 28. Whenever observations of any phase of any one of the meteorologi- cal elements have been made for a considerable time, it often is desirable to summarize the observations. Such a summary ordinarily contains four things, the average value, usually called the normal value, the greatest value, the least value, and the insumma- average departure from normal. The meaning of these and "aliens. S the method of determining their numerical values can be best illustrated by an example. The total number of thunder showers ob- served at Albany, N.Y., for successive years is shown on p. 22. The average number of thunder showers i 22. This is usually spoken of as the normal number of thunder showers per year. The largest num- ber is 35 in 1910, and the smallest number is' 7 in 1890. The year 1910, with 35, shows a departure from normal of 13. The departure for 1909 is 1, for!908 is 9, etc. When these departures are averaged, the result is called the average departure from normal. Its value here is 6. Thus the normal value 22, the greatest value, 35, the least value, 7, and the average departure from normal, 6, form a convenient summary of the observations and give a complete picture of what may be expected at Albany as regard the number of thunder showers during a year. This, by the way, is an 22 METEOROLOGY YEAR NUMBER OF THUN- DER SHOWERS YEAR NUMBER OF THUN- DER SHOWERS 1884 24 1898 26 1885 22 1899 22 1886 14 1900 28 1887 14 1901 30 1888 9 1902 28 1889 12 1903 22 1890 7 1904 28 1891 9 1905 21 1892 32 1906 32 1893 23 1907 25 1894 23 1908 31 1895 18 1909 21 1896 17 1910 35 1897 22 illustration of irregular variation, and most of the observations which are summarized in this way are examples of irregular variation only. 29. Graphical representation. The variation of a quantity, whether periodic, irregular, or both, can be readily pictured by plotting its various Graphical values to scale. This is called graphical representation, and represen- the curve which represents the variation is called the graph. The values are plotted by choosing equal distances on one of the axes, usually the horizontal or X-axis, to represent time and by laying off equal distances on the other axis to represent the values of the quan- tity. The points are then located in accordance with the observations and are connected by a broken line or a continuous curve. The following examples will illustrate this method : Place : Williarastown, Mass. JUNE 12, 1907 JUNE 9, 1906 JUNE 12, 1907 JUNE 9, 1906 Temperature Temperature Temperature Temperature 2 A.M. 48 64 6 P.M. 70 68 4 A.M. 44 63 8 P.M. 62 66 6 A.M. 43 63 10 P.M. 55 65 8 A.M. 51 68 Midnight 50 64 10 A.M. 62 73 2 A.M. 47 63 Noon 68 80 4 A.M. 44 62 2 P.M. 70 83 6 A.M. 43 60 4 P.M. 71 72 INTRODUCTION 23 Figure 4 represents the typical undisturbed daily variation in tem- perature. Figure 5 represents the typical variation disturbed by a thunder shower between 3 and 5 o'clock in the afternoon. 70 65 60 55 50 45 40 ^~-~ s William: 1 town, Mass. / s \ June 12,1 907 / N \ 1 \ / \ \ 7 x k y ^-*, 24 6 8 10 12 2 4 6 A.M. NOON P.M. 10 12 2 4 MIDN. A.M. FIG. 4. Typical Undisturbed Daily Variation in Temperature. N Many other examples of graphical representation will be found scat- tered through the book. Usually it is the variation of some quantity FIG. 5. Daily Variation Disturbed by a Thunder Shower. with the time that is represented. If any two quantities are so related that, as one changes, the other changes also, the relation can be rep- resented graphically in this way. THE THREE METHODS OF INVESTIGATION 30. There are three methods of reaching conclusions and deriving general laws. These are known as the inductive, the deduc- The induc- tive, and the experimental method. In the inductive method tlve method - fa.nts and observations are generalized to get the underlying law. 24 METEOROLOGY It proceeds from particulars to the general. All general statements based upon meteorological observations are thus examples of the induc- tive method. Meteorology is usually presented largely from the induc- tive standpoint. In the deductive method one starts with general principles or laws and determines what ought to take place in a particular instance. This must The deduc- then be compared with the observed facts and an agreement tive method. stamps the whole work as correct. In Chapter IV, C, will be found a good example of deductive reasoning. We will start there with two fundamental facts, determine what ought to be the pressure and wind direction at different points on the earth's surface and then compare this with the results of observation. The experimental method is essentially the laboratory method. Here the conditions of the experiment are changed and the results noted. In The expert- ^is way new facts, relations, and laws may be found, and all mental assumptions and hypotheses may be rigidly tested. The experimental method is of only limited application in mete- orology because all of the phenomena furnished by nature are on such a stupendous scale that man is powerless to change the conditions. THE PLAN OF THE BOOK 31. The first chapter contains introductory material and a discussion of the composition, pressure, and height of the atmosphere. The second The con- chapter is devoted to the study of the heating and cooling of tents of the the atmosphere, as this is of such fundamental importance. The meteorological elements are then considered in order in Chapters III to V inclusive. Chapter VI is given up to the study of the different kinds of storms. Two chapters are then devoted to the practical side of meteorology, namely, weather bureaus and their work and weather prediction. Part II, consisting of five chapters, is devoted to special subjects not always included in meteorology proper. They could be omitted without destroying the unity or completeness of the book. These subjects are climate, floods and river stages, atmospheric elec- tricity, atmospheric optics, atmospheric acoustics. Since the book is intended primarily as a text-book, the syllabus at the beginning of each chapter, the marginal topics, and the questions, topics The aim of for investigation, exercises, and references at the end of each the book. chapter have been added. It is also hoped that the intelligent reader who has not had special training in this science will be able to find INTRODUCTION 25 here a clear and concise picture of the modern aspects of the subje* This volume does not pretend to be a compendium or compilation of existing knowledge. It is hoped, however, that in the references to the literature the way has been pointed out to a more complete knowledge of the subject on the part of him who desires it. QUESTIONS (1) How is a natural science defined? (2) Mention several natural phenomena and state the natural science to which each belongs. (3) Enumerate the natural sciences. (4) What is the province of each ? (5) Why are physics and chem- istry fundamental ? Qtf Define meVy, (7) Of what does it treat ? (8) Why is met'y onejafJ^he oldest sciences j* (9) What is the origin of the word m^JejL? (10) Wher? where, and by whom was the first treatise written? (11) What did it contain ? (12) Give the dates of the four periods in the history of met'y. (13) What brought about each new period ? (14) State the characteristics of each period. (15) Name the two lines of usefulness of me^'y. (16) Discuss the question of the financial saving. (17) What is the duty of schools and col- leges as regards the public ? (18) Why should the public be educated in met'y ? (19) Along what three lines does a natural science off er^ training? (20) State the character of each line of training. (2i) State some of the varied in- fluences and effects of the weather, (22) State the relation of jnet'y to physics. (23) State its relation to the other natural sciences^ (24) Define the atmos- 1 V phere. (25) What is the origin of the word ? (26) What are the character- a* istics of air ? (27) State some of the physical projDer^ies of gases. (28) What are the four major constituents of pure dry air ? (29) State % approximately the percentage composition, (30) Tlow much hydrogen is found in the atmosphere ? (31) Name the four rare gases jn the atmosphere, (32) Give the proofs that air is a mechanical jnixture of its Components. (33) How constant is the composition of the air ? (3%) How much GO 2 does exhaled air contain ? (35) Why is the composition so constant ? (36) What change takes place in the composition at great heights? (37) Why? (38) What facts tend to justify this theoretical conclusion ? (39) Name the minor constituents of the atmosphere. (40) How much water vapor is preseiS?? (41) Why is water ''vapor "important ? (42) How many organic particles are in the atmosphere? (43) What is included under this head? (44) fre the inorganic particles visible? (45) Describe Aitken's dust counter. (46) State the working principle. (47) How is a deter- mination made?' (48) How numerous are dust particles? (49) Does the number vary? (50) What are the sources of atmospheric dust? '(51) What are some of the effects of this dust ? (52) What is the nature afid use of ozone? (53) What are the offices and activities of the atmosphere as a whole? (54) What are the functions of the nitrogen ? (55) What are the functions 'of the oxygen ? (56) How is it lost to and gained by*the atmosphere ? (57) What are thefunctions of CO 2 ? (58) How is it lost to and gained by the atmosphere ? (59) Is the amount constant everywhere ? (60) How are CO 2 and O held in equilibrium? (61) Describe briefly the atmos^iere of the other heavenly bodies in our solar system. (62) What determines the amount ? (63) What was the early history of the atmosphere ? (64) What changes have taken place in geological times? (65) Is the atmosphere changing now as regards bulk and composition? (66) What is the probable future of the atmosphere? (67) Define gravity. (68) What is its direction ? (69) State some of Its effects 26 METEOROLOGY on matter. (70) Define geosphere, hydrosphere, and atmosphere. (71) State the origin of the words. (72) What is the shape of the hydrosphere? (73) Why does the atmosphere exert a pressure? (74) How much is the pres- sure? (75) Why does the pressure change with elevation? (76) Why is there no theoretical limit to the atmosphere? (77) What is meant by the " sensible " height of the atmosphere? (78) Describe the five methods of determining it. (79) What heights have been found? (80) How high have balloons and kites ascended? (81) At what height is the middle layer located? (82) Define the met'l or weather elements. ,(83) Name the six met'l elements. (84) How may the condition of the atmosphere at any time and place be de- scribed? (85) Define weather and climate. (86) Have the met'l elements constant values? (87) Name and define the two kinds of variation or change. (88) Illustrate by means of examples the various kinds of change or variation. (89) What four things are computed when observations are summarized? (90) Define what is meant by normal values. (91) What is graphical repre- sentation? (92) How is a graph constructed? (93) Name the three methods of deriving general laws. (94) Describe and illustrate the inductive method. (95) Describe and illustrate the deductive method. (96) Describe and illus- trate the experimental method. (97) What is the relative importance of these three methods in met'y? (98) What is the order of the subjects treated in this book? (99) What is the aim of the book? TOPICS FOR INVESTIGATION (1) The financial saving caused by the Weather Bureau. (2) The relation of weather to disease. (3) The relation of weather to crime. (4) The extent to which meteorology is taught in the schools and colleges. (5) The agreement of the furious determinations of the composition of the air. (6) The history of the discovery of argon. i (7) The history of the discovery of the rarer gases in the air!^ (8) The dust of the atmosphere. (9) Changes in the earth's atmosphere during geological times. (10) Meteors. (See any text-book on astronomy.) (11) Determination of the maximum and minimum of graphs. (The usual method is to draw a line connecting the middle points of the chords parallel to the X-axis and terminated by the curve. ) % PRACTICAL EXERCISES (1) Copy frojn some source or compile a few tables showing the relation between the prevalence of certain kinds of disease and the weather. Plot the graphs representing these relations. (2) If physical and chemical apparatus is available, experiments may be performed to show the composition and pressure of the atmosphere, and the properties of gases. (3) Write out a definite, complete description of the weather at several dif- ferent times and places. (4) Copy from some source or compile observations of some quantity sub- ject to irregular variation only and summarize them in accordance with section 28. (5) Plot several graphs from series of observations of some kind and state the kinds of variation depicted. INTRODUCTION 27 REFERENCES * , : Observations of Atmospheric Dust," W. B. Bulletin 11, p. 734 ARRHENIUS, Lehrbuch der kosmischen Physik, Composition of the ati pp. 473-490. KASSNER, Das Welter und seine Bedeutung fur das praktische Leben, to practical life, pp. 114-144. MOORE, JOHN W., Meteorology, 2d ed., The Influence of Weather or pp. 407-457. RAMSAY, The Gases of the Atmosphere; the History of their Discovery. 1 For further information concerning the books and for further references see Appen- dix IX. CHAPTER II THE HEATING AND COOLING OF THE ATMOSPHERE THE NATURE OF MATTER, HEAT, TEMPERATURE, AND RADIANT ENERGY, 32 THE SOURCES OF ATMOSPHERIC HEAT, 33 INSOLATION Amount, 34. Variation with latitude and time of year, 35-37. Distribution over the earth, 38. THE INTERRELATION OF MATTER AND RADIANT ENERGY Reflection, 39. Transmission, 40. Absorption, 41, 42. Actinometry, 43, 44. Behavior of the ocean as regards reflection, transmission, and absorption, 45. Behavior of the land as regards reflection, transmission, and absorption, 46. Behavior of the atmosphere as regards reflection, transmission, and absorption, 47. CONDUCTION AND CONVECTION Conduction, 48. Convection, 49. Convection in the atmosphere, 50, 51. Evidences of convection, 52. TEMPERATURE GRADIENTS Vertical temperature gradient, 53, 54. The isothermal layer, 55, 56. Inversion of temperature nocturnal stability, 57. Diurnal instability conditions of convection, 58. How THE ATMOSPHERE is HEATED AND COOLED, 59 THE NATURE OF MATTER, HEAT, TEMPERATURE, AND RADIANT ENERGY 32. In order to have a clear conception of those fundamental processes which are operative in the heating and cooling of the atmosphere, one must understand something about the nature of matter, heat, tempera- 28 i* HEATING AND COOLING OF THE ATMOSPHEI ture, and radiant energy. This subject will be briefly sketched r the reader must be referred to text-books on physics for a more treatment. A piece of wood, for example, may be divided into several portions. These portions may in turn be subdivided, and the process may be con- tinued until the resulting portions are scarcely visible in a Molecule powerful microscope. The natural question at once arises and as to whether this process of subdivision could be carried on indefinitely provided the mechanical means were at hand for accomplishing it. The answer is that there is a smallest particle which can exist and still retain the properties of the substance in question. This smallest par-J tide is calle^the moleculeT "M^^uile^^ atojmsl which may be like or unlike, few or many. There are known to , eighty differdfiridncfioi : atoms corresponding to the numberjoL^o-called elements. It is now thotfghtTthat the atom in turn is composed of per- haps thousands of particles which may even, under certain conditions, escape from the atoms themselves.jrr~Haese particles are spoken of as corpuscles or electrons. A4oms are held together by the force of chemiTfal | affinity to form molecules; Molecules are held together by cohesion or| adhesion to form masses-, iand~aLl-maoooa -exertra gravitational attraction! pea-each, .other. The whole of space is supposed to be filled with a substance called luminiferous ether, having certain properties that account for the facts which are observed. The planets have moved through it for ages, without showing appreciable retardation ; it must, then, be practically frictionless. Waves are transmitted with tremendous velocities ; it must, therefore, be highly elastic. It has the same prop- erties at all points and in all directions ; it must, then, be homogeneous. Ether waves come to us from the remotest depths of space; it must, therefore, be all-pervading. We picture, then, the whole universe as filled with this frictionless, highly elastic, homogeneous, all-pervading substance. The molecules are not in contact with each other, but are embedded in this ether as dust particles are suspended in dusty air or as sand parti- cles are suspended in muddy water. These molecules are Heat and not at rest but are in constant motion, colliding with their temperature, neighbors, rebounding, and quivering. Since the molecules are in motions possess energy, and this energy of the molecules is called heatl, Health* -'.it a form of energy. The r. !v possesses is thus simply the sum total of the kinetic er 30 METEOROLOGY molecules. A careful distinction must be made between heat and tem- perature. Temperature is determined by the velocity of motion on th part of the individual molecules. If a body is heated, the molecules mov more rapidly and hit harder. There is now a more active motion on the part of the molecules and the temperature is thus higher. The sum total of the energy of the molecules is also greater and th body thus possesses a greater amount of heat. The following numerical data for hydrogen gas under normal conditions will give a more concrete and definite picture of the construction of mattej : The number of molecules per cubic centi- meter is 21 X 10 18 , each molecule with a diameter of 5.8 X 10~ 8 centimeter, and a mass of 4.6 X 10~ 24 gram. These molecules are moving with an average velocity of 1.9 X 10 5 centimeters per second over a path whose average length is 9.7 X 10" 6 centimeter, before colliding with another molecule. The average number of collisions per second is 9,480,000,000. Whenever a molecule collides with another and rebounds and quivers, j it becomes the centeFof an ether wave which spreads out spherically inl Radiant every direction. These waves have various lengths, that is,' energy. distances from crest to crest, and various intensities. They all proceed with a velocity of 186,330 miles per second in ether in which no material molecules are embedded. If a body is heated, a larger num- ber of waves is sent out and shorter waves are emitted as well as the longer ones. Terrestrial bodies at ordinary temperatures are emitting waves having lengths from, say, 0.001500 to 0.000270 centimeter. The waves sent out by the sun are shorter, having lengths varying from 0.000270 to 0.000019. Those between 0.000270 and 0.000075 are spoken of as the infra-red or heat rays ; those between 0.000075 and 0.000036 are light waves running from red to violet ; those between 0.000036 and 0.000019 are called the ultra-violet rays. The limits are by no means definite. The figures have been given simply to indicate approximately the limits between which the great majority of the wave lengths lie. The energy conveyed by ether waves is spoken of as radiant energy. AH bodies, then, which are emitting ether waves are sending^but radiant energy. The statements above have been made as if they were all undoubted facts. Many, however, are merely highly probable hypotheses. The probability of the correctness of an hypothesis increases with the number and complexity of the facts of observation which are explained and correlated by it, and as it becomes more and more evident that no other hypothesis can explain as much. OO)/ N'G OF THE ATMOSPHERE THE SOURCES OF ATMOSPHERIC HEAT 33. ..There are uiree sources.. from which heat might be supplied to the Litmosphere ; n;tmeJy^:Uhe_jSujv,the, earthls-dntenor, ancL.the-.ata.i>; and |.Qthex.hfiaYenlyJaQdifia^ . There are' two ways of showing that Th ,, three the amount of heat supplied to the atmosphere by any other possible urce than the sun is relatively negligible. The heat sup- s plied by the earth's interior would be the same at the poles as at the equator, the same by night and during the winter as by day and during the summer. Thus the changes, at least, in the temperature ^ but the of the atmosphere cannot be accounted for by heat from the sun negii- earth's interior. Again, the stars shine by day as well as by ^ night and on all portions of the earth's surface.' Thus again, the changes in temperature cannot be explained on the basis of heat received from the stars. Direct measurements have also been made of the amount of heat received from these sources, and the general conclusion is that the total amount of heat supplied by all other sources than the sun is not suf- ficient to change atmospheric temperatures by 0.25 Fahrenheit. It is the sun, then, which controls absolutely the heating of the atn - ;* and we must look to it for an explanation of all the varied \.:' in atmospheric temperatures. INSOLATION 34. Amount. The radiant energy, that is the t energy in the form of | ether waves, received from the sun is given the special name of insolation, [ Some idea of the amount of insolation can be gained by con- Definition of sideringthe size and condition of the sun itself. The sun is an insolation, immense globe, having a diameter of 866,000 miles and an average surface temperature of 10,000 degrees Fahrenheit. More than 325 worlds like ours could be strung like beads on a string around the equator of the sun. The best pictures of the intense incan- T Illustrations. descence of the sun's surface can be gained by realizing that each square yard of the -Tin's surface is constantly emitting 140,000 horse power of energy. The earth receives but a minute portion (one two-billionth part) of this, and yet the energy received in the course of a year would be sufficient to melt a layer of ice 241 feet thick covering the whole earth. 35. " ' latitude and time of year. The amount oM>- tioi sun at any particular time and place depe thro arncss to the sun. The amount of rad ; aru Tierg 32 METEOROLOGY The three factors which de- termine the amount of insolation received. ved from any hot body varies inversely as the square of the distance of the point in question from the body. Thus the amount of insolation varies inversely as the square of the distance frofn the sun. (2) Directness of the rays. When the sun's rays fall upon &r surface obliquely they are spread out over a larger area than when they fall perpendicularly. The result is that the amount of energy received by each unit of surface is far less for oblique than for perpendicular rays. The accompanying figure will illustrate the principle. Let AB be the width of a beam falling perpendicularly upon the surface MN. The total amount of energy will be concentrated upon the space whose width is CD. Suppose a beam of the same width AB falls obliquely upon the surface. Its energy will be spread out over a surface having a width of CE. The amount of energy received FIG. 6. Etoergy Received by Perpendicular . ,. r ^T> ank Oblique incidence. by each unit of area of CD will be seen at once to be far greater than that received by each unit of area of the surface CE. It can be readily demonstrated by trigonometry that the amount of energy re- ceived by a unit area in the case of oblique incidence equals the amount received by perpendicular incidence multiplied by the sine of the angle of elevation of the sun. By angle of elevation is meant the distance in degrees of the sun above the horizon. (3) Duration. The amount of energy received from a radiating body is, of course, directly proportional to the duration of the radiation. 36. In order to understand the variations in these three factors which determine the amount of insolation received, one must study the revolution of the earth about the sun. The earth's path around the sun is not an exact circle, but an ellipse with the sun in one focus. This path or orbit of the earth lies in a plane called the plane of the ecliptic, and the earth completes the circuit in 365^ days. The average distance of the earth from the sun is 92,900,000 miles, but the actual distance varies about 1,500,000 miles in either direction in the course of a year. The earth is nearest to the sun on January 1 and at its greatest distance on July 1. The axis of the earth is not perpendicular to the plane of the ecliptic, but makes an angle of 66? with it, and this a.\is remains parallel to itself as the earth revolves about the sun. Figure 7 illustrates this revo- The char- acteristics of the earth's revo- lution about the sun. THE HEATING AND COOLING OF _THE ATMOSPHERE 33 Changes in distance. mtion. The north pole and the northern hemisphere of the earth are turned most directly towards the sun on June 21 and away from the sun on Dec. 21. Figure 8, which indicates the position of the earth on June 21 and Dec. 21, shows this change in presentation to the sun's rays. Since the earth is nearly 3,000,000 miles nearer the sun on January 1 than on July l,fnore insolation must be received in January than in July. The difference amounts to about 7 per cent. The change in the presentation of the earth to the sun's rays causes an apparent Changes in migration of the sun directness, northward from Dec. 21 until June 21 and southward from June 21 until Dec. 21. The sun is directly overhead at noon on March 21 at the equator, on June 21 at the tropic of Cancer, on September 23 at the equator again, and on December 21 at the tropic of Capricorn. This migration of FIG. 7. The Revolution of the Earth around the Sun. JUNE 21 DECEMBER 21 FIG. 8. Til- Presentation of the Earth to the Sun on June 21 and December 21. the sun through 47 (2X23J) causes decided changes in the directness of the sun's rays and in the length of the day. The elevation of the sun at noon for any place in the northern h^rrji .->' METEOROLOGY is < can be shown to be 90 - <#> on March 21 and September 23, 90 - + 23J on June 21, and 90 - < - 23i on December 21. Thus the noon elevation of the sun changes by 47 during the year, and this causes a great change in the insolation received at any one point. The length of the day also changes markedly throughout the year. In the latitude of New York State the length of the day varies from a little Changes in more than fifteen hours during the summer to a little less than Duration. n j ne ft ours during the winter. The accompanying table gives the greatest possible duration of insolation for various latitudes. 41 Latitude Duration 12 h 17 13 hr 49 16 h, (. 63 20 hr - 66 30' 24 hr - 67 21' Imo. 69 51' 78 ft 2 mo. 4 mo. 90 6 mo. 37. Since the three factors that determine the amount of insolation received have different values for different latitudes and times of year,! it follows that the insolation received varies with both time and! latitude. For the same place it is different at different times of year ; for the same date it is different at different places, that is, different latitudes, on the earth's surface. Figure 9 shows the amount of insolation received for the various Variation latitudes at five differ- with lati- ent times between the 20th of March and the 21st of June. The unit of insola- tion is the amount received in a day at the equator on March 21. On the 21st of June, for example, the amount of insolation Deceived at the North Pole is greater than that received at the equator. As far as nearness to the sun is concerned, the amount received at both the equator and the pole would be the same ; as far as directness is con- cerned, at the equator the sun at noon has an elevation of 66j while at the pole its elevation is but 23i. Due to this cause, much more energy would be received at the equator. The duration of insolation, how- ever, is twelve hours at the equator and twenty-four boars at the pole. When the combined effect of the three factors is determined it is found that the amount of energy received at the pole is greater than that received at the equator. Figure 10 shows the amount of energy received at three different lati- jjj 20 30 10 .60 60 80 50 FIG. 9. Variation in the Insolation with Latitude at five Different Dates, (after WIENER.) HEATING AND COOLING OF THE ATMOSPHERE 35 tudes on the earth's surface < two maxima, one on the 2L S( ptember t These are varia the dates when the sun ^^ is directly over head at noo the insolation thus falls directly. The March max is greater than the Septembe because the earth is nearer 1 sun at that time. The du and directness are the sai both cases. The amount solation received at the Pole is zero until the 21st of fl when it rises rapidly to a ma? on the 2la/t of June, then TV/ luring the y 5t of March tion Lime. , JAN. n and FEB . most MAR . imum APRIL ir one MAY ,/ JUNE $ the XT JULY PtylOn AUG. ae in SEPT. f in- OCT - North r u DEC ' Larch, JAN . mum drops FlQ -ear. At the equator there are and the other on the 23d of p P P *p r j-^ o *. ea oo o \ \ \ FEB. MAR. APRIL MAY JUNE JULY AUG. SEPT. OCT. NOV. DEC. JAN. \ f ~- _ -~ .. - -^ ^ ^- -, s^ } ) \ S ^** .. ^s' s\ f l E ( '( \ V p p p 1*. 0> 00 10. Annual Vtfnatior 9 fed i in the Insola- ent Latitudes. back toytftrQ again on the 2 3d Of tion Received at thu^Diffei September. 38. Distribution ove the earth is shown by the values of insolation for fo ent latitudes. The unit is t in a day at the equator on distribution of insolation over ying table, which gives ^rent dates and six differ- nt of insolation received table. 21st of March. LATITUDE ~*~S 20 40 60 90 - 90 March 21 1.000 0.934 0.763 0.499 0.000 0.000 June 21 0.881 1.040 1.103 1.090 1.202 0.000 Sept. 23 0.984 0.938 0.760 0.299 0.000 0.000 Dec 21 .. .... 0.942 0.679 0.352 0000 0000 1 284 Since the amount of insolation varies with both latitude and time, this cannot be expressed in the ordinary way by means o a ' Figure 11, aowever, from DAVIS'S Elementary Meteorology, represents tiis relation. The time is plotted along the hori- zontal axis aid latitude along the other axis. The amount of tion of in- insolation foi any definite time and latitude is indicated by receded. the length of ti? perpendicular to the plane of the axes. The surface which pisses through the ends of these perpendiculars thus represents f he vacation in insolation. The curves given in FTP- 9 36 METEOROLOGY represent the intersection with this surface, of planes perpendicular to the time axis at the dates in question. The curves given in Fig. 10 are the intersection with this surface of planes passed perpendicularly to the latitude axis at the latitudes in question. In all of the foregoing sections, 35 to 38 inclusive, it is the distribution of the insolation at the outer limit of the atmosphere which has been FIG. 11. Variation in the Insolation with Latitude and Time. (From DAVIS'S Elementary Meteorology.) considered. The distribution over the earth's surface would be the same provided the earth had no atmosphere or provided the earth's atmosphere allowed all of the insolation to pass through it. The question as to what portion is absorbed in the atmosphere and what portion actually reaches the earth's surface will be discussed in sections 43 and 44. THE INTERRELATION OF MATTER AND RADIANT ENERGY Reflection. Whenever ether waves strike a materiil medium and are turned back, they are said to be reflectel. This is entirely analogous to the reflection of a sound onvater wave from a rigid barrier. There are two kinds of reflation ; regu- lar or mirror reflection, and irregular or diffuse Bflection. In the case of regular reflection the angle made oy the incident ray with the perpendicular to the reflecting surface is /qual to the angle 39- Definition. The two kinds of re- flection. THE HEATING AND COOLING OF THE ATMOSPHERE 37 ' made by the reflected ray. In the case of diffuse reflection the reflected rays pass off in every direction and are scattered. Burnished metal and ground glass are examples of these two kinds of reflectors. The best reflector known is burnished silver, and this reflects abj^98 per cent of the incident rays when these are perpendicular to the surface. The insolation from the sun on reaching the earth falls upon a variety of things, such as pure air, dusty air, clouds, water, etc. These p eflect i on things arranged in order of reflective power are as follows : of insoia- water, snow, cloud, dusty air, earth, pure air. Of these pure air reflects practically nothing, while snow and- water reflect from 30 to 50 per cent. 4. Transmission. Whenever ether waves are allowed by a body- to pass through it they are said to be transmitted^ A water analogy would be the passage of an ocean wave through the meshes of a fish Definition net suspended in it. The best transmitter known is rock salt. All transmission is selective in character ; that is, each body allows waves of certain lengths to pass through it "more readily than others. In the case of the atmosphere, the longer waves are the ones which selective are most readily transmitted. Glass transmits well the light transmis- waves, but does not transmit as readily the longer and shorter s ether waves. I^ock salt is the best transmitter because it transmits well ether waves of practically all lengths. The various things upon which insolation may fall arranged in order of excellence as regards The ^ ans _ transmission are as follows : pure air, dusty air, water, snow, mission of cloud, earth. Of these the earth transmits practically noth- msolation - ing, while pure air transmits more than 90 per cent of the insolation which falls upon it. 41. Absorption. Absorption takes place whenever an ether wave enters a body and is destroyed by it. A water analogy for this process can also be given. Suppose a pond to be covered with logs which are in contact with each other. If a water wave strikes these logs, the first ones are set in motion, and the wave loses intensity and soon ceases to exist. The logs grind against each other and by friction are soon brought to rest. The energy of the water wave has Absorption been transformed into energy of motion on the part of the logs, of insoia- The best absorber known is carbon. The various things upon * which insolation may fall, arranged in order of excellence as regards absorption, are as follows : earth, snow, cloud, water, dusty air, pure air. 42. There are four effects of absorption. (1) It may heat the body. In this case the energy of the ether waves has been used up in exciting 38 METEOROLOGY a more vigorous motion on the part of the molecules. (2) It may cause vision. Whenever ether waves of certain wave lengths enter the eye The four an( ^ ^ u P on the rods and cones of the retina, they impart effects of a stimulus which results in vision. In the human eye absorption. waveg of greater length than 0.000075 centimeter or shorter than 0.000036 centimeter produce no effect. The color perceived ranges from red to violet, depending upon the length of the wave. (3) It may cause chemical reaction. When ether waves are absorbed by certain compounds, the molecular activity which is excited is so great that the molecules break apart and new compounds are formed. This is the basis of photography. Certain salts, usually silver salts, on the photographic plate are decomposed by the light rays, and thus a perma- nent impression on the photographic plate is made. But a small amount of ether energy is used up in this way, however. A very large amount of insolation is consumed by plants, as described in sec- tion 15, when the green cells using the energy of insolation com- bine the sap and CO 2 of the atmosphere, making the complex molecules from which its own tissues are built up. The longer ether waves are usually more efficacious in causing heat, the inter- mediate ones in producing vision, and the shorter ones in causing chemical decomposition. (4) It may cause change of state. If inso- lation falls upon a body of water, a large amount of energy is used in evaporating the water, that is, in changing it from the liquid to the gas- eous state without changing its temperature. Energy used in this way is called latent heat. 43. Actinometry. 1 An actinometer is an instrument for measuring the insolation received from the sun. Actinometry treats of the use Thefunda- of this mental . ment in making formula. the measure- ments. The fundamental formula in actinometry is FIG. 12. Showing the Contrast in the Thickness of fj = Co?. H here represents Air passed through by Vertical and Oblique Rays. the amount of energy re- ceived on a unit of surface at right angles to the sun's rays, a is the percentage transmitted by the atmosphere when the rays fall vertically. Its value would be unity for perfect transmission. I is the thickness of the atmosphere, considering the vertical thickness as unity. C is^a 1 For a bibliography of the articles on solar radiation see Bulletin of the Mount Weather Obs.. Vol. 3, part 2, pp. 118-126. 1 THE HEATING AND COOLING OF THE ATMOSPHERE 39 constant, ordinarily known as the " solar constant." The thickness of the layer of air through which the sun's rays pass increases rapidly as the sun nears the horizon. Figure 12 shows well the contrast in the length of path for vertical and oblique rays. The following table gives the various values of / for the different altitudes of the sun : Altitude of sun 5 10 20 30 50 70 90 Thickness of the atmosphere in units 35.5 10.2 5.56 2.90 1.99 1.31 1.06 1.00 It will be seen that the thickness of air traversed by the sun's rays when rising or setting is more than 35 times the thickness when the 1 (jr sun is directly overhead. The term a 1 is equal to one if I is zero or if a is equal to one. a equaling one means that the atmosphere is transmitting all of the in- solation. / equaling zero corresponds to a zero thickness of the atmosphere,, that is, to no atmosphere at all. In either i ng of the case H would equal C. C is then the amount of energy re- solar con ~ Stflnt* ceived by a given unit area at right angles to the sun's rays, provided there were no absorption in the atmosphere, or provided the atmosphere wore not present. If the elevation of the sun above the horizon be known, I can at once be computed from the table given above. Absolute or relative values of H can be^ietermined in several different ways. Seven ways will be sirfaply mentioned here, and the reader must be re- me t ho( i s of fericd to special articles on the subject for a more detailed determining treatment. The various methods make use of calorimetrical received? 7 apparatus, a pyrheliometer, a thermoelectric couple, a bolom- eter arrftngqment, chemical decomposition, a black bulb thermometer in vacuo, or photographic paper. The black bulb thermometer will be considered later HJL section 70. The method by means of photographic paper, although not especially sensitive, is so ingenious as to be worthy of a few words of explanation. The. blackening of an ordinary piece of photogra phic paper when exposed to the rays of the sun depends upon the intensity of the sun's rays and the time of exposure. The ratio of the times of exposure required to produce the same amount of blacken- ing for two different values of intensity gives the ratio of those intensities. Thus relative values of H can be readily determined. If H has tveeii determined for two different values of I, by means of the equation given above, C and a can be computed. There The method are two ways of getting different values of I One is by mininTc taking simultaneous observations from a mountain top and a and a. 40 METEOROLOGY near-by valley, another is to take observations at different times of day, thus getting different thicknesses of atmosphere. Numerous measurements have been made under the most varied con- ditions and at many places. The different values for C are not in good The values agreement, but the values ordinarily used at present are of C and a. ^ wo or three calories per square centimeter per minute. A calorie is one one-hundredth of the amount of heat required to raise the temperature of one cubic centimeter of water from the freezing to tho boiling point. When the sky is clear, the vahie of a, the trans- mission coefficient, is about 75 per cent. Under various condi- tions, howeve^^alues as low as even 10 or 20 per cent have been observed. The best modern values of the solar constant are without doubt those which have been ob- tained since 1904 by Mr. C. G. Abbot, Director of tH e Astro- physical Observatory of the Smithsonian In- stitution. His results are in much better agreement and would seem to indicate a 13 AUGUST, 1 1888 Temp: max. 210 C.. jmin. Bel. Jiumidity. M, Sky very cle ir VI VII VIII IX X XI NOON I II HI JV V FIG. 13. - Insolation on Mont Ventoux. (From HANN'S Lehrbuch der Meteor ologie,) 7 MONTlPEU iv/n i rc.ui.ie.n, fu 1 13 AUGUST, 1888 Tem'p- iJx. 3ltO C., ml "' Rel.j humidity; 70^. Sky Wind: east, moderate. Pressure- j 761. 7 mml VI V.Il Vlli IX X XI NOON I II III IV V FIG. 14. Insolation at Montpellier. ( From HANN'S Lehrbuch der Meteor ologie.') The effect of atmos- pheric ab- sorption on the insola- tion re- ceived dur- ing the day. value a little less than 2 (between 1.9 And 2.0). Figures 13 and 14 show well the effects of atmospheric .of ^orption. The amount of insolation received on a surface having an area of a square centimeter and kept constantly at right angles to the sun's rays is here given for the same day, August 13, 1888, on Mont Ventoux with an elevation of 2000 meters, and at Montpellier with art elevation of 40 meters. If thn atmos- phere had transmitted all the incident insolation, the value VD COOLING OF 1 : 3JE ATMOSPHERE 41 i rp al to the solar constant at both places from the n :.: intii sunset. The graphs show well the gradual rise in value of the insolation during the morning and the gradual falling off during the afternoon, due to the varying thickness* of the at- mosphere through which the rays were coming. The largest value on Mont Ventoux is 1.6, while the largest value at Montpellier is 1.2. This difference is due to the absorption of the 1960 meters of air which is their difference in elevation. It will also be noticed that the maximum values were received several hours before noon. The reason for. this is that the air is always less transparent in the afternoon than during the morning. The accompanying table from Angot' gives the amount of insolation received at the earth's surface in the course of a year at various latitudes, first on the assumption that the atmospher^ transmit? a-lL--the e^-ct of the insolation, and secondly on the assumption that 60 per of atmos- cent is transmitted. The unit of insolation here \fsed is the so^tiorfon amount of insolation received at the equator during oi\e day, the insoia- on the 21st of March. It will be seen at once that the "dim- Reived at inution in the amount of insolation at the pole is much more different marked than at the equator. It will also be seen from Figs. 13 and 14 and the table that not more than one third of the in- solation received from the sun actually reaches the earth's surface on any particular day. The transmission coefficients for various places on the earth's surface during any given day probably vary from 10 per cent to perhaps 90 per cent. Sixty per cent would seem to be a large average for the whole earth, but even with that figure, as seen from the table, barely more than one third of the insolation reaches the earth's surface. a = 1.0 a = 0.6 EQUATOR 10" 20 30 40 50 60 70 80 POLE 350.3 170.2 345.5 166.5 331.2 155.1 307.9 137.6 276.8 115.2 239.8 90.6 199.2 67.4 166.3 47.7 150.2 33.5 145.4 28.4 45. Behavior of the ocean as regards reflection, transmission, and absorption. The ocean reflects about 40 per cent of the insolation which falls upon it, and transmits the remainder to considerable The ocean depths, but eventually absorbs all the insolation which changes but is transmitted. The ocean changes extremely little in tern- perjure* 5 " perature between day and night, and there are five reasons for between day this : (l) So large an amount, namely, 40 per cent, of the in-. " solation is reflected and thus lost as far as heating the water is concerned. (2) The insolation which is absorbed is transmitted to considerable 42 METEOROLOGY depths, and thus it is not a thin surface layer, but a layer of considerable depth that is involved. (3) A considerable amount of evaporation takes place from the surface of the ocean, and the insolation is used in causing this change of state and not used in causing a rise of temperature on the part of che water. (4) It requires a larger amount of heat to raise the temperature of a given quantity of water than any other substance. Tlie technical expression for this is that the specific heat of water is larger tl.an that of any other substance. For this reason a very small rise in temperature takes place as the result of absorbing a considerable amount of insolation. (5) The water of the ocean is in continual motion. Thus again it is not a small surface layer which is involved, but a considerable amount of water. On account of these five reasons, the temperature of the ocean rises very little during the day. Because it is in continual mo- tion, transmits readily, and has a high specific heat, it cools but little dur- ing the night. The behavior of the ocean may then be briefly sum- marized by stating that its temperature change between day and night is extremely small, never amounting to more than one or two degrees. 46. Behavior of the land as regards reflection, transmission, and absorption. Dry ground reflects but a few per cent of the insolation which falls upon it, transmits practically none, and thus absorbs nearly all. The rise in temperature of dry ground under insolation perature is very great. In the first place, it absorbs nearly all of the Ch oundbe- ms l a ti n > an d, being. opaque, this takes place in a thin sur- tween day face layer. Its specific heat is far les? than that of water, and, being a solid, there can be no mixture as in the case of the ocean. Since the ground is a good Absorber, it is also a good radiator and thus for similar reasons it cools rapidly at night. If the ground is wet or covered with vegetation, the change in temperature between day and night is much less than for dry ground, but always greater than for the ocean. 47. Behavior of the atmosphere as regards reflection, transmis- sion, and absorption. Pure air reflects but a minute quantity of the The chan e mso ^ a ^^ on which falls upon it, absorbs as little, and transmits in tempera- practically all. As a result it changes temperature but I'tmo^here very little between day and night. Dusty air reflects more between and absorbs more, but is still one of the best transmitters. The atmosphere contains more dust, water vapor, and carbon dioxide near the earth's surface, and thus the change in tem- perature between day and night increases with nearness to the earth's sur- face. The reflection and absorption on the part of the atmosphere are THE HEATING AND COOLING OF THE ATMOSPHERE selective. In the case of reflection it is the shorter waves which are most effectively scattered, while in the case of absorption it is the longer ether waves which are chiefly concerned, and this is particularly the case when the amount of water vapor and carbon dioxide is large. The temperature change in the atmosphere between day and night due to its behavior as regards insolation would probably be larger than the temperature change of the ocean, but never as large as that of the ground. The contrast between the temperature of the ground and the tempera- ture of the air above it is well brought out by the following typical example. The observations were made in 1895 at Tiflis with a latitude of. 41 43' and an elevation of 410 meters. The trastbe- averages of the temperature observations (centigrade) made tw lace with a very small change in elevation. Figure 19 shows the actual ertical temperature gradient obtained at Berlin on the 19th of October, 893. bv m^ano '- f " Kn oo, T-I^O ooo ens i O n took place at 10.18 50 METEOROLOGY 6000 5000 4000 .3000 \ A.M., the balloon reached the highest point at 1.45 P.M., and descended to the ground at 4.20 P.M. The figure thus represents the vertical temper- ature gradient between 10.18 in the morning and 1.45 in the afternoon. _^_ r _^ r __^ r _ T _^__ The average value of the vertical temperature gradient The average for ail times of value. dayt for all gea . sons of the year, and for all places and weather, is ordina- rily considered to be 1 Fah- renheit for 300 feet, or 0.6 Centigrade for 100 meters. In order to understand the change in the vertical tem- perature gradient The three ^ . characteris- during the day, tic gra- three character- dients. istic gradients must be first considered. These three gradients are the average vertical tempera- ture gradient which occurs usually about 9.00 in the morning and 8.00 in the evening, the gradient which exists at the time of minimum or lowest temperature, and the gradient which occurs at the time of maximum or highest temperature. These three gradients are represented graphically in Fig. 20 for a typical October day. The maximum temperature during this day is assumed as 60*F., the minimum temperature as 30 F., and the average temperature for the day as 45 F. This average temperature would occur about 9.00 in the morning and 8.00 in the evening. The straight line B in the figure represents the average vertical temperature gradient and is plotted with a fall of 1 F. for 300 feet. Thus at a height of 9000 feet above the earth's surface the temperature would be 30 lower than at the earth's surface. The gradient C, which represents the vertical tempera- ture gradient at the time of the maximum temperature, follows the 2000 1,0.00 -28 -20' -10 C 10 TEMPERATURE. CENTIGRADE FIG. 19. Graph Illustrating Vertical Temperature Gradient. (From ASSMANN'S Wissenschaftliche Luftfahrten.) THE HEATING AND COOLING OF THE ATMOSPHERE 51 average vertical temperature gradient closely in the upper atmosphere, but departs from it more and more as the earth's surface is approached. The gradient A, which represents the vertical temperature gradient at the time of the minimum temperature, also follows the average gradient P4OUU 11700 10800 9900 9000 8100 fc Hi 7200 Z g 6300 gj 5400 Ul 4500 3600 2700 1800 900 B \ 1 A \ Y \ V ^ \ A \ - \ V \ \ 3 s \ \ \ \ s S v. D\ e> A \ 3 \ \ \ Y i \ \ x \ \ ^ \ \ \ \\ \ \ \, \ \ \N \ \ > \ ^ \ \ \ \\ \ \ 2 5 s \ ^ 10 20 X) 40 TEMPERATURE, FAHRENHEIT FIG. 20. Temperature Gradients. 50 closely in the upper atmosphere but has a sharp bend at low altitudes, usually less than 2000 feet. The normal change in the vertical tem- perature gradient during the day is as follows: At the time of the min- imum temperature it approximates to the form A. It then The daily straightens out and at about 9.00 in the morning is following variation in closely the average vertical temperature gradient, which is te mpera- a represented graphically by the straight line B. It then ture gra- curves in the other direction, and by the time* the maxi- mum temperature is reached it has assumed the form C. It then 52 METEOROLOGY straightens out and by about 8.00 in the evening has reached the average vertical temperature gradient again. It then swings over until, when the time of minimum temperature is again reached, it has approximately the form A. The daily range of temperature is defined as the difference between the highest _and lowest temperature which occurs during the day. It will be seen in the figure That the daily range decreases rapidly with elevation. At the earth's surface it was taken as 30 ; at the height of a mile, as indicated at H, it is only 10, while at the height of two miles, as indicated at G, its value is about 6. It is not necessary to take observations at great elevations in order to ascertain these facts. The observations taken at the top and base of the Eiffel Tower in Paris, at an elevation of 302 and 2 meters respectively, show the same results. The daily range decreases by 62 per cent in winter and by 45 per cent in summer for this difference in elevation of only 300 meters ; and for 12 hours during the night in winter and for 8 hours during the night in summer the temperature at the top averages higher than at the base. The average vertical fremperature gradient must not be confused with the adHh a *' n rfllif of cooling of unsaturated rising air. The one is 1 Fajirenheii fatSDQJeet, while the other is 1.6 Fahrenheit for batLTraTe ^00 ^ f The vertical temperature gradient states that on of cooling the average there is a decrease in temperature of 1 F. with withThe 6 eac h 300 feet of elevation. According to the adiabatic rate vertical tem- of cooling, rising air by expansion cools, when not saturated gradient. with moisture, 1-6 F., for every 300 feet. Should air with a temperature of 30 rise, it would cool adiabatically due to ex- pansion 1.6 for 300 feet and the behavior of a quantity of air rising with this initial temperature is shown graphically by D in Fig. 20. The be- havior of quantities of air rising with initial temperatures of 45 or 60 is shown by E and F respectively. H 55. The isothermal layer. The facts which have just been stated concerning the vertical temperature gradient, its average value, and the daily change in it apply only to the lower por- verticai tem- tion of the atmosphere, say, from the earth's surface to a height gradient be- ^ two or three miles. Between the height of two miles and tween three six miles or somewhat more, the characteristics of the vertical miles!* temperature gradient are quite different. Here it is much/ more regular and the average value is somewhat greater than! near the earth's suriace. The changes in the gradient between day and' THE HEATING AND COOLING OP THE ATMOSPHERE 53 night, summer and winter, and with the weather, are all extremely small. Of course, at any given height, the temperatures are lower in winter than in summer, but the gradient is almost exactly the same. It is also true that, at the same height, the temperatures are usually higher when the weather is fair than when a storm or area of low pressure is present, but the gradient changes very little with the weather. After a height of six miles, or in some cases much more is reached, the temperature seems to remain constant with elevation, or may, indeed, increase slightly with altitude. This layer in which the tern- The iso _ perature remains so nearly constant is usually spoken of as thermal fthe isothermal layer or the warm stratum of the atmosphere. layer * 'This subject has been investigated especially by Teisserenc de Bort, Assmann, and in this country by Humphreys and Rotch, the director of the Blue Hill Observatory near Boston. In an article in the Monthly Weather Review of May, 1908 Professor Rotch writes : " This in version of temperature was first discovered by M. Teisserenc de Bort with the sounding balloons sent up from his observatory at Trappes, near Paris, France, in 1901, and almost simultaneously by Professor Assmann from similar German observations. Since then almost all the balloons which have risen more than 40,000 feet above central Europe (that is, near latitude 50) have penetrated this stratum, without, however, determining its upper limit. Teisserenc de Bort early showed that its height above the earth, to the extent of 8000 feet, varied directly with the barometric pressure at the ground. Mr. Dines gives the average height of the isothermal layer above England as 35,000 feet, with extremes of nearly 50 per cent of the mean. Ob- servations conducted last March by our indefatigable French colleague, Teisserenc de Bort, in Sweden, just within the Arctic Circle, show that the minimum temperature occurred at nearly the same height as at Trappes, namely, 36,000 feet, although Professor Hergesell, who made use of sounding balloons over the Arctic Ocean near latitude 75 N., during the summer of 1906, concluded that the isothermal stratum there sank as low as 23,000 feet. " During the past three years the writer has dispatched 77 sounding bal- loons from St. Louis, Mo., U.S. A., latitude 38 N., and most of those which rose higher than 43,000 feet entered the inverted stratum of temperature. This was found to be somewhat lower in summer, but the following marked inversions were noted last autumn: October 8, the minimum temperature of 90 F., occurred at 47,600 feet, whereas at the maxi- mum altitude of 54,100 feet the temperature had risen to 72 ; October 54 METEOROLOGY 10, the lowest temperature of - 80 was found at 39,700 feet, while - 69 was recorded at 42,200 feet, showing a descent of nearly 8000 feet in the temperature inversion within two days. The expedition sent out jointly by M. Teisserenc de Bort and the writer, on the former's steam yacht Otaria, to sound the atmosphere over the tropical Atlantic during the summer of 1906, launched sounding balloons both north and south of the equator within the tropics, and although some of these balloons rose to nearly 50,000 feet, they gave no indication of an isothermal stratum. In fact, the paradoxical fact was established that in summer it is colder 10 miles above the thermal equator than it is in winter at the same height in north temperate regions. This results from the more rapid decrease of temperature in the tropics and the absence of the numerous temporary inversions which, as Mr. Dines has pointed out, are common in our re- gions below 10,000 feet. If, therefore, as seems probable, the isothermal or relatively warm stratum does exist in the tropical and equatorial regions, it must lie at a height exceeding 50,000 feet, from which height, as the data quoted show, it gradually descends toward the pole, at least in the northern hemisphere." This isothermal layer exists, then, at a great height over the equator, has an average height of about six miles over the middle latitudes, and comes still nearer to the earth's surface in polar regions. Its average tempera- ture during the summer is about 60 F., while its average winter tem- perature is about 71 F. When the weather -is fair and is c6ntrolled by an area of high pressure, it is at a greater height and about 14 F. colder, as an average for all the quadrants of a high than when the weather is stormy and an area of low pressure is dominant. In this respect its temperate is the opposite of the air temperature between a height of two and six miles. The isothermal layer also begins at a lower altitude in winter than in summer. Its upper bounding surface has never been determined. 56. Many articles dealing with the isothermal layer will be found in the periodical literature since 1900, and the reader must be referred to A possible these for a full treatment of the subject. No single univer- expianation sa jiy accepted explanation of all the facts in connection thermal with the isothermal layer has yet been given. In what fol- iayer. i ows a p OSS ible explanation of some points is presented. Since the height of the isothermal layer is always more than five miles above the earth's surface, the amount of water vapor present must be extremely small, if not entirely negligible ; furthermore the amount of carbon dioxide is much less than at the earth's surface. Thus of the THE HEATING AND COOLING OF THE ATMOSPHERE 55 three ingredients which are the chief causes of the absorption and radia- tion on the part of the atmosphere the dust alone remains. At a height of five miles or more the dust which exists in the atmosphere can come from but one source. The dust blown up from the earth or that which comes fro^the evaporation of ocean spray or from volcanoes cannot, except Hinder unusual conditions, penetrate to this height. The dust which exists above five miles must come primarily from the meteors. This is put into the atmosphere at a height of 50 miles or more and as it slowly settles through the atmosphere it must cause the distribution of dust to be uniform below this height. Now the one process which is op- erative in heating and cooling the atmosphere at a height of more than five miles is the absorption of radiation bythe dust particles and the* radiation of heat by them to space above and to the earth below. No account is here taken of the adiabatic changes in temperature of the whole layer due to its rising or falling. The temperatures would be different,' but the layer would still remain isothermal. It can be shown mathematically that the radiation received by a body at a given distance above an in- finite plane (and the earth's surface may be considered as such) is inde- pendent of its height above that plane. Thus the amount of insolation from the sun and the amount of radiation from the earth absorbed by the dust paticles will be independent of the distance above the earth's sur- face, ^ne should thus expect that each given volume of air having in it the same number of dust particles would gain by absorption the same amount of heat and would lose at night by radiation the same amount of heat. But the density of the air grows steadily less with elevation. We have thus the same amount of heat applied to a smaller quantity of air, and one would thus expect an increase of temperature with elevation and decreasing density. The dust particles are probably not the only absorbers and radiators. Due to the action of ultra-violet light and the aurora borealis, the ozone content of the upper air may be fairly large. Ozone absorbs certain wave lengths readily, and the other gaseous con- stituents of the atmosphere may play a small part. ^57. Inversion of temperature nocturnal stability. It was a fact of early observation that on the still, clear nights of winter and during the clear, frosty nights of the late spring or early fall, the tempera- Definition ture was somewhat higher on a mountain top or at a small of mversion - elevation above the earth's surface than on the earth's surface itself. When the temperature increases instead of decreases with elevation, it is spoken of as an inversion of temperature. It was formerly thought that inversions of temperature were of comparatively rare occurrence. It is 56 METEOROLOGY now known, however, that they occur on practically all nights. When an inversion of temperature occurs, the vertical temperature gradient has the form of A in Fig. 20. During the night the atmosphere is entirely stable. If a quantity of air with a temperature of, say, 30 on a night when an inversion of Wh the at temperature exists should be pushed up a short distance, mosphere is it would expand and cool at the rate of 1,6 F. for 300 feet, stable at an( j fi n( j itself cooler than its surroundings. It would then immediately drop back again or, more accurately stated, it would never have spontaneously commenced to rise at all. Thus at night the atmosphere is entirely stable. 58. Diurnal instability conditions of convection. During the day the atmosphere may find itself in a state of unstable equilibrium. Suppose when the vertical temperature gradient has the form mo h s y phere a is C in Fi S- 20 > a quantity of air with a temperature of 60 unstable commences to rise. It will expand and grow cool at the rate during th of lQ o F for 30Q feet> ^ g ghown by the gtraight Hne F in the figure it will continually find itself, in spite of the cooling, warmer than its surroundings, and will continue to rise. It will rise until it has cooled to the temperature of its surroundings ; or, expressed graph- ically, it will rise until the straight line F intersects the vertical tempera- ture gradient, C. The conditions for convection may now be summarized. (1) The atmosphere must be heated at the bottom. (2) The atmos- 1 ne tnree conditions pnere must be heated at one point more than at surround- ing places. (3) If a quantity of air which starts to rise is to continue its ascent, the average vertical temperature gradient must be greater than 1.6 for 300 feet. How THE ATMOSPHERE is HEATED AND COOLED 59. The present chapter may be summarized and' the principles which have been stated may be illustrated by considering the processes which are operative in the diurnal heating and cooling of the atmosphere near the earth. In the heating of the atmosphere five processes must be considered. These are in order : the absorption of insolation, the absorption of the The five radiant energy from the earth, conduction from the earth, heating pro- mixture by means of the wind, and convection. As was seen in the study of actinometry under the most favorable con- ditions, when the sun's rays fall vertically, only 90 per cent of the insola- THE HEATING AND COOLING OF THE ATMOSPHERE 57 tion is transmitted and the remaining 10 per cent is absorbed. If the rays do not fall vertically and under unfavorable atmospheric condi- tions, a much larger percentage of the insolation is absorbed by the atmosphere. The dust, water vapor, and carbon dioxide are the chief absorbers, and they increase in quantity with nearness to the earth's surface. As a result a rise in temperature of perhaps a degree would occur in the upper atmosphere and a rise in temperature of perhaps 3 or 4 would occur near the earth's surface. The earth becomes warmed by absorbing the insolation which reaches it and radiates its heat in the form of long ether waves to space. These longer ether waves are absorbed even more readily than the shorter ether wavds which constituted the insolation. This constitutes the second source of heating, and here again the amount of absorption and the result- ing heating would be greater nearer the earth's surface. In connection with conduction it was seen that only a layer 2 or 3 feet deep of the atmos- phere would be heated by being in contact for many hours with the warm ground. This process, then, would be effective only in heating the layer of air close to the ground. The wind, however, so thoroughly mixes the various layers of air which are in contact with each other that the heat imparted by conduction to the 2 or 3 feet nearest the earth's surface would be distributed over perhaps two or three thousand ^feet near the ground. Since this lower layer is heated by conduction, convection )rtit itloT iftti( would take place, and this circulation would transfer heat fro surface to altitudes of from one to several miles. ? As a resultof these five heating processes the temperature of the air is raised but little at great altitudes above the earth's surface, and the amount of heating increases rapidly as one approaches the earth's surface. The vertical temperature gradient C in Fig. 20 has thus been explained. There are three processes operative in the cooling of the atmosphere, namely, radiation to the cool ground and to space, conduction, and mixture by the wind. The air, filled with dust, water vapor, The three and carbon dioxide, radiates heat as well as it absorbs it. The cooling pro- ground at night cools rapidly and the air loses heat by radi- c ation both to the open sky and to the cooling ground. Conduction, again, would cool the two or three feet of air in immediate contact with the earth's surface, and the wind, although on an average less at night than during the day, would mix this layer with the two or three thousand feet of air above it. Since convection is not operative at night, the marked cool- ing must take place very near the earth's surface. The vertical tempera- ture gradient A at the time of lowest temperature has thus been explained. 58 METEOROLOGY QUESTIONS (1) Describe the molecular structure of matter. (2) How many kinds of atoms are there? (3) Describe the structure of the molecule and atom. (4) Describe the luminiferous ether. (5) What properties must it have ? (6) How are molecules related to each other? (7) Define heat and temperature/ (8^ How are ether waves caused? (9) Define radiant energy." (10) Dis- tinguish between an hypothesis and a fact. (11) Name the three possible sources of atmospheric heat. (12) Prove in two ways that all other sources of heat than the sun are relatively negligible. (13) Describe the size and condi- tion of the sun itself. (14) Give illustrations of the amount of energy received from the sun. (15) Define insolation. (16) Upon what three things does the amount of insolation received depend? (17) Explain how the amount of insolation received depends upon the obliqueness of the rays. (18) Describe the earth's ormfr ground the sun. (19) What is the effect of the change in dis- tance of the earth from the sun? (20) What are the two effects of the change in presentation of the earth to the sun's rays? (21) How great is the change in directness during the year ? (22) State the change in duration at various places during the year. (23) Explain why insolation varies with latitude and time. (24) How may this relation of insolation to latitude and time be expressed graphically ? (25) Define reflection. (26) Describe the two kinds of reflection. (27) Arrange the substances upon which insolation may fall in order of excellence as regards reflection. (28) Define transmission. (29) What is meant by selec- tive transmission ? (30) Contrast rock salt and glass as transmitters. (31) Ar- range the substances upon which insolation may fall in order of excellence as re- gards transmission and absorption. (32) Give the water analogy for the three processes, reflection, transmission, and absorption. (33) What are the four effects of absorption? (34) Define actinometry. (35) State the fundamental formula in actinometry. (36) What is represented by each letter in the formula? (37) Define the solar constant. -(38) In how many ways may absolute and relative values of H be determined? (39) Describe the photograph paper method of de- termining relative values of H. (40) Describe the method of determining C and a. (41) What values have been found for C and a? s(42) State the effect of atmospheric absorption on the insolation received at the earth's surface in dif-* ferent latitudes. (43) What is the behavior of the ocean as regards reflection, transmission, and absorption? (44) Why is the temperature rise during the day extremely small? (45) Why does the ocean cool but little during the night? (46) State the behavior of dry ground as regards reflection, transmission, and absorption. (47) Why is the rise in the temperature of dry ground under in- solation so large? (48) .What is the effect of dampness or vegetation on the temperature change? (49) State the behavior of the atmosphere as regards reflection, transmission, and absorption. (50) Name the three chief absorbing components of the atmosphere. (51) How do the temperatures of the air and the ground compare during the day and at night? (52) How do the tempera- tures of the air and the ocean compare during the day and at night ? (53) Define conduction. (54) Name several poor and several good conductors. (55) To what height would the air be heated by conduction alone? (56) Describe and explain convection in a liquid. (57) Where should convection be expected to take place in the atmosphere? (58) State the difference between convection in a liquid and in a gas? (59) What is meant by the adiabatic rate of cooling of rising air? (60) Name the three evidences of local convection. (61) De- scribe and explain the mirage. (62) Describe and explain dust whirlwinds. (63) Describe and explain cumulus clouds. (64) What is meant by vertical THE HEATING AND COOLING OF THE ATMOSPHERE 5<) mperature gradient? (65) Name the three ways of obtaining the vertical mperature gradient. (66) Why are mountain observations no longer used? 7) Describe the use of balloons for obtaining observations. (68) Describe le use of kites for making meteorological observations. (69) With what does le vertical temperature gradient vary? (70) What is the numerical value of ae average vertical temperature gradient? (71) Describe the three character- ;tic gradients which occur during the day. (72) Define daily range. (73) How may the daily range be represented graphically ? (74) How may the adiabatic rate of cooling be represented graphically ? (75) What is meant by the isothermal layer? (76) At what height does it occur? (77) What are its characteristics ? (78) Define inversion of temperature. (79) Explain why the atmosphere is un- stable during the day. (80) State the three conditions of convection. (81) Name the five processes which are operative in the heating of -the atmosphere. (82) Describe in detail the effect of each process. (83) Name the three/processes operative in the cooling of the atmosphere. (84) State the effect of each of the three processes. % TOPICS FOR INVESTIGATION (1) The nature and characteristics of the atom. (2) The amount reflected, transmitted, and absorbed by the different things upon which insolation may fall. (3) The methods of obtaining the value of the solar constant. (4) Changes in the temperature of the ocean between day and night. (5) Changes in the temperature of the ground between day and night/ (6) The method of computing the adiabatic rate of cooling of air. (7) Mountain observatories. (8) Balloon ascensions for scientific purposes. (9) The use of kites in meteorology. (10) The variation in the vertical temperature gradient. (11) The isothermal layer. PRACTICAL EXERCISES (1) Contrast the amount of insolation received at several different places and at several different times by working out the exact value of the three factors. (2) Determine by the photographic paper method the transmission coefficient of the atmosphere on some cloudless day. (3) Observe carefully, or better photograph, some cumulus clouds. (4) Note the time when the cumulus clouds disappear in the late afternoon. (5) If possible, compare the observations of temperature made on some near-by mountain top and at some valley station. REFERENCES ASSMANN, BBRSON, and others, Wissenschaftliche Luftfahrten (3 vols.). ROTCH, Sounding the Ocean of Air. For recent articles and references on the isothermal layer see : GOLD, E., and HARWOOD, W. A., The present state of our Knowledge of the upper Atmosphere as obtained by the Use of Kites, Balloons, and Pilot Balloons, 8, 54 pp., London, 1909. The Mount Weather Bulletin (particularly articles by Humphreys). See Appendix IX for other books on upper air investigation. CHAPTER III THE OBSERVATION AND DISTRIBUTION OF TEMPERATURE THE DETERMINATION OF TEMPERATURE Thermometry, 60, 61. Thermometers, 62, 63. The real air temperature, 64. Thermometer shelter, 65. Sling thermometer, 66. Ventilated thermometer, 67. Thermographs, 68. Maximum and minimum thermometers, 69. Black bulb thermometer, 70. Thermometers for special purposes, 71. THE RESULTS OF OBSERVATION The observations, 72. Normal hourly, daily, monthly, and yearly temperature, 73-76. Diurnal, annual, and irregular variation, 77, 78. Temperature data, 79, 80. Differences of temperature with altitude, 81. Temperature differences over a limited area, 82. THE DISTRIBUTION OF TEMPERATURE OVER THE EARTH Construction of isothermal charts, 83. Isothermal lines for the year, 84, 85. Isotherms for January and July, 86, 87. Poleward temperature gradient, 88. Thermal anomalies, 89. Annual range of temperature, 90. Extremes of temperature, 91. Other temperature charts, 92. Polar temperatures, 93. THE TEMPERATURE OF LAND AND WATER Ocean temperatures, 94. Lake temperatures, 95. River temperatures, 96. Temperature below the surface of the land, 97. THE DETERMINATION OF TEMPERATURE 60. Thermometry. Thermometry, as the word indicates, has to do with the determination of temperature, and the instrument Definition. used is called a thermometer. 1 There are three systems of ther- 1 6tp/Jir) = heat; ^rpov = measure. 60 OBSERVATION AND DISTRIBUTION OF TEMPERATURE 61 mometry or thermometric scales, namely, the Fahrenheit, the Centigrade, and the Reaumur. The two standard temperatures in each system are the same, namely, the melting point of ice and the boiling point of water. In each case the ice and water must be pure and the pressure must have the standard value. These two standard grade, and temperatures are numbered differently in the three systems, In the Fahrenheit system the melting point is numbered 3u and the boiling point 212,; in the Centigrade system the melting point is numbered and the boiling point 100 ; in the Reaumur system the melt- ing point is numbered and the boiling point 80. The inter- vals between the two standard temperatures are thus 180, reiationTe- 100, and 80 respectively and in the ratio of 9 to 5 to 4. The tween the following formula thus indicates the relation between the three systems where F, C, and R denote the three corresponding temperatures : F - 32 = C = R 9 54' F 32 represents the F. interval above the melting point. Since the C. and R. thermometers have the melting point indicated by zero the intervals are expressed directly by C and R. A temperature expressed in any one of the three systems can be immediately changed into the cor- responding temperature in the other systems by means of this formula. 1 For example, Let F = 60 ; the equations then become : 60 32 C _ R, 28 __ C ._ / . /nr _ i cs nr | # _ 1 94 2 ~9~ ~5 ~4' 5 4 ! ( All temperatures below the zero of the scale are indicated by a minus sign and are read " so many degrees below zero." Thus 7 degrees below zero Fahrenheit is indicated by 7 F. Jjl OO /"Y 1 The formula ^= = may be modified so as to make the mental computation of 9 5 the Centigrade temperature for the Fahrenheit easier. C = |(F-32) = (F- 32) -555... = (F-32) (i + A-J + iJo *). The rule would be : subtract 32 from the Fahrenheit temperature ; take one half of it ; add to this one tenth of itself, one one-hundredth of itself, etc. Thus for 60 Fahrenheit : 60 - 32 = 28 * of 28 = 14 C=14 1.4 .14 .01 etc. C = 15.55 + 2 In Appendix II is given a table for converting Centigrade into Fahrenheit and vice versa. In Appendix III, a graphic comparison of the two scales is given. 62 METEOROLOGY The inven- tion of the thermom- eter. 6 1 . The thermometer was invented by Galileo and Sanctorius of Padua in 1590. Sanctorius had been a pupil of Galileo, and later became pro- fessor of medicine at the same university. Figure 21 repre- sents the thermometer used by Sanctorius for determining the feverishness of his patients. It consists simply of a glass globe opening into a narrow tube and partly filled with fluid. This is then inverted and dipped into a vessel of fluid. The bulb is taken in the hand and the feverishness is indicated by the height at which the liquid stands in the tube. If this early form of the thermometer was invented by Galileo and adopted by Sanctorius, or whether it was the combined invention of Galileo and Sanctorius, is uncertain. Within a few years, however, the instrument was inverted, and it then became a thermometer of essentially the same form as at present. For nearly 100 years after Confusion f . J and uncer- the invention of the thermometer all taintyfor was CO nfusion and uncertainty, for many years. . various fluids were used and various systems of graduation were employed. In order to get a constant temperature various devices were used. Some used the temperature of the water of a certain spring as constant. The tem- perature of the subcellar of the Observatory of Out of this chaos the three systems of ther- mometry emerged, because the founders of the systems made instru- ments of such high quality and in such large numbers that they came to be recognized as standard instruments. The Fahrenheit thermometer originated at Dantzig about 1714. The two great improvements introduced by Fahrenheit were the use of mer- The Fahr- cury as the fluid, and the use of two known temperatures for enheit ther- the graduation. The reason for choosing 32 as the melting mometer. p Om t o f ice and 212 as the boiling point of water is uncertain. It is known that he had traveled in Iceland, and it may be that the zero of his thermometer scale was the lowest temperature which he had ever experienced. It may be that he used the ^temperature of the human body for standardizing his instrument, intending to have this 100. The average temperature of the l '>ody, however, is only 98.6. A mixture of salt and snow ! by the Italian thermometer n akers for getting the zero their instruments. The temper- FIG. 21. The Original Thermometer of Sanc- torius. Paris was also used. OBSERVATION AND DISTRIBUTION OF TEMPERATURE 63 ature of this is nearly zero, and it may be that Fahrenheit used this for determining the zero. Fahrenheit was also a maker of astronomical instruments, and it may be that in order to use his machine for dividing circles for graduating the stem of the thermometer, it was necessary to have the interval between the melting and boiling points 180. The Centigrade thermometer was invented by Celsius and Linnseus at the University of Upsala in Sweden in 1742. The two fixed points are num- bered and 100 respectively. In the original thermometers The Centi _ the freezing point may have been numbered 100 and the boil- grade ther- ing point 0, so that a distinction should be made between the ' Celsius and the Centigrade thermometer. Within a few years, however, the Centigrade system of numbering the fixed points came into general use. The Reaumur thermometer was invented in 1731 by the French physicist bearing that name. The Fahrenheit and Centigrade scales are both used in scientific work. The Fahrenheit scale is used popularly in all The R g au _ English-speaking countries. The Centigrade thermometer is mur ther- used popularly in France and portions of Germany. The Reau- * mur thermometer is used popularly in Russia and portions of Germany. No thermometer is used popularly in the country in which it was invented. 62. Thermometers. A thermometer consists essentially of a glass capillary tube of uniform diameter or bore called the stem, which is at- tached to a bulb. The bulb is usually cylindrical in form and of about the same diameter as the stem. This is a matter tion and of convenience, however, and many good thermometers have ^ncipie O f nearly spherical bulbs. The bulb and part of the stem are ather- filled with some fluid, usually mercury, and above this there is z a vacuum. The principle of a thermometer is that for a given increase in^temperature_the fluid in the bulb expands many times (in the case of mercury seven times) as much as the glass itself^ Thus if the bore is uniform, to equal increments of temperature will correspond equal changes in the length of the column of fluid in the stem. The thermom- eter proper is attached to a case in order to protect it from injury and to facilitate its being attached to various objects. The presence of the milk glass at the back of the thermometer is to furnish a better reflector for reading the instrument. The curved front of the stem serves to magnify the liquid within, thus making it more easily visible. Figure 22 represents thermometers of various forms for determining temperature. 1 1 The electrotypes for this figure and for many other illustrations of meteorological apparatus appearing in this book were very kindly furnished by Mr. Henry J. Green, 1191 Bedford Avenue, Brooklyn, N. Y. in onlj) a w of the Weather Bureau stations. 76 METEOROLOGY THE RESULTS OF OBSERVATION 72. The Observations. The stations of the U. S. Weather Bureau are of three kinds: regular stations, cooperative stations, andjspecial stations. The regular weatrieT~t)ureau stations, of which kinds of 6 u there are from 180 to 200, are located ordinarily in the larger S. Weather cities, since they usually publish a daily weather map which must be distributed as soon as possible. There are from 3600 to 4000 cooperative stations, and from 300 to 500 special sta- tions. The observations of temperature taken at a regular station are the real air temperature at 8 A.M. and 8 P.M., the highest and lowest temperatures during the preceding 12 hours, and a continu- Theinstru- ous thermograph record. The instruments for determining ments used and the these temperatures are, a good mercury thermometer, maxi- temperature mum an( j minimum thermometers of the regular Weather taken at Bureau form, and a Richard Freres thermograph. These each kind instruments are located in a thermometer shelter which is or- dinarily placed from 6 to 10 feet above the roof of some high building in the city. At a cooperative station the highest and lowest temperatures during a day are determined, and also the reading of the maximum thermometer just after it has been set. The purpose of taking this observation is to make sure that the maximum ther- mometer has been set and also to give the real air temperature at the time of observation. Maximum and minimum thermometers only are necessary for these observations, and the shelter which contains the thermometers is ordinarily located from 5 to 10 feet above sod in the open or on the north side of some building. Special stations take those observations for which the stations were established. 73. Normal hourly, daily, monthly, and yearly temperature. From these observations of temperature which are made at the various Weather Bureau stations, certain normal temperatures and other tem- tion between P era ture data may be computed. At the very outset the average, three words " average," " normal," and " mean " must be mean* 1 ' " carefully defined. By an average is meant simply the sum of a number of observations divided by the number of the observations. If the observations have been extended over a sufficient length of time so that the accidental irregulariti iminated by taking the average, then the average value may be spoken of as a normal. Usually at least twenty years of observations are required ' OBSERVATION AND DISTRIBUTION OF TEMPERATURE 77 for a normal. The word " mean " is used by various writers to cover both average and normal. If a good continuous thermograph record for at least twenty years is available, the normal hourly temperatures for the various days of the year may be computed. For example, the normal 9 A.M. Normal temperature for October 28 would be found by averaging the hourly tem- twenty or more values which had been recorded for this partic- peri ular hour on the date in question. Similarly, the normal hourly tem- peratures for all the hours of all the days in the year may be computed. Usually this is not done for every day in the year, but the days are grouped by months. The average temperature for a day is found by averaging the 24 values of hourly temperature observed during that day. This requires a ther- mograph, and since thermographs have not been in general The use even at regular Weather Bureau stations for many years, methods of various combinations have been sought such that the average fh^av^ra"? temperature for the day might be computed from the tern- daily tem- peratures observed at certain definite times during the day. peri Some of the various combinations which have been used are the fol- lowing: \ (8 A.M. + 8 P.M.) ; I (7 A.M.+ 2 P.M. + 9 P.M. + 9 P.M.) ; 4 (maximum + minimum). Of these \ (7 A.M. + 2 P.M. -f- 9 P.M. + 9 P.M.) was formerly used by many Weather Bureau stations for comput- ing the average daily temperature. At the present time J (maximum + minimum) is used at all U. S. Weather Bureau stations for computing the average daily temperature. The reasons are : the apparatus required is simple, the computation is very easy, and in the long run the average daily temperatures computed in this way approximate very closely the average daily temperatures found by taking one twenty-fourth of the sum of the twenty-four hourly values. The normal daily temperature is found by averaging the average dailies for the date in question for a suffi- cient number of years to eliminate the accidental irregularities. If the normals are based on twenty years of observations, it will be formal found that there is not an even transition from day to day, daUy tem- but jumps in temperature of even two or three degrees occur. * It might seem that the time were not sufficiently long to eliminate acci- dental irregularities. It is found, however, that these abrupt changes do not disappear even if the normal is based on a hundred years of obser- vations. It may be that these abrupt changes are not due to the fact that the normal is based on too short a time, but to the fact that there are actual abrupt advances and recessions of temperature which take 78 METEOROLOGY place on nearly the same date each year. Ordinarily, however, the nor- mal daily temperatures are " adjusted " before they are published. That is, the irregularities are smoothed out and there are no jumps in temper- ature from day to day. The accompanying table gives the adjusted nor- mal daily temperatures for every day in the year at Albany, N.Y. This station at Albany, N.Y., has been chosen for most of the data in this book because the record is a long one, the observations have been well taken, and it is typical of New England and the Middle Atlantic States. 1 The larger cities, as New York and Boston, are too near the ocean, which is always a disturbing factor, to be typical. In the case of an average ADJUSTED NORMAL DAILY TEMPERATURES AT ALBANY, N.Y. (Temp. F.) DATE JA . FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. OCT. Nov. DEC. 1 24 22 27 38 53 64 71 72 66 57 44 32 2 24 22 27 39 53 65 71 72 66 57 44 32 3 24 22 28 40 54 65 71 72 66 56 43 31 4 24 22 28 40 54 65 71 72 66 56 43 31 5 23 22 28 41 55 66 71 71 66 55 42 30 6 23 22 29 41 55 66 71 71 65 55 42 30 7 23 22 29 42 55 66 71 71 65 54 42 30 8 23 22 29 42 56 66 72 71 65 54 41 30 9 23 22 30 43 56 66 72 71 64 53 41 29 10 23 23 30 43 57 67 72 71 64 53 40 29 11 23 23 30 44 57 67 72 70 64 52 40 29 12 23 23 30 44 58 67 72 70 64 52 40 28 13 22 23 31 45 58 68 72 70 63 52 39 28 14 22 23 31 45 58 68 72 70 63 51 39 28 15 22 23 32 46 59 68 72 70 63 50 39 28 16 22 24 32 46 59 68 72 70 62 50 38 27 17 22 24 32 47 60 68 72 69 62 50 38 27 18 22 24 33 47 60 69 72 69 62 49 37 27 19 22 24 33 48 60 69 72 69 61 49 37 27 20 22 24 33 48 61 69 72 69 61 48 37 26 21 22 25 34 48 61 69 73 69 61 48 36 26 22 22 25 34 49 61 69 73 68 60 48 36 26 23 22 25 34 49 62 70 73 68 60 48 35 26 24 22 26 35 50 62 70 73 68 60 47 35 25 25 22 .26 35 50 62 70 73 68 59 47 35 25 26 22 ' 26 36 51 63 70 73 68 59 46 34 25 27 22 26 36 51 63 70 73 68 58 46 34 25 28 22 27 36 52 63 70 72 67 58 45 34 24 29 22 37 52 64 70 72 67 58 45 33 24 30 22 38 52 64 71 72 67 57 44 33 24 31 22 38 64 72 67 44 24 1 At the Albany station, Mr. George T. Todd is local forecaster and Mr. Herbert E. Vail is assistant. It is to the courtesy, kindly interest, and willing assistance of these gentlemen that the data for Albany in this book are due. OBSERVATION AND DISTRIBUTION OF TEMPERATURE 79 daily temperature, a departure from normal of 10 is common, of 20 is unusual, and of 30 is almost certainly record breaking. 74. The average monthly temperature is found by averaging the daily temperatures for the various days of the month. An average Average and monthly temperature is practically independent of the way normal in which the average daily temperatures from which it is tempera- computed were found. A normal monthly temperature is tures - found by averaging the average monthlies for a sufficient number of ( \ AVERAGE AND NORMAL MONTHLY AND YEARLY TEMPERATURES AT ALBANY, N.Y. (Temp. F.) YEAR JAN. FEB. MAR. APR. MAY JUNE JULY Atra. SEPT. OCT. Nov. DEC. AN. 1874 28 22 30 36 56 66 70 66 63 49 36 26 46 1875 14 15 25 38 56 65 69 69 58 47 31 36 43 1876 29 24 29 42 55 70 73 72 60 47 40 17 47 1877 16 29 30 46 56 67 72 72 64 49 41 32 48 1878 22 24 38 52 56 64 74 70 64 53 38 28 49 1879 18 20 30 42 60 66 71 68 60 56 38 29 47 1880 30 28 33 50 66 71 75 71 65 52 39 25 50 1881 19 26 37 47 65 65 74 73 71 55 43 39 51 1882 26 31 38 46 55 69 74 72 65 56 41 30 50 1883 21 28 29 46 58 71 73 74 61 51 44 30 49 1884 23 32 36 47 59 71 71 72 67 51 39 28 50 1885 23 15 24 46 57 67 73 68 61 51 41 31 46 1886 21 23 33 51 59 66 72 71 64 53 40 24 48 1887 21 25 28 43 65 69 77 69 60 50 38 27 48 1888 15 22 26 50 58 69 71 71 61 46 41 30 47 1889 31 20 37 44 62 68 72 70 64 49 43 35 50 1890 31 31 31 47 57 68 71 71 62 51 38 20 48 1891 25 28 32 49 57 68 69 71 67 50 39 37 49 1892 24 26 30 46 56 71 73 72 62 51 38 26 48 1893 17 22 31 44 58 70 72 72 59 54 39 26 47 1894 27 21 40 48 60 70 75 69 67 53 36 29 50 1895 23 19 30 47 62 73 70 72 67 47 41 32 48 1896 20 25 27 50 64 68 74 73 62 48 44 26 48 1897 25 26 35 48 59 65 75 70 63 53 39 30 49 1898 24 28 42 46 58 70 75 73 67 53 39 29 50 1899 23 22 31 48 60 71 73 72 61 53 40 32 49 1900 26 26 28 48 58 70 74 75 68 58 42 29 50 1901 24 19 33 50 59 70 76 73 65 52 34 27 48 1902 23 24 40 48 57 64 70 68 63 51 43 23 48 1903 24 27 43 48 62 63 71 65 64 53 36 23 48 1904 15 17 31 44 63 69 72 69 61 49 35 20 45 1905 21 18 33 46 59 67 74 69 63 52 38 32 48 1906 32 22 28 47 59 69 73 73 66 52 39 24 49 1907 22 17 37 43 53 66 73 69 64 48 39 32 47 1908 25 20 35 46 61 69 75 70 66 54 40 29 49 Normal 23 24 33 46 59 68 73 71 64 51 39 29 48 80 METEOROLOGY years to eliminate accidental irregularities. It should also be equal to the average of the normal daily temperatures for that month. The average yearly temperature is found either by averaging the aver- age monthly temperatures for the year, taking account of the number of days in each month, or by averaging the average daily tem- nonnai peratures for the year. A normal yearly temperature is yearly tem- f oun d by averaging the average yearlies for a sufficient num- ber of years. It should also equal the average of the normal monthlies, taking count of the number of days in each month, or the aver- / J 70 65 60 55 50 45 40 35 30 25 20 15 10 5 n \ 7** / \ / i \ \ i \ \. / 1 \ / NORM AL YEA LY \ / \ A P -NORM/ ORMAL L MON1 DAILY HLY V / \ \ / \ \ / / / \ \ ^z 7 JAN. FEB. MAR. APRIL MAY JUNE JULY AUG. SEPT. OCT. NOV. DEC. FIG. 36. Graphical Representation of the Station Normals of Temperature at Albany, N.Y. age of the normal dailies for all the days in the year. The table on page 79 gives the average monthly temperatures and the average yearly temperatures for several years, and also the normal monthly and yearly OBSERVATION AND DISTRIBUTION OF TEMPERATURE 81 temperatures for Albany, N. Y. It will be seen in the case of the average monthly temperatures that a departure from normal of 3 is common, of 5 is unusual and of 7 is almost certainly record breaking The de- partures in winter are usually larger than those in summer. In the case of an average yearly temperature, a departure from normal of 0.5 is common, of 1.5 is unusual, and 2.5 almost certainly record breaking. In the table on pages 82 and 83 are given the normal monthly and annual temperatures for 20 cities in the United States and twenty-five for- eign places together with certain facts in connection with their location. DEC. MID. 3 I 6 i 9 NOON SUNhlSE SUNSET FIG. 37. Thermo-isopleths, Centigrade, at Berlin, Germany. (After BORNSTEIN.) The data for the United States have been taken from the publications of the United States Weather Bureau while the data for the foreign places have been derived from HANN'S Lehrbuch der Meteor ologie. 75. The normal daily, monthly, and yearly temperatures are often spoken of as station normals of temperature, and these may Graphical be expressed graphically, as is shown by Fig. 36. representa- 76. If the normal hourly temperatures for the various Jon normals hours of each day in the year are known, these may also be of tempera- expressed graph' lly, jjs is shown for the city of Berlin, Ger- many, in Fig. 37. Months are here plotted along the F-axis and hours 82 METEOROLOGY 5 rH LENG RE IN Y o> S q CD oq rH rH (N 00 Oi SSSSS OfNOrHO CO CO CO ( 1 OOifNrHOO tO t> l> rfH CO CO tO tO tO tO CD Oi rH i> 00 tOOl>l>CO CO tO"*Tt< to CO CJi 00 00 O CO CO CO ^ CD O5 00 CO O CO lO 1C d 2 d 1 o3 3-B CO GO CO CO'* Oi CD CO 00-0 OJ O CO (N 00 O5 00 CO Tt< O CD CD to CD 00 G5OCDCD-* O5 >O CD t- i I COCO 00 CO > SCOOOCDO t-t- t^oo t^- Oi rH rH Oi tO O CO Oi rH oq q oq i> o CD CO CO CD 00 iO >O tO to 00 OrH rj< t>Oi Oi rH tO t^ CO 0000 I> i I (N tO OiCO tOOiMcO-* COrHrH fV| iyi iyi |V| |V| CO 00 00 tO rH O O O O O COO5COO5 t^ (M rHrHCO. rH Oi tO^f CO Tf CO CO rH oa &/) e3 l CD Oi O I s " 00 Oi O CD co l> tO CO CD CD Oi CO 00 O t>. CD 1> CO ^ l> to CO tO OO !> 1>00 tO to TtKN OiOOtO 00 t> 00 tO to Oi CN CO to (NCOCO oot-oo t> CO rH TjH O *-H tOTf oo i> oo to to CO CO Oi CO tO o CD' to 06 oo 00 CD t^ >O tO 00 tO CO CD CO t^ O CO CD CO q ^ CD co to it>-*oio 00 T^ q cocorH 10 o o o o o 00 rH OCO CO rH CO CO CO CO X&6& OBSERVATION AND DISTRIBUTION OF TEMPERATURE 83 C^J i-H CO CO CO CO CO CO CO CO CO CO CO IO i-H CO i-H FH (N CO CO CO CO CO i>t>ocoto go^ co iO CO CO OCO to tO OO os toococo to to to to 1-H r-H CO R S8S8S CO CO rH CO iO OiO CO OI> (N CQ 00 O co Th co t^ oi I>l>l>COtO tOCO GO"tf CO^ CO COOi i 1 OH> O 8S 1 t^.i-H>i-HO gg^p^ to to to to to toco I>OOCO iO tO 'CO CO GO GO 1 CO CO cO CO i rHCOOOCi CO (N COO'* I>COCO1>CO TflO5COi-lTH CO CO CO < CO CO t-H 00 (N J t^OCO^cO iObi-Hcq (M Oi(N-*OOrH 1> i- t O CO t^oo to 10 to 00 00 fr- OflOO O3 t- Oi l> t- t-CD iO I-H CO tO 'f CO i>oocoioto COCOOiiMCO i>i>i>oooo OcO 00 CO fMCOcO 1>I>CO CO l> CO CO tO i-H (N l^ CO CO tO GO tO to co to to to l>. rH tO rH CO '^S CO tO GO I-H CO * to to co co o oo 05 tO tO CO CO CO OOCQl^CO CO^OOi-H to <* to-* to to O5 (N O 00 CO (M iO GO O coTtiuoioto OiCO CO tOrJHCO i-H i-H i-H (M 00 t^ CO tO O tO COCO G5 t^- t> O ''t 1 to CO I-H OOOO 1-1 (N CM O^-IO^H O ^O(MrH(MTt< o o o o o i-H tOCO o o o o o . CO O5 i-H O O5 o o o o o o o o o o i-HTfl tO(M(M OOGOO(M(N GO O5 T}H l> T-H I-H lOCO i^?-0 I IO T^ CM Oi o o o o o o o o o o OOCO 00 CO GO CO ^ to i-H CO tO o o o o o i 1 rH i i GO I-H O CO ^ CO o o o o o i-H OS O5 to !> rt 1, 'I' Il 20 o o M CD 43 c3 ^ ^ up* !Jj! 84 METEOROLOGY Graphical representa- tion of nor- mal hourly tempera- tures. of the day along the X-axis. Lines joining points having the same values of .temperature have also been drawn, and these are ordinarily known in Europe as thermo-isopleths. They are also sometimes called chronoisotherms. This method of graphical representation was first introduced by Lalanne about 1843. In order to note the relation between the time of minimum temperature and the time of sunrise, dotted curves representing the time of sunrise and sunset have been added. From this figure the normal tem- perature at any hour of any day may at once be found, also the time of the minimum temperature each day, the time of the maximum tempera- ture, and the normal value of daily range. Similar charts for other cities will be found in various books and publications for Greenwich in SCOTT, Elementary Meteorology, p. 48 ; for Miinchen in ARRHENIUS, Cosmische Physik, p. 556 ; for Aachen in Meteor ologische Zeitschrift, April, 1904, p. 179 ; for Baltimore in FASSIG, The Climate and Weather of Balti- more, p. 62 ; for Chicago in WILLIS L. MOORE, Descriptive Meteorology, page 183. 77. Diurnal, annual, and irregular variation. The graph which repre- sents the daily variation in temperature is found by plotting to scale the The daily normal' hourly temperatures. If the values of the normal variation in hourly temperatures are not known, an idea of the form temperature. ^ ^ curve mav j^ obtained by noting the variation on some typical day, that is, some day when the other meteoro- FIG. 38. Thermograph Record showing Typical Daily Variation of Temperature at Albany, N.Y., October 14-18, 1908. (U. S. Weather Bureau.) logical elements have remained constant or followed as nearly normal courses as possible. Figure 38, which is a copy of the ther- mograph record at Albany, N.Y., for the five days ending October 18, OBSERVATION AND DISTRIBUTION OF TEMPERATURE 85 1908, slows a series of very typical daily variations. Figure 4 also shows a typical daily variation in temperature at Williamstown, Mass. The lowt st temperature usually comes at about the time of sunrise, and the highest temperature from 2.00 to 4.30 P.M., depending upon the season of the year. The maximum occurs early in winter and later in summer. The average temperature for a day occurs at about 9 A.M. and 8 P.M the rise during the morning and early afternoon is sharp, the curve being convex ; the drop during the afternoon and night is long and slow, giving a concave curve. This curve, which represents the daily variation, varies slightly with the time of the year, the elevation, and the immediate surroundings. It varies markedly with latitude and with nearness to the ocean. The fact that the highest temperature does not occur at noon when receipts of energy from the sun are largest, but several hours later, needs explanation. In Fig. 39 let A represent the insolation received from the sun on a horizontal surface dur- Why the ing some day in September. The curve, which represents highest tem- the receipts, begins at about 6 A.M., rises rapidly to a maxi- JoeVnot mum at noon, and drops down again to nothing at 6 P.M. occur at This represents the amount of energy received by a hori- zontal surface provided the earth had no atmosphere^ or provided the atmosphere transmitted all of the insolation which falls upon it. No question is here raised as to where the insolation is absorbed; whether in the at- mosphere, at the surface of MID. 2 4 6 8 10 NOON 2 4 e s^Vo MID. the ground, Or in a body of FlG - 39. Diagram Illustrating the Energy re- ...... ,, ceived and given off by the Earth during a Day. water upon which it may fall. The amount of energy given off by the earth to space depends 'upon the temperature of the earth ; thus the largest amount would be given off at the time of highest temperature and the least amount at the time of lowest temperature. The graph which represents the energy given off is indicated by B. During the early hours of the afternoon, although receipts have already commenced to fall off, it will be seen that they are still slightly in excess of expenditures, that is, a rise in temperature is still continuing and will continue until the receipts and expenditures become equal, which takes place at the time of highest temperature. 78. The curve which represents the annual variation in temperature is found by plotting the normal daily temperatures to scale, and this is 86 METEOROLOGY pictured for Albany, N.Y., in Fig. 36. The time of lowest temperature is during the last part of January, and the time of highest temperature is The annual during the last part of July, in each case nearly forty days variation in after the time of least and greatest receipts of energy from ure " the sun. The rise from February until July is slow and regular, and the fall from the last of July until the first of February is of the same nature. The reason that the highest and lowest temperatures do not come at the time of greatest or least receipts of energy is the same as that given for the time of occurrence of the maximum temperature during the day. The annual variation in temperature varies slightly with elevation and with the immediate surroundings of the station, and markedly with near- ness to the ocean and with latitude. Figure 40 gives the annual varia- tion at five different places on the earth's surface. The contrast be- tween these five curves shows we 11 the effect of latitude. At St. Ann Trinidad, which has a tropic* location, there are two maxim and two minima and an extremely small change during the year. With increasing latitude there is but one maximum and the change during the year increases steadily. Werchojansk with its high latitude and continental location has an immense change in temperature during the year. The graphs repre- senting the annual variation in temperature can be constructed for many other places by plotting the data given in section 74. The irregular changes of temperature are sometimes greater than the daily variation ; that is, the temperature may fall steadily during the day t instead of rising, or it may rise during the night instead of fluctuations falling. The irregular variations, however, are never larger in tempera- than the annual variation. ture. 79. Temperature data. From the temperature observa- tions which have been made at the different weather bureau stations, FIG. 40. Annual Variation in Temperature at five different Places. (1) St. Anns, Trinidad; (2) Palermo; (3) Berlin; (4) St. Petersburg; (5) Werchojansk, Siberia. OBSERVATION AND DISTRIBUTION OF TEMPERATURE 87 various temperature data in addition to the normals already cussed may be computed. Among these are the following : (1) Average and normal daily range for the various months and for the year. The daily range has been denned as the differ- ence between the highest and lowest temperatures during the day. , This range may be found, and normals may be computed, in exactly the same way as for observations of temperature. The accompanying table gives the average fully dis- Tempera- ture data which may be com- puted from the observa- tions taken at Weather Bureau stations. AVERAGE AND NORMAL VALUES OF DAILY RANGE OF TEMPERATURES AT ALBANY, N.Y. (Temp. F.) YEAR JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. OCT. Nov. DEC. AN- NUAL 1874 1875 1876 1877 1878 20.3 17.7 18.5 174 17.8 19.5 17.1 15.1 17.4 17.7 17.5 14.8 15.2 16.8 18.5 18.7 21.7 22.4 20.8 19.8 18.9 22.5 19.2 17.4 18.4 20.7 20.1 18.1 20.0 18.5 23.3 18.7 19.7 20.6 15.6 19.1 16.8 17.3 15.5 15.1 15.3 16.1 11.5 14.2 16.5 15.3 15.2 14.6 18.2 18.8 17.7 169 173 17.9 17.2 14.2 17.8 18.9 18.1 17.6 17.8 18.4 11.2 12.5 165 1879 20.2 17.4 16.2 16.7 21.5 17.7 18.4 16.8 16.2 17.7 14.6 17.7 17.6 1880 16.6 18.6 14.6 18.3 18.0 17.2 15.8 16.8 14.8 15.7 13.0 10.8 15.8 1881 16.5 14.1 11.1 14.4 16.6 15.7 15.0 15.5 14.3 15.8 12.3 12.5 14.5 1882 15.5 16.4 13.8 16.4 15.9 16.8 16.0 16.7 13.8 15.4 12.7 10.8 15.0 1883 13.7 14.3 15.5 15.8 17.1 16.5 16.9 17.5 17.0 15.0 13.4 13.8 16.4 1884 16.3 13.3 11.5 13.9 16.1 20.0 15.6 16.9 16.5 17.8 14.9 14.7 15.6 1885 1G.7 20.3 17.4 22.2 21.1 21.5 20.9 17.3 21.1 17.4 12.2 14.5 18.6 1886 15.2 17.6 14.7 20.1 20.6 18.8 21.3 22.1 20.1 19.3 16.9 16.5 18.6 1887 21.5 17.3 16.7 18.6 22.2 19.9 17.8 18.5 18.5 16.4 16.0 12.4 18.0 1888 15.9 19.6 16.3 20.0 18.3 20.4 22.3 19.5 16.8 14.5 14.3 14.0 17.6 1889 13.2 13.9 14.6 18.9 21.6 18.1 16.8 20.9 16.8 18.3 14.3 14.9 17.0 1890 14.9 15.2 14.9 21.8 19.4 20.9 20.7 17.1 15.7 18.1 14.1 14.1 16.8 1891 14.5 14.0 16.0 18.5 22.1 20.7 18.0 17.2 17.7 16.4 15.1 14.7 17.1 1892 13.9 14.4 13.8 18.4 15.6 17.9 20.9 17.5 18.5 16.6 11.1 10.9 15.8 1893 13.6 15.2 14.3 17.5 18.3 19.4 21.6 20.0 17.3 17.6 15.0 13.9 17.0 1894 14.5 16.5 16.7 17.4 19.1 19.4 20.9 19.8 18.6 15.3 12.4 13.2 17.0 1895 15.9 19.0 14.4 17.5 20.9 20.1 18.9 19.7 21.5 17.5 15.2 16.8 17.8 1896 13.5 15.7 16.2 20.8 20.9 20.0 19.2 19.6 19.5 15.2 14.7 14.4 17.5 1897 14.4 15.8 16.4 19.9 19.2 19.6 17.2 19.4 21.5 22.9 13.8 12.4 17.7 1898 15.5 14.6 17.5 16.6 15.9 19.2 20.7 2(^.7 20.2 15.7 14.2 14.9 17.0 1899 17.2 14.4 12.9 19.6 20.6 21.7 19.3 20.3 19.5 16.5 13.1 13.0 17.3 1900 17.7 16.2 17.4 19.1 24.3 21.7 22.0 20.5 19.5 18.0 13.0 14.5 18.7 1901 15.4 13.9 14.9 16.2 16.5 20.2 19.7 16.9 18.8 20.0 13.4 15.6 16.8 1902 15.1 14.0 15.5 17.1 19.7 18.6 17.2 19.7 18.1 17.0 16.5 18.1 17.2 1903 15.0 15.8 16.7 19.4 24.6 15.5 19.1 17.1 20.7 16.5 15.0 15.8 17.6 1904 17.8 18.3 14.7 17.6 21.2 19.8 18.9 19.3 18.0 17.5 14.2 15.6 17.7 1905 13.6 17.8 18.7 19.1 19.9 19.7 18.8 19.6 18.5 19.9 17.2 14.4 18.1 1906 14.9 19.9 14.0 19.7 21.4 20.4 19.2 20.0 21.1 18.4 13.5 15.0 18.1 1907 17.0 18.0 17.2 17.1 17.9 20.1 20.1 21.0 15.0 18.9 13.1 12.2 17.3 1908 17.6 16.7 16.2 18.8 18.6 22.2 20.6 19.6 22.4 21.5 13.8 14.5 18.5 Sums 564.5 575.6 545.4 630.8 687.6 676.6 665.2 661.6 640.8 600.9 491.3 500.7 603.8 Normal 16.1 16.4 15.6 18.0 19.6 19.3 19.0 18.9 18.3 17.2 14.0 14.3 ,17.3 88 METEOROLOGY value of the daily range for the various months, and for the year for several years, and also the normal values for Albany, N.Y. It will be seen that the range is greater in summer and smaller in winter, with a maximum in May and a minimum in November. (2) Monthly extremes of temperature. By monthly extremes of tem- perature are meant the highest and lowest temperatures which have been observed during the month in question. (3) Yearly extremes. (4) Absolute highest and absolute lowest for the various months and for the year. By absolute highest and absolute lowest are meant the very highest and very lowest temperatures which have ever been observed. (5) Variability. By variability of temperature is meant the difference between successive daily averages. The average value of the varia- bility for the various months and the year may be determined, and also normal values. The accompanying table gives the normal values for the various months and the year for several stations in the United States. The average variability for the various months and for the year and the normal values are also given for Albany, N.Y. The maximum of variability occurs in January and the minimum in August, one being more than twice the other. (6) Freezing days. By a freezing day is meant a day on which the temperature falls to 32 F. or below at some time during the day. (7) Ice days. By an ice day is meant a day on which the temperature remains below 32 throughout the whole day. (8) Days above 90. (9) Days above 100. (10) Zero days. By a zero day is meant a day when the temperature falls to zero or below at some time during the day. (11) Temperature on special days. The various normals of temperature for such special days as July 4, December 25, etc., may be computed. The table on page 91 gives for Albany, N.Y., the number of days above 90, the number of days above 100, and the number of zero days for a number of years. At the regular stations of the U. S. Weather Bureau the following tem- perature records are kept constantly filled in and computed to date : monthly mean and departure from the normal ; monthly mean maximum and minimum (to tenths) ; absolute maximum and date (each month) ; absolute minimum and date (each month) ; greatest daily range, mean daily range ; absolute monthly range, mean variability (var. to tenths) ; lowest maximum, highest minimum ; number of days with maximum 32 or below, 90 above ; number of days with minimum 32 or below, zero or below ; daily mean temperature (whole degrees) ; daily maximum temperature (whole degrees) ; daily minimum temperature (whole degrees) ; mean hourly temperature (to tenths). OBSERVATION AND DISTRIBUTION OF TEMPERATURE 89 c^cooo os o *o >o c^ co oo totoco o oo co ~* to* co* *o * co *!* 10 u CO t^- CO OOOt^-O t^-Tfito OOcO OO CO C5 tO Tl* Q CO OQ O^ OO CO CO "^ t^ CO tO I s * t^* tO *O *O tO tO t"* > CO^C^ to O5 CO OO O5COCO CO *O CO ^ THt> ^< C^ H CO ^O Oi CO ^~^ CO CO l^* tO ^d^ 01 CO 00 ^^ t > CO I s * CO o CO* co co o^ o^ o^ OQ ^ SsJO-1' I 1 ^l | ^jGO* | 3~H GO* <^i>~( GO* GO* O^ O^ "**** C^l CO *O C^ ^t"tO ^t* ^O i^ Co OO C^ CiQ Ol CO _/ .!-,- .J-J ..'._ ^ ^^J ^ p C_! COO^* t^t^-OQOS OiCOOO T^cOO O tOtO to to tOcdcO tOtO(M* tO rfCO >O tO E fe J3fl ^ pcooq torn t>. co os co co'o-ioQ 06^*0^ odt"~co* oo'oo'to i> toto to I s - <4 % I> CO CO CO (M iH C* O CO N (N O O l> O t- O5 CO 06 oo co co* to* 10* to t^ b % o $ . . . . . . . . ycf -.-. I . . w . __. (^ t_i 03 w * -> . C^ . JS^ 4^ _ -g -p 's-g-aijS^' ' '1 'ajq ' rl 80 *! 88 *! !^^*" 90 METEOROLOGY AVERAGE AND NORMAL VALUES OF VARIABILITY OF TEMPERATURE AT ALBANY, N.Y. (Temp. F.) YEAR JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. OCT. Nov. DEC. AN- NUAL 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 6.6 6.8 7.1 7.4 7.6 8.6 6.2 8.2 6.5 5.5 6.3 7.0 7.5 5.8 6.6 5.2 5.9 5.0 5.2 3.9 3.1 3.9 2.8 4.2 4.7 3.5 5.3 3.5 4.1 5.1 5.3 4.5 3.1 4.8 3.1 4.2 3.6 3.5 3.5 2.7 3.1 3.5 3.4 3.6 3.7 1.7 2.4 3.9 4.0 4.7 4.4 4.0 4.9 3.9 3.0 4.6 4.9 3.4 4.8 5.2 5.0 5.3 3.8 5.4 3.6 7.3 9.3 6.8 6.8 4.9 4.5 7.8 5.3 5.0 4.9 44 44 5 5 7.2 8.1 6.5 5.6 5.3 3.3 3.1 4.2 3.5 4.5 4.4 5.2 5.1 6.6 6.9 3.3 3.5 5.0 3.3 2.9 2.6 3.8 7.1 5.6 6.3 5.3 6.5 8.1 6.5 6.2 4.6 4.7 4.5 4.8 5.0 5.4 4.7 4.7 7.2 6.5 7.7 8.4 6.8 8.5 7.5 6.7 7.0 8.4 7.0 6.3 5.2 6.8 4.3 5.3 5.2 6.1 4.1 3.6 3.2 4.7 4.2 5.0 4.5 4.5 4.3 3.9 4.0 3.7 ] 3.3 3.3 3.5 4.9 3.3 3.6 2.5 3.5 2.7 4.2 2.9 2.8 2.8 2.3 3.1 4.2 2.6 2.6 3.8 4.3 4.9 4.5 4.3 4.5 3.8 4.5 6.5 5.4 4.5 4.5 4.8 5.3 5.1 4.0 i 5.8 4.6 7.5 5.7 8.1 8.1 6.5 4.4 4.6 4.2 4.2 4.8 4.0 3.6 3.3 4.1 3.7 2.9 4.3 & <& 3.8 3.5 5.3 3.7 6.9 7.0 5.2 4.6 9.3 7.0 10.5 5.8 6.4 7.8 7.3 7.0 8.2 8.5 6.2 5.8 3.9 6.8 5.5 5.0 5.6 4.3 5.5 3.2 4.5 5.4 4.1 4.6 4.5 4.6 4.9 4.2 4.7 3.8 4.4 2.8 4.2 3.5 4.0 3.5 3.7 2.6 3.5 4.2 4.8 4.7 3.5 3.5 4.9 2 7 4.5 4.5 5.0 3.3 5.6 6.4 4.7 4.8 5.6 8.6 6.0 4.5 9.6 5.6 5.6 5.3 4.8 5.5 5.0 7.5 6.7 6.4 7.6 4.5 6.8 5.1 4.5 4.9 4.7 3.2 3.7 3.0 3.4 3.0 2.5 7.0 4.6 4.6 3.8 5.8 7.4 6.0 6.0 5.1 5.1 7.4 6.4 7.1 7.1 7.4 7.5 6.8 8.7 7.4 6.7 7.4 7.3 6.8 5.2 4.7 6.4 5.1 4.8 4.9 7.2 6.8 8.6 9.1 7.3 4.6 3.9 4.6 7.1 6.3 5.6 5.7 4.8 4.3 5.0 6.2 5.9 6.4 4.4 3.6 4.0 2.9 3.8 3.9 4.6 4.5 3.3 4.4 5.2 4.4 3.4 3.9 6.6 3.7 5.2 3.9 3.5 4.6 5.5 5.7 4.1 4.6 3.4 5.0 3.7 3.3 3.9 3.2 4.4 3.7 3.7 3.4 4.7 3.1 3.8 4.3 4.1 3.4 3.7 4.0 3.7 3.7 1.9 3.7 3.5 2.8 3.5 2.9 4.1 2.6 2.3 2.6 3.7 3.4 3.6 3.9 3.7 5.0 4.4 5.5 4.6 5.0 3.8 4.5 5.7 3.9 5.9 4.9 4.2 5.5 4.8 5.2 4.8 4.4 5.2 3.8 5.7 4.5 5.4 5.3 5.3 6.3 4.0 3.4 4.3 3.3 5.7 4.0 5.1 6.0 4.4 3.3 4.0 6.1 6.0 5.8 5.2 8.0 7.9 5.5 7.4 6.2 8.7 4.9 6.4 5.2 44 4.7 5.2 4.6 5.0 14 5.4 4.9 5.2 5.2 5.1 Sums Normal 256.6 7.3 243.5 7.0 193.1 5.5 150.0 4.3 157.6 4.5 137.2 3.9 120.1 3.4 110.8 3.2 157.5 4.5 162.3 4.6 173.1 4.9 227.1 6.5 173.9 5.0 80. It has often been found desirable to compute a normal for a sta- The compu- ^^ on a ^ which observations have been taken for but a few tation of a years, and the best method of procedure is the following : from 1 insuffi- Choose some near-by station which has a well-determined dent obser- normal and at which observations have been made during the same interval. Assume that the normal at the station in question will bear the same relation to the normal at the chosen station OBSERVATION AND DISTRIBUTION OF TEMPERATURE 91 THE NUMBER OF ZERO DAYS, DAYS ABOVE 90, AND DAYS ABOVE 100, FOR ALBANY, N.Y. 1 H PQ i H > O O 8 H 1 | S w PQ o 1 CS3 > fl M 1 H 1 <3 o I 1 QQ O tsa 5 <3 1 I 1874 13 2 1886 9 8 1898 7 14 1 1875 34 2 1887 8 13 1899 10 15 1876 12 16 1888 16 8 1900 6 28 1877 7 3 1889 1 1 1901 6 17 1878 11 5 1890 5 7 1902 6 3 1879 15 3 1891 3 9 1903 11 6 1880 4 11 1892 6 15 1904 22 6 1881 8 8 1893 7 12 1905 7 7 1882 3 9 1894 7 16 1906....:. 12 5 1883 3 9 1895 10 1^ 1907 16 8 1884 7 5 1896 12 19 1908 8 9 1885 18 5 1897 4 7 1909...... 5 8 1910 11 5 &s the average for the few years at the station in question bears to the average for the same period at the chosen station. The normal deter- mined in this way is usually much more reliable and correct than one determined from too short a period of observations. 1 81. Differences of Temperature with Altitude. The thermometer shelters at which the various temperature observations have been made all have definite locations ; in a city usually on the roof of a The differ _ high building, in the country usually a few feet above the sod ences in or on the north side of some building. The natural question j^a^wj?" at once arises as to whether the observations would have been mometer different if the elevation of the shelter above the ground had piadTat been different. In other words, what are the temperature different differences in the small height of, say, 100 feet above the alti earth's surface. In the layer of air within five feet of the earth's surface marked differences in temperature will be found. During the day, when convection is operative and when wind velocities are large, the difference will be a comparatively small one, not more than a degree at most. The air in immediate contact with the ground is, of course, the warmer. At night the temperature differences are more i Monthly Weather Review, April, 1910. 92 METEOROLOGY marked and will perhaps average as high as 2 or 3 Fahrenheit, with maximum values of even 5 or 6. The layer of air in immediate con- tact with the ground is, of course, the colder. The layer of air from five to one hundred feet is so thoroughly mixed by the wind at night, as well as during the day, that a very small temperature difference will be found, probably not more than a degree at most, unless the air is held by natural or artificial barriers. Above 100 feet the regular vertical temperature gradient may be expected. As a general conclusion, then, a thermom- eter shelter should not be placed within five feet of the ground nor in a location where the air would be held stagnant by means of artificial or natural barriers. The temperatures observed at heights of from five to one hundred feet will probably be nearly the same. 82. Temperature differences over a limited area. The question here arises as to whether the observations of temperature would be different if the thermometer shelter were placed at different points ture differ- within a small area immediately surrounding the point in { l ues ^^ on - ^ smau< or limited area may be roughly defined as a square mile of surface in the form of a circle or square. During the daytime, on account of convection and the higher values of wind velocity, no appreciable difference in temperature over such a limited area will be found unless it is of unusual topography or the air is held stagnant by natural or artificial barriers. On some par- ticularly favorable days, namely, those with plenty of sunshine and a low wind velocity, the lower points, particularly those in narrow valleys, may be a few tenths of a degree Fahrenheit warmer than the upper parts of the area. At night the layer of air next to the ground grows cold and denser, and drains like water into the valleys and places of small eleva- tion. If the wind is unable to remove these pockets of cold air, a marked variation in temperature over a limited area will be found. For every limited area there will be a critical value of wind velocity, which for most areas is probably not far from three miles per hour. As long as the wind velocity remains larger than three miles these pockets of air will be removed and mixed with the air at other points, and no variation in temperature will be found. As soon as the wind velocity sinks below this critical value, a variation will begin to be manifest, and it is the valley station and those of low elevation which are ordinarily the coldest. Since the question of the variation in temperature depends upon the interplay between the drainage of colder air and the ability of the wind to remove those pockets of cold air, the variation will depend not only upon the elevation, but also upon the openness of the valleys, OBSERVATION AND DISTRIBUTION OF TEMPERATURE 93 their direction, the roughness of the surface, and the direction from which the wind comes. The average difference of temperature between the warmest and coldest places in the limited area will probably average about 4 throughout the year, and will show at times differences as much as from 10 to 15. The location of the shelter in a limited area will thus make a great difference with the temperature observations ; and if records are to be of value, the limited area surrounding each station should be critically investigated for several years. The chief factor which determines the variation in temperature over a limited area, particularly at night, is without doubt the interplay between the drainage of colder air into the valleys, and the ability of the wind to remove these pockets of cold air. A second factor is the nature of the surface. That this plays an important part has been well shown by the research work of Professor Henry J. Cox, of the U. S. Weather Bureau and others in connection with the minimum temperatures and frosts observed over the cranberry marshes of Wisconsin. 1 It has been found that cultivation, drainage, and sanding are very efficacious in pre- venting destructive frosts. The minimum temperatures observed have sometimes been 10 F. and even more, higher than over near-by un- treated marshes. Here there is essentially no difference in elevation and the wind velocity can be assumed to have been the same. The differences in temperature were brought about entirely by the nature of the surface. In causing temperature differences the surface acts in two ways. Differ- ent surfaces heat unequally during the day and at night they cool off at / very different rates. Since on frosty nights the air is always particularly quiet, its tempeiature is determined almost entirely by the temperature of the surface upon which it rests. THE DISTRIBUTION OF TEMPERATURE OVER THE EARTH * 83. Construction of Isothermal Charts. Observations of temperature have been made at many stations in all parts of the world, and some of the records are long ones. From these observations the normal monthly and annual temperatures may be computed. The tion of the stations are, however, at different elevations above mean sea mah> fortiie level, and in order to make the observations comparable with land sta- one another it is necessary to reduce them all to mean sea level. t Since the average vertical temperature gradient is 1 F. for 300 feet, the reduction could well be made by using this factor and adding 1 F. for 1 See Bulletin T, U. S. Weather Bureau. 94 METEOROLOGY each 300 feet of elevation. As a matter of fact the reduction factors used by different investigators in preparing charts are very different. Buchan used 1 F., for 270 feet, while others have used values all the way to 1 F. for 500 feet. The normal temperatures of the air over the ocean have been com- puted from temperature observations made by vessels while at sea. At The method P resen t nearly all vessels take meteorological observations, of determin- and these are reported to the weather bureaus of the various temperature countries as soon as port is reached. These observations are over the grouped by months and by squares, each square being gener- ally 5 on a side. For some squares of the Atlantic Ocean thousands of observations are made during a single month. Although these observations are not always simultaneous, yet, since the daily variation in temperature over the ocean is so small, the normals are nearly as correct as those for the land stations. These normal temperatures may be charted on a map and lines drawn through those places which have the same temperature. These lines are called ordinarily isothermal lines or simply isotherms. 1 struction of The chart is often spoken of as an isothermal chart. The isothermal g rs ^ isotherms for the world were drawn by Humbolt in 1817. These isotherms were improved by Kamtz in 1831 and Mohlmann in 1841 as new data were gathered. Dove in 1852 was the first to chart the normal monthly temperatures. 84. Isothermal lines for the year. The normal yearly temperatures have been used in the construction of chart I, 2 and this chart thus repre- isotherms sents the isotherms for the year. There are several charac- for the teristics of these isothermal lines which deserve careful con- sideration and explanation. (1) Highest temperature at the equator and lowest temperatures at the poles. The first and most obvious fact in connection with the isothermal Hi best ^ mes ^ or ^ ne y ear * s ^ e exigence of a hot belt near the equator temperature and low temperatures at the two poles. The northern part of ^oVamUow- South America, most of Africa, India, and a portion of Aus- est at the tralin arc surrounded by a line marked 80 F. This means poles ' that all points within this closed curve have a normal annual temperature of 80 F. or more. This is the hot belt, and a line passing through its center is often spoken of as the heat equator or the ther- mal equator. The temperature at the north pole is F., while the 1 From the Greek : foot = equal ; &tpivt\ ~ heat. 1 See end of book for the 50 charts. OBSERVATION p DISTRIBUTION OP TEMPERATURE 95 isothermal lines near the south pole have been omitted on account of insufficient data. The explanation of the existence of this hot equatorial belt and the low polar temperatures is for the tem- the well-known fact that the equator receives jnuch more Pf rature insolation from the sun than the polar regions. The ratio between the equator and pole is 347 to 143, and this is sufficient to account forvthe temperature differences. (2) The deflection of isothermal lines from parallels of latitude. All places with the same latitude receive the same amount of insolation from the sun. It would thus be expected that all places on the same parallel of latitude would have the same tempera- tionofiso- ture, and that the isothermal lines would be parallel to the thermal parallels of latitude. This is, however, by no means the case, and the chief cause of the deflection is the existence of ocean It is beyond the scope of this book to discuss fully the causes for the existence and the direction of ocean currents. The factors which cause ocean currents and determine their direction are : tem- perature differences between different parts of the earth ; W hicif deter- the permanent wind system of the earth ; the evaporation mine the which is greater at some parts of the ocean than at others ; and^kec- the inflow, which is also greater at certain places than at tionof others ; varying degrees of saltiness and thus density ; the ^ c earth's rotation on its axis ; the configuration of the coast line. Chart II presents the general scheme of the ocean Currents. In the Atlantic Ocean the Gulf Stream, consisting of warm water which has made the circuit of the Gulf of Mexico and emerged between The cen- Florida and Cuba, together with a considerable quantity of erai scheme water which has passed northward outside of the West Indies, of ocean sets diagonally across the Atlantic Ocean towards England and Scandinavia. The return current flows southward along the coast of Europe and along the northern part of Africa and back again along the equator. Another return current flows southward along Greenland and Newfoundland. The northern Pacific Ocean possesses a similar set of ocean currents. In the South Atlantic and South Indian and South Pacific oceans there is an oval circulation turning in a counterclockwise direction. On the European side of the Atlantic Ocean the isothermal lines over England and Scandinavia are carried far to the north by The effect the Gulf Stream and even bend backward on themselves. O f ocean The cool return current along the coast of Spain and northern c Africa carries the isothermal lines towards the equator. The 96 METEOROLOGY result is a fan-shaped spreading out of the isothermal lines over Europe. On the North American side of the Atlantic the cold return current from Greenland carries the isothermal lines southward, while the warm water coming up from the Gulf carries them northward, and the result is a crowding together of the isotherms. The same difference in temperature will be found in one half the distance on the North American side of the Atlantic as on the European. Numerous other illustrations of the deflec- tion of isotherms by ocean currents can be noticed by comparing the scheme of ocean currents with the isothermal lines for the year. Al- though the ocean currents are the chief cause of deflection, they are not the only causes, a-sjlivpirsifry n'f aiir-fara, whether land or water, vegetation^ covered or barren, and the genera,! winr| system, also play a part. (3) Regularity in the southern hemisphere. It will be noticed that the- isothermal lines are much more regular in the southern hemisphere than . in the northern. The reason for this is that the southern in the south- hemisphere is largely a water hemisphere, while the northern ernhemi- hemisphere has a diversified surface consisting of both land and water. A water surface tends to even out temperature irregularities and also to equalize the temperature between equator and pole. (4) The heat equator north of the geographical equator. It will be noticed that the central line of the hot belt lies, on the whole, north of the The heat geographical equator. The reason for this is not far to seek, equator Since the southern hemisphere is largely a water hemisphere tiuTgeo^ the equatorial portion has been cooled arid the polar regions graphical have been warmed by the exchange of water between the equator and poles. In the case of the northern hemisphere this is not so easily possible, since it Is largely a lancl^surface. For this reason the equatorial belt of high temperature lies north of the geographical equator. (5) The hot belt not of the same width and temperature throughout. It will be noticed that the hot belt inclosed by the isothermal line 80 is widest over South America and Africa. It narrows markedly The hot belt . not of the m crossing the Atlantic Ocean and disappears entirely over same width the p ac i nc Ocean. The reason for this is again the tendency of the ocean to equalize equatorial and polar temperatures. 85. Chart III represents the isothermal lines for the year for the United States. 86. Isotherms for January and July. In the construction of Charts IV and V the normal January and July temperatures have been used, and OBSERVATION AND DISTRIBUTION OF TEMPERATURE 97 these charts thus present the isothermal lines for January and July. The first three characteristics noted in connection with the isothermal lines for the year are the same for the isothermal lines for January and July. These three characteristics are : the hot thermTfor belt at the equator and the low temperatures at the poles ; the J anuaf y and deflection of isothermal lines from parallels of latitude ; the regularity in the southern hemisphere. In addition there are two new characteristics which deserve attention and explanation. (1) The hot belt and all isothermal lines migrate north and south in the course of the year. On the January chart it will be seen that the highest temperatures oecur in South Africa and in Australia, while on the July chart it will be seen tfrat the highest temperatures tion of iso- occur over the southern part of North America, Northern * hermal Africa, and Australia. There has thus been a decided migra- tion of the hot belt between January and July. If any other isothermal line is considered it will be found that it too hasiigrated. O*n the Pacific the hot belt migrates from 15 to 20 of latitude and on the Atlantic it migrates still less, andtonly on the western portion of this ocean does it cross the geographical equator to the southern hemisphere. On the con- tinents it shifts over a somewhat greater distance. In Africa it moves from about 23 north to 20 south latitude. In America the migration is from about 35 nbrth to 15 south. The average /however, is less than 47, which is the amount the sun migrates in the course of the year. Three things may be noted in connection with' this migration : (a) the migration is less than the migration of the sun ; (6) the hot belt lags behind the sun in its migration ; (c) the migration is greatest on land and least on the ocean. (2) The highest and lowest temperatures on land. On the January chart the highest temperatures are in the northern part of Africa and in Australia, while the lowest temperature is in north central Siberia On the July chart it will be seen that the highest an d i ow est temperature - are in central North America, North Africa, and tempera- Arabia. The truth of the statement that the highest and on'Tam' lowest temperatures occur on land is thus demonstrated. The reason for thl is not far to seek. The, land, as compared with the ocean, is always radical in its behavior as regards temperature changes. During the day and during thesummer it heats to a high temperature, during the night and during the winter it cools correspondingly low. Thus the highest and lowest temperatures are always to be expected on land. 98 METEOROLOGY 87. Charts VI and VII represent the July and January isotherms for the United States. 88. Poleward temperature gradient. The diminution in tempera- ture in going from equator to pole is often spoken of as the poleward temperature gradient. By examining the isothermal charts for the year, and for January and July, the three following generalizations may be formed : (1) the poleward tempera- ture gradient is larger in winter than in summer; (2) the poleward temperature gradient is larger in the northern hemisphere than in the southern hemisphere; (3) the poleward temperature gradient is larger on land than over the ocean. It will The char- acteristics of the pole- ward tem- perature gradient. 120 160 g 160 120 0" 40 SO" 120" :^ 120* 160 ISANOMALOUS TEMPERATURE LINES FOR JANUARY ( K) 40 80 120" FIG. 41. Isanoixialous Temperature Lines for January (Temp. F.). (After BATCHELDER.) be seen later that to the first of these, namely, the larger value of / poleward temperature gradient in winter, is due the increased wind s velocity and the greater violence of storms during the wint ?r. It is also an interesting fact that the poleward temperature gradient is about 800 times smaller than the vertical temperature gradient. That is, it would be necessary to travel poleward 800 miles to get the same diminution in temperature which would be gained by ascendi ig one mile. Thermal 89. Thermal anomalies. The average temperature of a anomalies, given parallel of latitude may be found by knowing the actual temperature at equally distant intervals along this parallel and finding OBSERVATION AND DISTRIBUTION OF TEMPERATURE 99 their average. The difference between the temperature of a place and the average for its parallel of latitude is called t,he thermal anomaly. These thermal anomalies for January and July are pictured in Figs. 41 and 42. On the January chart it will be seen that the north Atlantic ISANOMALOUS TEMPERATURE LINES FOR JULY FIG. 42. Isanomalous Temperature Lines for July (Temp. F.). (After BATCHELDER.) is 40 above the average of this latitude, the north Pacific is 20 above the average, while the central part of Asia and the central part of North America are 30 below the average for their latitude. If the July chart is examined, it will be seen that the center of North America and of Asia are above their latitudes in temperature, while the Pacific and the Atlantic Oceans average below their latitudes. These statements may be summarized as a general law: in summer the continents are above the average in tempera- ture, while the oceans are below ; during the winter the continents are below the average in temperature, while the oceans are above. ^ Figure 43, which represents the isothermal lines for Spain and Por- tugal for January and July, illustrates this general law for a small area. JANUARY JtJLJ FIG. 43. Isothermal Lines for Spain and Portugal for January and July (Temp. F.). 100 METEOROLOGY 90. Annual range of temperature. Figure 44 represents the annual range of temperature for the earth's surface. Many different kinds of annual range may be computed. The kind represented here is the difference between the January and July normals. The greatest value of range is 120 F. in north central Siberia. The northern part of North America comes next with 80 F., while South America, South Africa, and Australia have 30 each. It will be noticed that the greatest value of range always occurs on land, and that it is roughly proportional to the amount of The char- acteristics of the annual range of temperature. 120 160 160 120 80 40 LINES OF EQUAL ANNUAL RANGE OF TEMPERATURE FIG. 44. Annual Range of Temperature (Temp. F.). (After CONNOLLT from DAVIS'S Elementary Meteorology.) land surrounding the place in question. This is but another illustra- tion of the well-known principle that a land surface is radical in its temperature behavior as compared with an ocean surface. A land surface becomes excessively warm in summer and correspondingly cold during the winter, while a water surface is more conserva- tive in its behavior. 91. Extremes of temperature. The lowest temperature ever observed on the earth's surface is 90.4 F. at Werchojansk (or Verkhoyansk) in north central Siberia. This temperature was observed on January 15, 1885. The highest The highest and lowest tempera- tures ever observed in -the world. I OBSERVATION AND DISTRIBUTION OF TEMPERATURE 401 temperature ever observed was 127.4 F. at Ouargla in Algeria. It was observed July 17, 1879. North central Siberia and the northern portion of Africa are thus the coldest and hottest places in the world respectively. The lowest temperature ever observed in the United States is 65 F. at Poplar River in Montana. It was observed January 1, 1885. The highest temperature ever observed in the United States is 119 F., and this was observed at Phoenix, Arizona. The and lowest isothermal lines for the United States, for January and July, tem P era - indicate the portions of the country which are hottest and served in coldest respectively. jj 6 ^ s nited In making these statements only those temperatures which have been observed in regular series at weather bureau sta- tions have been taken into account. A temperature of 122 F. is reported for Death Valley, Cal., during the summer of 1891, and a temperature of 130 F. is said to have been observed at Mammoth Tank, Cal., on August 17, 1885. Since the thermometers may not have been properly sheltered, these temperatures are not usually considered as authentic. Temperatures as high as 154 F. have been reported from parts of the Sahara, and a temperature as low as 96 F. is reported from the Arctic regions of North America. At every place abnormally high and abnormally low temperatures have occurred if long periods of time are considered. Unusually cold winters and unusually hot summers have also been described. At many European cities where the records are long ones, these abnormalities make very interesting reading. Space does not permit, however, a full treatment of this subject. 92. Other temperature charts. All the temperature data men- tioned in section 79 may be charted, provided the data are available for many stations in all parts of the world or in a given country. Other tem _ Figures 45, 46, and 47 represent the highest temperatures ever perature observed in the United States, the lowest temperatures ev^r c observed in the United States, and the variability of temperature for January in the United States. 93. Polar temperatures. Charts VIII and IX represent the normal temperatures for the north polar regions 1 for January and for North July. It will be seen that during July the north pole is the polar tem- coldest part of the northern hemisphere. During January J north central Siberia is the coldest part of the northern hemisphere. Both 1 For isotherms for the north polar regions see Meteorologische Zeitschrift, 1906, p. 111. 102 ^ - METEOROLOGY of these facts are somewhat anomalous, since for a short time during the summer the North Pole receives more insolation than any other place in the northern hemisphere, and during the winter it receives less than any other place in the northern hemisphere. These seeming anomalies can, OBSERVATION AND DISTRIBUTION OF TEMPERATURE 103 however, be easily explained. During the summer the North Pole is the coldest part of the northern hemisphere, in spite of its value of insolation, for three reasons : first, these large values of h , e No ^ h x , ., i oi6 coldest insolation last for a very short time ; secondly, the polar in summer. 104 METEOROLOGY regions are covered with snow and ice, which reflect about 30 to 40 per cent of the insolation, which is thus lost as far as heating is concerned ; thirdly, the temperature cannot be raised markedly above 32 F. until all the snow and ice is melted, and this never occurs. OBSERVATION AND DISTRIBUTION OF TEMPERATURE 105 There are two reasons why north central Siberia surpasses the polar regions for cold during the winter ; first, the exchange of water between the equator and pole warms the polar regions. This is not possible in the case of a land surface. Secondly, north central trai Siberia Siberia is a land surface, while the polar regions is largely a coldest * water surface, or a water surface covered with ice and snow. Land is always radical in its temperature behavior, and thus cools to a greater extent than a water surface. THE TEMPERATURE OF LAND AND WATER 94. Ocean temperatures. The normal annual temperature of the surface of the ocean varies from between 80 and 90 F. at the equator to about 28 F. in the polar regions. The temperature of the The tem _ bottom of the ocean varies from 36 in the equatorial regions perature of to about 28 in the polar regions. Thus, at the equator the ^a^J^ temperature change between the surface and bottom is from of the between 80 and 90 to 28 ; in the polar region the tempera- ocean * ture is 28 throughout. The change in temperature between day and night is extremely small, not amounting to more than a degree or two at most. The temperature change between winter and summer is also small at the equator and in the polar regions, but greater in middle latitudes. At New York the temperature change between summer and winter is from about 70 to 30 F., and this is perhaps the greatest yearly change anywhere in the Atlantic the tempera- Ocean. On account of the ocean currents the difference in temperature between places but a few hundred miles apart may be considerable. Salt water freezes at 27 F. Thus all harbors north of 50 north latitude are frozen shut in winter, and ice forms all over the polar seas, seldom, however, to a depth of more than 5 or 6 feet. Special thermometers must be used for determining the temperature of the ocean water, particularly at considerable depths. The thermometer must not be influenced by pressure and must ist indicate the temperature at the required depth, regardless of special ther- the temperature of the layers of water through which the ^ed^ thermometer must be raised in drawing it to the surface. For a more detailed treatment of this subject the reader must be referred to special works on ocean temperatures. 95. Lake temperatures. In the case of a deep lake in the middle latitudes, the temperature of the surf ace water in summer will be between 106 METEOROLOGY 60 and 80, while at considerable depths the temperature will be 39 or above. With the coming of winter the surface water cools and thus becomes heavier and sinks to the bottom. This process con- perature of tinues until the temperature of the lake becomes 39 through- a lake from ou ^ 39 jp j s ^he temperature of water at its maximum torn at dif- density. As the temperature falls below this, the water again ferent times expands and becomes lighter. Thus, as the surface water cools below 39, it is now lighter than the water below and remains at the top. It cools finally to 32 and ice forms and the thick- ness of the ice grows greater and greater during the winter. Thus, in winter the temperature of the surface of a lake next the ice will be 32 F., while the bottom will have a temperature of 39. With the coming of spring the ice melts and the surface layers become warmed until the tem- perature again becomes 39 throughout. From this point on the warmer layers remain at the top until finally in late summer the surface may have heated up to 60 or even 80 F. The temperature of the bottom, if a lake is deep, will remain 39 ; if shallow, the lake may have become heated throughout, and the bottom temperature may be somewhat above 39. 96. River temperatures. If a river is deep and slow flowing, its River tem- temperature behavior will be the same as that of a lake. In peratures. winter the temperature of the water underneath the ice will be 32, and the temperature of the bottom 39 or slightly lower. If the river is shallow and rapidly flowing, the water will be so thoroughly mixed that the temperature will be the same throughout. 97. Temperature below the surface of the land. The ground is a very poor conductor of heat, and for this reason the daily variation in The annual temperature does not penetrate to a greater depth than two or variation three feet, and requires many hours to reach even that depth, penetrate The annua l variation does not penetrate to a depth of more more than than fifty feet and requires nearly six months to reach that depth. Thus, at a depth of thirty or forty feet in the ground the highest temperatures will be experienced in winter and the lowest temperatures in summer. Below fifty feet comes the layer of invariable temperature. This will have a thickness from perhaps fifty feet to several hundred feet, and this temperature is always the normal annual temperature for the of hivari- place in question. Caves are often located in this layer of able temper- invariable temperature, and the temperature of the air in such caves is always the normal yearly temperature for the place. Some cellars are also sufficiently deep to be located in this layer. They OBSERVATION AND DISTRIBUTION OF TEMPERATURE 107 will have a constant temperature throughout the year, and its value will be the normal annual temperature for the place. If the normal annual temperature is below 32, there will be a layer which remains constantly frozen throughout the year. Below the layer of invariable temperature, the temperature increases at the rate of 1 for every 52 feet (although this varies from 40 to 100 feet). These observations have been obtained from tunnels and from deep mines. These extend, however, not more than a tureen-* 1 mile below the surface of the earth, which is a mere scratch crease with compared with the radius of the earth. It would be impossible from astronomical and geological considerations for the temperature to increase at this rate all the way down to the earth's center. The temperature of spring water gives an idea of the depth of the layer through which the water percolates and finally emerges as a spring. If the water comes through the layer of invariable temperature, the temperature of spring water would remain the same perature of throughout the year and would have the normal annual tern- s P rin s perature for the place. If the layer through which the water passes is below this, the temperature would remain the same throughout the year, but have a higher value than the normal yearly temperature of the place. If the spring is a shallow one, there would be a change in tem- perature between summer and winter, but the average would again be the normal yearly temperature for the place in question. QUESTIONS (1) Define thermometry. (2) Name the three systems of thermometry. (3) How are the fixed points numbered in each system? (4) State the interre- lation between the three scales. (5) What is the best method of making a mental computation of the Centigrade temperature corresponding to a Fahren- heit? (6) When and where was the thermometer invented? (7) Describe the early form of the thermometer. (8) Describe the origin of the Fahrenheit, Centigrade, and Reaumur thermometers. (9) Describe the appearance and state the working principle of a thermometer. (10) Describe the various steps in the construction of a thermometer. (11) Name the essentials in a good thermometer. (12) What are the advantages and disadvantages of hav- ing a large bulb? (13) What is the ordinary form of the bulb and the reason for it ? (14) Why should the graduation be placed on the stem of a thermometer itself? (15) Name the inaccuracies in determining a temperature, (16) What is meant by the error of parallax? (17) What is the effect of atmos- pheric pressure on a thermometer? (18) What is the cost and accuracy of a thermometer? (19) Why is it hard to determine the temperature of a gas? (20) How should a thermometer be placed to give approximately the real air temperature? (21) What are the three methods of determining the real air 108 METEOROLOGY temperature? (22) Describe the thermometer shelter of the U. S. Weather Bureau. (23) Describe the other types of thermometer shelter. (24) De- scribe the sling thermometer. (25) Describe in full the ventilated thermometer. (26) What are the two kinds of thermographs in ordinary use? (27) Describe the Draper thermograph. (28) Describe the Richard Freres thermograph. (29) What is the working principle in each case ? (30) What is the purpose of maximum and minimum thermometers? (31) Describe the maximum and minimum thermometers of the regular Weather Bureau form. (32) Describe the Six maximum and minimum thermometer. (33) How are maximum and minimum thermometers set? (34) Describe the black bulb thermometer. (35) What is the purpose of the black bulb thermometer? (36) Name some special purposes for which thermometers of various forms have been adopted. (37) State the three kinds of U. S. Weather Bureau stations. (38) What tem- perature observations are made at each? (39) What instruments are used? (40) Where are these instruments located? (41) Where is the shelter located? (42) Distinguish between average, mean, and normal. (43) How are normal hourly temperatures computed? (44) How is the average daily temperature computed? (45) How are the average and normal yearly temperatures com- puted ? (46) What is meant by station normals of temperature ? (47) How may these be represented graphically? (48) What are thermo-isopleths. (49) How is the graph which represents the daily variation found ? (50) Explain the time of occurrence of the highest temperature. (51) Describe the annual variation in temperature. (52) Describe the irregular variation in temperature. (53) What temperature data beside normals may be computed? (54) Define varia- bility of temperature. (55) How are normal temperatures computed for stations where the record is a short one? (56) State the differences in tem- perature in a thermometer shelter placed at different altitudes. (57) Describe the temperature differences over a limited area during the day. (58) Describe the temperature differences over a limited area at night. (59) What observa- tions are used for the construction of isothermal charts? (60) How are iso- thermal charts constructed? (61) State the characteristics of the isothermal lines for the year. (62) What are the causes of ocean currents and their direc- tion? (63) Illustrate the effect of ocean currents on isothermal lines. (64) Explain the characteristics of the annual isothermal lines. (65) State the char- acteristics of the isotherms for January and July. (66) Describe the migration of the hot belt. (67) What is meant by poleward temperature gradient ? (68) What are its characteristics? (69) Define thermal* anomaly. (70) What are its characteristics? (71) What are the characteristics of the annual range of temperature ? (72) In what regions of the world have the highest and lowest temperatures been observed? (73) In what parts of the United States have the highest and lowest temperatures been observed? (74) Name other tempera- ture charts which can be constructed. (75) Describe polar temperatures during the summer and during the winter. (76) Why is the north pole the coldest place in the northern hemisphere during the summer ? (77) Why is north central Siberia colder than the north pole in winter ? (78) Describe the temperature of the ocean. (79) State the diurnal variation of this temperature. (80) De- scribe the annual variation of this temperature. (81) Does the ocean water freeze? (82) Describe the temperature changes which take place in the water of a lake between summer and winter. (83) Describe the condition of a river as regards temperature during the summer and in winter. (84) To what depths does the daily variation of temperature penetrate? (85) To what depths does the annual variation of temperature penetrate? (86) What is meant by the layer of invariable temperature? (87) What evidences are there of its exist- OBSERVATION AND DISTRIBUTION OF TEMPERATURE 109 ence? (88) What is the temperature of the ground below the invariable layer of temperature? (89) To what depths have observations been made? (90) Upon what does the temperature of the water of a spring depend? TOPICS FOR INVESTIGATION (1) The early history of the thermometer. (2) The details of the methods used in constructing thermometers in large numbers. (3) The inaccuracies in determining a temperature and the methods of eliminating them. ' (4) Thermometer shelters their construction, use, and accuracy. (5) Thermometers for determining the temperature of the earth. (6) Thermometers for determining the temperature of deep water. (7) The methods of computing the average daily temperature and their accuracy. (8) Temperature differences over a limited area. (9) Ocean currents. (10) Abnormalities of temperature and seasons at various places. (11) Ocean temperatures and the thermometers used in determining them. (12) Lake temperatures. (13) The te perature below the earth's surface. PRACTICAL EXERCISES (1) Draw the three thermometer scales side, by side and of such a length that the fixed points fall together. Number the points of division so that corre- sponding temperatures to the nearest degree can be read off. (2) If physical apparatus is available, study critically one or two good ther- mometers. Determine their fixed points ; test the uniformity of the bore ; determine the sluggishness ; determine the effect of sudden changes in tempera- ture, etc. (3) Compare the indications of thermometers of several different forms in the open, thermometers in a thermometer shelter, a sling thermometer, and a ventilated thermometer under the most varied conditions. (4) Determine the transmission coefficient of the atmosphere on several dif- ferent days by means of the black bulb thermometer in vacuo. (5) Determine the inaccuracies and behavior of a thermograph by checking its indications by means of a thermometer. (6) Compute the station normals of temperature for some station and represent them graphically. (7) Determine how well the different methods of computing an average daily temperature agree. (8) Plot the daily and annual variation in temperature for several stations in different parts of the world and in each case explain the characteristics of the graph. (9) Work up all or some of the temperature data mentioned in section 79 for several stations. Those stations may be chosen in which the student has a particular interest. (10) Contrast two near-by stations. One should be chosen in a large city and the other in the country near by. (11) Compare the observations made at the base and top of some tower. i 110 METEOROLOGY (12) Investigate the limited area surrounding a given station for temperature differences. (13) If the observations can be obtained, construct temperature charts showing for a certain country or for the world some of the temperature data mentioned in section 79. (14) Determine the temperature behavior of certain springs during the year and then determine the depth of the layer through which the water must come. (In the case of nearly half of the above problems, if the results were carefully worked out, they would be worthy of publication in some meteorological maga- zine.) REFERENCES For the history, description, illustration, and use of apparatus for determining temperature, see : ABBE, Meteorological Apparatus and Methods, Washington, 1888. Pages 11 to 107 treat of thermometers and thermometry. BOLTON, HENRY C.. Evolution of the Thermometer. The apparatus catalogues of such firms as : Henry J. Green, 1191 Bedford Ave., Brooklyn, N. Y. Julien P. Friez, Belfort Observatory, Baltimore, Md. Queen & Co., 8th and Arch Sts., Philadelphia, Pa. Negretti & Zambra, 38 Holborn Viaduct, London. James J. Hicks, 8-10 Hatton Garden, London, E. C., England. C. F. Casella & Co., 11-15 Rochester Row, Victoria St., London, S. W. R. Fuess, Diintherstrasse 8, Steglitz bei Berlin, Germany. Wilh. Lambrecht, Gottingen, Germany. Max Kohl, Chemnitz, Germany. For the exposure and care of thermometers, see : HAZEN, HENRY A., Thermometer Exposure, Professional Papers of the Signal Service, No. XVIII, 1885. Instructions for cooperative observers, U. S. Weather Bureau. See also the various guides for observers mentioned in Appendix IX in group (2)B. For meteorological observations, usually somewhat summarized, consult : (1) The serial publications of the U. S. Weather Bureau. (A} Daily Weather Map. (B) National Weather Bulletin, weekly during the summer, monthly during the winter. (C) Climatological Reports. These were issued monthly at 44 section centers until July, 1909. Since then they have been combined with Monthly Weather Review. (Z)) Weather Bulletins, weekly during the summer at 44 section centers. (Discontinued in 1908.) (E) Snow and Ice Bulletins, weekly during the winter. (F) Monthly Weather Review with annual summary and index. (G) Mount Weather Bulletin. (H ) Annual Report of the Chief of the Weather Bureau. (2) The periodical publications of the weather bureaus of the various countries. For a list of these see BARTHOLOMEW'S Physical Atlas, Vol. Ill (Atlas of Meteorology). (3) The serial publications of many private stations and observatories. For example : Meteorological Observations of the Massachusetts Agri- OBSERVATION AND DISTRIBUTION OF TEMPERATURE 111 cultural Experiment Station, published monthly since January, 1889 ; Observations at Blue Hill Observatory by Professor A. L. Rotch, pub- lished since 1886 in the Annals of the Harvard College Observatory. (4) Scientific magazines and special publications. The best way to locate these is to consult some digest of meteorological literature; as, Fort- schritte der Physik (part three). For temperature normals for various places, see : BUCHAN, ALEXANDER, Report on Atmospheric Circulation. HANN, Lehrbuch der. Meteorologie. HANN, Handbuch der Klimatologie. VAN BEBBER, Handbuch der Meteorologie. Report of the Chief of the Weather Bureau, particularly for 1891-1892, 1896- 1897, 1897-1898, 1900-1901, 1901-1902. Temperature and Relative Humidity Data, Bulletin O, U. S. Weather Bureau, WILLIAM B. STOCKMAN. Climatology of the United States, Bulletin Q, U. S. Weather Bureau, ALFRED JUDSON HENRY. The Daily Normal Temperature and the Daily Normal Precipitation in the United States, Bulletin R, U. S. Weather Bureau, FRANK H. BIGELOW. Report on the Temperatures and Vapor Tensions in the United States, Bulletin S, U. S. Weather Bureau, FRANK H. BIGELOW. Summary of the Climatological Data for the United States by Sections (106 are to be issued). For isothermal and climatological charts, see : BARTHOLOMEW, Physical Atlas, Vol. Ill (Atlas of Meteorology), Prepared by Bartholomew and Herbertson and edited by Alexander Buchan, 1899. BUCHAN, ALEXANDER, Report on Atmospheric Circulation (Report on the scien- tific results of the voyage of H. M. S. Challenger}. HANN, Atlas der Meteorologie, 1887, A section of the Berghaus Atlas, but can be bought separately. HILDEBRANDSSON, H. H., ET TEissERENC DE BORT, Les bases de la meteorologie dynamique (2 vols. have appeared), Paris, 1900-1907. Summary of International Meteorological Observations, Bulletin A of the U. S. Weather Bureau. ELIOT, SIR JOHN, Climatological Atlas of India, Edinburgh, 1906. Russia, Atlas climatologique de I' empire de Russie, St. Petersburg, 1900. BLODGET, LORIN, Climatology of the United States, Philadelphia, 1857. Isothermal Lines for the United States, 1871-1880, by A. W. GREELY. Professional Papers of the Signal Service, No. II, 1881. GREELY, GEN. A. W., American Weather, New York, 1888. Report of the Chief of the Weather Bureau, particularly for 1896-1897, 1897-1898, 1900-1901, 1901-1902. Climatic Charts of the United States, U. S. Weather Bureau, Washington, D. C., 1904 (W. B. 301). Climatology of the United States, Bulletin Q, by A. J. HENRY, 1906 (W. B. 361). For temperature differences over a limited area, see : Monthly Weather Review: July, 1905, XXXIII, p. 305. August, 1906, XXXIV, p. 370. August, 1908, XXXVI, p. 250. CHAPTER IV THE PRESSURE AND CIRCULATION OF THE ATMOSPHERE A. THE OBSERVATION AND DISTRIBUTION OF PRESSURE THE DETERMINATION OF ATMOSPHERIC PRESSURE Atmospheric pressure, 98. Mercurial barometer : history, construction, corrections, 99-101. The aneroid barometer, 102. Barographs, 103. Other so-called barometers, 104. THE RESULTS OF OBSERVATION The Observations, 105. Normal hourly, daily, monthly, and yearly pressure, 106. Diurnal, annual, and irregular variation, 107-109. Barometric data, no. THE VARIATION WITH ALTITUDE Reduction to sea level, in. Barometric determination of altitude, 112, 113. Variation with altitude, 114. THE DISTRIBUTION OF PRESSURE OVER THE EARTH Construction of isobaric charts, 115. Isobars for the year, 116. Vertical section along a meridian, 117. Isobaric surfaces, 118. Isobars for January and July, 119. Other pressure charts, 120. B. THE OBSERVATION AND DISTRIBUTION OF THE WINDS THE DETERMINATION OF THE DIRECTION, FORCE, AND VELOCITY OF THE WIND Wind direction, force, and velocity, 121. Wind vane, 122. Anemoscope, 123. Velocity estimations, 124. Anemometers, 125-127. 112 PRESSURE AND CIRCULATION OF THE ATMOSPHERE 113 THE LOCATION OF OBSERVATORIES Effect of surroundings, 128. Hill and mountain observatories, 129. THE RESULTS OF OBSERVATION The observations, 130. Prevailing wind direction; wind roses, 131. Normal hourly, daily, monthly, and yearly velocity, 132. Diurnal, annual, and irregular variation, 133-135. Wind data, 136. Prevailing winds of the world, 137. Other wind charts, 138. C. THE CONVECTIONAL THEORY AND ITS COMPARISON WITH OBSERVED FACTS THE CONVECTIONAL THEORY General convectional motion, 139. Arrangement of isobaric surfaces in a general convectional circulation, 140. Conditions of steady motion, 141. Barometric gradients, 142. Relation of wind direction to pressure gradient, 143. Effects of the earth's rotation on wind direction and pressure, 144-147. Buys Ballot's law, 148. THE COMPARISON OF THE CONSEQUENCES OF THE CONVECTIONAL THEORY WITH THE OBSERVATIONS OF PRESSURE AND WIND, 149 D. A GENERAL CLASSIFICATION OF THE WINDS THE CLASSIFICATION OF THE WINDS, 150 PLANETARY WINDS Typical system, 151-155. Trade winds, 156. Doldrums, 157. Horse latitudes, 158. Prevailing westerly winds, 159. Upper currents, 160. TERRESTRIAL WINDS Definition, 161. Annual migration of the winds, 162. Subequatorial and subtropical wind belts, 163. CONTINENTAL WINDS Definition, 164. Monsoons, 165. Other land effects, 166. LAND AND SEA BREEZES, 167 114 METEOROLOGY MOUNTAIN AND VALLEY BREEZES, 168 ECLIPSE, LANDSLIDE, TIDAL, AND VOLCANIC WINDS Eclipse winds, 169. Landslide and avalanche winds, 170. Tidal winds, 171. V6lcanic winds, 172. CYCLONIC STORMS, 173 WINDS OF OTHER PLANETS, 174, 175 A. THE OBSERVATION AND DISTRIBUTION OF PRESSURE THE DETERMINATION OF ATMOSPHERIC PRESSURE 98. Atmospheric pressure. The second meteorological element to be considered is the pressure of the atmosphere. We are probably less con- scious of atmospheric pressure and its changes than of any of ^ e ^ ner meteorological elements. It is true that great tant mete- changes in atmospheric pressure do produce marked physio- dement" 1 logical effects, but we are absolutely unconscious of the ordi- nary changes in atmospheric pressure from day to day. In meteorological work and weather forecasting, however, the pressure, with the possible exception of temperature, is the most important of the elements. The atmosphere has mass and is acted upon by gravity, and thus possesses weight and exerts a downward pressure. The pressure of the The atmos- atmosphere is simply the weight of the column of air above pheric pres- the station in question, extending to the limits of the atmos- wdghtof 6 phere. Atmospheric pressure thus diminishes with elevation the atmos- above the earth's surface because there is a less quantity of air to exert a downward pressure. Since the atmosphere is a gas, the pressure, as in the case of all fluids, is exerted in every direction. If there were no temperature differences and thus no winds, the pressure would be the same at all points on a level surface, as for example,* at all points on the hydrosphere. This is not Atmosphenc . pressure is the case, however, and the pressure is different at dinerent not con- points at the same level and is also constantly changing at the same station. It is desirable, therefore, to have instru- ments for determining the pressure of the atmosphere. The instrument for determining atmospheric pressure is called a ba- PRESSURE AND CIRCULATION OF THE ATMOSPHERE 115 rometer; 1 and there are two kinds of barometers, those employing a fluid and those without fluid. Since mercury is the fluid ordinarily used, such barometers are usually called mercurial barometers, The two Barometers which do not use a fluid are called aneroid 2 ba- kinds of rometers. The pressure of the atmosphere might be ex- l srs * pressed in poundals per square foot, or in dynes per square centimeter. As a matter of fact, however, it is expressed in terms of The pres- the length of an equivalent or balancing mercury column, sure is ex- A pressure of thirty inches thus means that the pressure of terms' of" 1 the atmosphere is the same as the pressure exerted by a inches of column of mercury thirty inches long. 99. Mercurial barometer: history, construction, corrections. The history of the mercurial barometer 3 dates from a series of experi- ments made by Torricelli in 1643. It was a remark of Galileo Torriceiii's Galilei of Pisa, the father of experimental science, when it was experiment called to his attention that water would not rise in a pump " more than eighteen cubits above the level of a well, that nature probably did not abhor a vacuum above that height, which attracted the attention of Torricelli, who was then his pupil and later his successor in the chair of Philosophy and Mathematics at Florence. Not being satisfied with this explanation, he instituted a series of experi- ments which led to the invention of the barometer in 1643. His most famous experiment, as pictured in Fig. 48, consisted in filling a glass tube The more than thirty inches long with mer- tion of the cury, covering the open end, and then inverting it over a vessel containing mercury. When the open end was uncovered, the mercury immedi- ately fell to a height of about thirty inches regard- less of the length of the glass tube. Torriceiii's explanation was that it was the pressure of the atmosphere which was supporting the column of mercury. This explanation received full confirma- * . tion a few years later, in 1648, when Pascal per- suaded his brother-in-law Perrier to ascend the Puy de Dome near Clermont with a Torricellian barometer. The diminution in length of 1 j3dpos = weight; ptrpov = measure. 2 d = without; vepbs = fluid. 3 For an historical account of the barometer, see Quarterly Journal of the Royal Meteor- ological Society, No. 59, July, 1886, p. 131 ; or Meteorologische Zeitschrift, 1894, p. 445. FIG. 48. Torriceiii's Experiment. 116 METEOROLOGY the mercury column during the ascent supplied final proof that it was the pressure of the atmosphere which was supporting the column of mercury. 100. A mercurial barometer of the Fortin form as used at the present time consists essentially of a glass tube more than thirty-four inches long, filled with mercury, Description . of a mercu- and inverted over a vessel containing mercury rial barom- anc j ca n e( j the cistern. The mercury used in the tube must be pure, and it must have been previously boiled in order to extract all air and moisture. An air trap is often inserted in the tube to prevent the ascent of any air or moisture into the Torricellian vacuum above the mercury column. The whole is inclosed in a brass case for protection, in order to make it somewhat portable, and to enable it to be suspended vertically. The brass case is cut away at the top, exposing the glass tube containing mercury, and is provided with a scale and vernier for reading the height of the mercury in the tube. A barometer is usually suspended from the top and held vertically by means of a ring with three set screws at the bottom. The back board is provided with two translucent windows at the bottom and top to illuminate the cistern and top of the mercury column, particularly at night, when a light is placed back of them. Figures 49 and 50 represent the barometer as a whole and a sectional view. The cistern is made up of a glass cylinder F, which al- lows the surface of the mercury q to be seen, and a top plate G, through the neck of which the barometer tube t passes, and to which it is fastened by a piece fcription of 6 " of kid leather, making a strong but flexible the cistern joint. To this plate, also is attached a small ivory point h, the extremity of which marks the commencement or zero of the scale above. The lower part, containing the mercury, in which the end of the barometer tube t is plunged, is formed of two parts i, j, held together by four screws and two divided rings. To the lower piece j is fastened the flexible bag N, made of kid leather, furnished in the middle with a socket k, which FIG. 49. res t s on the end of the adjusting screw 0. Those parts, Barometer. with the -glass cylinder F, are clamped to the flange B by of a barom- eter. PRESSURE AND CIRCULATION OF THE ATMOSPHERE 117 means of four long screws P and the ring R ; on the ring R screws the cap S, which covers the lower parts of the cistern, and supports at the end the adjust- ing screw 0. G, i, j, and k are of boxwood ; the other parts of brass or German silver. The screw serves to adjust the mercury to the ivory point, and also, by raising the bag, so as to completely fill the cistern and the tube with mercury, to put the in- strument in condition for transportation. A thermometer is attached to the middle of the barometer to indicate the tempera- ture of the instrument. A good barometer costs from $30 to $200. 1 01. There are two steps in reading a barometer. In the first place the screw at the bottom of the instrument must be turned until the surface of the . The two mercury in the cistern has been steps in brought to the end of the scale or Beading a . . barometer. to the end of the ivory point which serves as the end of the scale. This can be done very exactly because the ivory point is reflected in the mercury in the cistern. The surface is raised until the point and its image appear just to touch. The vernier is then set tangential to the top of the mercury column and the scale reading observed. There are three correc- tions to be applied to this reading of the length of the mercury column: (1) the meniscus correction; (2) the tern- ine tnree perature correction ; (3) the grav- corrections ity correction. Since mercury does not wet glass, capillary ac- tion will depress the mercury column and give it a rounded top, called the meniscus, as shown in Fig. 51. In reading a barom- e the zero of the vernier is placed FIG. 50. Cross-section of the tangential to the upper surface of the . em of a Barometer. to be applied. 118 METEOROLOGY meniscus. A correction must thus be applied in order to determine what the reading would be if capillary depression did not exist, and The menis- ^ ne mercury column were cut square across in- cus correc- stead of rounded. This correction is usually applied by the maker by moving the scale the proper amount. If this has not been done, the correction is usually combined with the temperature correction and furnished by the maker of the instrument as a table of corrections. The pressure inside a warm room is the same as out of doors, whatever the temperature may be there. Ordinary The tem- buildings are not sufficiently air-tight to permit perature differences of pressure between the inside and correction. ^g outside to exist for more than a few FIG. 51. The T , Meniscus. moments. If a barometer were taken from a warm room into the cold outside air, both the mercury and scale would contract and the reading would become different, although the pressure would be the same. It is necessary, therefore, to reduce all readings of a barometer to a standard temperature. 32 F. or C. are considered standard temperatures, and the accompanying table gives the corrections to be applied to a mercury barometer with a brass scale for various temperatures and pressures. Temp. F. 10 20 30 40 50 60 70 80 90 100 Pressure in Inches 26 +0.068 +0.020 -0.027 -0.074 -0.121 -0.167 + 0.044 - 0.003 - 0.050 - 0.097 - 0. 144 27 +0.070 +0.021 -0.028 -0.077 -0.125 -0.174 + 0.046 -0.003 -0.052 -0.101 -0.150 28 +0.073 +0.022 -0.029 -0.080 -0.130 -0.180 + 0.047 -0.003 -0.054 -0.105 -0.155 29 +0.076 +0.023 -0.030 -0.082 -0.135 -0.187 + 0.049 -0.004 -0.056 -0.109 -0.161 30 +0.078 +0.024 -0.031 -0.085 -0.139 -0.193 + 0.051 -0.004 -0.058 -0.112 -0.166 31 +0.081 +0.024 -0.032 -0.088 -0.144 -0.200 + 0.053 -0.004 -0.060 -0.116 -0.172 PRESSURE AND CIRCULATION OF THE ATMOSPHERE 119 The value of gravity is not the same at all points on the same level sur- face. Thus a column of mercury of the same length would The gravity not give the same pressure at all points on a level surface. It correction, is necessary, therefore, to reduce the readings to a standard value of gravity. The value of gravity at 45 north latitude is considered standard, and the accompanying table gives the corrections to be ap- plied for various latitudes. Latitude 90 80 70 60 50 40 , 30 20 10 + 0.07 +0.04 -0.01 -0.06 -0.08 inches Correction + 0.08 + 0.06 + 0.01 - 0.04 - 0.07 A good mercury barometer will indicate the pressure accurately to the hundredth of an inch, and will give a fair approximation to the thou- sandth. There are, however, several sources of error which Accuracy depend upon the accuracy of construction. Some of these and sources are the accuracy of the scale, the correct adjusting of the ivory c pointer, the purity of the mercury, the excellence of the vacuum above the mercury column, etc. There are various forms of mercury barometers. A modified form is sometimes used on shipboard ; and other special forms, such as the siphon barometer, and others, have been devised. For a full treat- Modified ment of these the reader must be referred, however, to special forms of the treatises on this subject. 1 102. The aneroid barometer. The aneroid barometer, as the name implies, is a fluidless barometer, and is thus much more portable than the mercury barometer. It was invented by Vidi in 1848. It consists essentially of a so-called vacuum box about an inch O f an aner- and a half in diameter and one quarter inch thick, made of oid barom - eter. German silver with corrugated top and bottom. The air has been exhausted and it is hermetically sealed. It is kept from collapsing by a strong leaf spring which extends over the vacuum box. If the pres- sure increases, the box is pressed together against the action of the spring; and conversely, if the pressure decreases, the elasticity of the spring causes the box to expand slightly. These small motions are magnified by a sys- tem of levers and communicated to a pointer which moves over a dial which is graduated to inches to correspond to a mefrcury barometer. Figures 52 and 53 represent an aneroid barometer and its internal con- struction. 1 Both the English and metric barometric scales are used. Sometimes both are put on the same barometer. In Appendix III a graphical comparison of the two scales is given. 120 METEOROLOGY FIG. 52. An Aneroid Barometer. FIG. 53. -T- The Internal Construction ' . \M<>roid Ham:' PRESSURE AND CIRCULATION OF THE ATMOSPHERE 121 The meniscus and gravity corrections do not exist in connection with the aneroid barometer. The instrument is, however, slightly affected by temperature. In order to compensate for tern- ? wthe perature changes a small amount of air is often left by the tions are maker in the vacuum box, and such an instrument is usually marked with the word " compensated " on its face. An aneroid barometer is at best an inaccurate instrument as compared with the mercurial barometer. The great advantage lies in^tsjortability. If it is not jarred unduly and is fre- The accu _ quently compared with a mercury standard, its indications racy of an can be trusted to a tenth of an inch and may give a a fair approximation to the hundredth. The cost of a good aneroid barometer varies from $5 to $40. The words fair, storm, change, rain, very dry, and the like, often found on the face of an instrument, are meaningless. FIG. 54. The Richard Freres Barograph. 103. Barographs. For many purposes it is desirable to have a con- tinuous record of barometric pressure and the instrument for keeping this continuous record is called a barograph. The Richard Descri tion Freres form of barograph (Fig. 54) is the one ordinarily used O f the Rich- by the U. S. Weather Bureau and the one commercially on the market. It consists of a battery of from six to ten of the vacuum boxes of the aneroid barometer placed one above the other. The 122 METEOROLOGY reason for the large number of vacuum boxes is to lessen the effect of the irregularities in any one, and to make the instrument more sensitive. The motion of these vacuum boxes is communicated by a system of levers to an arm which carries the V-shaped trough which contains the non-freezing glycerine ink. As the pressure changes this pen moves up and down. The details in the construction of the pen and recording mechanism have been fully described in connection with the thermo- graph in section 68. There are other more accurate forms of barographs in use for the pur- pose of research and in special observatories. The reader must again be referred to special treatises on the subject for the full description of these. 104. Other so-called barometers. In addition to the barometers just described, there are two instruments, usually called barometers, \ Description which deserve a passing mention, of the mouth One is hardly more than a scientific iter * toy and the other merely masquer- ades under the name of barometer. The one, to translate its German name, is called a " mouth-barometer " and, as represented in Fig. 55, consists of a bulb usually filled with a colored liquid which does not readily evap- orate. The bulb is attached to a graduated stem with a moderately small bore. The fluid does not entirely fill the bulb, but an air space is left and the stem protrudes into the bulb in such a way that the air which constitutes this bubble cannot enter the stem in any position of the instrument. The size of the air bubble and thus the height of the fluid in the stem depends upon the temperature and pressure. If the temperature could be kept constant, the height of the fluid in the stem would depend upon the pressure alone and the instrument would thus be a barom- eter. Now the temperature of the human body is remarkably con- stant, and if the bulb of the instrument is held in the mouth, it can be assumed that the temperature is always the same. Such an instru- ment is cheap and very portable, and, when calibrated in terms of a mercurial barometer, it will give results comparable with those ob- tained with an aneroid barometer. The instrument which merely masquerades under the name of barom- eter is usually designated by its makers as " the cottage barometer " FIG. 55. The Mouth- Barometer. :\ PRESSURE AND CIRCULATION OF THE ATMOSPHERE 123 or " the signal service barometer " or sometimes as " the chemical weather glass." It is a sealed glass tube about six or eight inches long and half an inch in diameter filled with a clear liquid which has in it a flocculent, sometimes partly crystalline, tionof^he^ substance. The amount of this substance, its appearance, composition and position in the tube are supposed to indicate the com- ing weather; and it is, of course, advertised as an infallible chemical guide to the coming weather. It is usually mounted in the same case with a thermometer and sold for 50 cents or much less. The actual composition and action of the instrument is this : The clear fluid is nearly always alcohol and the substance in it consists of equal parts of nitrate of potash, camphor, and ammonium chlorid. More of this mixture has been added than the alcohol can dissolve and the excess appears in the tube as the solid substance in the clear liquid. It is a thick tube and is hermetically sealed so that it is not affected in the least by changes in pressure and thus is no barometer at all. It is affected simply by temperature changes. As the tem- perature rises, more of the substance goes into solution. As the tem- perature falls, more of the . substance must come out of solution and appear as solid. Now the rapidity of temperature changes, un- equal temperatures on different sides of the tube, the direction and the amount of the light falling upon it, may make a difference in the form, amount, and position of the substance which comes out of solution as the temperature drops, and this probably accounts for the varied appear- ances of the instrument. Now, temperature changes alone are no in- dicator of the kind of weather that is coming or of the characteristics of a coming storm. Thus this instrument is neither a barometer nor an in- dicator of the coming weather. THE RESULTS OF OBSERVATION 105. The observations. At the regular stations of the U. S. Weather Bureau the atmospheric pressure is determined at 8 A.M. and 8 P.M. by means of a mercurial barometer, and a continuous barograph record is also kept. The instruments used for taking these observations are a good mercurial barometer taken at the and a Richard Freres barograph. These instruments are BtfreatMsta- lo-jsted ordinarily in the office part of the Weather Bureau tfons and Station and not in the thermometer shelter or in the open. m ents S< used. They are more conveniently and safely located within the building, and the atmospheric pressure is the same within the build- 124 METEOROLOGY ing as out in the open. No building is sufficiently air-tight to per- mit differences of pressure to exist for more than a few moments except perhaps in the case of a tornado. No pressure observations are required of the cooperative and special stations. 1 06. Normal hourly, daily, monthly, and yearly pressure. These normals of pressure are computed in exactly the same way as the corre- . spending temperature normals. In order to obtain a good The method 111 i of comput- normal hourly pressure, a barograph record for at least twenty ing pressure years is necessary. If, then, for any given date the pressures at any given hour for the last twenty years are averaged, the result would be the normal pressure at that hour for the given day. In this way the normal pressure for every hour of every day in the year might be determined. As a matter of fact, the days in a month are usu- ally grouped together, and thus the normal hourly pressures for a Janu- ary day or a February day, etc., are determined. If the average pressure for a day is to be determined, the barograph record is almost always used. One half of the pressure observed at 8 A.M. and 8 P.M., one third of the pressure observed at 7 A.M., 2 P.M., and 9 P.M., would give a rough approx- imation to the average daily pressure. Normals of pressure are not or- dinarily computed for any station where a barograph record has not been kept for several years. 107. Diurnal, annual, and irregular variation. If the normal hourly pressures are plotted to scale, the resulting graph represents the . . diurnal variation in pressure. If these normals are not of the daily available, a fairly good idea of the characteristics of the variation in diurnal variation can be formed by considering the change in pressure on some day when the other meteorological elements have been as normal as possible. Figure 56 illustrates a series of days during which the remaining elements remained un- usually normal. The general characteristics of the daily variation are these : the chief maximum usually occurs at about 10 in the morning, the chief minimum at 4 in the afternoon, a secondary maximum at 10 in the evening, and a secondary minimum at 4 in the morning. The pres- sure is thus subject to a double oscillation in the course of a day. The amplitude or amount of the daily variation is always small, never amounting to more than 0.2 of an inch, and in many places being of ten less than a tenth of this. It varies somewhat with the time of year, being usually greater in summer and somewhat less in winter. It varies also with latitude, its greatest values being found in 'the equatorial regions, while the amount grows steadily less with the higher latitudes. It is PRESSURE AND CIRCULATION OF THE ATMOSPHERE 125 also somewhat less on cloudy days than on days with plenty of sunshine. It is also somewhat va ries^i& nt greater for interior the season, stations than for ^on^tcT coast or island stations. For interior stations the secondary maximum and minimum become less promi- nent, while for coast and island stations the two are of equal prominence, or the night maximum and minimum may even become the most important. Elevation also plays an important part in the time of occurrence of the maximum and minimum. Approximate values of the amplitude or amount of the daily variation for various places are here _ *\ The amount given : Calcutta, and charac- India, lat. 24, teristics of 1 the varia- 0.116; Greenwich, tion at dif- lat. 52, 0.020; Dublin, 0.020; St. Petersburg, 0.012 ; Fort Con- ger, lat. 83, 0.010; Yuma, Arizona, 0.129; San Antonio, 0.117; Denver, 0.079; Al- bany, 0.074 ; St. Louis, 0.068 ; Philadr Ma, 0.061; San Franci 0.052 ; Bismarck, 0.038; k % 0.014. Figure 57 represents graphically the daily variation at Mexico City, San Francisco, St. Louis, New York, and Sitka, / (? f f ent places. 126 METEOROLOGY Alaska. By contrasting the values here Sitka Alaska (Lat.573) RE Due : OT 30.020, 23456789 1 'J 00 jL 2345 67 30.16' 80.14 80.12 30.10 80.08 80.06 80.04 30.02 30.00 29.98 29.96 29.94 o.oe' 0.04 0.02 0.00 0.02 0.04 0.06 10 Moo.l m 3 4 5 e (Lal',19-26; loTT' 56789 101112 FIG. 57. The Diurnal Variation in Pressure at Sitka, New York, St. Louis, San Francisco, New Orleans, and Mexico City. by moving waves of higher and lower pressure. given for various places, the truth of the foregoing state- ments as to the amount of the oscillation may be tested. Since the local time of occurrence of the maxima and minima is approxi- mately the same for all places, the diurnal changes in barometric pressure may be thought of as The change produced by in pressure considered waves of higher as caused anc j i ower pres- sure which move westward from the Atlantic Ocean, cross the continent, and pass off into the Pacific Ocean. The location and height of these waves have been computed by Dr. Oliver O. Fassig for each hour of the day for the western hemisphere. The results were found by using the daily variation in pressure, which had been determined from observa- tions at many stations in both North and South America, and drawing lines through those places which at the hour in question showed the same departure from the normal for the day. These charts will be found in Bulletin No. 31 of the U. S. Weather Bureau PRESSURE AND CIRCULATION OF THE ATMOSPHERE 127 or in the Monthly Weather Review for November, 1901. Two of them, for 10 A.M. and 4 P.M., the times of greatest and least pressure, are reproduced as Figs. 58 and 59. The numbers here represent depar- ture from the normal for the day expressed as thousandths of an inch. 1 08. The cause of this diurnal variation in pressure is not a tide in the atmosphere caused by either the sun or the moon, for if it were both the solar and lunar influence would be noticed and the corresponding periods detected. Temperature and the forma- tion of dew without doubt play a large Tem P era - FIG. 58. Diurnal Barometric Wave at 10 A.M., 75th Meridian Time. (FASSIG, U. S. Weather Bureau). part in causing this and dew as diurnal variation, but the exact way in which these two operate to produce the result cannot be satis- factorily stated. Due to the temperature change alone, the maximum would be expected at the time of least temperature or a little later, that is, at sunrise or a few hours after. v The minimum would be expected at the time of highest temperature, that is, from two to four in the afternoon, or somewhat later. Due to the formation of dew the largest amount of moisture is present in the atmosphere in the late afternoon and the least amount at the time of sunrise. The interaction of these two influen- ces o f temperature and dew, , - n however, cannot account for all the characteristics which have De-en ' t-rveu in oormoptmn with the daily variation. FIG. 59. Diurnal Barometric Wave at 4 P.M. 75th Meridian Time. , U. S. Weather Bureau.) 128 METEOROLOGY By means of harmonic analysis 1 this daily oscillation or variation in pressure may be separated into two components, one with a daily or twenty-four hour period, and another with a half-daily or tfoxftato two twelve-hour period. If this is done for many stations, choos- components ing those in equatorial regions as well as those in higher lat- ent* 1 eriods tudes, those in the interior of the continents as well as on the seashore or on islands, those located in valleys as well as on mountain sides and on mountain tops, it will be found that the twelve- hour periodic oscillation is remarkably regular and shows essentially the same characteristics everywhere, while the twenty-four hour oscilla- tion proves to be due almost entirely to local causes and is very different at different stations. Some have even gone so far as to ascribe a cosmic cause to this twelve-hour oscillation, that is, to ascribe it to some influence outside of the earth itself. It has also been thought that this twelve- hour oscillation may be simply a free oscillation of the corresponds atmosphere as a whole, considering it as an elastic body- to nothing This twelve-hour periodic oscillation is nearly as large as the twenty-four hour oscillation, and whenever this is the case and no reason in nature for a twelve-hour period can be found, it is ques- tionable whether this separation into two components can be considered as corresponding to anything in nature, or may not be merely a device of the mathematician. 109. The annual variation in pressure may be found by plotting to scale the normal monthly pressures. If this is done, it will be noted for the interior of continents that the pressure is somewhat higher variation^ /k nan ^ ne vear ly normal) in winter and less in summer ; for the pressure oceans the opposite is true, the pressure being somewhat cause* 8 higher in summer and lower in winter. The cause for this is not far to seek. During the winter, as was seen in con- nection with temperature anomalies, the continents are colder than the surrounding oceans. That means that the air is colder, denser, and heavier, and thus an increase in pressure during the winter is to be expected. During the summer the contrary is true. The continents are warmer than the surrounding oceans, the air is expanded and light, and the atmospheric pressure is correspondingly lower. This variation, again, does not amount ordinarily to more than a few tenths of an inch. The irregular variations in pressure, particularly in the temperate zones, are far larger than any of the periodic variations. Variations in 1 For an illustration see Monthly Weather Review, November, 1906. PRESSURE AND CIRCULATION OF THE ATMOSPHERE 129 pressure of an inch or two follow each other in rapid succession, and at irregular intervals. no. Barometric data. In addition to the normals described above, but few results are computed for the various stations from the Ooserva- tions of pressure. Among those sometimes computed, the following may be mentioned: (1) The daily range. From data which the barograph record the highest and the lowest pressure for may be the day may be determined. The difference between these f^nfthe ob- gives the daily range. From these values of range, normal servations of values for the day, for the month, and for the year may be P computed in the regular way. (2) Absolute range for the months and the year. By the absolute range in pressure for the months and for the year is meant the difference between the very highest and very lowest pressures observed during the period of time in question. (3) Frequency of irregular variations. The interval of time between successive, marked, irregular variations in pressure may be deter- mined from the records of the barograph, and the normal frequency of these irregular fluctuations can thus be determined for the vari- ous months and for the year as a whole. (4) Magnitude of irregular fluctuations. The magnitude of each irregular fluctuation can be deter- mined. If these are averaged in the regular way, the normal amount of the irregular fluctuations for the various months and for the year as a whole may be determined. None of these results is of any great interest or of far-reaching practi- cal importance. At the regular stations of the U. S. Weather Bureau there are three tables for pressure which are kept up to date. These contain : (1) Highest and lowest in inches and hundred ths (reduced to sea level) (the data are given for each month and the year as a whole) . (2) Mean station and absolute monthly range (inches and hundredths) (the data are again given for each month and the year as a whole). (3) Mean hourly pressure (inches and hundredths). THE VARIATION WITH ALTITUDE in. Reduction to sea level. Since the pressure of the atmosphere decreases with elevation, in order to compare pressures observed at differ- ent ^cations, it is necessary to take account of the elevation. The old rpl s was formerly done by determining tha difference be- way - t,,* ecu* the observed pressure and the normal yearly pressure for the sta- 130 METEOROLOGY tion in question. These differences could then 'be compare^: and thus the conclusion reached as to which place had the higher or lower pressure. At the present time all pressure observations are reduced to the same level; that is," to what the pressure would be if the observation had been made at sea level. In order to make this correc- way iTto ** * ^ lon ) ^ e weight of a column of air reaching from the station add the in question to sea level must be added to the observed the* column P ressure - Now the weight of this column of air is not a of air reach- constant, but varies with the pressure, with the tempera- sea leveL* ^ ure J an d slightly with the moisture in it. If the pressure is high, then the air is dense and heavy. If the temperature is high, it is correspondingly expanded and light. Moist air is lighter than dry air, when other conditions are the same. In order to deter- mine the value of this' correction, elaborate tables are ordinarily used. The best set of tables is probably the Smithsonian Meteorological Tables, published at Washington and carrying the number 1032 in the Smithso- nian Miscellaneous Collections. In this volume are contained the neces- sary tables for reducing barometric pressure to sea level, together with the explanation and derivation of the formulas used. In Appendix IV a short table is given, but this is intended not to serve for the reduction of well-taken observations, but simply to give a rough idea of the amount of the correction in a few instances. In a certain sense this reduction to sea level is fictitious. The mass of a column of air reaching down to sea level is added. The mass of this column of air is quite different from what it would be if there were no mountain or plateau and the station were actually located at sea level. For this reason, the older method by means of differences from normal might be better, particularly in the construction of weather maps. 112. Barometric determination of altitude. Since barometric pres- sure depends upon elevation, it is possible, by observing the barometric pressure at two different, near-by stations to determine their taneous ob- difference f elevation. In order to make a precise determina- servations tion, simultaneous observations of pressure and temperature and tenTer are necessarv - The moisture at the two stations is also some- ature at the times determined. The practical details in securing these two sta- simultaneous observations will at once suggest themselves. tions must . be obtained. If two observers and a sufficient number of instruments are available, it may be easily arranged to take the observations at some definitely appointed hour. For the pressure observations at the higher elevation, particularly if the station is located on a PRESSURE AND CIRCULATION OF THE ATMOSPHERE 131 mountain top, the aneroid barometer is ordinarily used. This is not as accurate as a mercurial barometer, but it is much more portable and convenient to carry. The temperature is ordinarily deter- mined by means of a sling thermometer. If a ventilated thermometer is available, it will give far better results. It curing the is ordinarily too troublesome to construct or to carry up a ^ons^*" thermometer shelter for protecting an ordinary thermometer. If two observers and the proper apparatus are not available, fair results can be obtained by taking observations at the base station, both before and after the ascent, and then determining by interpolation the proper value of pressure, temperature, and possibly moisture, at the time when the observations were taken at the summit. If the aneroid barometer is used, it should be compared both before and after the ascent with a mer- curial standard, in order to check its accuracy and to be sure that it has not become deranged by the jars of transportation. 113.. Let BQ and T indicate the pressure and temperature at the base station, that is, at the station whose elevation is known, and let B l and T l indicate the pressure and temperature at the summit or at the station for which the elevation is to be determined. The first and simplest method of computing the difference in eleva- tion is to take no account of the temperature and moisture, and allow 90 feet for each tenth of an inch of pressure difference. It has been found that the general average for all conditions There are of pressure and of temperature and of moisture is about 90 methods of feet to a tenth of an inch. Thus, if the difference in pres- computing . . . , a difference sure between two stations is two inches, it means a differ- in elevation ence in elevation of approximately 1800 feet. Whenever from the ob- servanons aneroid barometers are provided with a separate scale for ma de. indicating elevations, it is always divided so that a tenth of an inch corresponds to 90 feet. TEMP. F. Pressure 10 20 30 40. 50 60 70 80 90 100 22 inches 111 114 116 119 122 124 127 130 132 135 138 24 . 101 104 106 109 111 114 116 119 121 124 126 26 , 94 96 98 101 103 105 107 110 112 114 116 28 . 87, 89 91 93 95 98 100 102 104 106 108 29 . 84 86 88 90 92 94 96 98, 100 102 v 104 29.5 83 85 87 89 91 93 95 97 99 101 aos 30.0 81 83 85 87 89 91 93 95 97 99 1P1 30.5 80 82 84 86 88 90 92 94 96 98 100 31.0 78 xn 83 84 86 88 90 92 94 96 98 METEOROLOGY A better method of computing the difference in elevation is to take from the accompanying table the actual number of feet which correspond to a tenth of an inch difference in pressure for the average pressure and the average temperature at the two stations. The third way of computing the difference in elevation, which is per- haps a little more accurate than the use of the table given above is by means of the following formula : Difference in elevation 55,761 + 117 - 60) } / J This formula, as will be seen, contains nothing but the observations of pressure and temperature at the two stations. If the most exact possible use of the observations is to be made in determining the difference in elevation, the elaborate Smithsonian Meteorological Tables referred to above must be used. The formula is there derived, the necessary tables are given and their use explained. 114. Variation with altitude. The accompanying table shows ap- proximately, for average conditions, the barometric pressure which cor- The baro- responds to various elevations. These values, however, cor- metric pres- respond to average conditions, and are not exact enough for coirerponds ^ ne determination of elevation, particularly as the baro- to various metric pressure at any given station is constantly changing. altitudes. BAROMETRIC PRESSURE ALTITUDE BAROMETRIC PRESSURE ALTITUDE 30 inches Ofeet 21 inches 9,300 feet 29 inches 910 feet 20 inches 10,600 feet 28 inches 1,950 feet 18 inches 13,200 feet 27 inches 2,820 feet 16 inches 16,000 feet 26 inches 3,800 feet 15 inches 3.6 miles 25 inches 4,800 feet 1\ inches 6.8 miles 24 inches 5,900 feet 3| inches 10.2 miles 23 inches 7,000 feet T ^y inches 24.1 miles 22 inches 8,200 feet THE DISTRIBUTION OF PRESSURE OVER THE EARTH 115. Construction of isobaric charts. Isobaric charts are con- structed by making use of the normals of pressure which of construct- have been determined for various stations .in all parts of ing an iso- the world. These normals of pressure must first be reduced to sea level, and they are then charted on a map, and lines PRESSURE AND CIRCULATION OF THE ATMOSPHERE 133 are drawn connecting those stations which have the same value. These lines are called isobaric lines, or simply isobars. 1 1 1 6. Isobars for the year. Chart X shows the isobars for the world for the year. It is the normal annual pressures which have been here used for the construction of the chart. The following charac- The three teristics are at once evident : (1) The equatorial belt of low character- pressure and the belts of high pressure at 35 N. and 30 S. b^forthe" latitude. It will be noticed that along the equator there is a year for the belt of low pressure, the least'^pfessUre being 29.80 inches. world ' This belt is irregular, of varying width, and not of the same pres- sure throughout, and it lies somewhat on the north side of the equa- tor. The belts of high pressure are also not of the same pressure through- out. . The one at 35 N'. has its areas of highest pressure over the Pacific Ocean, the Atlantic Ocean, and central Siberia. In passing over Siberia it lies far north of .the equator. The southern belt of high pressure has its peaks of pressure over Ihe Pacific Ocean, the South Atlantic Ocean, and the Indian Ocean. From these two belts of high pressure, the pressure diminishes rapidly toward the poles ; in the southern hemisphere quite regularly; in the northern, however, with less regularity. (2) Regu- larity in the southern hemisphere. It will be noticed that the low pressure near the north pole consists of two depressions, one in the North Pacific near Alaska, and the other in the North Atlantic near Iceland, each with a central pressure of 29.70 inches. In the southern hemisphere, how- ever, the pressure drops much more uniformly and rapidly from the belt of high pressure toward the pole. (3) Lower in southern hemisphere than in-the northern hemisphere. It will also be seen that the diminution in pressure is much greater in the southern hemisphere than it is in the northern. 117. Vertical section along a meridian. A study of the distribution of pressure in a vertical section along a meridian, as shown in Fig. 60, will prove instructive. The north and south poles of the How the 30 _ earth and the equator are indicated by N., S., and E., respec- inch line is tively. At the equator, in .order to reach a barometric pres- sure of 30 inches, it is necessary to go a certain distance below sea level, arid the same is true at the north and south poles. At 35 N. and 30 S.. it is necessary to ascend a certain distance above the eartti's surface in order to find the 30-inch line. The heavy line marked 30 in the diagram thus indicates in the vertical set aon those places which would have a pressure of 30 inches. In order to locate the 29-inch line, it is necessary 1 From the Greek '. t 9381 J Pu y de Dome > France, 4800 ; Schnee- koppe, Germany, 5246 ; Sentis, Switzerland, 8215 ; Sonn- blick, Austria, 10,155; Wendelstein, Germany, 5669. The Eiffel Tower in Paris at an altitude of 990 feet has given very interesting meteoro- logical records. These records are of especial value because the slender form of the tower causes no disturbing influence in the condition of the air around it. THE RESULTS OF OBSERVATION 130. The observations. At all regular stations of the U. S. Weather Bureau continuous records of the wind direction and wind velocity are maintained. The instrument used for determining the wind direction is the wind vane with an electric contact maker. It will be remem- bered that by means of a sliding contact one or two of the four cir- PRESSURE AND CIRCULATION OF THE ATMOSPHERE 147 cuits N, S, E, and TF, are constantly kept closed. The current is sent through the instrument every minute by means of a clock, and on the revolving drum in the office below the wind direction is thus recorded to eight points of the compass. The instru- Th f obser - ment used for determining wind velocity is the Robinson wind at a cup anemometer. The contact is made after every mile r . e g ular sta - ' tion ; the of wind has gone by, and a spur is made on the revolving instruments drum to indicate that fact. Wind direction, wind velocity, u ^ e . d and amount of rainfall, and the duration of sunshine 1 are all location, recorded on the same revolving drum, which is usually spoken of as a triple register or meteograph and is located in the office part of the Weather Bureau. The wind vane and anemometer are usually exposed at the top of some high building at a considerable distance above the ground. The purpose of this is to give them as free an exposure as possible and thus prevent local influences from disturbing the wind direction or velocity. At the cooperative stations of the U. S. Weather Bureau the wind direction for the day is the only thing recorded. If the wind Observa has shifted during the day, the middle of the arc through tions at a which it has shifted during the twenty-four hours is deter- mined and the corresponding direction recorded. 131. Prevailing wind direction; wind roses. Since wind direction is not a mere number, normals cannot be com- puted in the usual way. For this rea- j. i Prevailing son, instead w j n( j a h- ec _ of the three tion ; two words mean, average, and normal, the single word " prevailing " is used. This distinction, how- ever, is not always rec- ognized. The prevail- ing wind direction may ^ 6g _ wind Rose for ^ 1909> at gyracuse> N Y be expiressed in two ways : either by means of a table, or graphically by means of what is JAN., 1909 1 One circuit is used for the double purpose of recording rainfall and sunshine. But confusion results, because it seldom rains while the sun is shining. 148 METEOROLOGY ALBANY JAN. ,1909, called a wind rose. If the prevailing wind direction, be it for a month or a year, or an indefinite time, is expressed by means of a table, this table contains simply the number of times each wind direction was ob- served. The number of calms must also be noted. The graphic representation of the table is known as the wind rose. The four directions north, south, east, and west are first drawn from a central W- ^ E point, and then the four intermediary directions. On these eight lines distances are laid off propor- tionately to the num- ber of times each of these wind directions was observed. If these points are con- nected by straight lines the resulting figure is called a wind rose. The number of calms may be expressed by a circle described about the center with a radius , proportionate to the The method * of construct- number of calms. In ing a wind tne following table is rose. . given the number of times the wind blew from each FIG. 69. Wind Rose for Jan., 1909, at Albany, N.Y. direction at Syracuse in the Mohawk Valley, and Albany in the Hudson River Valley during January, 1909. Figures 68 and 69 are the corre- sponding wind roses. The reason why east and west winds occurred at Syracuse and not at Albany is probably because the Mohawk Valley runs east and west, while the Hudson Valley runs north and south. In other respects the wind roses are quite similar. 132. Normal hourly, daily, monthly, and yearly velocity. Since the wind velocity is a mere number, the various normals may be computed SYRACUSE ALBANY N 5 NE 1 1 E 4 SE S 10 10 sw 3 5 W 3 NW 10 10 Calms 31 31 PRESSURE AND CIRCULATION OF THE ATMOSPHERE 149 in the usual way. In determining a normal hourly wind velocity, however, it is not customary to determine it for every hour of every day in the year, but for the various hours of ing the the months as a whole. For ex- ^^ es . ve " ample, it would be the 8 A.M., 9 A.M., 10 A.M., etc., wind velocities for Jan- uary which would be determined. 133. Diurnal, annual, and irregular vari- ation. The graph which represents the diurnal variation in wind velocity may be deter- mined by plot- ting to scale the normal hourly wind velocities. For New Eng- or D O I I 1. *n 10 9 8 7 s* ^ > \ Jft NUARY ^v ^ ^ * St L^ulf ^ *-> 11 89 D-191 ^ \ V k JUL-Y / ' *-^ *>s / ^ 17 16 15 14 13 12 11 10 9 8 I -( 89 9- 91 o)- -N EV\ rv OFK- S ^~* \ , Ais uJ VR f / -V "S, ^ / f~-+ "^ S *> K ^ YE AF [ ^ f s 4 5 s ^"N, ? / \ / \ ^w JL L\ / s ^ J? (Ar em orr eter 350 ft. abc ve ore un ) > 2 4 e 8 1012 246 81012 FIG. 70. The Daily Variation in Wind Velocity at New York, St. Louis, and San Francisco for January and July. 12 2 4 6 8 1012 2468 1012 NOON FIG. 70 a. The Daily Variation in Wind Velocity at New York, St. Louis, and San Francisco for January and July. land and the larger part of the United States the higher values of wind velocity occur during the day and the lower veloc- ities at night.- The maximum usually occurs between twelve and four in the afternoon, and the minimum at about the time of sunrise. The reason for this daily variation is to be found in convection. During the day the layers of air near the ground become heated ; and wind ve- as they rise, due to convection, i s a c a u *e. the colder air must come down to take the place of the rising air and bring with it the higher wind velocities of the upper atmosphere. The diurnal variation 122 4 6 8 10 122 46 8 1012 - NOON FIG. 70 b. The Daily Variation in Wind Velocity at New York, St. Louis, , and San Francisco for January and July. 150 METEOROLOGY of wind velocity is greatest on land and practically disappears over the ocean. It is less in winter than in summer and less on cloudy days than on clear days. The reason for these facts is evident, since convection on land, during the summer and on clear days, is far greater than under the opposite- conditions. Figure 70 shows graphically the change in wind velocity during the day at New York, St. Louis, and San Francisco for both January and July and illustrates well the truth of the statements which have just been made. There is also a slight diurnal variation in wind direction. This, how- ever, is so masked and changed by local conditions, particularly if the station is located near the seashore, where the land and sea variatio/in breezes blow, or near mountains, where the mountain and wind direc- valley breeze is felt, that the typical diurnal variation is cause"* hardly noticeable. In order to observe this diurnal variation in all its simplicity, it would be necessary to have a station surrounded on every side by practically identical conditions. If this were the case it would be found that the wind shifts slightly in a clockwise direction during the daytime and shifts back again in a counterclockwise direction during the night. The reason for this is again to be found in convection. The upper air currents always blow in a direction turned somewhat clockwise as contrasted with the surface winds. Thus, during the day, as these upper air masses come down, they will bring with them a wind direction turned slightly in a clockwise manner. The diurnal variation in wind velocity and in wind direction as described above occurs only at low altitudes. At high on the altitudes, at the level of the clouds for example, exactly the variation reverse is true; that is, the higher values of wind velocity occur during the night and the wind direction shifts in the opposite direction. If the prevailing wind direction for the various hours of the day and the normal hourly wind velocities for any station are known, they may be expressed graphically, as in Fig. 71, for the top of the Eiffel Tower at Paris. The wind direction during any tion of the hour of the day may be found by connecting the corre- ft!anhi Vari ' spending hour with the point 0, and the wind velocity by wind direc- noting the length of the line connecting the hour in ques- tion with the P int 0- li wil1 be seen tnat at this sli S nt elevation, 990 feet, a modification of the surface conditions is already apparent. PRESSURE AND CIRCULATION OF THE ATMOSPHERE 151 134. If the normal daily or the normal monthly velocities are plotted to scale, the graph will indicate the annual variation in wind velocity. In general, the wind blows harder in winter than in summer. The maximum usually comes in the very late winter, Feb- variation in ruary, March, or April, and the minimum during the summer, * y^' g July or August. The reason for this is twofold. In the first character- place, when the trees are covered with leaves and vegetation c^use* 1 * 4 is most luxuriant, wind velocities are lessened much more by friction than during the winter, when the trees are bare and the ground is snow-covered and frozen. In the second place, the poleward temperature gradient is much 8 p M greater in winter than in summer (see section 88), and it will be seen later that it is this difference in temperature between equator and pole which drives the convectional circulation which is at the foundation of the general wind system of the globe. The greater the difference in temperature between the FIG. 71. Diurnal Variation in Wind Direction and Velocity at the Top of the Eiffel Tower during June, July, and August. (After -ANGOT.) equator and pole, the more energetic will be this circula- tion, and thus the higher will be the wind velocities. The following table gives the normal monthly wind velocity for several stations in the United States. In Fig. 72 the results are shown graphically for four stations, Philadelphia, Chicago, Phoenix, and San Francisco. The graphs for Philadelphia and Chicago are typical for the northeastern and central portion of the country and illustrate the description of the annual variation which has just been given. In the Southern States and on the Pacific coast the character- istics are very different because the whole type of weather and weather control are different. 152 METEOROLOGY O AJ i oc 10 \ \ Chicago Philadelphia San Francisco Phoenix M.ONTK8 FIG. 72. The Annual Variation in Wind Velocity. PRESSURE AND CIRCULATION OF THE ATMOSPHERE 153 THE NORMAL WIND VELOCITY IN MILES PER HOUR FOR THE VARIOUS MONTHS AND FOR THE YEAR 3 1 i A I iio 3gg B- Is -s 3 - -' S a S 1 5 o M OQ I o fc 6 p i B 3J 3 Bismarck, N. Dak. 35 34 8 9 10 12 11 10 9 9 10 10 9 8 10 Charleston, S.C. 92 38 9 12 11 12 11 10 10 9 9 9 9 9 10 Chicago, 111. 274 15 18 18 19 19 17 14 14 14 16 17 18 19 17 Columbus, Ohio 222 9 10 10 9 8 7 7 6 7 8 9 9 8 El Paso, Tex. 133 10 12 14 13 12 10 7 9 9 9 9 10 10 Indianapolis, Ind. 164 12 11.7 11.5 12.1 11.3 9.9 8.9 8.2 7.4 8.3 9.4 10.4 11.5 10.0 Key West, Fla. 53 38 11 11 11 11 9 8 8 7 8 11 11 11 10 New Orleans, La. 121 39 8.8 9.3 9.4 8.9 7.6 6.1 6.1 6.1 7.4 7.8 8.3 8.8 7.9 Omaha, Neb. 121 37 9 9 10 11 9 8 7 7 8 8 9 9 9 Philadelphia, Pa. 184 11.1 12.0 11.6 11.3 10.2 9.4 9.0 8.4 8.9 10.3 10.4 10.9 10.3 Phoenix, Ariz. 56 15 3.4 4.0 4.5 4.6 4.6 4.4 4.5 4.0 3.8 3.5 3.3 3.2 4.0 Portland, Ore. 106 37 6 7 6 6 6 6 6 5 6 5 6 6 6 St. Louis, Mo. 317 10 12.0 12.1 12.4 12.0 10.9 9.0 8.8 8.0 9.2 10.4 11.8 11.8 10.7 St. Paul, Minn. 124 36 7.8 8.3 8.8 9.3 8.7 7.7 7.1 7.1 8.0 8.5 8.1 7.8 8.1 San Francisco, Cal. 204 21 7.3 7.5 9.1 10.4 11.1 13.0 13.5 12.6 10.4 8.1 6.9 6.9 9.7 Seattle, Wash. 151 7 1 73 74 68 64 6? 55 / t 8 54 56 70 70 64 The year 1908 is the last year included in these normals. There is also an annual variation in wind direction. This, however, depends so much upon the location of the place on the earth's surface that no general rules can be laid down. For New England the prevailing winds of winter are more northwest and north than during the summer, when they shift more to the south and wind direc- southwest. The reason in every case is the building up of ^ cause. areas of high pressure over the continents during the winter and of low pressure over the continents during the summer. It is to these changing areas of high and low pressure that the annual change in wind direction is due. 135. In addition to these regular diurnal and annual changes of wind direction and velocity, there are many irregular variations. In the first place, both wind direction and wind velocity are constantly changed slightly from moment to moment, in wind di- - We characterize this by saying that the wind is usually gusty, This is due entirely to the irregularities in the surface of the earth over which the air is moving. In addition, wind Direction and * 154 METEOROLOGY wind velocity are very different on different days. The reason for this change in direction and velocity is to be found in the various storms, which are to be discussed later (Chapter VI). 136. Wind data. In addition to the prevailing wind direction and the various normals in connection with the velocity which have been Extremes described above, very few results are computed from the ob- of velocity, servations of wind. Practically the only one of any great interest is the extremes of velocity which have occurred at various places. At nearly all stations in the United States wind velocities have gone above fifty miles an hour for a period of five minutes and at nearly every sea- coast station velocities above seventy miles for the same period have been recorded. The accompanying table gives the maximum wind velocity ever recorded at a number of stations in the United States. At the regular stations of the U. S. Weather Bureau there are five tables of wind data which are kept constantly up to date. These contain (1) Total movement in miles. (The data are given for the various months and for the year as a whole.) (2) Prevailing direction and average hourly velocity (velocity to tenths). (3) Maximum velocity, direction, and date. (4) Mean hourly wind velocity (miles and tenths per hour). (5) Prevailing wind direction (hourly). STATION MAXIMUM VELOCITY, MILES DIREC- TION DATE OF OCCURRENCE LENGTH OF RECORD IN YEARS LAST YEAR INCLUDED IN RECORD Abilene, Texas . . . 66 sw sw May 8, 1892 June 8, 1892 22 22 1908 1908 Albany, N.Y. . . . 70 w E Feb. 2, 1876 Oct. 23, 1878 34 34 1907 190$ Atlanta, Ga. . . . 60 JNW 1NW Feb. 16, 1903 April 8, 1907 29 29 1908 1908 Atlantic City, N.J. 72 NE Sept. 10, 1889 34 1908 Augusta, Ga. . . 52 NE July 28, 1893 37 1908 Baltimore, Md. 70 W June 20, 1902 37 1908 Binghamton, N.Y. 44 W Jan. 20, 1907 17 1907 Bismarck, N.D. 74 NW Mar. 10, 1878 33 1908 Block Island, R.I. 90 NE Oct. 27, 1898 28 1908 Boise, Idaho . . 55 SW May 13, 1900 9 1908 fNE Mar. 3, 1891 36 1908 Boston, Mass. . . . 60 N |NE Oct. 27, 1898 Oct. 24, 1901 37 37 1908 1908 I E Nov. 5, 1900 37 1908 Buffalo, N.Y. . . . 90 SW Jan. 13, 1890 37 1907 Burlington, Vt. . . 60 SE Jan. 20, 1907 22 1907 PRESSURE AND CIRCULATION OF THE ATMOSPHERE 155 STATION MAXIMUM VELOCITY, MILES DIREC- TION DATE OF OCCURRENCE LENGTH OF RECORD IN YEARS LAST YEAR INCLUDED IN RECORD Cairo, 111 . 84 w June 21, 1891 36 1908 Charleston, S.C. . . 96 E Aug. 28, 1893 37 1908 Chattanooga, Tenn. 60 W May 12, 1895 29 1908 Chicago, 111. ... 84 NE Feb. 12, 1894 37 1908 Cincinnati, O. . . . 52 NW May 23, 1901 37 1908 Cleveland, O. . . . 73 s Nov. 26, 1895 37 1908 Columbus, O. . . . 66 NW Jan. 20, 1907 29 1907 Davenport, Iowa . . 72 SW Aug. 7, 1872 37 1908 Denver, Col. . . . 68 NW May 1, 1902 36 1908 Des Moines, Iowa. . 64 SW April 1, 1892 29 1908 Detroit, Mich. . . . 76 SW Oct. 26, 1895 ' 37 1908 Duluth, Minn. . . . 78 NE Aug. 16, 1881 37 1908 El Paso, Texas . . . 78 W Mar. 5, 1895 29 1908 Galveston, Texas * . 84 NE Sept. 8, 1900 37 1908 Hatteras, N.C. . . 105 N July 17, 1899 33 1908 Havre, Mont. . . . 76 NW June 9, 1890 26 1908 Helena, Mont. . . . 60 f W 1 w Feb. 6, 1890 Dec. 25, 1890 27 27 1908 1908 Honolulu 55 SE Dec. 31, 1906 33 1908 Huron, S.D. . . . 72 NE Jan. 6, 1903 26 1907 Indianapolis, Ind. 60 W June 25, 1882 37 1908 Jacksonville, Fla. . . 75 SW Feb. 16, 1903 36 1908 Kansas City, Mo. 55 NW July 10, 1902 20 1908 Key West, Fla. . . 88 SW Oct. 19, 1876 37 1908 Lexington, Ky. . . 68 w Sept. 8, 1899 23 1908 Los Angeles, Cal. . . 48 NE Jan. 28, 1882 30 1907 Memphis, Tenn. . . 75 SW Mar. 9, 1901 37 1908 Milwaukee, Wis. . . 60 f SW 1 sw July 24, 1874 Oct. 16, 1880 37 37 1908 1908 Minneapolis, Minn. . 84 NW July 20, 1904 17 1908 New Haven, Conn. . 62 SE Oct. 21, 1904 35 1908 New Orleans, La. . . 60 E Aug. 19, 1888 37 1908 New York, N.Y. . . 80 N Mar. 20, 1899 37 1908 Omaha, Neb. . . . 64 NE July 13, 1905 37 1908 Philadelphia, Pa. . . 75 SE Oct. 23, 1878 37 1908 Phosnix, Ariz. . . . 48 SE July 25, 1903 12 1908 Portland, Me. . . . 60 f SE 1 sw Mar. 21, 1876 Dec. 2, 10, 1878 36 36 1908 1908 Portland, Oregon . . 55 ' S Mar. 25, 1897 37 1908 St. Louis, Missouri . 80 NW May 27, 1896 37 1908 St. Paul, Minn. . . 102 NW Aug. 20, 1904 37 1908 Salt Lake City, Utah 66 NW Nov. 15, 1906 34 1908 San Francisco, Cal. . 64 NE Nov. 30, 1906 37 1908 Savannah, Ga. . . . 76 NW Aug. 21, 1898 37 1908 Springfield, Mo. . . 64 W May 29, 1905 20 1908 Syracuse, N.Y. . . 66 S Mar. 24, 1907 5 1908 Washington, D.C. . 66 SE Sept. 29, 1896 37 1908 Yuma, Arizona . . 54 NW Mar. 17, 1894 30 1908 * Anemometer blew away 120 miles by estimation. 137. Prevailing winds of the world. Charts XI and XII, in addition to the pressure, indicate the prevailing wind direction for the world for 156 METEOROLOGY * January and July. In preparing these charts, that wind direction which predominated during the month in question was the acteristics of one cnarte d- If these charts are carefully compared, it the prevail- will be found that the following generalizations can be the world f ma de. i I* 1 the northern hemisphere the wind blows spirally inward toward areas or belts of low pressure, turning in a counterclockwise direction. From areas or belts of high pressure the wind blows spirally outward, turning in a clockwise direction. The results are the same for the southern hemisphere with the exception that the direction of rotation is the converse. In the centers of the areas or belts of high and low pressure calms occur. 138. Other wind charts. In addition to the two charts described above many other wind charts might be prepared. The prevailing wind direc- other wind tion for all the months in the year for the world as a whole charts. or or anv se p ara te country might be charted. Also, the normal wind velocity or the extremes of velocity for all the months in the year or for the year as a whole might be charted for the whole world or for any given country. C. THE CONVECTIONAL THEORY AND ITS COMPARISON WITH OBSERVED FACTS THE CONVECTIONAL THEOKY 139. General convectional motion. The general convectional mo- tion of the atmosphere can best be illustrated by means of a fluid analogy. In Fig. 73 let NES represent a long tank filled with fluid and x conve tionai circu- heated at the bottom and in ff s fl* nina the center. The layer of fluid next the bottom will be FIG. 73. Diagram illustrating Con- heated by conduction and will expand, vection in a Long Tank of Fluid. . : thus raising the layers of fluid above it and causing a slight bulging of the upper surface. Gravity acting upon this will cause the fluid to flow towards the ends of the tank, thus decreasing the pressure in the center and increasing it at the ends. This increased pressure at the ends will drive in the colder fluid toward the center, forcing the lighter expanded fluid to rise and thus starting the convectional circulation. The atmosphere, as was fully discussed in Chapter II, is heated pri- marily at the bottom, and the amount of heat supplied to the equator is PRESSURE AND CIRCULATION OF THE ATMOSPHERE 157 nearly three times the amount applied at the poles, the exact ratio being 347 to 143. One would thus expect a convectional circulation to take place between the equator and poles, the warm air rising at the equator, flowing poleward on the outside of the atmosphere, descending in the polar regions, and re- lationbe- turning toward the equator along the earth's surface. The continents are warmer than the adjoining oceans in summer * and colder in winter ; one would thus expect a convectional circulation between continents and adjoining oceans. Along^ the seashore the land is warmer than the water during the day and colder other con- than the water at night ; one would thus expect a convec- vectionai tional circulation between the land and the near-by water. circulatlons - For one reason or another a certain locality might be warmer than near- by places ; small local convectional circulations would thus be expected. As a matter of fact, as will be seen in later sections, all of these convec- tional circulations actually exist. The first gives rise to the most im- portant of the general winds of the world ; the second is the cause of the monsoons ; the third is the reason for land and sea breezes ; while certain cloud formations, thunder showers, and dust whirlwinds are evidences of the last. 140. Arrangement of isobaric surfaces in a general convectional circulation. If there were no temperature differences and no winds, the isobaric surfaces would be parallel to ^ -'""" ^^ the hydro- ^nli^ec 1 - 4 " ^" ^^> is sphere and at ti n before i convection the same dis- starte( j. tance above it. A vertical section along the meridian of these iso- baric surfaces would show ^_ ^ a series of lines parallel to the earth's surface. These are represented in Fig. 74 , , ,. ,. FIG. 74. The Meridional Section before Convection by heavy continuous lines. started. Suppose now that the at- mosphere is suddenly heated at the bottom and at the equator more than elsewhere. The air will expand and the isobaric surfaces will be warped upward, the upper surfaces showing an increasing amount of bulge. The vertical section along the meridian of these warped sur- 158 METEOROLOGY faces is represented by the dashed lines in Fig. 74. The excess of air at the equator will now, under the influence of gravity, flow off toward the poles, decreasing the sr . v equatorial pressure and increasing the pressure at the poles. This increased pressure at the \ 1 r\ > / ^ i i { '/ i x 29 in. In, FIG. 75. The Meridional Section after Convection was permanently established. The meridi- onal section after con- vection was permanently poles will force established. , i , , in the colder air along the earth's sur- face and cause the warm equatorial air to rise, thus starting the convectional circulation. Figure 75 shows a vertical section along a meridian of the isobaric surfaces, and the air circulation after the convectional circulation has become perma- nently established. A meridional view of the air circulation is shown more in detail in Fig. 76. One half of the atmosphere is below three miles of elevation, and 30 N. latitude marks the dividing line on the earth's surface, one half of the at- A meridi- onal view of mosphere being between it and the equator and one half between it and the pole. Thus, between 30 N. latitude and the equator the air would be rising, flowing poleward outside of the three- mile limit, dropping down to lower levels between 30 N. latitude and the pole, and returning equatorward below the three-mile limit. These last two diagrams illustrate the convectional circulation between equator and pole which would be expected on a non-rotating earth heated at the equator and colder at the poles. In all three of the above diagrams the vertical scale is much exaggerated as compared with the horizontal. 141. Condition of steady motion. The condition of steady motion can be stated in a single sentence. As long as in the atmosphere differ- FIG. -76. A Meridional View of the Air Circulation. PRESSURE AND CIRCULATION OF THE ATMOSPHERE 159 ences of temperature are maintained, a steady convectional circulation will continue. In the case of the earth, the equatorial portions are constantly maintained at a higher temperature tion of than other regions, and the atmosphere is heated at the stea . d y bottom. Thus a permanent convectional circulation be- tween equator and pole is to be expected. 142. Barometric gradients. The barometric gradient is denned in exactly the same way as the two temperature gradients which have already been described (sections 53 and 88). The vertical _ Definition temperature gradient was denned as the change in tempera- O f the baro- ture with elevation above the earth's surface; the poleward metr . ic temperature gradient was defined as the change in tempera- ture with distance in going from equator to pole. Similarly, the baro- metric or pressure gradient may be defined as the change in barometric pressure with distance. This is ordinarily expressed as a change of so many hundredths of an inch of pressure in 500 miles, or so many milli- meters per latitude degree (69.5 statute miles). It can be HOW found by noting the difference in pressure between two places ex P re ssed. and the distance between them, and then reducing this to the difference per 500 miles or per latitude degree. 143. Relation of wind direction to pressure gradient. The relation of wind direction to the barometric or pressure gradient can be expressed best by means of a diagram. In Fig. 77, let A, B, and C represent respec- tively three isobaric lines, with a pressure of 30.3, 30.0, and 29.7 inches respec- Relation of wind direc- tion to pres- sure gra- tively. Suppose that a mass dients and of air M is located on the jj^" 16 30-inch line. The pressure on the K face of this mass of air is slightly greater than 30 inches, because it lies between the 30-inch line and the 30.3-inch line. In the same way, the pressure of the L face of this mass of air is less than 30 inches because it lies between the 30-inch line and the 29.7-inch line. There is thus an unbalanced pressure on this mass of air tending to push it from the 30-inch line to the 29.7-inch line. This may be generalized by saying that air tends to move along 30.0 29.7 FIG. 77. The Relation of Wind Direc- tion to Pressure Gradients. 160 METEOROLOGY pressure gradients and at right angles to isobaric lines. If this princi- ple is applied to the air in an area of high and low pressure, about as illustrated in Fig. 78, it will be seen that the air should "highs" ^ move directly outward from areas of high pressure and directly inward toward areas of low pressure. The first recognition of the effect of the earth's ro- tation. FIG. 78. Air Motion about " Highs" and " Lows " on a Non-rotating Earth. 144. Effects of the earth's rotation on wind direction and pres- sure. The introduction into meteorology of the idea that the rotation of the earth might influence the direction of moving air masses was slow. The first use of this principle to explain observed wind directions was made by Hadley in 1735. It had been known, for a considerable time previous to this, that the trade winds did not blow directly towards the equator, but had an oblique movement. This had been observed especially in the case of the trade winds from 30 N. latitude to the equator which blow from the northeast. His explanation of the oblique movement was essentially as follows : A mass of a * r Carting from 30 N. latitude directly toward the equator would be passing over regions which would nave a greater tnewinas eas terly velocity of motion due to the earth's rotation, than the places from which it had come. As a result it would lag behind the rotating earth and thus, instead of moving due south, it would deviate to the right and become a northeast wind. takes in Hadley's explanation contains the germ of the truth, but is Hadiey's ex- erroneous in at least two directions. In the first place, it tacitly assumes that the effect of the earth's rotation would be felt only by air masses moving north or south, and not by masses the direc- f PRESSURE AND CIRCULATION OF THE ATMOSPHERE 161 moving in an easterly or westerly direction ; and in the second place, it is a natural corollary of Hadley's explanation that the wind velocity must increase as the equator is approached, for with the motion from the north must be compounded the motion from the east due to lag. Both of these are, however, mistakes, since the earth's rotation influences air moving in any direction, and the velocity does not increase as the equator is approached. The deflective effect of the earth's rotation on moving bodies on its surface was worked out later by various mathematicians, but its com- plete application to the motion of air masses, and the first Fen-el's rational explanation of the general wind system of the globe, work - was begun by Ferrel in 1856 He was a school teacher at Nashville, Tenn., and was a self-taught mathematician of remarkable originality and ability. It is probably not too much to. say that Ferrel's work caused a revolution in the science of meteorology. 145. The law which expresses the effect of the earth's rotation on moving air can be briefly stated as follows : If a mass of air starts to move on the earth's surface, it deviates to the right in the The law northern hemisphere (to the left in the southern hemisphere), which ex- and tends to move in a circle the radius of which depends effTcTof^e upon its velocity and the latitude of the place. The accom- earth's ro- panying table indicates, for several velocities and latitudes, * the radius of this circle. It will be seen that the earth exercises a deviating influence on air moving east or west of exactly the same kind as if the air moved north or south. RADIUS OF CURVATURE (IN MILES) FOR FRICTIONLESS MOTION ON THE EARTH'S SURFACE Latitude 5 10 20 30 40 50 60 70 80 90 77 20 miles an hour . 00 880 442 224 153 119 100 88 82 78 10 miles an hour . . oo 440 221 110 112 76 59 50 44 41 39 38 5 miles an hour . . oo 220 56 38 30 25 22 20 19 19 146. The effect* of this deviation to the right on the air masses which, due to convection, are moving from the equator toward the poles on the outsid^ of the atmosphere must now be considered. Instead of mov- ing directly from the equator poleward, the air masses will be deviated to the right in the northern hemisphere and become more and more a 162 METEOROLOGY west air current encircling the pole in a great whirl. It is a principle of mechanics that whenever a rotating body is not acted upon by out- side forces, the moment of momentum must remain a con- stant. The formula for the moment of momentum is SM VR, where M represents the mass of each particle of the rotating body, V the velocity of the particle, and R the radius, that is, the distance of the particle from the center of rotation. This product M VR must be summed up for all the particle of the rotating body, and thus %MVR represents the moment of momentum of the body. As this ring of whirling air about the pole approaches it, the mass remains constant, the radius is decreasing, anc ^ thus ^ ne velocity must steadily increase, and it has been computed that, if the velocities were not held down by fric- tion, they would amount to hundreds of thousands of miles per hour. A whirl of air with these high velocities must cause cen- trifugal force, and this centrifugal force will hold air away from the pole, thus causing a diminution in the barometric pressure. The amount of land at the north pole is much greater than at the south pole. One would thus expect wind velocities and the diminution in pressure to be larger at the south pole than at the north pole. An exact analogy to this process can be seen in the escape of water from a washbowl through a central vent. Due to some cause, the escaping water usually takes up a motion of rotation. As the water ap- proaches the vent to escape, the velocity of rotation becomes greater and greater, and the centrifugal force developed is often sufficient to hold the water away from the center and cause an empty core above the vent. The effect of the earth's ro- tation on the air masses ap- proaching the poles. Low polar pressures explained. An analogy. 30 in. FIG. 79. The Meridional Section of the Isobaric Surfaces in a Convectional Circulation on a Rotating Earth. The vertical section along a meridian of the isobaric surfaces on a rotating earth should thus have the appearance of Fig. 79. On a non-rotating earth, the pressure at the equator would be low and at the poles high. This was illustrated in Fig. 75. It is the centrifugal force due to the circumpola* whirls whi'-h has held the air away from the poles and turned the high pressure into the low pressure. The larger velociti :it the south pole are responsible for the lower barometric pressure there. The verti- cal section along a meridian of the isobaric surfaces on a rotating earth. PRESSURE AND CIRCULATION OF THE ATMOSPHERE 163 147. The effect of the earth's rotation on the motion of air in connec tion with areas of high and low pressure must also be considered. In Fig. 80 the dashed lines represent the air O fthe * motion on a non-rotating earth (see Fig. 78) ; the full e&T ^ h ' s r - lines represent the air motion on a rotating earth. The air motion deviation to the right in each case is evident. These diagrams apply to the northern hemisphere. In the southern hemisphere the direction of deviation is the opposite. FiG.,80. Air Motion about " Highs " and " Lows " on a Rotating Earth. 148. .Buys Ballot's law. About 1850 Buys Ballot, after a careful study of the air circulation about storms, generalized a law which was of great practical value. The usual statement of the law is Buys Bal- that if one stands with his back to the wind, the pressure on lot>s law - his left hand is lower than on his right. This law derived from observa- tions found its way into text books on meteorology, was used by the captains of vessels, and had a very wide practical application. This is its chief importance, for considered in connection with the preceding sections, it is simply the inevitable result of the rotation of The reason the earth on air moving in towards areas of low pressure. In for lts truth * Fig. 80, if one applies Buys Ballot's law, its truth becomes at once evident. 164 METEOROLOGY COMPARISON OF THE CONSEQUENCES OF THE CONVECTIONAL THEORY WITH THE OBSERVATIONS OF PRESSURE AND WIND 149. In the first two subdivisions of the present chapter, the material was presented from the inductive standpoint (see section 30). The instruments for making various observations were described, The mduc- ^e various observations taken were then stated, and from live method of gaining these observations several generalizations were made. It a^ouiTthe 011 was ^ oun( ^' ^ or instance, that there was a belt of low pressure pressure at the equator, belts of high pressure at 35 N. and 30 S. amfair^ 011 ^itude, and low pressure at the poles, the pressure at the motion. south pole being much lower, than at the north pole. It was furthermore seen that air tended to move spirally outward from areas of high pressure and spirally inward toward areas of low pressure, turning clockwise about areas of high pressure, counterclock- wise about areas of low pressure in the northern hemisphere. These conclusions were simple generalizations from observations and serve as a good example of the inductive method of gaining information. In the third subdivision of this chapter, the deductive method of reasoning was followed. Two general principles were used; namely, The deduc- that the atmosphere was heated at the bottom and more at tive method, ^e equator than elsewhere, and secondly, that the earth turned eastward on its axis. From these two general principles it was determined what the pressure distribution over the earth ought to be and what the air circulation about areas of high and low pressure ought to be. If these deductions from the two principles are tested, by com- paring them with the generalizations derived from the observations, The com- exact agreement will be found. This stamps the whole parison. process of reasoning, as well as the conclusions, as correct. The present illustration of the inductive and deductive method of reasoning is the best one to be found in the realm of meteorology. D. A GENERAL CLASSIFICATION OF THE WINDS ^ THE CLASSIFICATION OF THE WINDS 150. The chief characteristics of a general convectional circulation between equator and pole on a rotating earth have just been stated ; it The general now remams to treat in detail the various components of winds of this circulation. These, together with certain other winds, due in most cases to convection on a smaller scale, are usually spoken of as the general winds of the globe. These were first classified PRESSURE AND CIRCULATION OF THE ATMOSPHERE 165 by Dove, a German meteorologist, nearly one hundred years ago. He divided them into permanent, periodic, and irregular; the per- manent being those whose direction remained the same The Dove - throughout the year, the periodic being those whose direc- classifica- tion changed during a certain interval of time, the irregular being those which showed marked irregular changes in either direction or velocity. In the American Meteorological Journal for March, 1888, Professor W. M. Davis published a classification based upon another principle. This classification is also treated in the seventh The Davis chapter of his book, Elementary Meteorology, Ginn and ciassifica- Company, 1894. In this classification, the winds are divided taon ' into sun-caused, moon-caused, and earth-caused winds. In neither of .these classifications is any account taken of the importance, or intensity, or the amount of the earth's surface covered by these winds. In the following sections these general winds of the globe will be treated roughly in the order of their importance, although the treatment is based largely upon the Davis classification. In each case the position of the wind in both the Dove and Davis classification will be given. PLANETABY WINDS 151. Typical system. By planetary winds are meant those winds which would occur on any planet heated at the equator and turning eastward on its axis. Any migration of the point of appli- Definition cation of the greatest heat and all irregularities of surface of planetary are here left out of consideration. The typical system of wmds * planetary winds can be best illustrated by means of a series of diagrams. In section 146 it was shown that a planet heated at the Thepres- equator and turning eastward on its axis would develop a sure belts, belt of low pressure at the equator, belts of highjpressure in the tropical regions, and - cups of low pressure at the poles^ These deductions are also in exact agreement with the results of the- pressure observations made ih/different parts of the earth. One -would thus expect calms at the equator and in the tropical regions, and winds blowing outward from the belts of high pressure, turning to the right in the northern The air cir- hemisphere and ta the left in the southern. These are cuiation. illustrated in Fig. 81. The equatorial belt of calms is called the dol- drums; the tropical belts of calms are called the horse The names latitudes; the northeast winds blowing from the tropical of the belt of high pressure in the northern hemisphere and,Jbhe southeast winds blowing from the tropical belt of high pressure in the 166 METEOROLOGY HORSE LATITUDES HORSE LATITUDES PREVAILING WESTERLIES PREVAILING WESTERLIES The extent of this sur- face distri- bution. FIG. 81. Surface Distribution of the Planetary Winds; to 3000 ft. southern hemisphere toward the equator in each case are called the trade winds ; the southwest wind, blowing in the northern hemisphere from ' the belt of high pressure poleward, and the northwest wind blowing from the high pres- sure belt in the southern hemisphere poleward are called the prevailing wester- lies. This surface distribu- tion of the winds extends to an altitude of about 3000 feet. At this height, as was seen in connection with the merid- ional section of the isobaric surfaces (Fig. 60) the equatorial belt of low pressure disappears and the isobaric lines become practically straight with a slight droop at the poles. 152. On the outside of the atmosphere, that is, from a height of about 13,000 feet on, the air currents move from the equator toward the pole in both hemispheres, turning to the right in the northern hemisphere so as to become a southwest air current, and to the left in the southern hemisphere so as to become a northwest air current. These air currents approaching the poles in spiral paths give the impression of circumpolar whirls in each hemisphere. These are indicated in Fig. 82. Right at the equator the air moves from east to west, and the reason for this can be readily stated. The air, rising at the equator, in order to start its east air cur- circumpolar journey, goes into re- gions which have a larger eastward velocity of motion, due to the earth's rotation, than the earth's surface itself. As a result, these rising air masses lag behind and give the impression of an east wind. 153. From the tropical belt of high pressure to the pole in the northern hemisphere, both the upper air currents and the surface winds blow from the south- west. If the air is not to beroma massed at the pole, a return current The air cur- rents in the outer layer of the atmos- phere. rent at the equator FIG. 82. The Air Currents in the Outer Layer of the Atmosphere; 13,000 ft. on. PRESSURE AND CIRCULATION OF THE ATMOSPHERE 167 in an intermediate layer is necessary. The direction of the air motion in this intermediate layer from, say, 3000 feet to 13,000 is illustrated in Fig. 83. The air would start from the poles with a slight velocity, and would tend to be deviated toward the right in the The air mo- northern hemisphere and thus be- tioninthe come a northeast wind, but it is im- intermediate layer. prisoned between two layers of air, both of which have a high velocity from some westerly point. As a result, it is carried east- ward and becomes a northwest re- The exp i a _ turn current in spite of the deviation nation of its to the right caused by the earth's d rotation. After passing the tropical belt of high pressure the surface wind becomes a north- east wind and the intermediate layer then also obeys the deviating force of the earth's rotation and becomes a northeast wind. 154. The north-south component of the circulation of the atmosphere can be illustrated by projecting it upon the plane of the meridian. This is shown in Fig. 84. On - FIG. 83. The Air Circu- lation in the Interme- diate Layer; 3000 to 13,000 ft. the outside of the atmos- \ The atmos- pheric cir- phere, that is, from 13,000 cuiationin feet on, the motion is section, northward, while in the in- termediate layer, from 3000 to 13,000 feet, the motion is southward. The surface winds move southward from the tropical belt of high pressure to- ward the equator, and northward from this belt toward the pole. At the equator there are rising air currents and in the tropical belt of high pres- sure descending air currents. The circulation in the southern hemisphere is exactly the same. 155. This general circulation of the atmosphere must be modified in two ways, as will be seen later, to adapt it to the actual earth, and it is also invaded by storms, principally in the region of the The typical prevailing westerlies which cause much confusion and re- system must duce wind velocities to a marked extent. If the earth did not rotate on its axis, the air motion would be a EQUATOR FIG. 84. The North-south Component of the Circulation of the Atmosphere. 168 METEOROLOGY simpler direct north and south motion, as illustrated in Fig. 78, and not an oblique circulation as at present. The effect of the deflec- tive force due to the earth's rotation is to increase some- onthewinds w^ at tne win( i velocities; but by causing the circulation to if the earth be oblique, it also retards the exchange of air between rotating equator and pole and thus causes greater temperature dif- ferences to exist between equator and pole than otherwise would. In the Dove system all these winds would be classified as per- manent, in the Davis classification they would be considered sun-caused. 156. Trade winds. The trade winds blow from the tropical belts of high pressure toward the equator ; as a northeast wind in the northern Description hemisphere and a southeast wind in the southern hemisphere, of the trade They cover nearly half of the earth's surface and their name is derived, not from their importance to commerce, but from the steadiness with which they blow. They blow with moderate to brisk velocity and particularly over the ocean have an unchanged direc- tion for perhaps weeks at a time. They carry but few clouds by day, and by night the sky is usually cloudless. As they approach the equator the velocity steadily increases, and the amount of moisture contained in them also becomes greater. The trade winds were first described by Halley in 1686, and since that time many observations have been made of them. Their thickness is about 13,000 feet. This information is gained by noting the motion of the cirrus and cirro-cumulus of the trades clouds which occur at this level or higher, from kite and and how balloon observations, and also from the observations made found. on mountain tops which exceed this elevation. In some cases the smoke coming from volcanoes has been observed to drift from the southwest in the air currents of the outer layer of the atmosphere, while the lower clouds drifting in the trade winds moved from the northeast. 157. Doldrums. The doldrums are the equatorial belt of calms. The air, laden with moisture and at a high temperature, brought in by Description ^ ne trade winds, here loiters and finally rises to commence of the its poleward journey on the outside of the atmosphere. ms ' This rising air cools, reaches the dew point, and then yields cloud and precipitation. The doldrums are thus characterized by light, baffling breezes, frequent calms', overcast sky and heavy rains, often in the form of thunder storms and squalls. 158. Horse latitudes'. In the tropical belts of high pressure, at 35 N. and 30 S. latitude, are located the horse latitudes. These stand PRESSURE AND CIRCULATION OF THE ATMOSPHERE 169 in marked contrast to the doldrums, although the wind is light and variable and tne velocity small. Calms are frequent. The sky, however, is nearly always clear and the amount of moisture The horse in the atmosphere is small^ This is due chiefly to the fact latitudes - that here we have to do with descending instead of ascending air currents. 159. Prevailing westerly winds. In the northern hemisphere, from the tropical belt of high pressure poleward, the wind blows from the south- west or some westerly quarter. In the southern hemisphere The pre _ the wind direction is from the northwest. Both these vailing west- winds are spoken of as the prevailing westerly winds. In the erly wmds * southern hemisphere the velocity is larger, sometimes amounting to a gale, and for this reason, particularly among sailors, thesQ winds are often spoken of as the " brave west winds," and the region of their occur- rence in the southern hemisphere called the " roaring forties." Sailing vessels, in going from Europe to Australia often make the outward jour- ney around the Cape of Good Hope and the return journey around Cape Horn, thus using the prevailbp westerlies which encircle the south pole. The prevailing westerlies are more invaded by storms than any other of the permanent winds, and in many parts of the United States the succession of storms is so rapid that the prevailing westerly wind only makes itself manifest in the general averages. 1 60. Upper currents. The upper currents move from the equator as a southwest air current in the northern hemisphere and a northwest air current in the southern hemisphere, tionofmo- spirally poleward^ and are often spoken of as the anti-trades. tion and The velocities are high and much greater in winter than in i s ti cs of the summer. The motion of these air currents has been fre- quently determined by the drifting of smoke from lofty vol- canoes, by the cirrus clouds which occur at these levels, by observa- tions on high mountains, and by kite and balloon observations. /v TERRESTRIAL WINDS \^ 161. Definition. The typical system of planetary winds must be modified in two ways. In the first place, the axis of the The planet- earth is inclined 23i to the plane of the ecliptic, which ary system causes a change in the presentation of the northern hemi- mu ^nj e s sphere to the ra\ " of the sun. On account of this the sun modified in migrates 47 in the course of a year, being farthest north on the 21st of June and farthest south on the 21st of December. As will be 170 METEOROLOGY seen later, this migration modifies the wind system. In the second place the earth's surface is not uniform, but is very cHversified, being composed of both land and water and with numerous mountain ranges. Definition ^his diversity of surface also affects the typical system of of terrestrial planetary winds. By terrestrial winds are meant the typical planetary system modified by taking account of the first condition ; namely, that the sun migrates 47 in the course of a year. 162. Annual migration of the winds. This migration of the sun through 47 causes a migration of the heat belt. The heat belt-migrates less than the sun and lags behind it from four to six weeks. The migra- The migration of the equatorial belt of high temperature sun, heat causes the equatorial belt of low pressure, of which it is the belt, equa- cause, to migrate. This migration in turn lags behind the pressure heat belt and migrates over a smaller distance. The mi- belt, and gration of the equatorial belt of low pressure causes the system. permanent wind system of the globe to migrate. The migration of the wind system lags some two months behind the migration of the sun, and on the average covers a distance of The amount 5 or 6. The following table, which gives for the Atlantic of the mi- and Pacific oceans the boundaries of the trade winds and doldrums for March and September, makes clear this migra- tion of the wind system. ATLANTIC OCEAN PACIFIC OCEAN March September March September N.E. Trades 26 N. - 3 N. 35 N. -11N. 25 N. - 5 N. 30 N. - 10 N. Doldrums 3 N. - 11 N. - 3 N. 5 N. - 3 N. 10 N. - 7 N. S.E. Trades 0N.-25$. 3N.-25S. 3N.-28S. 7N.-20S. t 163. Subequatorial and subtropical wind belts. This migration of the wind system over 5 or 6 in the course of a year gives rise to three belts on the earth's surface which require particular equatorial consideration. These are called the subequatorial and the and sub- subtropical belts, and they are illustrated in Fig. 85. In ^ e case ^ *ke nor thern subtropical belt, when ^the horse latitudes are farthest north, most of the belt is covered J the northeast trade winds. When the horse latitudes are fartlu^t south, most of the belt is covered by the prevailing we winds found lies- Thus places in this belt in the course of a year will in each experience calms, southwest winds, and northeast win The corresponding subtropical belt in the southern hemi- sphere will experience calms, southeast trade winds, and northwest PRESSURE AND CIRCULATION OF THE ATMOSPHERE winds. ^Ln the case of the subequatorial belt a further complication arises. When the doldrums are farthest north, the southeast trade winds from the southern hemisphere cross the equa- tor. After crossing the equator, they come under the deflective influence of the earth's rotation and become a southwest wind. Thus, places in the subequa- torial belt north of the equator experience calms, northeast winds, and southwest winds in the course of a year. A place south of the equator in this belt will experience calms, southeast trade winds, and northwest winds. 171 PREVAILING I I WESTERLIES "*T*1 CALMS s.w. WINDS N.E. WINDS' CALMS .E. WINDS WINDS CALMS N.W. WINDS S.E. WINDS TRADE WINDS PREVAILING WESTERLIES FIG. 85. The Subequatorial and Subtropical Wind Belts. CONTINENTAL WINDS 164. Definition. The typical system of planetary winds which must, exist on every planet heated at the equator and turning eastward on its axis must be modified in two directions in order to fit Tne two it more exactly to our earth as it actually exists. 'The first modifica- one of these two modifications has already been discussed. f^caiV^- The' sun migrates 47 in the course of a year and, as a tem of plan- result of this migration, the equatorial jylt ofjrigh tem- etary winds * perature, the equatorial belt of Iowj3e.ssure, and the wind system of the world migrate north and south, lagging behind the sun and each migrating over a less distance. This migration of the wind system gives rise to three belts known as the subequatorial and the two subtropical belts, where calms and winds blowing of terrestrial from two opposite directions in the course of a year exist, and The planetary winds thus modified are called the terrestrial winds. The second modification is due to tne fact that the earth's surface is not uniform. It is not an all land surface, or an 172 METEOROLOGY all water surface, but is highly diversified, consisting of par^mnd and part water. In addition the surface of the ocean has not the same temperature along a parallel of latitude due to the presence of ocean currents. The terrestrial winds modified by taking account of this diversity of surface are called continental winds. In the chapter on temperature, the fact was emphasized that the continents during the winter are colder than the adjoining oceans under the same latitude, while in summer they are warmer than the of diversi- adj acent oceans. In view of the convectional theory, one ties in the would expect the continents to become areas of low pressure during the summer and of high pressure during the winter. This tendency must strongly modify and distort the belts of pressure running round the world. The oceans, due to the presence of hot or cold ocean currents, have not the same temperature along a parallel of latitude. As a result one would expect highs to build over the colder parts of an ocean while lows would form over the warmer parts. As a result of this the belts of pressure would again be modified or broken up. The general tendency, then, of diversities in the surface would be to form peaks and depressions and to break up continuous belts. In section 119, when the isobars for January and July were being con- sidered, it was found that there were eight areas of high pressure and six areas of low pressure to be considered. Six of these highs and three of these lows either persist throughout the whole year or are so prominent during a large portion of it that they make themselves felt in the annual averages and appear on the chart of isobars for the year. These peaks and depressions are often spoken of as the " permanent highs and lows " or as " the centers of action " because they exert such a marked influence, as will be seen later, on the wandering areas of high and low pressure which constitute our storms and determine the weather to such a marked extent. This tendency of diversities in the surface to break up the belts of pressure modifies to such an extent the typical planetary wind system that it is sometimes claimed that it does not really exist on the poleward side of the tropics, but is replaced by an air circulation about these highs and lows. The eight highs may be divided into two groups. There are five which A stud of are l ca t e d over the ocean and in the region of the high the eight pressure belts at 35 N. and 30 S. latitude. They are also nearly permanent in position and persist throughout the year. They are also located on the eastern side of the ocean. They are located in the North Pacific opposite California, in the North Atlanti^^ PRESSURE AND CIRCULATION OF THE ATMOSPHERE 173 of Spain, in the South Pacific, in the South Atlantic, and in the Indian Ocean. In each case, they are located near the place where a cold ocean current flowing towards the equator crosses the belts of high pressure and to this is without doubt due the breaking up of the belt and the building of the peak of pressure. These facts have just (1911) been brought out by Humphreys in a valuable article in the Mount Weather Bulletin and for further details the reader must be referred to this article. The other three highs are located over land surfaces, namely, over Siberia, over North America, and over South Africa. In each case they appear during the winter and disappear during the summer. The Siberian high builds to such an immense height during the winter that its influence is felt in the yearly averages and it also appears on the chart showing the isobars for the year. These are without doubt due to the low temperatures to which the continents fall during the winter. During the summer, when the land surface heats up, they disappear. The six lows may also be divided into two classes. Four of them are located in the equatorial belt of low pressure. They are in South Africa, South America, Australia, and India. In each case they appear during the summer and disappear during the winter. They are due without doubt to the excessive heating of the land surface during the summer. The other two lows are located near Iceland and just south of Alaska, in each case over the ocean. The Iceland low persists throughout the year, while the Aleutian Islands low disappears during the summer. These are the two depressions into which the polar cap of low pressure breaks up. These are without doubt due, as Humphreys has also pointed out, to the Gulf Stream and the Japan current. The warm water brought to the far north by the Gulf Stream stands in sharp temperature contrast to the perpetual ice and snow of Greenland and Iceland. As a result a low persists over the ocean throughout the year. In the case of the ocean near the Aleutian Islands the same thing is true during the winter. During the summer, both Alaska and the coast of Asia heat markedly. As a result the temperature contrast and also the low disappear. Charts XIII and XIV represent the air circulation of the Atlantic Ocean for January-February, and July- August. The vari- Thg ^ dr _ ous components of the planetary wind system are clearly dilation of evident. The trade winds, the prevailing westerlies, the doldrums, and the horse latitudes are all easily recognized. The; ted, however, by lines that run parallel to latitude 174 METEOROLOGY lines ; but a distinct tendency is shown to blow in a spirafflirectio around these areas of high and low pressure. 165. Monsoons. This tendency of the continents to become the centers of high pressure during the winter with spirally outflowing wincls, Definition of and the centers of areas of low pressure during the summer a monsoon. w ^h spirally inflowing winds, is not usually sufficient to cause a complete reversal of the wind direction between winter and summer. The temperature differences and the resulting pressure differences are too slight to bring about this result. The usual result is an increase or decrease in wind velocity, and a slight change in wind direction between summer and winter. There are several instances, however, when a complete reversal is brought about, and these (winds which change direction with the season^ are spoken of as monsoons. The word is of Arabic origin and means " season." Monsoons are particularly notice- Where mon- able in connection with those places which are located in the soons occur, subtropical or subequatorial wind belts. The monsoonlike changes in the wind direction brought about by the migration of the wind system when aided by continental influences readily develop into pronounced monsoons. The most typical monsoon in the world occurs in India. During the winter an immense peak of high pressure forms over southern Siberia, with a central pressure of 30.50 inches. The wind blowing and de^ 86 spirally outward from this area of high pressure causes north- scription of east winds over India. The air coming over southern Siberia monsoon an d across the Himalaya Mountains is dry, and this is spoken p of as the dry or winter monsoon. During the summer the equatorial belt of low pressure migrates to the north of India, and the wind blowing spirally inward towards this belt of low pressure causes southwest winds to blow over India. The air coming from the ocean is moisture-laden, and when forced to rise in this area of low pressure and also by the Himalaya Mountains, it causes drenching rains over India. This is called the southwest or wet monsoon. The coming of the south- west monsoon has been vividly described by many authors. For several weeks previous, the air has been nearly calm and the breezes have been light. Then the wind begins to blow from the southwest, at first fit- fully, but gradually attaining greater velocity and steadiness. The dark clouds on the horizon and the breaking of the waves on the shore announce the coming of the steady southwest wind. Soon, with light- ning flashes and thunder, the rain, which descends in torrents, perhaps lasts for two or three weeks and marks the beginning of the monsoon. The PRESSURE AND CIRCULATION OF THE ATMOSPHERE 175 FIG. 86. Pressure and Wind Distribution over India during January. (From SALISBUKY'S Physiography.) clouds then break away, and wind blows steadily and freshly from the southwest, and rains are frequent during the whole of the monsoon period. Figures 86 and 87 indicate the pressure and wind distribution over India during January and July. Monsoon winds are also noticeable in the peninsula of Spain and Portugal. Figures 88 and 89 illustrate the pressure and wind direction during January and July. It will be seen that the penin- The mon _ sula is the seat of low pressure with inflowing winds during soons of the summer and the seat of high pressure with outflowing winds during the winter. Australia also v exhibits a well- and of marked monsoon, as illustrated by Fig. 90. Other portions of the world also exhibit monsoon tendencies, but not to such marked extent as those countries just mentioned. 29.80 FIG. 87. Pressure and Wind Distribution over India during July. (Fiom SALISBURY'S Physiography.) 176 METEOROLOGY 1 66. Other land effects. The various results due to the fact that the earth is not a water surface throughout, but has a diversified surface The seven made up of part land and water, may be briefly brought land effects, together and summarized here : (1) The belts of pressure are broken up into peaks and depressions which are called the permanent highs and lows. (2) The continents are the cause of the monsoons which FIG. 88. Pressure and Wind Direction in Spain and Portugal during January. FIG. 89. Pressure and Wind Direction in Spain and Portugal during July. have just been described. (3) The diurnal change in wind velocity and wind direction is brought about by the presence of land. During the day the * SUMMER WINTER FIG. 90. Wind Direction in Australia during January and July. wind velocities are greater than during the night. The wind direction also changes slightly during the day and shifts back again during the night. The cause of this is convection, which causes the upper air during the day to come down to the earth's surface, bringing with it its higher velocity and changed direction. Over the ocean the diurnal change in either wind velocity or direction is barely noticeable. (4) The conti- N PRESSURE AND CIRCULATION OF THE ATMOSPHERE 177 nents also cause an annual change in wind direction. This is brought about by the fact that they become centers of high pressure during the winter and low pressure during the summer. (5) Mountain chains cause changes in wind direction and velocity. This is not only true on a small scale in connection with local barriers, but the longer -a'hd higher mountain chains of the world influence the general winds of the world both as regards direction and velocity. (6) Wind velocity is much reduced by friction when the air is passing over land. This is proven by the fact that the normal wind velocity over the ocean is nearly twice what it is over the land. (7) The Arctic winds, if they exist, are without doubt the result of a land surface. Only few observations have been made in the polar regions, but they would seem to show that the barometric pressure does not decrease steadily as the pole is approached, but that it reaches a minimum in about 70 or 80 north latitude, and then in- creases slightly toward the pole. In the polar regions, winds blowing from the north instead of the southwest have also been observed. These would be the inevitable result of an increase of pressure toward the pole. If these observations are trustworthy, the explanation is without doubt to be found in the increasing percentage of land near the north pole. This would decrease the wind velocity so that the highest values of wind velocity and thus the least pressure would be found, not at the north pole itself, but perhaps 20 or 30 from it. The observations, however, are not sufficiently numerous or trustworthy to make an elaborate explanation or treatment of the subject necessary. / LAND AND SEA BREEZES 167. At the seashore the land is heated more than the adjacent water during the day and at night by rapid cooling becomes colder than the adjacent water. This gives rise to a small convectional circulation known as land and sea breezes. The sea breeze tion of land ordinarily begins about ten or eleven o'clock in the morning JJ 1 * "* and blows gently but with increasing velocity from the ocean toward the land. It does not usually penetrate more than ten or fifteen miles inland, and it reaches its maximum velocity ordinarily at two or three in the afternoon. At about sunset the sea breeze dies out, and it is replaced during the night by the land breeze, which blows from the land toward the ocean. The velocity of the sea breeze and its regularity are usually more marked than in the case of the land breeze. The lars^ 1 J sea breezes, as has been fully shown hv nhsprvations made 178 METEOROLOGY by kites and captive balloons, do not ordinarily extend to a greater height than, say, a thousand feet. The phenomenon, therefore, of land and sea breezes is confined to a space of fifteen miles either side of the coast line, and to the lower thousand feet of atmosphere. The explanation of the land and see breezes may be thus stated : During the early hours of the morning the land heats rapidly under sunshine and becomes warmer than the adjacent water. nation of The air expands, flows off aloft over the ocean, thus increas- land and sea m g fa e pressure slightly over the ocean, and decreasing it over the land. This increased pressure over the ocean drives in the air in the form of the sea breeze. At night the land and the layer of air next to it cool rapidly, and the converse process takes place, the air being forced to move by the increased pressure from the land toward the ocean. It has been often observed that, when the sea breeze first makes its appearance, it starts some distance from the land, as is shown by the rippled surface of the water, and then slowly beats its way in toward the shore. This can be readily explained. As the air over the land becomes heated during the early morn- ing hours, it tends to expand laterally as well as to flow off above over the ocean. This lateral expansion would tend to hinder the coming of the sea breeze near the shore. Furthermore it is the increased pressure over the ocean which drives in the sea breeze, and, as this is first felt at a considerable distance from the shore, it is there that one ought to expect the first beginnings of the sea breeze. In many places the sea breeze is not felt as a breeze blowing directly from the ocean during the day, and the land breeze at night does not blow di- rectly from the land, but they are com- pounded with the pre- vailing wind and thus merely cause changes in the direction or velocity of the pre- vailing wind. On the shores of Long Island the prevailing wind direc- tion is west, and as is shown in Fig. 91, the sea breeze changes this to a southwest wind by day and the land breeze is combined with it and Why the sea breeze starts some dis- tance from the land. Land and sea breezes combine with the pre- vailing wind. PREVAILING WEST WIND AT NIGHT FIG. 91. The Effect of Land and Sea Breezes on the Pre- vailing West Wind on the Shores of Long Island. PRESSURE AND CIRCULATION OF THE ATMOSPHERE 179 becomes a northwest wind during the night. On the coast of Cali- fornia and Chile, where the prevailing wind direction is west, that is, from the ocean toward the land, the sea breeze during the day causes a marked increase in the wind velocity, while the land breeze during the night reduces it to practically a calm. This is said to be partic- ularly pronounced in the portions of Chile near Valparaiso. Here the wind velocity during the day may amount almost to a decided gale, hindering walking, making the transaction of business unpleasant, and almost cutting off intercourse between vessels in the harbor and the shore. At night an almost complete calm reigns. MOUNTAIN AND VALLEY BREEZES 1 68. The so-called mountain and valley breezes are well known in all mountainous countries, but they are particularly noticeable in long, narrow valleys which emerge on a large open plain below* . . . They are periodic, sun-caused winds which are particularly ofthemoun- well developed on still clear days. During; the night, from tain and val ~ 1 . .ley breeze, a few hours after sunset until early morning, the mountain breeze flows down from the mountains through the valley to the plain below. It is not only easily detected, but at times it attains a moderate velocity; and those who camp at night in the open in mountainous countries soon learn to build camp fires on the downhill side, so that the mountain breeze may blow the smoke away from their tents. During the day the valley breeze is felt as a gentle breeze blowing up through the valley and up the mountain slopes. It is usually hardly noticeable and is never as well developed as the mountain breeze. During the night, the layer of air next the ground becomes cooled by radiating its heat to the colder ground and to the sky, and also by con- duction to the cold ground. This layer of cold, and thus The cause dense, air drains, just as water would, into /the valleys and ofthemoun- places of less elevation than surrounding regions. As a * result, depressions are filled at night with pockets of colder air which have drained into them from the surrounding slopes. In a long, narrow valley this drainage of colder air makes itself felt as a mountain breeze: If a glacier is located in the valley, the layer of air next it may be cooled sufficiently to cause a mountain breeze even during the daytime. . As the mountain breeze moves down the valley the air is compressed and thus is heated to the extent of 1.6 for every 300 feet of descent. If the motion down the valley is slow so that there is sufficient time for the air 180 METEOROLOGY to radiate its heat to the ground and sky, or to lose it by conduction, it may reach the open plain below as cold or even colder than the air over the plain. If, however, the descent has been rapid, the air will probably reach the plain with a higher temperature than the air over the plain, and to this fact may be due the often observed result that frosts are less severe on a plain opposite the point of entering of a long valley. During the day the air at the bottom of a valley is heated more than the air along the mountain slopes, particularly as one side of a valley The cause is nearly always in x ^ - ____________ > _-^* of the vai- shade. As a result of \-~-"""' _ ^ --. ../ eeze ' this heating, the iso- \* / baric surfaces are warped upward. . ~* In Fig. 92, which shows a cross section of a valley, the normal ^_ _^ isobaric surfaces are represented by f u jj lines, while the upward bulging of these surfaces is indi- cated by dotted lines. As a re FIG. 92. Cross Section of a Valley, showing I A .* .J. 'TIT- r^.u the Isobaric Surfaces. suit of this upward bulging of the isobaric surfaces, the air flows off toward the side; and, when it meets the sloping sides of the valley, it is deflected upward, thus causing valley breezes which flow up the mountain slopes during the day. ECLIPSE, LANDSLIDE, TIDAL, AND VOLCANIC WINDS 169. Eclipse winds. When, during an eclipse of the sun, the shadow of the moon falls upon the earth, the sun's radiant energy is withheld from a belt which represents the path of the eclipse. As a acteristks result of this, the air ought to be colder here than in the and cause of surrounding regions, and thus a belt of high pressure should wfnds? be developed along the path of an eclipse of the sun.. The air would gently descend in this belt of high pressure and flow outward in both directions from it along the earth's surface. The few observations which have been made in connection with eclipses would seem to show that these conditions are actually realized. But' few observations have, however, been made; and the winds due to an eclipse are of such seldom occurrence, of such little importance, and of such slight intensity that they are hardly worth considering. They would be classified as irregular sun-caused winds. PRESSURE AND CIRCULATION OF THE ATMOSPHERE 181 170. Landslide and avalanche winds. Masses of air are often pushed ahead of landslides and avalanches in sufficient quantity to cause destruction even at considerable distances. The rush of air ahead of an avalanche is sometimes felt at a distance and ava- of a mile, and destruction has been caused at a distance of a la * che quarter of a mile. These winds are ordinarily classified as irregular earth-caused winds. In the case of an avalanche wind, how- ever, it is indirectly of solar origin, for it is the energy of the sun which has evaporated the water and caused the circulation of the air which has deposited the snow on the mountain side to build the avalanche. 171. Tidal winds. In some gulfs and bays, where the rise and fall of the tide is considerable, a slight motion of the air away from the gulf or bay when the tide is coming in and toward the gulf or bay when the tide is going out has been observed. This teristics and would seem to be caused by the actual raising of the atmos- c *^ e . f phere by the incoming water as the tide comes in and by the falling of the atmosphere to take the place of the out-going water as it recedes. These tidal winds, if they exist, would be compounded with the land and sea breeze in such a way that it would require careful observations at many stations surrounding the gulf or bay to be sure of their presence. Such breezes, if they exist, would be classified as periodic moon-caused winds. 172. Volcanic winds. It has been observed in connection with cer- tain volcanoes while in eruption that there seems to be a rising air column above the volcano accompanied by inflowing breezes from Volcanic all directions. There are two causes for the rising of air over winds - the volcano. In the first place it may be caused by the explosive action of the eruption; and secondly, the heating of the air by the hot lava and other material ejected from the volcano may cause convectional motion. A volcano in violent eruption is sometimes capped by a thundershower with violent thunder and lightning. These volcanic winds are of telluric origin and are irregular as to time of occurrence. CYCLONIC STORMS 173. The general winds of the earth have now been considered in detail, and, as a result of what has been said, it ought to be possible to state for any portion of the earth exactly what changes in wind direc- tion and wind velocity ought to take place in the course of a day or a year. In the northeastern portion of the United Stafes, for example, 182 METEOROLOGY the prevailing wind direction should be from some westerly quarter, since this portion of the country is located in the region of the prevail- ing westerlies. The wind ought to blow more from the sou thwest in summer and from the northwest in winter, regards di- when the extensive area of high pressure builds over North veiociy^f d America due to the low temperatures. The wind velocity the winds in ought to be higher in winter than in summer, reaching its easterruSart max i mum m February or March and its minimum in August of the or September. The velocity ought to be somewhat greater States 1 during the daytime than at night, and there should also be a slightly diurnal change in wind direction. If a place is located near the seashore or in a mountainous region, the land and sea breezes and the mountain and valley breezes ought to make them- selves felt. If careful observations of the changes in wind direction or in wind velocity are, however, made for any period of time, it will be at once noticed that the irregularities both as regards There are _. . _ . . . numerous directiojAnd velocity are numerous. For example, during irregu- t /ne wi^H^Bheii the wind ought to be blowing from the northw^^^^h moderate velocity, the wind will perhaps be blowing from ^he^HBFwith high velocity accompanied by driving snow. This fcay continue during a whole night and, instead of increasing, die day comes. All of these irregular and unlooked-for wind directions and wind velocities are due to the presence of ies are so-called storms, of which there are four different kinds : due to extratropical cyclones or the lows of our weather map ; storms, of . . which there tropical cyclones, which are sometimes called hurricanes in arejour tne West i nc ii es or typhoons in the China Sea; thunder showers; tornadoes. These four storms have been ar- ranged in order of size, for the extratropical cyclones sometimes cover an area a thousand or more miles in diameter, while the tornadoes, although more violent than any of the others, cover but an extremely small area. These are usually spoken of as cyclonic storms, because the air in each case is usually in motion in a spiral direction. These will be fully considered in one of the following chapters. WINDS OF OTHER PLANETS 174. A careful consideration of those changes in the general winds of our own earth which would result from a change in the conditions upon which they depend will serve as an introduction to the consideration of the wind system which ought to exist on the various planets. PRESSURE AND CIRCULATION OF THE ATMOSPHERE 183 If the earth turned more rapidly on its axis, the deflective force would be greater, and the circulation between equator and pole would be more oblique than at present. As an example, the trade The effect winds, in the northern hemisphere, instead of blowing from on the wind the northeast would blow from a still more easterly point. systemo ? a *j * more rcipid The effect of having a more oblique circulation would be to rotation of cause higher wind velocity, to retard the exchange of air t between equator and pole, and thus to accentuate the differences in temperature which already exist. If the earth's axis were inclined to the plane of its orbit at a greater angle than 23.5, the migration of the wind belts would be more pronounced, the subtropical and subequatorial belts would be wider, and the tendency to monsoons would be system of a felt over a larger portion of the earth's surface. cUnation" If the earth's surface were entirely a water surface instead of being a diversified land and water surface, land and sea breezes, and mountain and valley breezes would be absent^tJ^belts of pressure would be uniform around the world^B^Hpons on n e e w j nd would be unknown ; the daily change in wincHj^fcction system of a and wind velocity would no longer exist, and theyearly surface! change in wind velocity would be very much less. The factors, then, upon which the characteristics of the wirfft s depend are the rotation of the earth on its axis, the location belt of highest temperature, the migration of this belt of highest temperature during the year, and the character- istics and diversity of the surface. If these facts were known termine the in detail for any planet, it ought to be possible to determine ^ tem> by deduction the characteristics of the wind system which should be present. 175. Mercury and probably Venus turn the same face toward the sun always, and thus- rotate on their axes very slowly, Mercury making a complete rotation in eighty-eight days and Venus in two hundred and twenty-five days. As a result the center of the illuminated hemisphere ought to become an systems of area of low pressure with air blowing spirally inward along the surface of the planet, rising in this area of low pressure, flowing outward on the outside of the atmosphere, and descending again in an extensive area of high pressure which would form on the dark hemisphere of the planet. In the case of Mars, since the surface features are but seldom obscured by haze or cloud, no observations are available 184 METEOROLOGY for determining the characteristics of the air circulation on that planet. Since it turns on its axis in about the same time as the earth, and has an axis inclined at about the same angle to the plane of its orbit, it ought to have a system very similar to ours. In the case of Jupiter, Saturn, Uranus, and Neptune, the planets are probably continually cloud-covered, and, particularly in the case of Jupiter and Saturn, there are belts parallel to the equator which are easily observable. These are probably due to the atmospheric circulation, but since no facts are known concerning the surface of these planets, it is impossible to form any definite deductions as to what wind system ought to exist. QUESTIONS (1) Why is pressure an important meteorological element? (2) Why does the atmosphere exert a pressure? (3) Is the atmospheric pressure constant? (4) How is the pressure of the atmosphere measured and expressed? (5) Give a detailed history of the invention of the mercurial barometer. (6) Describe a mercurial barometer as used at present. (7) Describe in detail the cistern of a mercurial barometer. (8) What are the two steps in reading a mercurial barometer. (9) Name and treat fully the three corrections to be applied to the reading of a mercurial barometer. (10) Describe an aneroid barometer. (11) What corrections must be made to the reading of an aneroid barometer? (12) Compare the accuracy of an aneroid and mercurial barometer. (13) What is a barograph? (14) Describe the Richard Freres barograph. (15) Describe the construction and action of the so-called " mouth barometer." (16) Describe the construction and action of the " chemical weather glass." (17) What causes the changes in its appearance? (18) What observations of pressure are taken at the various Weather Bureau stations? (19) What instruments are used and where are they located ? (20) How are the pressure normals computed ? (21) Describe the diurnal variation in pressure. (22) With what does the diurnal variation in pressure changer (23) Treat fully the various explanations of this diurnal variation. (24) Describe the annual variation in pressure and state its cause. (25) What other barometric data may be computed from the observations of pressure? (26) Describe the old method of reducing pressure observations in order to compare them. (27) Describe the modern way of reducing pressure observations to sea level. (28) What observations are neces- sary to determine altitude by means of the barometer ? (29) How are these observations secured in practice ? (30) Describe in detail the four methods of computing the difference in elevation from the observations made^ (31) How are isobaric charts constructed ? (32) Describe the chief characteristics of the isobars for the year. (33) How is the vertical section along a meridian con- structed? (34) Describe the course of the thirty-inch line. (35) Why do differences in pressure exist over the earth's surface? (36) What is meant by an isobaric surface? (37) What is the normal form of an isobaric surface? (38) What is the form of isobaric surfaces over areas of high and low pressure? (39) Describe the chief characteristics of the isobars for January and July. (40) What other pressure charts might be constructed? (41) Define wind and state what three things can be measured. (42) Define windward, leeward, veering, and backing. (43) To how many points of the compass is the wind PRESSURE AND CIRCULATION OF THE ATMOSPHERE 185 direction determined? (44) What is the relation between wind velocity and the pressure of the wind ? (45) Describe the wind vane of the Weather Bureau form. (46) What are the advantages of this form of wind vane? (47) What is the anemoscope? (48) Describe the rotating cylinder form of anemoscope. (49) Describe the electrical contact form of anemoscope. (50) What is meant by a wind scale? (51) Describe the ten-point and the Beaufort wind scale. (52) State the three groups of anemometers. (53) Describe a simple deflection anemometer. (54) Describe two pressure anemometers. (55) Describe the Robinson cup anemometer. (56) How is this instrument tested and its constant determined? (57) Describe other forms of anemometers. (58) How can small wind velocities be determined? (59) What are the chief effects of surroundings on wind direction and velocity ? ^ (60) Describe the influence of a valley. (61) Describe the effect of buildings, (62) Describe the effect of the nature of the surface. (63) What is the effect of altitude? (64) What are the advantages of mountain observatories? (65) What observations of wind are taken at regular Weather Bureau stations? (66) State the instruments used and their location. (67) Define prevailing wind direction. (68) How may prevailing wind direction be expressed? (69) Describe the construction of a wind rose. (70) How are normal wind velocities computed? (71) De- scribe the daily variation in wind velocity and state its cause. (72) Describe the daily variation in wind direction and state its cause. (73) How may the diurnal variations in wind direction and velocity be expressed graphically? (74) Describe the annual variation in wind velocity and state its cause^ (75) Describe the annual variation in wind direction and state its cause. (76) What is the cause of irregular variations in wind direction and velocity? (77) What other wind data may be determined from the observations? (78) State the different characteristics of the prevailing winds of the world. (79) What other wind charts may be prepared? (80) Describe the general convectional motion in a long tank heated at the bottom. (81) Why may a convectional circulation between equator and pole be expected? (82) What other convectional circu- lations would be expected? (83) Illustrate by means of two diagrams the ar- rangement of the isobaric surfaces in the general convectional circulation just before its beginning and after it has become permanently established. (84) What is the meridional view of the air circulation ? (85) State the condition of steady motion. (86) What is meant by a barometric gradient? (87) How is it expressed? (88) What is the relation of wind direction to pressure gradient and isobaric lines? (89) What is the reason for the relation? (90) Describe the air motion about highs and lows. (91) State the historical rise of the recog- nition of the effect of the earth's rotation on air motion. (92) What was Hal- ley's explanation of the direction of the trade winds? (93) Describe Ferrel's work. (94) State the effect of the earth's rotation on the. air motion. (95) State the effect of the earth's rotation on the air masses moving poleward on the outside of the atmosphere. (96) Explain in full the cause of the polar caps of low pressure. (97) Illustrate by means of two diagrams the meridional section of pressure on a rotating and on a non-rotating earth. (98) Illustrate by means of two diagrams the air motion about highs and lows on a rotating and non-rotat- ing earth. (99) What is Buys Ballot's law? (100) What is the reason for its truth? (101) Contrast the inductive and the deductive method of gaining information as illustrated in this chapter. (102) What is meant by the general winds of the globe ? (103) What is the basis of the Dove classification? (104) What is the basis of the Davis classification? (105) Define planetary winds. (106) Illustrate by means of a diagram the air circulation near the earth's sur- face. (107) Illustrate by means of a diagram the air circulation in the upper 186 METEOROLOGY layer of the atmosphere. (108) Illustrate by means of a diagram the air motion in the intermediate layer. (109) Illustrate by means of a diagram the atmos- pheric circulation in cross section. (110) What are the effects of the earth's rotation? (Ill) Describe the trade winds. (112) Describe the doldrums. (113) Describe the horse latitudes. (114) Describe the prevailing westerlies. (115) Describe the upper currents. (116) In what two ways must the planet- ary system of winds be modified? (117) Define terrestial winds. (118) What causes the yearly migration of the wind system? (119) What are the charac- teristics and amount of this migration? (120) Describe the subequatorial and subtropical wind belts. (121) Illustrate by means of a diagram the kind of winds found in each belt. (122) Define continental winds. (123) What is the effect of a diversified surface on the belts of pressure? (124) What effect does the building of peaks of pressure and depressions have on the air circulation? (125) What is a monsoon? (126) Describe the monsoons of India. (127) State in full the cause. (128) At what other place on the earth's surface are monsoons felt? (129) State the various effects of land on wind direction and wind velocity. (130) Treat fully the so-called Arctic winds. (131) Describe the land and sea breezes. (132) What is the explanation of land and sea breeze? (133) Why does the sea breeze start over the ocean. (134) In what way is land and sea breeze combined with other winds? (135) Describe the mountain and valley breezes. (136) Explain in detail the valley breeze during the day. (137) Explain in detail the mountain breeze during the night. (138) Describe in detail the air motion caused by eclipses, landslides, tides, and volcanoes. (139) How are these various winds classified in each system of classification? (140) What is the cause of irregularities in the general winds of the world. (141) Name the four storms. (142) What would be the effect on the wind system if the earth turned more rapidly on its axis? (143) What would be the effect on the wind system if the earth's axis were inclined at a greater angle? (144) Name the factors upon which the characteristics of the wind system of a planet depend. (145) Describe the probable wind systems of the various planets. TOPICS FOR INVESTIGATION (1) The history of the mercurial barometer. (2) Modified forms of mercurial barometers. (3) The various forms of barographs. (4) The construction and action of chemical weather glasses. (5) The diurnal variation in barometric pressure. (6) The barometric determination of altitude. (7) Wind scales. (8) Anemometers. (9) Mountain observatories and their work. (10) Ferrel' s contributions to meteorology. (11) The monsoons of India. PRACTICAL EXERCISES (1) Test carefully an aneroid barometer. This might include its accuracy for a given range of pressures, the effects of temperature changes, and the effects of jars due to transportation. (2) Study a chemical weather glass. PRESSURE AND CIRCULATION OF THE ATMOSPHERE 187 (3) Learn to read a mercury barometer and to reduce the observations by applying all the corrections. (4) Plot the graphs showing the annual variation in pressure at several sta- tions and explain the peculiarities of each graph. (5) Determine a difference in elevation by means of a barometer. (6) Work up some or all of the barometric or pressure data mentioned in section 110 for several stations. Those stations may be chosen in which the student has a particular interest. A barograph record of considerable length is almost essential. Extremes of pressure and the frequency and magnitude of the irregular variations would be the most interesting. The attempt should be made to draw general conclusions by contrasting different places, summer and winter, etc. (7) Determine the wind velocity on the stillest nights of winter. (8) Construct several wind roses for different places for the same month or year and for the same place for different months. (9) Work up some or all of the wind data mentioned in section 136. (10) Perform the experiment illustrated in Fig. 70. (11) Investigate some valley for mountain and valley breezes. REFERENCES For the description, illustration, construction, and use of apparatus to measure pressure and wind, see : ABBE, Meteorological Apparatus and Methods (Washington, 1888) ; pp. Ill to 179 and 183 to 336 treat of pressure measuring and wind measuring instru- ments respectively. MOORE, JOHN W., Meteorology, 2d ed., pp. 120 to 138 and 265 to 292. PLYMPTON, The Aneroid Barometer (D. Van Nostrand Company, New York). WALDO, Modern Meteorology, pp. 59 to 135. WILD, H., Ueber die Bestimmung des Luftdruckes, Rep. fur Met. Ill, N. 1, pp. 1-145. Report on the Barometry of the United States, Canada, and the West Indies, Report of the Chief of the Weather Bureau, 1900-1901, Vol. II. MARVIN, C. F., Circulars A, D, F, Instrument Division of the U. S. Weather Bureau. Instructions for Cooperative Observers (U. S. Weather Bureau, Washington). Apparatus catalogues of various firms. See p. 110. (See also the various guides to observers mentioned in Appendix IX, in group 2(B).) For observations of pressure and wind consult the publications mentioned on p. no. For normal values of pressure and wind, see : BUCHAN, ALEXANDER, Report on Atmospheric Circulation. HANN, Klimatologie. Report of the Chief of the Weather Bureau particularly for 1891-1892, 1896- 1897. Climatology of the United States (Bulletin Q of U. S. Weather Bureau by A. J. HENRY, 1906). Summary of the Climatological Data for the United States, by Sections (106 are to be published). For charts of pressure and wind, see : BARTHOLOMEW, Physical Atlas, Vol. III. 188 METEOROLOGY BUCHAN, ALEXANDER, Report on Atmospheric Circulation (Report on the Scien- tific Results of the Voyage of H.M.S. Challenger). HANN, Atlas der Meteorologie, 1887. HILDEBRANDSSON, H. H., ET TsissERENC OB BORT, Les bases de la meteor- ologie dynamique, Paris, 1900-1907. Segelhandbuch der Deutschen Seewarte ; Hierzu ein Atlas (for the Atlantic, Pacific, and Indian Oceans). Summary of International Meteorological Observations (Bulletin A of the U. S. Weather Bureau). ELIOT, SIR JOHN, Climatological Atlas of India, Edinburgh, 1906. Russia, Atlas climatologique de V empire de Russie, St. Petersburg, 1900. BLODGET, LORIN, Climatology of the United States, Philadelphia, 1857. GREELY, American Weather. Report of the Chief of the Weather Bureau for 1900-1901, Vol. II. Climatic Charts of the United States (U. S. Weather Bureau). Climatology of the United States (Bulletin Q of the U. S. Weather Bureau). For the daily change in barometric pressure, see : ALT, EUGEN, Die Doppeloszillation des Barometers insbesondere im Arktischen Gebiete. (Inaug. Diss.), 22 pp., 1909. ANGOT, H., "Etude sur la Marche Diurne du Barometre," Ann. Bu. Cent. Met., Paris, 1889. COLE, FRANK N., The Daily Variation of Barometric Pressure, W. B. Bulletin No. 6, 1892. FASSIG, OLIVER O., The Westward Movement of the Daily Barometric Wave, Bulletin No. 31 of U. S. Weather Bureau or Monthly Weather Review for November, 1901. HANN, JULIUS, " Theory of the Daily Barometric Oscillation," Quart. Journ., Vol. 20, p. 40. HANN, JULIUS, Untersuchungen liber die tdgliche Oscillation des Barometers^ Wien, 1889. WAGNER, ARTHUR, "Die Temperatur verhaltnisse in der freien Atmosphare," Beitrage zur Physik der freien Atmosphare, Band III, Heft 2/3. WILD, HEINRICH, Repetorium fur Meteorologie, Vol. VI, No. 10 (La marche diurne du barometre en Russie par M. Rykatchew). For the barometric determination of altitude, see : Smithsonian Meteorological Tables. See section 111 of this book. WHYMPER, How to Use the Aneroid Barometer (John Murray), London. WILSON, Topographic Surveying (Wiley and Sons). For a classification, description, and explanation of the winds and the general circulation of the atmosphere, see : ABBE, CLEVELAND, The Mechanics of the Earth's Atmosphere, a Collection of Translations. Second Collection ; Washington, 1891. Third Collection ; Washington, 1910. BIGELOW, FRANK H., Report on the International Cloud Observations (Washington, 1900). BRILLOUIN, MARCEL, Memoires originaux sur la circulation ginerale de I'atmos- phere (Paris, 1900). COFFIN, J. H., The Winds of the Globe. DAVIS, American Meteorological Journal, March, 1888. FERREL, A Popular Treatise on the Winds. The preface of this book contains a list of his previous articles on the same and kindred subjects. CHAPTER V THE MOISTURE IN THE ATMOSPHERE A. THE WATER VAPOR OF THE ATMOSPHERE EVAPORATION . Water vapor, 176. Latent heat, 177. Amount of evaporation, 178. Distributlon-of the water vapor, ro,. THE CONDITION OF .THE ATMOSPHERE AS REGARDS MOISTURE Capacity of air for water vapor, 180. Saturation, 181. Humidity ; absolute and relative humidity, 182. Dew point, 183. Problems, 184. THE DETERMINATION OF THE MOISTURE OF THE ATMOSPHERE Hygrometers for determining absolute humidity, 185. Hygrometers for determining relative humidity, 186, 187. Dew point hygrometer, 188. Psychrometer, 189. Recording hygrometers, 190. THE RESULTS OF OBSERVATION The observations, 191. Normal hourly, daily, monthly, and yearly values of absolute and relative humidity, 192. Daily and annual variation in absolute humidity, 193, 194. Geographical variation in absolute humidity, 195. Daily and annual variation in relative humidity, 196, 197. Geographical variation in relative humidity, 198. Other moisture data and charts, 199. THE EFFECT OF WATER VAPOR ON THE GENERAL CIRCULATION, 200 B. DEW, FROST, FOG DEW Condensation, 201. Dew, 202, 203. The part played by latent heat, 204. Conditions for the formation of dew, 205. 189 190 METEOROLOGY FROST Frost, 206, 207. Prediction of frost, 208, 209, 210. Protection from frost, 211. Frost observations, frost data, and charts, 2x2. FOG The nature of fog, 213. Fog observations, fog data, and charts, 214. C. CLOUDS THE CLASSIFICATION OF CLOUDS Early history, 215. The international system, 216. The thirteen cloud forms, 217. The sequence of cloud forms, 218. THE OBSERVATION OF CLOUDS AND CLOUDINESS AND THE RESULTS OF OBSERVATION Height of clouds, 219. Direction and velocity of motion, 220. Cloudiness, 221. Sunshine records, 222, 223. Observations of clouds, cloudiness, and sunshine, 224. Normal values, data, and charts, 225, 226. THE NATURE OF CLOUDS Nuclei of condensation, 227. Size and constitution of cloud particles, 228. Haze, 229. THE FORMATION OF CLOUDS Introduction, 230. Condensation in warm winds blowing over cold surfaces (method i), 231. Condensation in ascending currents due to convection (method 2), 232. Condensation in forced ascending currents (method 3), 233. Condensation caused by diminishing barometric pressure (method 4), 234. Condensation in atmospheric waves (method 5), 235. Condensation caused by radiation (method 6), 236. Condensation due to conduction (method 7), 237. Condensation by mixing air (method 8), 238. Condensation by diffusion of water vapor (method 9), 239. Conditions that favor a clear sky, 240. D. PRECIPITATION THE KINDS OF PRECIPITATION Rain, 241. Snow, 242. Hail, 243. Ice storms, 244. Rain-making, 245. <_ Cooling produced by precipitation, 246 THE MOISTURE IN THE ATMOSPHERE 191 THE DETERMINATION OF PRECIPITATION AND RESULTS OF OB- SERVATION The measurement of rainfall, 247. The measurement of a snowfall, 248. Observations of precipitation, 249. Normal values and precipitation data, 250-256. THE DISTRIBUTION AND EFFECTS OF PRECIPITATION Geographical distribution of precipitation, 257-260. Other precipitation charts, 261. Variation in the amount of precipitation with altitude, 262. Relation of rainfall to agriculture, 263. Relation of rainfall and forests, 264. Effects of snowfall, 265. A. THE WATER VAPOR OF THE ATMOSPHERE EVAPORATION 176. Water vapor. The two terms, the moisture of the atmosphere and the water vapor in the atmosphere, are synonymous, and Definition of both refer to the water in invisible gaseous form which is water Vfl p r - always present in the atmosphere, but in amounts varying at the earth's surface from almost nothing to 4 per cent as a maximum. Evaporation The change from the solid or liquid state to this invisible gase- and conden- ous state is called evaporation ; its opposite is condensation. Water vapor is supplied to the atmosphere from a variety of sources. Nearly three fourths of the earth's surface is a water surface, and the oceans, lakes, and other bodies of water supply by of t ^ e sou evaporation the larger portion of the water vapor to the water vapor atmosphere. Other sources of water vapor are the surface of mosp e here. the ground which is nearly always moist, the leaves of plants and vegetation in general, and the air exhaled from the lungs of animals. Water vapor is lighter than air, the ratio being 100 : 62. That is a cubic foot of water vapor at a certain temperature and _ p Relative under a certain pressure has .62 of the mass of a cubic foot lightness of air at the same temperature and under the same pres- sure. Moist air is thus lighter than dry air because a por- tion of the dry air has been replaced by water vapor, which is lighter. 177. Latent heat. When evaporation takes place from a water surface, the molecules near the surface break away from the Definition attractions of their neighbors and make their way into the air of latent above. This breaking of the bond of adhesion between the molecules requires an expenditure of energy; and the energy used up in 192 METEOROLOGY evaporation, that is, in separating the molecules, is called latent '. This energy may be supplied from two sources. The molecular energy The two f the body itself may be used up and in this case the body sources of becomes cold ; or radiant energy in the form of ether waves latent heat. may ^ use( j up directly to supply the latent heat. The amount of latent heat involved in the change from the solid to the liquid, or from the liquid to the gaseous state, is large. If the Centigrade Values of system of measuring temperatures is used and the unit of latent heat. h ea t j s defined as the amount of heat required to raise the temperature of a gram of water 1 C., then the amount of latent heat required to change a gram of ice to a gram of water at the same tempera- ture is 79, and the number of heat units required to change the gram of water to a gram of water vapor at the temperature of boiling water is 536. If the Fahrenheit system of measuring temperature is employed and the heat unit is defined as the amount of heat required to raise one gram of water 1 F., then the latent heat in changing from the solid to the liquid state is 143, and from the liquid to the gaseous state 966. If evaporation takes place at a lower temperature than the boiling point of water, a larger amount of latent heat is required. For example, in the Fahrenheit system of thermometry it requires 1092 heat units to change a gram of water at 32 to water vapor at the same temperature. This latent heat makes itself felt in many different ways. One reason why the ocean rises so little in temperature under the direct rays of the sun is because so large an amount of the radiant energy of Illustrations . . . of the part the sun is used up in causing evaporation instead of in. heat- played by j n g the water. One reason why the air temperature at the latent heat. . J . north pole is so low in summer in spite 01 the large amount of insolation received is because the surface is a snow and ice surface and the insolation is used up as latent heat in melting the snow and ice instead of in raising the temperature of the air. After it has rained in summer, the air usually remains cooler for a considerable time. The reason is that after a rainfall the ground and all surfaces are wet, and the drying of these surfaces uses up the radiant energy of the sun, and this prevents a rapid rise in the air temperature. 178. Amount of evaporation. The amount of evapora- depends upon a great variety of things : such as the tion depends nature of the surface from which the evaporation is taking place, the amount of water vapor already in the atmosphere, the temperature, the velocity of the wind, and the baro- metric pressure. Each surface has its own rate of evaporation. Areas THE MOISTURE IN THE ATMOSPHERE 193 covered with vegetation under the same conditions evaporate about one third more water than a free water surface. In measuring evapora- tion, a free water surface is usually taken as the standard. The more water vapor there is already in the atmosphere, the slower the evapora- tion, while high temperature and high wind velocity favor evaporation. Evaporation is least when the barometric pressure is largest. If the barometric pressure is higher, the number of molecules in each cubic foot of air resting upon the evaporating surface is larger and the molecules of water find more hindrances and greater difficulty in making their way into the air above. Thus the higher the pressure, the slower the evaporation. The amount of evaporation is usually determined by exposing large pans of water in the open and measuring the amount evaporated in a given time. These evaporating pans should have a large of area and considerable depth, so that the ing the temperature will be approximately the same evaporated as the temperature of the surrounding areas. Values of the amount evaporated have also been deter- mined by measuring the evaporation from inclosed tanks, reservoirs, etc. Relative values of the amount evaporated may be determined by means of an inge- nious little instrument pictured in Fig. 93, and The pi che called a Piche evaporimeter. It consists evaporim- essentially of a long glass tube graduated to cubic centimeters or cubic inches. A piece of rough paper is held against the open end of this glass tube by means of a brass spring and plate. The rough paper is kept wet, and as the water evaporates from it, it is replaced by water from the glass tube. The amount evaporated from the rough paper in a given time can thus be determined by simply reading the amount of water remaining in the glass tube. By means of this piece of apparatus, which is easily managed and quickly read, relative values of the amount of evaporation on days of different types and at different times of the year, and in different places, may be readily deter- mined. In order to determine by means of this instrument the absolute amount evaporated, it must be standardized in terms of a free water surface near it, under the same conditions. o 194 METEOROLOGY Automatically recording evaporimeters have also been recently devised. 1 The amount of evaporation varies all the way from a few inches to several hundred inches per year in dry, hot countries. The amount of The amount evaporation is greater during the day than at night and evaporated, greater during the summer than in winter, except under very special conditions. Unless the atmosphere is to become unduly filled with water vapor, the amount which condenses from it in the form of precipitation must equal the amount which evaporates in it. Thus for the earth as a whole the amount of evaporation must equal the amount of precipitation. 179. Distribution of the water vapor. The water vapor which is supplied to the atmosphere by evaporation is distributed by three pro- Water vapor cesses > diffusion, convection, and wind. Diffusion is is distri- the process whereby the molecules of water vapor, due to diffusion their own motion, make their way slowly from point to convection, point between the molecules of the air. It is at best a slow process, and due to this cause alone, water vapor would be transported but a few feet in many hours. Air rising due to con- vection carries the water vapor along with it, and in this way it is dis- tributed throughout the lower three to five miles of the atmosphere. The wind and air currents complete the distribution by transporting the water vapor immense distances and scattering it widely. THE CONDITION OF THE ATMOSPHERE AS REGARDS MOISTURE 1 80. Capacity of air for water vapor. By the capacity of air for Definition water vapor is meant the amount of water vapor which a of capacity, given quantity of air can hold. The capacity depends, It depends upon the temperature only. With increasing temperature on temper- the amount of water vapor which the air can hold in- y ' creases rapidly ; in fac/t, it increases at an increasing rate. 181. Saturation. If a given (Quantity of air contains all the water vapor which it can hol*d; in ojther words, if its capacity is entirely Definition of satisfied, then it is said to be in a state of saturation. The saturation, amount of water vapor which saturates a given quantity of air may be expressed in grams per cubic foot or grams per cubic nfeter, or it may be expressed in terms of the pressure which it exerts in milli- 1 For recent works on evaporation, see articles by BIGELOW in the Monthly Weather Review since 1905. THE MOISTURE IN THE ATMOSPHERE 195 meters or inches. The following table, which applies to saturated air, gives for the Fahrenheit scale the amount of moisture in grains per cubic foot, the pressure in inches, and the mass of the satu- Table of rated air in grains per cubic foot. It also gives for the values - Centigrade scale the amount of moisture in grams per cubic meter, the pressure in millimeters, and the mass of the saturated air in kilo- grams per cubic meter. TEMPERATURE VAPOR PRESSURE AMOUNT OF WATER MASS OF SATURATED VAPOR IN GRAINS AIR IN GRAINS DEGREES F. INCHES PER Cu. FT. PER Cu. FT. -30 0.010 0.12 650 -20 0.017 0.21 634 -10 0.028 0.35 620 0.045 0.54 606 +10 0.071 0.84 593 20 0.110 1.30 580 30 0.166 1.97 568 40 0.246 2.86 556 50 0.360 4.09 544 60 0.517 5.76 533 70 0.732 7.99 521 80 1.022 10.95 509 90 1.408 14.81 497 +100 1.916 19.79 487 TEMPERATURE VAPOR PRESSURE AMOUNT OP WATER MASS OF SATURATED VAPOR IN GRAMS AIR IN KILOGRAMS DEGREES C. MM. PER Cu. METER PER Cu. METER -30 0.38 0.44 1.45 -20 0.94 1.04 1.40 -10 2.15 2.28 1.35 4.57 4.87 .30 +10 9.14 9.36 .25 20 17.36 17.15 .20 30 31.51 30.08 .15 +40 54.87 50.67 .11 In order to compute the amount of water vapor or saturated air which may be present in any given space, it should be held in mind that 7008 grains constitute one pound avoirdupois. Sup- niustration pose a room 15 feet square and 10 feet high were filled with saturated air at a temperature of 70 F. Its volume would be 2250 cubic feet and it would thus contain 2.6 pounds of water vapor and 167 pounds of moisture-laden air. 182. Humidity ; absolute and relative humidity. Humidity is defined as the state of the atmosphere as regards moisture. If the air were 196 METEOROLOGY absolutely dry, its humidity would be spoken of as zero. It is the humidity, as much as the temperature, which adds to the uncomfort- Definition of ableness of a sultry day in summer. In fact, the human Humidity.- body feels three things: temperature, wind, and moisture, and not, as is sometimes popularly supposed, temperature only. The cold on a windy day in winter is more penetrating than still cold, because^ the wind drives the cold air through the clothing body feels 11 m * con tact witn the skin. The cold is also more penetrat- temperature, ing on a damp day than on a dry day. The reason is because amTwind. ^ ne moisture makes the clothing a better conductor and thus lessens the heat of the body. A moist, hot day in summer is much more oppressive than a dry, hot day, because the moisture in the atmosphere prevents that free evaporation of the perspiration from the human body which cools it. *If an instrument could be invented which could indicate the feelings of the human body, it would be necessary to take account of temperature, wind, and moisture in its construction. Absolute humidity is defined as the actual quantity of moisture present in a given quantity of air.j It may be expressed as a certain . number of grains per cubic foot or a certain number of grams absolute and per cubic meter. \By relative humidity is meant the ratio relative o f ^he ac tual amount of water Vapor present in the atmos- phere to the quantity which could be there if it were satu- rated.) Relative humidity is always expressed in per cent. 183. Dew point. If the temperature of a quantity of air contain- ing moisture is lowered, a temperature will be finally reached when the Definition of given quantity of air is saturated with moisture and is con- dew point, taining all the moisture that it can hold. This temperature is spoken of as the dew point, and any further reduction in temperature must result in the condensation of some of the moisture in the form of dew, frost, fog, cloud, or precipitation. 184. Problems. The four quantities, absolute humidity, relative The inter- humidity, dew point, and temperature, are so linked together relation of by the table given in section 181, that if any two of these d C e P6 ohit rC> ^ our Quantities have been observed or determined, the other absolute hu- two may be obtained by computation. The following relative and examples will make clear the meaning of these four terms humidity. and their interrelations. Example I. If the temperature is 70 F. and the dew point is 50 F., find the relative and absolute humidity. From the table it is seen that air at 50 F. can hold 4.09 grains per cubic foot. THE MOISTURE IN THE ATMOSPHERE 197 This same amount of moisture was present at 70 F. The absolute humidity is thus 4.09 grains per cubic foot. This simply says that if air containing 4.09 grains at a temperature of 70^F. is cooled down to 50 F., it is containing all the moisture it can, it is saturated, and has reached the dew point. Air at 70 F. can contain 7.99 grains per cubic foot. The relative humidity is thus or 51 %. Example II. If the temperature is 40 F. and the relative humidity (R. H.) is 70 %, find the absolute humidity (A. H.) and the dew point. Air at 40 F. can hold 2.86 grains per cubic foot. Thus .70 = ^5\ A. H. = 2.00. 2.86 From the table it is seen that 2.00 grains saturates air with a temperature a little above 30 F. The dew point thus lies between 30 and 31 F. Example III. If the absolute humidity is 4.09 grains and the relative humidity is 80 %, find the dew point and temperature. From the table it is seen that 4.09 grains saturates air at 50 F. 50 F. is thus the dew point. .80 = - ^5 - . Possible amount = 5.11 grains. Possible amount The air must have a temperature between 50 F. and 60 F. to hold this amount. Interpolation would give a value of about 56 F. Since there are six possible combinations of four things taken two at a time, there are six possible examples. Only three have been solved, but the solution of the other three would be along the same lines as indicated above. The formula R. H. = - - and the table are sufficient to solve all six. Possible Amount THE DETERMINATION OF THE MOISTURE OF THE ATMOSPHERE 185. Hygrometers for determining absolute humidity. It has just been shown that the four quantities, temperature, absolute The four humidity, relative humidity, and dew point, are so linked jjJ^J^! 8 ' together by means of a table that if any two of the four are ture, reia- known, the other two may be determined by computation. |^ e ^ b u s ^! d " The methods for determining the real air temperature have lute humid- already been fully discussed in Chapter III. The neces- *.* sary apparatus is either a thermometer in a thermometer ail be deter- shelter, a sling thermometer, or a ventilated thermometer. " e n d s ^ The experimental methods and the necessary apparatus for suitable determining the other three quantities must now be care- appa* 118 - fully considered. These instruments for determining the moisture 198 METEOROLOGY of the atmosphere are called, in general, hygrometers 1 or moisture measurers. Absolute humidity is determined ordinarily by means of the so-called chemical hygrometer. This consists usually of two U-tubes containing calcium chlorid (CaCl 2 ) and sulfuric acid (H 2 S0 4 ). Anhy- The chemi- (j rO us phosphoric acid is perhaps better than sulfuric acid, eter for A known quantity of the moisture-laden air is drawn through determining these U-tubes so slowly that the moisture is absorbed and humidity. thus by weighing the tubes both before and after the given quantity of air was drawn through, the amount of moisture in it, and thus the absolute humidity, may be determined. In perform- ing the experiment care must be taken to have the air pass so slowly through the U-tubes that all of the moisture will be absorbed, and furthermore, certain precautions are necessary in order to prevent any moisture from gaining access to the tubes except from the air which has passed through them. Another method of determining absolute humidity is to inclose a given quantity of the moisture-laden air in a glass vessel methods of and then to extract the moisture by chemical means. By determining measuring the diminution in pressure, or if the pressure humidity. is kept the same, by noting the diminution in volume, the absolute humidity may be determined. 1 86. Hygrometers for determining relative humidity. Relative humidity is determined ordinarily by means of the hair hygrometer. . This consists essentially of a long human hair, from which struction of the oil has been extracted by soaking it in alcohol or a weak alkali solution. A weak solution of caustic potash (KOH) or caustic soda (NaOH) is ordinarily used. The hair thus treated changes its length with changes in moisture, and it has been found by experiment that these changes in length are nearly propor- tional to changes in relative humidity. As shown in Fig. 94, one end of the hair is fastened rigidly to the frame, while the other passes over a cylinder and is held taut by a weight. An index is attached to the cylinder, which moves over a dial graduated from to 100 rtrumtnt" 1 " P er cen ^> an d this indicates the relative humidity. The may be instrument may be standardized by comparing it with ardized. some other accurate hygrometer or by determining the and 100 per cent points. The point is verified by exposing the hygrometer to air which has been entirely desiccated by chemical = moist; ptrpov = measure. THE MOISTURE IN THE ATMOSPHERE 199 means. The instrument must be left a sufficient time in this moisture- free air to take up a constant reading. The 100 per cent point may be verified by exposing the hygrometer to com- pletely saturated air. Air may be completely saturated by blowing live steam into it and then cooling it down to an ordinary tempera- ture, or by spraying moisture into it by means of an atomizer. The hair hygrometer is also affected somewhat by temperature changes, but as these are slight they are ordinarily neglected. At best, the its hair hygrometer is an inaccurate accurac y- instrument. If recently standardized, its in- dications may be trusted perhaps to 2 or 5 per cent. If this is not the case, its indications are not reliable to within 10 or 15 per cent. Another form of hygrometer which works on the same principle is in very common use. The outer case is usually circular . . Another and has a diameter of an inch or form with two. The working part consists th f saine - ~ . principle. of a fine piece of spring copper FIG. 94. The Hair Hygrometer. which has been bent into the form of a spiral spring. This is coated with some hygrometric material. A bamboo preparation is often used for this purpose and it is generally colored red. This hygrometric material changes length with moisture changes, and thus causes the spiral spring to wind up or unroll. These motions are communicated to a pointer which moves over a dial graduated to indicate relative humidity. 187. There are two other instruments for determining relative humidity, but they are scientific toys rather than accurate meteoro- logical instruments. One is the so-called weather house. TWO scig- This consists of a little box usually in the shape of a house tifictoy 8 ^ and ordinarily provided with two openings in which are figures of a man and of a woman. It is so arranged that if the air is dry, the woman appears ; while if the air is damp, the man appears. These two figures are held by a twisted strand of hygrometric material which winds up or unwinds with moisture changes, and this causes the motion of the figures. It has been found that a solution of cobalt chlorid of the proper 200 METEOROLOGY strength will turn pink if the air is damp, and bluish in color if the air is dry. A piece of filter paper or cheese cloth saturated with a solution of the proper strength thus becomes a rough indicator of the relative humidity. An attempt has been made by standardizing, the strength of the solution and arranging a scale of color, to make the instrument give quantitative results. 1 88. Dew point hygrometer. The simplest possible instrument for determining the dew point consists of a bright tin cup, provided with The con- a thermometer as a stirring rod and partly filled with water, struction If ice water or cold water is added to the water in the cup, o?th* C dew ^ s temperature will be steadily reduced and there will come point hy- a moment when the outside of the cup will become coated with moisture. This means that the layer of air in con- tact with the cup has been cooled below the dew point, and the moisture has been deposited in the form of dew. The reading of the thermometer will thus indicate the dew point. A more accurate deter- mination can be made if the water in the cup is now allowed to rise in temperature until the moisture on the outside disappears. The average of the two temperatures will give a very close approximation to the dew point. This simple method of determining the dew point has been modified in various ways. The cooling may be produced by causing ether in the cup to rapidly evaporate, and black glass has also been often used instead of the bright metal. In all the instruments, however, the fundamental principle remains the same. 189. Psychrometer. The psychrometer l is an instrument which indicates directly neither absolute humidity, relative humidity, nor dew Description P om t, but all three of these may be indirectly determined of a psy- by means of its indications. It consists of two identical thermometers attached to a frame. One remains in its ordinary condition, while the bulb of the other is covered with a linen jacket to which is attached a wick which extends down to a vessel of water. The instrument is pictured in Fig. 95. The evaporation from the wet bulb thermometer cools^it and thus causes a difference in the readings of the two thermometers. The underlying principle is this : the larger the amount of moisture in the atmosphere the less the evap- The under- oration, and thus the smaller the difference between the lying prin- indications of the two thermometers. This difference, how- ever, is affected by pressure and by wind, as well as by the moisture present in the atmosphere. The higher the barometric pres- 1 i]/vxpfc = cold ; p-trpov = measure. THE MOISTURE IN THE ATMOSPHERE 201 sure, the larger the number of molecules in each given volume, and, as a result, the smaller the evaporation. The effect, however, of the pressure on the rate of evaporation is so slight that it is ordinarily . neglected. If the air in contact with the wet bulb thermometer affected by stagnates, it will soon become filled with moisture and further P ressu 5 e evaporation will cease. It is therefore necessary to main- tain a constant supply of fresh air in contact with the wet bulb ther- FIG. 95. The Psychrometer. FIG. 96. The Whirled Psychrometer. mometer. This may be done by rapidly whirling the instrument, or by blowing air against it by means of a fan, or by placing it in HQW ^ a tube and drawing air rapidly through the tube. Ordinarily, effect of the thermometer is whirled rapidly, and such an instrument is illustrated in Fig. 96. The instrument is practically use- less below 32 F. When the water in the wick^and linen jacket freezes, evaporation still takes place, but the supply can no longer be Difficulties kept good. Furthermore, the compression caused by the below 32 F ' formation of the ice around the bulb of the thermometer often causes it to indicate a temperature from one-half to a degree too high. Psychrom- eter readings are for this reason often discontinued after the tempera- 202 METEOROLOGY ture goes below 32 F. The accompanying table gives for various temperatures and various differences between the wet and dry bulb thermometer, the relative humidity and the dew point. For the reduc- tion of a long series of observations, a more elaborate table will be found more convenient. DIFFERENCE OF READINGS OF DRY AND WET BULBS TEMPERATURE OF AIR FAHRENHEIT -10 10 20 30 40 50 60 70 80 90 100 , D.P. . . -22 -7 5 16 27 38 48 58 69 79 89 99 1 R.H. . . 55 71 . 80 86 90 92 93 94 95 96 96 97 oJ ID. P. . . -76 -18 -1 12 24 35 46 57 67 77 87 98 2 R.H. . . 10 42 60 72 79 84 87 89 90 92 92 93 o D.P. . . -39 -9 7 21 33 44 55 66 76 86 96 31 R.H. . . - 13 41 58 68 76 80 84 86 87 88 90 4 ! D.P. . . -22 1 17 30 42 53 64 74 85 95 4 L R.H. . . 21 44 58 68 74 78 81 83 85 86 fi !D.P. . . _ _ _ -18 7 24 37 49 61 72 82 93 6. R.H. . . 16 38 52 61 68 72 75 78 80 c D.P. . . _ _ _ _ -8 16 31 45 57 68 79 90 8< R.H. . . - - - - 18 37 49 58 64 68 71 74 D.P. . . _ _ _ _ 4 25 40 53 65 77 87 in 10 R.H. . . - - - - - 22 37 48 55 61 65 68 19 D.P. . . _ _ __ -16 17 '35 49 62 74 85 12 R.H. . . - - - - - 8 26 39 48 54 59 62 14 ID. P. . . _ _ _ _ 5 28 45 58 70 82 14 R.H. . . - - - - - - 16 30 40 47 53 57 D.P. . . _ _ _ _ _ -20 20 39 54 67 79 16 R.H. . . - - - - - - 5 21 33 41 47 51 1 o D.P. . . _ _ _ _ _ _ 8 33 50 63 76 io R.H. . . - - - - - - - 13 26 35 41 47 f\f\ D.P. . . _ _ _ _ _ _ _ -13 25 45 60 73 20 R.H. . . - - - - - - - 5 19 29 36 42 oo D.P. . . _ _ _ _ _ _ _ 15 39 56 69 22 R.H. . . - - -. -' - - - - 12 23 32 37 O A D.P. . . _ _ _ _ _ 32 51 66 24 R.H. . . ~ ~ ~ ~ ~ ~ " " 6 18 26 33 190. Recording hygrometers. It is just as desirable to keep a The record- con ^ muous record of the moisture of the atmosphere as of ing hygrom- the other meteorological elements. It is the relative humidity of which the continuous record is ordinarily kept, and it is ac- humidity. complished by means of a recording Richard Freres hygrom- THE MOISTURE IN THE ATMOSPHERE 203 eter. As pictured in Fig. 97, this consists of a strand of hair fastened at each end and held taut in the middle by means of a hook. The FIG. 97. The Recording Hygrometer. (The case has been removed.) changes in length of this strand are magnified by a system of levers and communicate to a pen which rises and falls over the revolving drum. THE RESULTS OF OBSERVATION 191. The observations. At the regular stations of the U. S. Weather Bureau, psychrometer observations are made at 8 A.M. and 8 P.M., the hours at which observations of the other meteorological Theobserva- elements are made. In addition, a continuous record of tions taken, relative humidity, by means of the Richard Freres recording hygrom- eter, is usually kept. At the cooperative stations no observations of moisture are required. At a great many foreign stations the hair hygrometer is used instead of the psychrometer for temperatures below the freezing point. 192. Normal hourly, daily, monthly, and yearly values of absolute and relative humidity. Since absolute humidity and relative humidity are expressed by mere numbers, the various average and The aver _ normal values may be computed in exactly the same way as ages and the corresponding averages and normals for the other meteor- ^computed ological elements. In computing the normal hourly values intheregu- it is customary not to compute the value for each hour of every day in the year, but to compute the value for each hour of the dav for the various months of the year. 204 METEOROLOGY 193. Daily and annual variation in absolute humidity. The daily variation in absolute humidity may be shown graphically by plotting The char- *o SGS ^ e the normal hourly values. If these are not known, acteristics a fairly good idea of the characteristics of the variation may variation^ ^ e obtained by noting the actual change in the absolute absolute humidity on some day when the other meteorological ele- ] ments have followed as closely as possible a normal course. The maximum usually occurs in the late afternoon and the minimum at the time of sunrise. During the day evaporation has gone on rapidly from all \\ater surfaces and from the damp ground, and in- creased activity on the part of the plants and animals during the day has also added a large amount of moisture to the atmosphere. As a result the absolute humidity is greatest in the late afternoon or early evening. During the night, a large quantity of water vapor comes out of the atmos- phere in the form of dew, and evaporation is also less. As a result, the minimum is reached about the time of sunrise. During the summer months in some warm, moist, low-lying countries, there is a secondary minimum in the middle of the afternoon which causes two maxima, one during the morning and the other in the early evening. The reason for this is to be found in convection, which be- comes very energetic during the afternoon and carries up the warm, moisture-laden air, replacing it with dryer air from above. The result is a decrease in the absolute humidity which gives rise to the secondary minimum and the two maxima. 194. The annual variation in absolute humidity is found by plotting to scale the normal monthly values. The maximum usually occurs in The charac- ^ ne ^ a ^ e summer and the minimum during the winter, teristics of The reason for this is the increase in evaporation during variation ^ ne summ er due to the higher temperatures and the fact in absolute that the ground is not frozen. Plant life is also much lty ' more luxuriant. In the accompanying table will be found a series of values of absolute humidity for the various months illustrations ^ ^ e ^ ear anc ^ ^ or ^* e vear as a whole for various stations in the United States. In Fig. 98 some of these results are shown graphically and illustrate the points just mentioned. 195. Geographical variation in absolute humidity. The ofthe la en n geographical variation in absolute humidity is closely corre- erai wind lated with the general wind system. The largest values occur absolute** 1 ' 1 a ^ ^ e e Q ua to r > because temperatures there are highest and humidity. the air is relatively calm. The amount decreases through _ MOISTURE IN THE ATMOSPHERE 205 NORMAL ABSOLUTE HUMIDITY IN GRAINS PER CUBIC FOOT FOR THE VARIOUS MONTHS AND FOR THE YEAR STATION LENGTH OF RECORD fc o & u & ANNUAL Albany, N.Y. } f 15 15 1.07 1.23 1.15 1.27 1.55 1.71 2.35 2.47 3.66 3.86 5.26 5.51 5.87 6.26 5.43 5.70 4.65 4.86 3.08 3.28 2.12 2.22 1.41 1.55 3.13 3.33 Bismarck, j 8 A.M. 15 0.46 0.46 0.78 1.82 2.79 4.20 4.91 3.84 2.96 1.92 1.08 0.70 2.16 N. Dak. I 8 P.M. 15 0.55 0.59 1.07 2.28 3.25 4.86 5.28 4.02 2.85 2.39 1.32 0.79 2.44 D ' T^ "k 8 A.M. 5 1.24 1.59 1.82 2.21 2.77 3.09 2.80 2.91 2.66 2.08 1.87 1.59 2.22 jjOISC, .LClcl.no 8 P.M. 5 1.46 1.61 1.49 2.07 2.61 3.09 2.93 3.21 2.72 2.58 2.15 1.79 2.35 "R ncrf art TVT f\ SQ 8 A.M. 15 1.18 1.18 1.50 2.21 3.39 4.88 5.67 5.50 4.58 3.14 2.19 1.50 3.08 JjUbLOU, ivltibo. 8 P.M. 15 1.30 1.30 1.70 2.42 3.53 4.98 5.93 5.79 4.82 3.19 2.24 1.60 3.23 Buffalo, N.Y. 8 A.M. 15 1.17 1.07 1.41 2.18 3.26 4.86 5.79 5.43 4.31 2.91 1.98 1.39 2.98 8 P.M. 15 1.25 1.20 1.54 2.24 3.34 4.91 5.96 5.43 4.44 3.06 2.08 1.48 3.08 Charleston, 8 A.M. 2.80 3.00 3.62 4.61 6.18 7.67 8.38 8.43 6.98 5.08 3.75 2.93 5.29 S.C. 8 P.M. 3.06 3.28 3.95 4.70 6.38 8.27 8.75 8.86 7.63 5.33 4.17 3.32 5.64 Chicago 111. 8 A.M. 15 1.14 1.14 1.55 2.30 3.51 5.15 5.82 5.36 4.25 2.99 1.89 1.35 3.04 8 P.M. 15 1.31 1.48 1.80 2.57 3.53 5.14 6.07 5.96 4.85 3.14 2.11 1.60 3.30 Columbus, 8A.M. 15 1.43 1.35 1.82 2.62 3.89 5.40 5.98 5.65 4.54 3.04 2.17 1.61 3.29 Ohio 8 P.M. 15 1.56 1.56 2.04 2.86 4.07 5.53 5.96 5.89 4.56 3.21 2.29 1.80 3.44 Denver, Col. 8A.M. 15 0.82 0.91 1.13 1.64 2.38 3.16 3.75 3.38 2.40 1.65 1.11 0.95 1.94 8 P.M. 15 1.12 1.16 1.33 1.75 2.49 2.99 3.82 3.29 2.47 2.04 1.35 1.18 2.08 El Paso, Tex. 8 A.M. 15 1.41 1.46 1.42 1.52 2.01 2.97 4.94 4.87 3.87 2.62 1.71 1.42 2.52 8 P.M. 15 1.44 1.40 1.18 1.14 1.51 2.23 4.08 4.20 3.61 2.39 1.72 1.49 2.20 Galveston, 8A.M. 15 3.76 4.08 4.78 6.16 7.20 8.69 9.92 8.97 8.17 6.30 4.87 4.07 6.42 Tex. 8 P.M. 15 3.98 4.31 5.10 6.26 7.20 8.42 8.87 8.99 7.98 6.41 5.12 4.35 6.42 Havre, Mont. 8 A.M. 15 0.74 0.68 0.95 1.67 2.63 3.37 4.00 3.30 2.57 1.77 1.21 0.94 1.99 8 P.M. 15 0.86 0.87 1.35 1.86 2.76 3.54 3.60 3.50 2.98 2.20 1.56 1.11 2.18 Helena, Mont. 8 A.M. 15 0.77 0.85 1.09 1.52 2.17 2.74 3.06 2.81 2.31 1.69 1.22 0.99 1.77' 8 P.M. 15 0.88 0.96 1.25 1.82 2.22 2.92 2.90 2.45 2.35 1.81 1.45 1.14 1.85 Indianapolis, 8 A.M. 15 1.39 1.36 1.75 2.70 4.00 5.36 5.69 5.61 4.52 2.96 2.01 1.62 3.25 Ind. 8 P.M. 15 1.62 1.56 2.00 2.91 4.28 5.71 6.23 5.86 4.71 3.17 2.27 1.89 3.51 Key West, Fla. 8 A.M. 15 5.94 6.26 6.34 6.62 7.50 8.84 8.87 8.87 8.95 8.38 7.03 6.34 7.49 8 P.M. 15 6.18 6.22 6.38 6.62 7.60 8.57 8.72 8.99 8.79 8.02 7.16 6.30 7.46 Los Angeles, 8A.M. 15 2.46 2.77 3.15 3.46 4.08 4.52 5.17 5.34 4.71 3.91 2.90 2.49 3.75 Cal. 8 P.M. 15 3.32 3.38 3.73 3.87 4.27 4.71 5.28 5.53 5.20 4.77 3.99 3.56 4.30 New Orleans, 8 A.M. 15 3.42 3.72 4.36 5.21 6.59 8.07 8.69 8.53 7.43 5.43 4.21 3.55 5.77 La. 8 P.M. 15 3.59 3.97 4.59 5.35 6.47 7.87 8.60 8.34 7.40 5.60 4.54 3.84 5.85 New York, 8 A.M. 15 1.42 1.39 1.68 2.48 3.67 5.13 6.14 6.06 5.19 3.38 2.27 1.65 3.37 N.Y. 8 P.M. 15 1.59 1.56 1.89 2.66 3.94 5.45 6.39 6.38 5.45 3.57 2.46 1.75 3.59 OmaVin Nph ^ A.M. 15 0.88 0.95 1.46 2.45 3.77 5.36 6.07 5.96 4.12 2.67 1.60 1.15 3.04 mana, iNeb. g p M 15 1.16 1.21 1.76 2.71 4.06 5.52 6.36 6.17 4.48 2.83 1.91 1.52 3.31 Philadelphia, ( 8 A.M. 15 1.49 1.40 1.70 2.58 4.02 5.56 6.41 6.18 6.28 3.32 2.32 1.70 3.58 Pa. | 8 P.M. 15 1.54 1.63 1.99 2.72 4.18 5.38 6.46 6.36 5.31 3.52 2.48 1.78 3.61 Phrpni-jr AnV i ^ A-M- 8 1.96 2.12 2.01 2.11 2.26 2.55 5.19 5.38 4.07 2.83 2.17 1.68 2.86 .mx, Ariz, j g p M 8 2.20 1.90 1.98 1.91 2.09 2.18 4.27 4.80 4.04 2.48 2.47 1.90 2.72 -p . . 8 A.M. 1.01 1.04 l.?5 2.04 3.31 4.61 5.43 5.36 4.30 2.78 1.87 1.21 2.86 Jrortiandj iVle. 8 P.M. 1.01 1.10 1.54 2.18 3.44 4.73 5.65 5.61 4.44 2.93 1.89 1.28 2.98 PortlnnrJ Orp> 8A.M. 15 2.18 2.33 2.54 2.80 3.39 3.80 4.21 4.62 4.27 3.58 2.86 2.44 3.25 jruitidnci, v^re. 8 P.M. 15 2.31 2.49 2.62 2.91 3.43 4.17 4.21 4.49 4.19 3.78 3.15 2.61 3.36 Qf T A/r 8 A.M. 16 1.46 1.45 2.03 2.95 4.66 6.24 6.65 6.33 5.11 3.34 2.19 1.70 3.68 OC. AjOUlS, IVlO. 8 P.M. 16 1.62 1.72 2.36 3.41 4.79 6.20 6.76 6.43 5.50 3.51 2.39 1.90 3.88 ^t "Ponl IVTinn " A.M. 15 0.65 0.66 1.05 2.06 3.12 4.63 5.47 5.10 3.84 2.45 1.38 0.95 2.62 ot. i aui, iviinn. o 15 0.75 0.90 1.32 2.28 3.24 4.76 5.55 5.15 3.81 2.62 1.51 1.08 2.75 Salt Lake City, 8 A!M! 15 1.19 1.37 1.54 1.75 2.41 2.67 2.98 2.98 2.36 1.98 1.66 1.35 2.02 Utah 8 P.M. 15 1.48 1.56 1.81 2.01 2.51 2.62 2.96 3.12 2.55 2.31 1.94 1.60 2.47 San Francisco. ( 8 A.M. 15 3.04 3.12 3.35 3.46 3.80 4.03 4.26 4.51 4.46 4.22 3.85 3.23 3.79 Cal. ! 8 P.M. 15 3.28 3.28 3.44 3.56 3.87J4.00 4.42 4.69 4.48 4.14 3.81 3.39 3.86 Seattle, Wash, fA-JJ; 13* 13* 2.42 2.48 2.42 2.49 2.51 2.47 2.77 3.2313.63 2.72 3.3313.85 4.07 4.34 4.36 4.31 3.98 4.14 3.42 3.54 2.80 2.00 2.57 2.70 3.18 3.27 Washington, B.C. 8 A.M. 8 P.M. 15 15 1.49 1.63 1.52 1.62\ 1.98 2.14 2.764.315.87 2.96i4.-".' B.76J6.81 5.31 B.95l.92|5.68 3.50 2.33 3.61 2.44 1.66 1.77 3.69 3.88 The last year included in these normals is 1903. The time used is Eastern Standard Time. 206 METEOROLOGY the trade wind belts both on account of the lower temperatures and on account of the mixing caused by the larger wind velocities. In the horse latitudes the moisture is slightly less, both because of \ San Francisco 8P.M. Washington 8P.M. 8A.M. New Orleans 8P.M. St. Louis 8P.M. Bismarck 8P.M. 2 d MONTHS FIG. 98. The Annual Variation in Absolute Humidity. lower temperatures and because we have to do here with descending air currents which are always relatively dry. In the region of the prevailing westerlies the amount of moisture decreases steadily both THE MOISTURE IN THE ATMOSPHERE 207 on account of the still lower temperatures and the rapid motion of the air. It must not be supposed that the absolute humidity is the same for all places which have the same latitude or lie in the same wind belt. (The various factors which determine the amount of absjolujfcejnimidity are the temperature, distance the absolute from the ocean, the inclosure by mountains, and the altitude? * On the whole, the higher the temperature, the larger the amount of moisture in the atmosphere. The amount decreases with distance from the ocean and is markedly less if the place is surrounded by mountains. It also decreases markedly with altitude and practically disappears at a height of ten miles. 196. Daily and annual variation in relative humidity. The daily variation in relative humidity is found by plotting to scale the normal hourly values. The minimum usually occurs during the The chajv early hours of the afternoon and the maximum just before actenstics sunrise. During the morning the amount of moisture in the vitiation In atmosphere rapidly increases, but with the rising tempera- reiativ ture the capacity of the air for moisture increases so much more rapidly that the relative humidity decreases and reaches a T7\\j 4 , S 10 rft j \ 8 ofl 2 4 l 2 4 / g' 10 jt j M : ( o P.M. 21 21 84 76 84 75 81 68 75 55 75 64 79 58 79 55 83 56 83 60 81 63 81 69 83 76 81 64 Salt Lake City, Utah 22 74 68 58 48 48 39 34 34 40 51 52 71 52 San Francisco, f 8 A.M. Cal. 1 8 P.M. 19 19 87 75 86 72 85 70 85 69 87 72 89 72 92 77 93 79 88 73 86 71 85 71 H 88 73 Seattle, Wash, { f p - JJ' 88 81 86 74 85 66 85 58 86 59 85 56 85 51 87 54 89 63 90 74 88 81 88 82 87 67 * Washington, ( 8 A.M. D.C. 1 8 P.M. 15 15 77 69 75 66 75 65 70 59 75 68 76 71 77 72 80 74 81 76 80 72 79 69 76 77 69 The time is expressed in Eastern Standard Time. Ordinarily the last year included in Q i-> .^f-m n 1 o I'Q 1 QflQ Tf -fV>o a+a+irvn liaa a * f.hp. last. Vfia.r IS 1903. the normals is 1908. If the station has a *, the last year is 1903, 210 METEOROLOGY is less in the trade wind belts, and reaches a minimum of about 70 per cent in the horse latitudes. It then increases again in value, and in the polar regions is between 80 and 90 per cent, thus surpassing the value at the equator. Locally, it is influenced by the same four factors which determine the absolute humidity. 199. Other moisture data and charts. In connection with mois- ture, the average and normal values for the various months and for the year are the only results which are computed from the ob- made of servations taken. In the case of the U. S. Weather Bureau, the obser- the re l a tive humidity at 8 A.M. and 8 P.M. are the only data which are kept constantly computed to date. If these normal values are known for a country or for the world, corresponding charts may be prepared. Charts XV and XVI give the relative humidity for the United States for January and July. THE EFFECT OF WATER VAPOR ON THE GENERAL CIRCULATION 200. As was stated in section 176, water vapor is lighter than air under the same conditions of pressure and temperature, the ratio The mois- of mass being 62 to 100. Thus, moisture-laden air is tureacts lighter than dry air. Now the amount of moisture in each temperature cubic foot of air at the equator is normally six times the in causing amount in the same volume of air at the pole. This excess ferencesand of moisture at the equator will thus cause the air at the wind - equator to be lighter than at the pole, and will thus operate in the same way as the higher equatorial temperatures fo accentuate the pressure differences between the equator and pole, and to cause a general circulation of the atmosphere. B. DEW, FROST, AND FOG DEW 201. Condensation. Condensation is the opposite of evaporation; Definition. ^ * s ^ ne passage of the moisture of the atmosphere from the invisible gaseous form, water vapor, into some visible, 'forml^of 11 solid or liquid form. When condensation takes place, condensa- the water vapor takes the form of dew, frost, fog, cloud, rain, snow, or hail. There are thus seven forms of con- densation, and each of these forms must receive careful consideration. THE MOISTURE IN THE ATMOSPHERE 211 There are two ways in which the condensation of some of the water vapor in a given quantity of air may be brought about: by com- pression or by cooling. Of these, the first can be performed in the laboratory, and never occurs in nature. Con- densation by cooling is the only^ process which occurs in given quan- nature. Suppose a vessel contains two cubic feet of air at a temperature of 70 F., and suppose, furthermore, that it densed by contains 7 grains of moisture per cubic foot. Let this air be compressed until its volume is only one cubic foot, and suppose that the temperature has been held at 70 by cooling the con- taining vessel. This cubic foot of air will now contain 14 grains of moisture, and by referring to the table in sec- tion 181, it will be seen that a -cubic foot of air at 70 can tion b 7 com - contain only 7.99 grains. Therefore the excess of moisture, namely, 6.01 grains, must come out of the air in some form. In nature this can never occur, because there is no means by which the compressed air may be kept at a constant temperature. Descending air currents are so warmed by compression that the capacity for water vapor in- creases much more rapidly than the increase in the amount of moisture in each cubic foot. Thus, descending air currents soon become dry. If a given quantity of air is cooled, it will soon reach a temperature at which the amount of moisture in it saturates it, and any further cooling will cause moisture to come out of it in some one of the seven The three forms of condensation. This cooling of the air in nature ways in may be brought about by three processes. It may be caused maybe* by ^xpansion, and in this case the air grows colder with- cooled in out the addition or subtraction of heat from any outside nature * objects ; secondly, it may be cooled by mixture with colder _airj thirdly, it may be cooled by conducting its heat to surrounding objects which are colder or by radiaWng its heat to space or surrounding objects. ^pfQ2. DewT" In the early morning, more particularly in summer, shortly after sunrise, the temperature rises rapidly and the amount of moisture added to the atmosphere by evaporation from Desc ription bodies of water and the moist ground also increases rapidly, pf how dew Activity on the part of plants and animals also adds large quantities of water vapor to the atmosphere* As a result, all during the day, the amount of moisture in the atmosphere, that is, the number of grains in each cubic foot, is steadily increasing. In the early evening, just after sunset, the leaves of trees and plants, the grassy covering of the ground, the ground itself, and in short all material objects lose their 212 METEOROLOGY heat rapidly by radiation to space and become colder than the overlying layer of air. The layer of air loses its heat by conduction, and to a certain extent by radiation, and soon reaches the saturation point. Any further reduction in temperature will cause the excess moisture to come out on any solid object as dew. Dew has been studied for a little more than a century, but the first correct explanation was given by Wells in 1814. It was essentially the explanation just given above. The formation of water drops on the outside of an ice pitcher in a warm room on a summer's day is also an illustration of this process, ice pitcher The ice pitcher corresponds to the ground which has become illustration. co \^ } <} ue to radiation. The layer of air next it corresponds to the layer of air near the surface of the ground, and the water which collects on the outside of the pitcher corresponds to the dew. 203. There are three sources of the moisture which go to make up dew. A large part of it comes from the lo_wer layer of the atmosphere th itself. Some of it comes from the leaves^of trees and plants. sources of The moisture exudes from these, and, since evaporation is the mois- no t possible, it collects on them and adds to the supply of 11UTC* dew. Furthermore, a certain amount of jnojsture comes up from the subsoil by capillary action, and as soon as it evaporates it again condenses on the blades of grass and the leaves of plants above. Various estimates and measurements have been made of the amount of dew. The figures usually given are these : in a well-watered region, The amount during a single night, the dew would, perhaps, amount to of dew. 01 O f an mc h o f wa ter, and the total amount of dew which occurs in the course of a year might amount to an inch. In desert regions, of course, the amount is practically nothing. 204. The part played by latent heat. When evaporation takes place, energy is required to separate the molecules, and this energy is called latent heat. Conversely, when condensation takes Condensa- . tion liber- place, this latent heat again makes itself felt. It may either ates latent heat the surface upon which the condensation takes place, or retard its rate of cooling by adding to its supply of heat. The small daily range of temperature in a well-watered region is an illustration of the part played by latent heat. During the day, high Latent heat temperatures are prevented because so much energy is lessens daily used up as latent heat in evaporation and not in heating the air. At night, cooling is retarded by the liberation of so much latent heat. THE MOISTURE IN THE ATMOSPHERE 213 205. Conditions for the formation of dew. There are two condi- tions for the formation of dew ; a clear sky and absence of wind. The reason for this is because a large drop in temperature during the night is essential for the formation of dew, and both clouds and wind prevent this. Clouds prevent it by hinder- forThTfor- ing that free radiation of heat to space which is essential for * S B a large drop in temperature. Wind prevents it by mixing clear sky the layer of air in contact with the ground which has be- come cold by conduction with the warmer layers above. When dew is deposited, an inversion of temperature nearly always occurs. More dew forms in valleys than on hilltops, and there are two reasons for this. It is in the valleys that the ponds, lakes, streams, and water courses are usually located, and these add a large amount of moisture by evaporation to the atmosphere. Further- forms in W more, at night the cold and thus denser air drains into the vall eys than valleys from the surrounding hilltops. In riding over a hilly road at night, the increase in moisture and the lower temperatures on descending into the valleys is always plainly evident. FROST 206. Frost. Frost is dew which has formed with the temperature below 32 F. It consists of small, translucent, frozen drops placed side by side, and also of feathery, spinelike forms. It is The appear- sometimes thought that the translucent, frozen drops are ance of frost< due to the moisture which has exuded from the leaves of plants and frozen there. It may be the dew which had formed before the tem- perature went below 32 F. The moisture which comes from the atmos- phere and from the ground usually has a feathery, spinelike form. In winter, the moisture which comes up from the subsoil usually forms ice crystals" just beneath the surface of the ground, raising the surface and giving to the ground a spongy character. The word frost is really used in two different senses. It is used to designate the deposit of frozen dew which may form at any time of year, provided the temperature is below 32 F. and the dew point Light and has been passed. It is also used to designate the occurrence killing of those temperatures, particularly m the late spring and y aufcann, which are destructive to vegetation. When used in this "t sedjf two kinds of frosts may be distinguished ; light frosts or white frosts, and killing frosts. If the temperature does not fall much 214 METEOROLOGY lower than 4 below the freezing point, only the very tenderest vegeta- tion is killed and the frost is spoken of as a light or white frost. If the temperature goes below 28 F., even the hardier forms of vegetation will be killed, and the frost is then spoken of as a killing frost. 207. Two processes operate to produce the cooling which may result in the destructive frosts of late spring and early autumn. These are, first, the importation of colder air, and secondly, the radia- of the de- tion of heat from the ground and the cooling of the air next structive ^ o ft by conduction and radiation. 'On some occasions, a late spring strong northwest wind will import cold, dry air, thus hold- and early j n g down the maximum temperature during the day, and causing a low temperature. If the wind dies down during the night and the sky becomes clear, it takes but little radiation to cause sufficient cooling to produce a frost. On other occasions, there seems to be but little importation of cool, dry air. The sky is quite clear and there is almost no cloud. Radiation is excessive, and the resulting large drop in temperature may cause a frost. While both of the above processes are usually active, it is generally easy to see which predominates. As in the case of dew, a clear, still night is neces- sary for frost formation. These conditions are most likely to be fulfilled on the first or second night following the passage of an area of low pres- sure and the transition of the weather control to an area of high pressure. As will be seen later, this facilitates both the importation of colder air during the day and the radiation at night, the two processes which cause the low temperatures required for a frost. 208. Prediction of frost. A frost is predicted by the U. S. Weather Bureau in exactly the same way as any other temperature. The ob- The u s. servations of the various meteorological elements are made Weather at many stations at 8 A.M. and 8 P.M., and these are distrib- method of u ^ ec ^ ^y telegraph to all stations where predictions are predicting made. From the weather map which is prepared from these observations, and from the sky appearance, the probable clearness of the sky during the following night, and the probable wind velocity must be estimated. From these, chiefly, the probable drop in temperature is determined. In this way, the probable mini- mum temperature of the following morning and thus the likelihood of the frost is determined. 209. There is a widespread opinion that if the frost po^ct delayed until the maximum temperature of the preceding dl^knd Uie dew point at the time of the maximum have been determined, that THE .MOISTURE IN THE ATMOSPHERE 215 a much more accurate estimation of the probable minimum temperature of the following morning can be made. It is generally supposed that the temperature will drop rapidly until the dew point is reached and then any further drop will be much retarded SSion from by the liberation of latent heat. In other words, the dew the maxi- point will serve as a guide in determining the probable ^"aturTand amount of drop. As a matter of fact, in the northeast por- dew point tion of the United States the air is so dry and the dew point before^ 7 lies so low on all those days when a frost is imminent, that the dew point is seldom reached and plays practically no part in deter- mining the minimum temperature. The drop is, also, far from constant, and must be carefully estimated for each individual case, taking into account the probable characteristics of the afternoon and night. 210. If, after the probable minimum temperature in the thermometer shelter has been estimated, it is desired to determine the probable tem- perature of low-growing vegetation in the open, at various points in a limited area surrounding the station in question, f er e nC e~ De _ three things must be taken into account : First, that plant tween the temperatures go below the real air temperature because the temperature plants are in the open, without such a hindrance to radiation in a ther- as is the shelter about a thermometer. On the average, Jheite^and vegetation in the open will go about 2 F. below the real air the temper- temperature and in extreme cases 3 or 4 below. Second, vegetation vegetation is located near the ground and not at the height of the instruments in the thermometer shelter. The difference in tem- perature between the surface of the ground and the thermometer shel- ter, say from 5} to 6 feet above the surface of the ground, will average perhaps 3 or 4, and in extreme cases might amount to 6 or 7. Third, the variation in temperature over a limited area may amount to several degrees. In fact, for a limited area it may easily amount to 6 or 7 and in certain cases may be much more than that. Thus, the temperature of vegetation in the open, near the ground, in the coldest portion of a limited area, may be expected to average from 10 to 14 lower than the estimated minimum in the thermometer shelter, and under extreme circumstances may differ by from 5 to 10 more. 211. Protection from frost. - Tender vegetation is usually protected from frost by covering it with cloth or paper or some such Ordinar covering. The radiating surface is thus transferred from protection the ground or the leaves of the plants to the covering, and it is thus protected against frost. 216 METEOROLOGY It has often been proposed to cover a whole village or even a portion of a state with a smoke layer by building smudge fires. If concentrated The use of action of this kind could be secured, and a smoke layer of a smoke sufficient thickness could be produced, the radiation would be transferred from the surface of the ground to the surface of the smoke layer, and temperatures from 5 to 6 higher than would otherwise occur might be maintained underneath the smoke layer. Experiments are just beginning to be made on such a large scale as to make possible definite conclusions. 212. Frost observations, frost data, and charts. The observations 'made of frosts are the dates of the first light frost and the first killing frost in the autumn, and the last light and the last killing vations 86 frost in the spring. From these observations, the normal taken and dates of the first and last light and killing frost may be deter- niined. If these dates are known for a whole country or for the whole world, they may be indicated on charts. The departure from normal, however, in the case of these dates, may amount to three weeks in either direction. Charts XVII and XVIII depict, for the United States, the normal dates of the first killing frost of the autumn and the last killing frost of the spring. FOG 213. The nature of fog. If a layer of air of considerable thickness falls below the temperature of saturation, the moisture is unable to come The nature out as dew or frost at the bottom of the layer, but condenses of fog. as f O g Thjg condensation takes place on the dust and other particles in the air. A fog particle, therefore, is a minute water drop, about 0.001 inch in diameter, with a dust particle or some other particle for its nucleus (see section 227). Whenever fog occurs, the relative humidity is usually very high, at least 90 per cent. There are a few cases on record, however, when a dense fog has existed with a relative humidity even below 60 per cent. In these very exceptional cases there must have been something which prevented the evaporation of the fog parti- cles, as a coating of oil or the like. The time of year when the maximum number of foggy days occurs is very different in different parts of the world. On the Atlantic coast of The time of North America, from Maine northward, fogs occur chiefly maximum during the summer. In the southern part of the United fogginess. states, particularly inland, fogs occur chiefly during the winter. In New England and the middle Atlantic States, particularly THE MOISTURE IN THE ATMOSPHERE 217 inland, fogs are of two entirely different kinds. One is most prevalent during the late winter and early spring, while the other occurs chiefly during the late summer and early autumn. One is caused by the trans- portation of air and the other by radiation. In fact, all fogs are caused in one or the other of these two ways. In the late winter, t/ warm moisture-laden air may come up from the south over p0 rtation S " the cold snow-covered regions at the north. As a result, a fog f of layer of air of considerable thickness is cooled below its dew th"radia- point, and condensation 'in the form of fog takes place, Again, due to rain or a continued thaw, the air may have become warmed and moisture-laden. The wind suddenly changes to the northwest, the temperature drops rapidly, and a layer of air of considerable thickness may go below the dew point and the moisture come out in the. form of fog. In both cases, the fog has been caused by the transportation of air, and this transportation is brought about by the presence of storms. The formation of fog in the autumn is en- tirely different. During the day, the temperature rises to considerable heights and evaporation is considerable. The nights are growing longer and the drop in temperature is considerable. At night the temperature often goes below the dew point for a considerable thickness of air, and the moisture comes out in the form of fog. These fogs always commence during the night and are thickest at sunrise. They usually disappear dur- ing the forenoon. The disappearance of these fogs, and in fact all fogs, is brought about by the wind or a rise in temperature. The wind may mix the fog-laden layer of air with a dryer layer, thus causing its disappearance. A rise in temperature may so increase the capacity of the air for vapor that the fog particles evaporate and become invisible water vapor. It has been sometimes stated that the city fogs are entirely different from the country fogs. This is only true in the sense that combustion products play a larger part in their formation, thickness, and fo g continuance. The number of fogs in London has steadily increased with the growth of the city and the increase in the combustion products. This is shown in the accompanying table. The increase YEAR NORMAL ANNUAL NUMBER OF FOGGY DATS 1871-1875 50.8 1876-1880 58.4 1881-1885 62.2 1886-1890 74.2 Very recent data would seem to show that the number of foggy days has reached its maximum and is now decreasing. x 1 See M eteorologusche Zeitschrift, March, 1911, p. 135. f 218 METEOROLOGY in the dust and smoke, due to condensation, furnishes more nuclei for the condensation of water vapor and also facilitates the radiation of heat during the night, and thus the formation of fog. Furthermore, after the fog particle has once formed, it becomes coated with oil, and it is thus much harder for it to be evaporated when the temperature rises. In fact, the fog forms to such a thickness that the heating during the day is no longer sufficient to disperse the fog entirely, and it continues to grow thicker and thicker on successive days and only disperses when a marked change in the weather occurs. 214. Fog observations, fog data, and charts. The only observa- tions made of fog are the day on which it occurs. From these obser- The obser- vations, the normal number of foggy days for the various vations months and for the year may be determined. If these ttuTresuits normals are known for many stations, they may be charted computed. for a given country or the world. C. CLOUDS THE CLASSIFICATION OF CLOUDS 215. Early History. Although the varying forms of the clouds must have been observed by the earliest peoples, no cloud names or Howard's cloud classifications have corne down to us from antiquity, classifica- The first classification was proposed by Lamarck, a French naturalist, in 1801. He distinguished six kinds of clouds, to which French names were given. Luke Howard in 1803 proposed the scheme from which the one in general use at the present time was des- tined to develop. His classification included seven forms in all, four type forms and three combinations of the type forms to indicate inter- mediate varieties. To the four type forms of cloud he gave the Latin names Cirrus, Cumulus, Stratus, and Nj/mbu_s, meaning respectively lock of hair, pile, layer, and storm cloud. The name in each case indi- cates the most striking characteristic of the cloud. The cirrus is the delicate, fibrous, hairlike cloud. The cumulus is the lumpy, piled-up form of cloud. The stratus is the level sheet or layer. The nimbus is the cloud from which precipitation is falling. After Howard's time many other systems of classification and nomenclature were proposed, some being 'mere modifications or extensions of the Howard system. By 1891 nearly twenty systems had been proposed and had been used more or less, so that there was great uncertainty, inexactness, and in- definiteness in the use of names. THE MOISTURE IN THE ATMOSPHERE 219 216. The international system. In 1891, Hildebrandsson of Upsala and Abereromby, an English meteorologist, presented a report on cloud classification to the International Meteorological Conference ... . . The adop- wmch was then in session at Munich. This report was tion of the adopted, and the system thus inaugurated has since found international almost universal acceptance. In this system thirteen names are used to designate the different kinds of clouds. Howard's four type forms were retained, four names for intermediate varieties were formed by combining the names for the type forms, two names were The thirteen formed by prefixing alto, meaning high, and three by prefixing cloud names. fracto, meaning broken or windblown, to the names for type forms, thus making in all thirteen names. The following is an arrangement of the four type forms in order of height above the earth's surface : Cirrus Cumulus Stratus l Nimbus Stratus The thirteen cloud names are : Type forms : cirrus, stratus, cumulus, nimbus Intermediate varieties : cirro-stratus, cirro-cumulus, strato-cumu- lus, cumulo-nimbus Alto forms : alto-stratus, alto-cumulus Fracto forms : fracto-stratus, fracto-cumulus, fracto-nimbus The underlying principle of the classification is evident from the arrangement as regards height. Other possible names would have been alto-cirrus, alto-nimbus, fracto-cirrus, cumulo-cirrus, strato- cirrus, cumulo-stratus, strato-nimbus, nimbo-stratus, cirro- nimbu,, nimbo-cirrus, nimbo-cumulus. But these are all cipie of impossible combinations, redundant, or unnecessary. For Ration! 81 example, alto-cirrus is unnecessary because cirrus is always high ; fracto-cirrus is unnecessary because cirrus is always a detached broken cloud. Strato-nimbus is unnecessary because a nimbus cloud is in the form of a layer. Strato-cirrus is considered the same as cirro- stratus, cumulo-stratus the same as strato-cumulus, etc. Cirro-nimbus is impossible because the two kinds of clouds never occur at the same altitude, and alto-nimbus is impossible because a nimbus cloud is always low. 1 Stratus is here considered simply as a cloud in a horizontal sheet or layer. When not near the earth's surface it is always spoken of as alto-stratus. 220 METEOROLOGY These cloud names may also be grouped as follows : Cirrus type : cirrus, cirro-stratus, cirro-cumulus methods of Stratus type : cirro-stratus, alto-stratus, stratus, fracto- grouping stratus, strato-cumulus doud^orms. Cumulus type : cirro-cumulus, alto-cumulus, cumulus, fracto-cumulus, strato-cumulus, cumulo-nimbus. Nimbus type : cumulo-nimbus, \l nimbus, fracto-nim- - bus The International Committee grouped them as follows : Upper clouds : cirrus, cirro-stratus Intermediate clouds : cirro-cumulus, alto-stratus, alto-cumulus Lower clouds : strato-cumulus, nimbus, fracto-nimbus Clouds formed by di- urnal ascending currents :- cumulus, fracto-cumulus, cumulo-nimbus High fogs : stratus, fracto-stratus In order to give more exact names to the cloud forms the international system is sometimes subdivided. This is usually done by adding Subdivision Latin adjectives or nouns to the thirteen names of the inter- of the sys- national system. An exact description of the variety to which the name is to apply should then be given or a refer- ence given to some place in the literature on the subject where such a description can be found. 217. The thirteen cloud forms. Cirrus (abbreviation Ci.) Description: The name Cirrus is applied to detached clouds gener- ally of white color made up of slender, delicate, irregularly curling fibers or wisps of cloud. This kind of cloud is variously described as made up of wavy sprays of cloud, irregularly branching structures, bundles of drawn-out filaments, long curving threads of clouds, or fluffy heads with long streamers. It sometimes takes the form of belts which cross the sky and converge in perspective towards one or two points of the horizon. It also takes the form of feathers or ribbons or delicate fibrous bands and is sometimes striated or rippled. It is often called popularly " cats' whiskers " or " mares' tails." In its less typical form it may be composed of gauzy, clotted masses, or almost CIRRUS CIRRO-CUMULUS CIRRUS CIRRO-STRATUS CUMULUS CUMULO-NIMBUS F RA CTO-CUMU LUS STRATO-CUMULUS CUMULUS CUMULUS . THE MOISTURE IN THE ATMOSPHERE 221 with absent of structure. It very seldom causes a halo. Tho direc- tion of motion is usually from the west, and no shadows are cause i by it. Transition forms: Thickening: Ci. Cu. or Ci. St. Thinning: Disappears. Height: Mean 28,000 ft., max. 46,000 ft., min. 9000 ft. Methods of formation: (4), (5); perhaps (6), (3), (8); possibly (7), (9). 1 Stratus (abbreviation St.) Description : Stratus was originally defined as lifted fog in a hori- zontal stratum. It is now applied to any low-lying horizontal cloud sheet of approximately uniform thickness. It is a flat, structureless cloud, usually of wide extent and causes what is called " gray weather." It is sometimes thin in places so that the under surface appears as parallel lines or rolls of clouds all around the horizon. In breaking up it sometimes appears in lenticular patches. Transition forms : Thickening : Fr. St. or Nb. Thinning : St. Cu. or A. St. Height: Mean 2100 ft., max. 6000 ft., min. 400 ft. Methods of formation: (1), (3); perhaps (7). Cumulus (abbreviation Cu.) Description: Cumulus clouds are thick, rounded lumps, whose sum- mits- are domes or turrets with protuberances and whose bases are flat. When viewed opposite the sun, they are white with dark centers. When viewed near the sun, they are dark with brilliant dazzling white edges. They cast dense shadows, appear in greatest abundance during the warm part of the day, and look like exploded cotton bales. They often appear to be arranged in rows parallel to the horizon, in which case the flatness of the base is particu- larly noticeable. Transition forms : Thickening : St. Cu. or Cu. Nb. Thinning : A. Cu. Height: Top : Mean 6000 ft., max. 15,000 ft., min. 2400 ft. Bottom : Mean 4400 ft., max. 12,000 ft., min. 1600 Methods of formation: (2), (3), 1 These refer to the nine methods of cloud formation to be discussed later 231-239). 222 METEOROLOGY Nimbus (abbreviation Nb.) Description : The Nimbus is the rain cloud and is a dense, dark sheet ^ of formless cloud from which precipitation is falling. It is widely extended, and when breaks occur an upper cloud area is usually seen. Transition forms : Thickening : Heavier nimbus Thinning : St., Fr. St., or St. Cu. Height: Mean 3600 ft., max. 18,000 ft., min. 200 ft. Methods of formation: (1), (2), (3). Cirro-stratus (abbreviation Ci. St.) Description: Cirro-stratus is a thin, whitish sheet or veil of cloud which gives the whole sky a milky appearance. It is widely extended, Cirro- always distinctly fibrous and streaky in appearance, and stratus. looks like a tangled web of short, curling fibers matted to- gether. It often produces halos around the sun and moon, is grayish or white in color, and may be flocculent, granular, or banded at times'. Transition forms : Thickening : A. St. Thinning : Ci. Height : Mean 22,000 ft., max. 42,000 ft., min. 7000 ft. Methods of formation: (1), (3), (8); perhaps (6), (7), (9). Cirro-cumulus (abbreviation Ci. Cu.) Description : Cirro-cumulus consists of small white balls or flakes of semi-transparent clouds without shadows or with very faint ones. Cirro- These are arranged in groups and often in rows. These have been likened to a thin sheet broken in many places and curled at the edges, to a flock of sheep lying down, to the foam in the eddying wake of a steamer, or to floating ice cakes on the surface of a river. The popular name is " mackerel sky." Transition forms : Thickening : A. Cu. Thinning : Disappears or Ci. Height: Mean 20,000 ft., max. 35,000 ft., min. 7000 ft. Methods of formation: (2), (3), (4) ; possibly (8), (9). THE MOISTURE IN THE ATMOSPHERE 223 Alto-stratus (abbreviation A. St.) Description: Alto-stratus is a thick veil of gray or bluish color, brighter near the sun, without fibrous structure. The sun and moon are sometimes faintly visible through it, and coronas but not halos are produced by it. When breaking up it sometimes appears in lenticular patches. It appears most often in the early morning and in winter. Transition forms : Thickening : St. Thinning : Disappears or Ci. Height: Mean 15,000 ft., max. 36,000 ft., min. 3800 ft. Methods of formation: (1), (3), (4), (6); possibly (9). Alto-cumulus (abbreviation A. Cu.) Description: Alto-cumulus is the name given to white or grayish balls of dense fleecy cloud with shaded portions which are detached, but frequently lie close together. The cloud balls are often Alto- grouped in flocks or arranged in rows, sometimes in two cumulus - directions, and the balls in the center of a group are generally the largest. Alto-cumulus is more flattened and disklike than the typical cumulus. Transition forms : Thickening : St. Cu. Thinning : Ci. Cu. Height: Mean 12,000 ft., max. 27,000 ft., min. 2700 ft. Methods of formation: (2), (3), (5). Fracto-stratus (abbreviation Fr. St.) Description: When a widely extended cloud sheet which would ordinarily be called stratus is torn by the wind or mountain Fracto- summits into irregular fragments, it is called fracto-stratus. stratus - (The remaining facts are the same as for stratus.) Fracto-cumulus (abbreviation Fr. Cu.) Description : Fracto-cumulus is the name given to a cumulus cloud which has been torn by high winds so that its margins are Fracto- raggoi and irregular. It is a flat, tattered, broken cumulus. cumulus - Sometimes the cloud appears to have been rolled instead of torn. (The remaining facts are the same as for cumulus.) 224 METEOROLOGY Strato-cumulus (abbreviation St. Cu.) Description : Strato-cumulus consists of large balls or rolls of dark cloud, flat at the base, often covering the whole sky and leaving only a strato- little blue sky here and there in the breaks. It is some- cumulus, times defined as stratus thickened here and there into cumulus or cumulus joined together with a common, flat base to make a layer. Transition forms : Thickening: Fr. S., N., or Fr. N. Thinning: Fr. Cu., or Cu. Height : Mean 7000 ft., max. 14,000 ft., min. 1000 ft, Methods of formation : (2), (3) ; possibly (8). Cumulo-nimbus (abbreviation Cu. Nb.) Description : Cumulo-nimbus is the thundercloud. It consists of heavy masses of cloud rising like mountains, towers, or anvils. Below is Cumuio- the dark, formless nimbus, or fracto-nimbus, cloud from nimbus. which rain is falling. At the top a veil or cap of fibrous texture called false cirrus is often seen. It is a detached cloud, but usually covers large areas. Transition forms : Thickening : Large Cu. Nb. Thinning: Cu. Height : Top : Mean 28,000 ft., max. 42,000 ft., min. 16,000 ft. Bottom : Mean 4400 ft., max. 8000 ft., min. 600 ft. Methods of formation : (2), (3). Fracto-nimbus (abbreviation Fr. Nb.) Description : When nimbus is torn into small patches, or if low, Fracto- ragged, detached fragments of cloud (scud) move rapidly nimbus. below the main mass, it is called fracto-nimbus. (The remaining facts are the same as for nimbus.) 218. The sequence of cloud forms. Many observations have been made to determine the probability that fair weather or precipitation will The usual follow the appearance of a certain cloud* form. With the sequence of exception of nimbus, however, which by definition is a cloud ms * from which precipitation is falling, the preponderance of probability is so small that it is of no value in forecasting. There is, however, a fairly regular sequence of cloud forms in pass- ing from a clear sky to stormy weather. First comes the cirrus ; this THE MOISTURE IN THE ATMOSPHERE 225 thickens to cirro-stratus or cirro-cumulus and then becomes alto-stratus or alto-cumulus ; next the nimbus appears which often becomes fracto- nimbus. In clearing off, the sequence is generally this : the nimbus usually breaks up into fracto-nimbus, disclosing often an upper cloud area of cirrus or cirro-cumulus. The upper cloud layer gradually dis- appears and the lower fracto-nimbus becomes strato-cumulus, fracto- cumulus, and finally cumulus. THE OBSERVATIONS OF CLOUDS AND CLOUDINESS AND THE RESULTS OF OBSERVATION 219. Height of clouds. Simultaneous observations of the altitude (angular distance above the horizon) and the azimuth (direction) of a cloud obtained from two stations a mile or two apart and connected by telephone, permit the determination of the methods of height of the cloud above the earth's surface and its distance determining from the two stations. It is necessary that the two stations ^clouds* be connected by a telephone in order to agree upon the portion of the cloud which is to be observed at a given moment. Cloud heights may also be determined photographically by taking simultaneous photographs from two stations, as before. Cloud heights may also be determined by a single observer if he is provided with an accurate map of the surrounding country. The necessary observations in this case are the altitude and azimuth of the cloud, the location of its shadow on the surrounding country, and the date and time of observation. The student of trigonometry will readily see how these observations can be utilized in computing the height and distance of the cloud. These three do not exhaust the possible methods of determining a cloud height, for nearly a dozen have been proposed. Long series of observations of cloud heights have been made at the Blue Hill observatory, near Boston, at Upsala in Sweden, and at Berlin. The results of these measurements show that the height of clouds varies all the way from zero t which is actual contact many O bser- with the earth's surface, to about ten miles in the case of the vations have lofty cirrus clouds. The greatest, least, and mean height and the of the various kinds of clouds have already been given in results. section 217. Clouds are It has also been found from these observations that the higher in average height of clouds in summer is larger than that in ^nhT winter. winter. 226 METEOROLOGY by estima- tion. A nepho- scope and its use. 220. Direction and velocity of motion. The direction and velocity of motion of a cloud is usually determined by estimation. A cloud Obtained * s s Pken of, for example, as moving rapidly or slowly from ordinarily the northwest. If the cloud is near the horizon, however, reliable estimations cannot be made. Exact determinations of the direction of motion and the velocity of motion of a cloud may be made at the same time that its height is determined, if the same cloud is observed twice at an interval of twenty or thirty minutes. Ordinarily, an instrument which is called a nephoscope, 1 and is pic- tured in Fig. 101, is used for determining the direction and velocity of motion of clouds. This consists essentially of a horizon- tal mirror of black glass, provided with a divided circle. A small compass box is usually attached to the mirror so that the zero of the scale may be set at either magnetic or true north. This is provided with an eyepiece held in a curved framework so that it can be moved up and down and at the same time remain at an unchanging distance from the center of the mirror. The mirror is usually provided with a series of concentric circles at equal dis- tances from each other and with a scale etched across its face. When a cloud is to be observed, the mirror is turned and the eyepiece placed at such a height that the image of the cloud is observed to follow the divided scale across the mirror. The direction of the cloud motion can then be determined by reading the horizontal divided scale, and if its height is estimated, its velocity of motion may be determined by observing the distance on the mirror passed over in a given interval of time. The results of such observations show that the clouds move much more rapidly than the surface winds. When surface winds are moving with a velocity of 50 miles an hour, it is not at all unusual for the cloud velocity to be 150 or even 200 mites per hour. Definition of 22i. Cloudiness. The term cloudiness refers to the cloudiness, amount of the sky covered by clouds, and has nothing to do with the kind of clouds. It is usually determined by a naked-eye = cloud; ffKoirtw = I look at. FIG. 101. A Nephoscope. Results. THE MOISTURE IN THE ATMOSPHERE 227 estimation, and for this estimation the sky is divided into two portions by a small circle at an altitude of 45. The amount of the upper clouds and of the lower clouds is estimated separately and the HOW esti- average determined. It is customary to estimate the mated - amount in fifths, since the sum of these estimations will in this case give at once the cloudiness in tenths for the whole sky. There are five adjectives used to express the degree of cloudiness. Cloudless is used when no clouds at all are visible. The sky is said to be clear if it is from to T B U covered with cloud. The term partly cloudy is used to designate an amount of cloudiness a( jje c tives between T % and T 7 ^ inclusive. The word cloudy designates used to a sky from ^ to all covered with cloud. If the sky is com- cioudTness pletely cloud-covered, it is usually spoken of as overcast. Very often but three of these terms are used in connection with cloudi- ness : clear, partly cloudy, and cloudy. In this case, cloudless is included in clear, and overcast in cloudy. The term fair is also sometimes used. It is synonymous with partly cloudy and allows T^-O of an inch of pre- cipitation, but no more. 222. Sunshine rec- ords. There are vari- ous instru- The three ments for kinds of determining sunshine recorders, the amount of sunshine, which is the converse of cloudiness. The three in most general use are called the burnt paper, the photographic, and the electrical con- tact sunshine recorder. The burnt paper sun- shine recorder, often called the Campbell-Stokes sunshine recorder, is illustrated in Fig. 102, and consists of a sphere of glass which focuses the The burnt rays of the sun on a piece of paper held in a curved ghTne 81 framework at the back. This faces south and is exposed recorder, so that the sun may shine upon it from sunrise until sunset. When- FIG. 102. The Burnt Paper Sunshine Recorder. 228 METEOROLOGY ever the sun is shining, the glass bulb focuses its rays on the paper and chars a track. If the sun goes under a cloud, it is evident on the chart by the absence of charring. The strip of paper is renewed every day, way FIG. 103. The Photographic Sunshine Recorder. and in this rate record of of sunshine is kept. The photo- graphic sun- shine recorder, often called the Jordan recorder, is illustrated in Fig. 103. It - consists of a light-tight cylindrical box (sometimes two) contain- ing a piece of sensitized photographic paper. The . sun's rays are allowed to fall upon this The photo- f graphic paper through a very small opening, and as the sun passes from rising to setting, a track is left on this photographic paper. Any interruption in the sunshine makes itself mani- fest as an interruption in this track. By changing the position of the small opening, records for a whole month may be kept on a single piece of paper. 223. The sunshine recorder used by the U. S. Weather Bureau is the electrical contact recorder. It is pictured in Fig. 104, and consists essentially of a black bulb thermom- eter surrounded by a glass jacket from which the air has been extracted. Two wires have been sealed in the stem of the thermometer. When the sun shines, the black bulb quickly absorbs its radiant energy, the mer- cury rises in the stem, contact is made, and in the office below the pen makes a step every minute in the line it is tracing on a revolving drum. If an accu- the duration sunshine recorder. The Weather Bureau sunshine recorder. FIG. 104. The U. S. Weather Bureau Sun- shine Recorder. THE MOISTURE IN THE ATMOSPHERE 229 the sun ceases to shine, the black bulb thermometer radiates its heat, the mercury drops, contact is broken, and the pen makes a straight line upon the revolving drum. In this way a continuous record of sunshine is kept. The amount of sunshine may be expressed in two ways : either as the absolute duration in hours and decimals of an hour, or as n owtne a percentage. The percentage is the ratio of the actual amount is duration of the sunshine to the total possible. 224. Observations of clouds, cloudiness, and sunshine. At the regular stations of the U. S. Weather Bureau, the kind of cloud and the cloudiness are observed at eight in the morning and eight in Theobserva- the evening. A continuous record of the sunshine is also kept tions made - by means of an electrical contact sunshine recorder. The black bulb thermometer, with its vacuum jacket, is exposed on the roof of the build- ing. The record is made on a revolving drum in the office, on which is recorded wind direction, wind velocity, and rainfall, as well as sunshine. At the cooperative stations, the average cloudiness of the day as a whole is the only thing which is recorded. 225. Normal values, data, and charts. From these observations of cloudiness and sunshine, the following average and normal values may be computed : the normal cloudiness at 8 A.M. and Tfae 8 P.M. for the various months and for the year, the normal values hourly sunshine, the normal monthly sunshine, and the whichaf e * computed, normal yearly sunshine. For these normals the annual variation in cloudiness and the diurnal and annual variation in sunshine may be determined in the regular way. In the case of the regular sta- tions of the U. S. Weather Bureau, the records which are kept constantly computed to date are : hours of sunshine and percentage of possible ; average cloudiness at 8 A.M. and 8 P.M. ; number of days clear, partly cloudy, and cloudy; hourly sunshine in percentages. 226. For the northeastern part of the United States and for the larger part of the earth's surface, the amount of cloudiness is a maximum during the day and a minimum during the night. The The charac _ reason for this is convection, which causes an increase in teristks of the amount of cumulus clouds by day. The annual varia- 2^ Annual tion in cloudiness depends more upon the locality. In variation in general, the summer time is freer from clouds than the winter time.- The following table, which gives the normal values of cloudiness for the various months and for the year for various stations in the United States, will make clear the characteristics of the variation. 230 METEOROLOGY NORMAL VALUES OF SUNSHINE (PERCENTAGE OP POSSIBLE) FOR THE VARIOUS MONTHS AND FOR THE YEAR h Q j 1 LENGTH RECOR 2 < >-9 1 a < % K E 3 >H < s 1 P 1-3 | | i ! 1 s ANNUA Albany, N.Y. . . . 7 43 59 55 60 60 65 61 68 64 55 39 34 54 Bismarck, N. Dak. . 10 51 56 53 56 59 60 69 66 63 60 49 46 57 Boise, Idaho . 5 33 45 53 55 64 76 87 82 79 63 45 45 61 Boston, Mass. 10 51 57 53 53 57 60 60 60 62 54 45 52 55 Buffalo, N.Y. 14 31 42 49 55 55 66 65 66 60 48 27 24 49 Charleston, S.C. 51 48 53 63 66 61 54 51 57 55 58 50 56 Chicago, 111. 10 46 53 51 62 63 68 71 68 63 61 42 38 57 Columbus, Ohio 10 35 44 44 59 61 65 71 70 65 60 39 32 54 Denver, Col. . 14 73 67 67 68 61 69 67 66 75 76 71 68 69 Galveston, Tex. 14 47 44 47 58 66 72 70 67 68 73 59 51 60 Helena, Mont. 10 43 48 53 57 54 60 73 74 61 62 42 41 56 Indianapolis, Ind. 7 43 47 40 52 57 61 72 64 67 63 46 39 54 Los Angeles, Cal. 7 68 73 69 70 60 67 78 76 76 75 79 80 73 New Orleans, La. 14 46 44 50 54 65 55 50 51 64 64 51 47 53 New York, N.Y. . 10 53 58 54 57 56 60 57 60 60 52 51 52- 56 Omaha, Neb. . . . 7 60 56 54 60 60 65 74 68 67 59 54 49 60 Philadelphia, Pa. 50 53 51 55 58 61 63 63 66 60 53 55 57 Phoenix, Ariz. . . 8 72 79 80 85 89 94 83 82 87 87 84 85 84 Portland, Me. 57 60 57 60 68 61 65 66 64 54 46 53 58 Portland, Ore. . . 14 31 36 41 43 49 69 61 49 44 23 20 22 41 St. Louis, Mo. . . 13 53 51 52 58 63 67 70 71 71 70 52 49 60 St. Paul, Minn. . . 8 49 55 49 58 55 59 66 59 61 52 44 44 54 Salt Lake City, Utah 14 43 44 52 59 64 79 81 77 79 69 55 44 62 San Francisco, Cal. 9 53 56 60 70 67 76 71 60 67 67 57 56 63 Seattle, Wash. . . 9 23 39 46 50 49 52 62 57 48 35 16 18 41 Topeka, Kan. . . . 60 56 57 59 61 63 72 73 64 66 51 54 61 Washington, D.C. . 13 48 50 48 53 55 63 64 55 68 60 51 54 57 The last year included in these normals is 1903. If these normal values of cloudiness and of sunshine have been deter- mined for many stations, over the earth's surface, charts may be pre- pared which will show these values. Charts XIX and XX show the normal sunshine in per cent for January and for July for the United States. Chart XXI gives the normal an- nual cloudiness for the whole earth. Lines connecting those places which have the same amount of cloudiness are called isonephs, and these were first drawn for the world by Teisserenc de Bort in 1884. It will be seen that the geographical variation of cloudiness is somewhat correlated with the general wind system. The average cloudi- ness at the equator is between" 50 per cent and 60 per cent. This drops down to a minimum of about 40 per cent at latitude 25 or 30 N. It then increases with increasing latitude to a maximum at about 65 N. The cloudi- ness of the United States and the world. THE MOISTURE IN THE ATMOSPHERE 231 with a value of between 60 per cent and 70 per cent, and then drops somewhat as the pole is approached. Cloudiness varies from place to place because of altitude, temperature, inclosure by mountains, near- ness to bodies of water, and the character of the storms. THE NATURE OF CLOUDS 227. Nuclei of condensation. There are several laboratory experi- ments which would seem to show that the condensation of water vapor always takes place on some material object or on the dust i n i abora . or water particles which are present in the atmosphere, tory experi- If a vessel contains a quantity of air from which all dust particles have been removed, the temperature may be lowered tion are many degrees below the dew point without causing any r condensation in the form of fog or cloud. The condensation takes place slowly and entirely on the sides and bottom of the containing vessel. The reduction in temperature is ordinarily brought about by quickly rarefying the air and thus cooling it by expansion. If, however, ordinary air which contains a large amount of dust is used instead of the dust-free air, as soon as the temperature is lowered below the dew point, it immediately becomes filled with a foglike condensation. It is accord- ingly supposed that a dust or moisture particle serves as the nucleus or center of condensation of the water vapor which forms clouds or fog. There are certain facts, however, which are not in accord with this supposition. The amount of sediment in rain water is extremely slight, yet each raindrop is composed of myriads of cloud particles, There . each one of which is supposed to have a dust particle for its but little nucleus. This seeming contradiction is explained by two sediment in rain water. considerations. In the first place, the dust particles in the atmosphere are probably extremely minute. It has also been found that ionized air permits condensation as well as dusty air. It has been furthermore found that the air is always in a more or less _ Ionized air ionized condition. This condition can be brought about acts the by electrical discharges, by ultra-violet light, by cathode rays, or even by the impact of the air against obstacles. Condensation, however, does not readily take place on ions. It has been found that the air must contain several times as much moisture as will saturate it, before condensation on the ions can be forced. It would thus seem that dust particles must play by far the larger part in serving as nuclei of condensation in the atmosphere. 232 METEOROLOGY 228. Size and constitution of cloud particles. Cloud particles vary in size from ^TF to T1I Vo f an inch m diameter. These results have Size of keen obtained by direct measurement and also by deduc- doud tions from certain optical phenomena in which the cloud particles play a part. It was formerly supposed that the cloud particles were hollow spheres. The reasons for this assumption wSe the fact that cloud particles remained suspended for cieTare such long times in the atmosphere, and furthermore certain solid, not optical phenomena were better explained on this assumption. Recent observations, however, have shown conclusively that cloud particles are solid throughout. The reason for the suspension in the atmosphere lies in their small size. It can be shown ^at a water drop YO^TF f an mc ^ m diameter would ordi- suspension narily fall at the rate of less than two inches per second in current? "* a "* a ^ a ^ or dinary pressures. It would thus require but a very feeble ascending air current to keep a cloud particle in sus- pension or even to cause it to rise. 229. Haze. Haze occurs both high up in the atmosphere and near Two kinds the earth's surface. The high haze pales out the blue color of haze. o f ^he s ky ^y ^ay an( j gi ves to it a whitish or washed-out appearance. At night it makes itself manifest by dimming the light . . of the stars and rendering the fainter ones invisible. In and expiana- this it acts as a very thin cirrus cloud, and thus thick haze tion of the j s o ften spoken of as cirrus haze. It is caused by minute high haze. ice particles in the upper atmosphere, and is thus a conden-. sation product. Haze near the earth's surface makes itself manifest by obscuring and rendering indistinct the outlines of distant objects. It is well known Description that dust pl a y s a large part in this, and some authorities and expiana- credit the cause of haze to the presence of an abnormal haze'ne'ar 8 amount of dust in the atmosphere. The fact that haze the earth's occurs chiefly in September and October, when the dry leaves are falling from the trees and being blown about by the increasing winds, would seem to lend color to this ex- planation. It has also been thought that haze is due to the con- densation of water vapor on these large dust particles near the earth's surface, but as haze often occurs when the air is far above the dew point, it would hardly seem that condensation on any such scale is possible. A good part of the indistinctness of distant objects is probably optical and not mechanical. That is, it is due to the THE MOISTURE IN THE ATMOSPHERE 233 mixing of layers of air, of different temperature or containing very different amounts of invisible water vapor, and thus having different refractive indices. THE FORMATION OF CLOUDS 230. Introduction. The clouds are formed by the condensation of water vapor in a quantity of air which has become supersaturated; that is, which has gone below its dew point and is containing HOW clouds more moisture than the given quantity of air at the tempera- are formed - ture in question can contain. This condition may be produced by the addition of water vapor to a quantity of air already near the point of saturation, or by lowering the temperature so that the air can no longer hold the water vapor which it formerly contained. It must be held in mind that when condensation occurs in the free atmosphere, the condi- tions are very different than they are in the usual laboratory experi- ments. When the amount of water vapor which saturates a given quan- tity of air is being determined in a laboratory, large objects and parti- cles are present upon which the condensation may take place. In the free atmosphere the condensation must take place upon the extremely minute particles or ions. This explains why air has often been found which contains more than the saturating amount of moisture, and yet without condensation. Valuable information could be gained by deter- mining the amount of water vapor which saturates air when condensa- tion must take place on minute particles. The amount would be much greater than that given in ordinary tables, and would probably be dif- ferent for particles of different sizes. There are nine processes which may lead to cloud formation. Eight of these produce condensation by cooling the air, and one by adding water vapor. These nine processes may be divided into three The ldne groups. The first three are the major processes in cloud processes of formation. The second group of three are those processes ^on nm^be" which are common, but yet are not such powerful cloud pro- divided into ducers as those in the first group. The three processes which form the third group are ordinarily unimportant and in- significant. These nine processes are numbered one to nine inclusive, and in the previous sections they have been referred to by these numbers. 231. Condensation in warm winds blowing over cold surfaces (Method 1). If warm, moisture-laden air blows over a cold surface, 234 METEOROLOGY f its temperature will be reduced, and it may be lowered below the dew point, in which case condensation must take place. There are four minor illustrations of this method of cloud or fog formation, condensa- The first one is the spring fogs while it is thawing. The tion is pro- ground is perhaps snow-covered and frozen, and the tem- peratures are well beloW the freezing point. The wind begins to blow from the south, and warm air is brought by transporta- tion and is carried over these snow-covered, frozen surfaces, minor iiius- The snow begins to melt and the air goes below its dew tnitions of ^ point, and a fog is the result. A second example of this pro- cess is the formation of fog near Newfoundland. A warm, moisture-laden wind from the south blows over the Gulf Stream against the icebergs and cold water brought down from the arctic regions. The result is a cooling of the warmer air below the dew point and the forma- tion of the many fogs which are prevalent in this region. This is especially true in summer, when the temperature contrasts are particu- larly marked. A third example is the so-called mountain cloud. This is particularly noticeable if the mountain is snow-capped. The air mov- ing over the summit is cooled below the dew point, and a cloud banner streams out from the mountain top in a direction opposite to the wind direction. This cloud usually has a certain definite length. It disap- pears, due to the fact that the air which had become cooled in passing the mountain top has become warmed again by mixture or by absorbing the sun's r.adiant energy sufficiently to go above its dew point and con- tain the moisture in the form of invisible water vapor. A fourth example of this process is the formation of the so-called frost work in winter. A warm wind blows against objects which are colder, than the air. The air in contact with them is cooled below the dew point, and the moisture is deposited on these objects in the form of frost work. This frost work grows out to windward against the air current. These forms are partic- ularly prevalent on the iron work of towers or buildings which are located on high mountains. The major example of this cloud-forming process is the formation of clouds when the wind is south. A south wind transports large quantities of warm, moisture-laden air from southern countries and exa^ieTof carries it over colder surfaces farther north. As a result, this cloud- immense quantities of air are cooled below the dew point, process an( ^ c l u ds result. Due to irregularities in the surface and friction, there is probably a certain amount of mixing on the part of the air, and also rising and falling. Other cloud-forming processes THE MOISTURE IN THE ATMOSPHERE 235 would thus probably be operative at the same time. These clouds are chiefly of the stratus and nimbus variety, and occur ordinarily about one half mile to two miles above the earth's surface. The v - A A Jvina ana reason for this particular height is twofold : the air near the height of earth's surface is retarded by friction so that the greater the the cloud ' height above the earth's surface, the more vigorous the transportation. Due to this cause, the most vigorous transportation, and thus condensa- tion, would take place at the greatest heights. But the amount of mois- ture in the atmosphere decreases with altitude. Thus, two causes are counterbalanced to produce a maximum of moisture transportation, and thus cloudiness at the height stated above. 232. Condensation in ascending currents due to convection (Method 2). When air rises, due to convection, it comes into regions of less barometric pressure, and consequently it expands. The expansion causes cooling at the rate of 1.6 F. for 300 ft. condensa- (0.993 C. for 100 meters). As a result the dew point is often passed, and the excess moisture must then condense in the form of cloud. Since convection occurs usually during the day, clouds formed by this process are generally daytime clouds, and the kind is some one of the cumulus forms. Two constants make possible the computation of the height at which a convection-formed cloud ought to appear. These two constants are the adiabatic rate of cooling, that is, the cooling due to the ex- The compu _ pansion as the air rises, and the lowering of the dew point due tation of the to expansion. The value of the first constant is 1.6 F. for ^hfchthe 300 feet. The other constant is introduced here for the first cloud will time. Suppose that a cubic foot of air contains a certain amount of moisture. If this air rises a certain height, it will have expanded and become two cubic feet ; and since the amount of moisture has remained unchanged, each cubic foot will now contain one half the moisture which it contained before. As a result, the dew point will have become lowered. This dropping back of the dew point, due to expansion, amounts to .33 F. for 300 feet, or .2 C. for 100 meters. Thus, as air rises, its temperature lessens 1.6 F. for each 300 feet, and the dew point is lowered .33 F. for each 300 feet. The result is that the temperature of the air approaches the dew point at the rate of the dif- ference, namely 1.27 F. for each 300 feet. The height at which a con- vection-caused cloud should be formed can thus be computed by divid- ing the difference between the temperature and the dew point by 1.27, and multiplying it by 300. For example, suppose on a summer after- 236 METEOROLOGY noon the temperature is 85 F. and the dew point has been found to be 70 F. The height of a convection-caused cloud would be - - 300 or 3543 feet. After a cloud begins to be formed, the air continues to rise, and it will continue to rise until its temperature has fallen to the temperature of its surroundings. In order to determine this, the vertical temperature gradient must be known. There are three clouds are things which favor further rise of the air after it has com- ^J^ y menced to become cloudy. These are the absorption of radiant energy, the latent heat of condensation, and the latent heat of fusion. As soon as the rising air becomes cloudy, it ab- sorbs the radiant energy of the sun and earth. Furthermore, the latent heat of condensation is supplied to the rising air, thus lessening the rate of cooling, and causing further rise. As soon as the freezing point is passed, the latent heat of fusion is added to the latent heat of conden- sation. As a result, all of these convection-caused clouds are usually extremely thick, because the rising air supplied with heat in these three ways must rise to considerable heights before it has cooled by expansion to the temperature of the surrounding air. In the case of a thunder- cloud, the thickness may amount to as much as five or even six miles. There are several illustrations of clouds formed in this way. The cumulus clouds, which occur particularly in summer, and generally during the hotter parts of the day, are usually convection- ofconvec- formed. This can be decided by noting whether they tion-formed disappear in the late afternoon, as soon as convection stops. A thunderstorm, as will be seen later, is convection-caused, and is simply an overgrown cumulus cloud. The clouds formed over certain islands are also good illustrations of this process. The ground warms more rapidly during the day than the ocean. This causes con- vection, the warm air being forced to rise by the cool air which comes in from the surrounding ocean. Thus certain islands are cloud-covered during the day, and the sky becomes clear again at night when convec- tion ceases. In some cases this process becomes vigorous enough to cause a thundershower each day. d as 233> C ncl ensation * n forced ascending currents (Method cent may be 3). When air is forced to rise, it will expand and become caused in cooler, and perhaps cloudy, in just the same way as in con- vection. . There are two ways in which air may be forced to ascend ; one is by passing over a mountain or hill or other barrier THE MOISTURE IN THE ATMOSPHERE 237 which forces it to rise, and the second is by being forced to rise in a storm center, that is, in an area of low barometric pressure. If air is forced to rise by a barrier, a bank of cloud parallel to the barrier ordinarily forms on the leeward side. This is a fairly common cloud in mountainous regions, and the position of the cloud is determined by the wind direction. Air forced to rise by tenstics of storm centers is one of the chief causes of the stratus and the clouds nimbus clouds which often cover large sections of the country. The formation of these immense cloud areas will be fully discussed in the chapter on storms. 234. Condensation caused by diminishing barometric pressure (Method 4). If, for any reason, the barometric pressure diminishes, the air will at once expand and become cooler. If the air is nearly saturated with moisture, that is, if it is near its fo JJ e J by dew point, it will require but little diminution in pressure diminishing to cause the air to become supersaturated, and condensation in the form of cloud will then take place. This is a process of intermediate importance in cloud formations, and gives rise chiefly to clouds of the cirrus or stratus variety. 235. Condensation in atmospheric waves (Method 5). When- ever a fluid flows over obstacles, waves are usually formed in it. There are many illustrations of waves formed in this way in nature. How The surface of a field of drifting snow usually becomes wavy, atmospheric Sea sand underneath the water also becomes wavy, due to the passing backward and forward of the water. A stream flowing over an irregular bed usually has waves formed on its surface. In just the same way, when air flows over a rough surface, waves are formed in it. These waves may have a height of only a few hundred feet from trough to crest, or at times they may attain a height of half a mile or even more. When the air rises, due to the passage of one of these waves, it expands, becomes cooler, and may go below its dew point. When it falls it is compressed, rises in tempera- The charac . ture, and usually is cloudless. The waves in the atmos- tenstics of phere would thus give rise to a series of clouds in the form of formecTby parallel bars. This is of very common occurrence, and all atmospheric the rippled or striated clouds are due to atmospheric waves. This appearance is particularly noticeable in connection with cirrus or stratus clouds. 236. Condensation caused by radiation (Method 6). A layer of air near its dew point may radiate its heat to space or the cold 238 METEOROLOGY ground, and go below its dew point and become cloudy. This is often noticed in the northeast portion of the United States on ^ ne s ^> clear, cold winter mornings. Excessive radia- a radiation- tion/ has taken place from the upper layers of the atmosphere, cloud* ^ nev nave cooled below the dew point, and a thin overcast- ing in the form of an alto-stratus cloud, is the result. It is visible in the early morning, and with the rise in temperature it soon disappears. 237. Condensation due to conduction (Method 7). This is a very minor factor in cloud production. A layer of air could lose suffi- cient heat by conduction to adjoining layers to go below its auction dew point and become cloudy. Suppose a layer of air is at could pro- a temperature of 40 and is saturated with moisture, and duce a cloud. suppose, furthermore, that the layer of air above or below it has a temperature of 30. By conduction, this layer of air might lose its heat to the adjoining layer, go below the dew point, and become cloudy. This process of cloud formation would probably produce some cirrus or stratus cloud form. 238. Condensation by mixing air (Method 8). If two quantities of saturated air of different temperatures are mixed, the result is always How mixing condensation. A numerical example will illustrate the saturated air truth of this. Suppose a cubic foot of saturated air with a tempera-* 1 temperature of 70 F., and another cubic foot of saturated tures pro- air with a temperature of 50 F. are mixed. The cubic foot duces cloud. ^ air ^^ ^ temperature of 70 F. contains 7.99 grains of water vapor per cubic foot, while the cubic foot of air with the temperature of 50 F. contains 4.09 grains. These values are ob- tained from the table in section 181. The resulting air must contain the average of these two values, namely 6.04. The resulting tempera- ture will be 60 F., and it will be seen from the table that air at 60 F. can contain but 5.76 grains per cubic foot. The excess must condense in the form of cloud. As a matter of fact, only a fraction of this excess would actually condense in the free atmosphere, for as soon as condensa- tion started, latent heat would cause a rise of temperature and thus enable the air to hold more moisture. The variety of cloud formed in this way would be of the stratus, perhaps cirrus, type. Historically this process of cloud formation was one of the first to be Historical considered, and in fact it was thought that it was one of importance. fae ma j or processes in cloud formation. It is now known, however, that it plays a very minor part in the formation of cloud. THE MOISTURE IN. THE ATMOSPHERE 239 239. Condensation by diffusion of water vapor (Method 9). Suppose a given quantity of air saturated with water vapor is situated between two layers of air containing a larger amount of moisture. By diffusion some of this moisture may pass to sionmay the layer of air in question, thus supersaturating it and cause < r on ~ causing cloudy condensation. Cirrus or stratus would probably be the cloud form produced. 240. Conditions that favor a clear sky. The converse or opposite working of all the methods of cloud formation just mentioned would favor a clear sky. One of the nine processes, however, has _ , ,, . . x- mi - Thecondi- no opposite or converse, and this is convection. The onus-,, tions that sion of this one would leave two major processes of cloud favor a clear formation. There are thus two major processes which favor a clear sky. These would be cold winds blowing over a warm surface and forced descending air currents. It is a well known fact of observa- tion that clear skies are very prevalent when the wind is north or north- west, and when the station is covered by an area of high barometric pressure, which causes descending and outflowing air currents. D. PRECIPITATION THE KINDS OF PRECIPITATION 241. Rain. Four of the seven forms of condensation have already been fully considered. These are dew, frost, and fog, which occur near the earth's surface, and clouds, which form high up in the Precipita- atmosphere. The remaining three, rain, snow, and hail, are tion - all included under the general term of precipitation, and they require the most vigorous condensation for their production. Raindrops are formed from the cloud particles. These are not all of the same size, and the larger ones will fall faster than the smaller ones, or, if they are being carried up by ascending air currents, The forma _ it will be the larger particles which are carried up less rapidly, tion of a Due to these motions many collisions must occur, and when- ever two cloud particles collide, they coalesce into a diminutive rain- drop. As soon as a raindrop begins to fall at all rapidly, it v^ 11 soon come into layers of air warmer than itself, and condensation of water vapor will then take place on the cold drop. A raindrop thus increases in size, due to collision and condensation, until it reaches the base of the cloud and begins its final fall to the surface of the earth. 240 METEOROLOGY But for two circumstances, all clouds would yield precipitation. In the first place, the velocity of the ascending air currents is often suffi- cient to hold stationary or even carry up drops of consider- cioud does able size. Then again, after a raindrop leaves the base of a not yield cloud, it at once begins to evaporate, and often disappears long before the earth's surface is reached. It is not at all uncommon to see a dark trail of rain depending from a cloud while no trace of precipitation reaches the earth below. It was formerly thought that electricity played a large part in preventing the formation of rain- drops by a cloud. It was known that all clouds are highly electrified, and it was, supposed that the small cloud particles, being charged with like electricity, would be kept from coalescing by electrical repulsion. It was accordingly supposed that raindrops could form only when the cloud particles had been discharged by a lightning flash or in some other way. Later experiments seem to show that in most cases elec- tricity helps rather than hinders, if it plays any part at all. In size, raindrops vary from very small to perhaps yV of an inch in the case of large pattering raindrops. A raindrop two or three times The size of larger than this could not exist, as it would separate into raindrops. parts due to the rapidity of its fall through the air. The size of raindrops is usually determined by allowing them to fall into a layer of flour. By allowing drops of measured sizes to fall into the same flour and noting the size of the dough balls formed, the size of the rain- drops can be determined. Raindrops are largest at the base of the cloud, and diminish in size, due to the evaporation, as the earth's sur- face is approached. Of the nine processes which may lead to cloud formation, only the The three three major processes are vigorous enough to cause suffi- cioud-form- cient condensation to produce precipitation. These are cess^which warm winds over cold surfaces, convection, and forced may lead to ascent by a barrier or storm. Thus whenever precipitation on> falls from a nimbus cloud, it has been formed by one or more of these three processes. \ 242. Snow. When condensation is sufficiently vigorous to cause precipitation while the temperature is below the freezing point, snow- The struc- flakes instead of raindrops are formed. The structure of ture of snowflakes may be carefully studied by catching them on a snowflakes. piece of bJack ^^ and o b serv i ng t h em through a magnify- ing glass. They have also been observed and photographed through a microscope. Many observers have sketched the varied forms of snow- THE MOISTURE IN THE ATMOSPHERE 241 flakes, but the most complete study of them has been made by Mr. Wilson A. Bentley, of Nashville, Vt., who has observed them for more than twenty years, and whose microphotographs have been reproduced in the Monthly Weather Review for 1901 and 1902. TJje six examples pictured in Fig. 105 are taken from this collectio#^Tf the tempera- tures are low, the snowflakes are always small, flat, and regular. They always have angles of 60 or 120, which are characteristic of crystal- lized water. They bear every evidence of having been formed about a single center by continuous condensation or by the addition of small cloud particles. Attempts have been made to correlate their varied appearance with the temperature, the type of storm, the rapidity of condensation, etc., but the effect of each factor which enters in has not yet been determined. If the temperature is very low at least below zero Fahrenheit fine ice needles ture on the are formed instead of snowflakes. If the temperature is of near the freezing point, particularly in the lower layers of the atmosphere, the snowflakes often mat together and form large clots. If the temperature is still higher, the snowflakes sometimes partially melt. Much of the rain which falls in the winter time probably left the cloud as snow and melted during the descent. 243. Hail. There are three kinds of hail, and each occurs at a dif- ferent time of year and is formed in a different way. Three kinds The hail which occurs during the winter consists of ofhail - small, clear pellets of ice of about the size of large raindrops in fact, they are frozen raindrops. Hail of this kind occurs when the temperature of the cloud where the raindrop is formed is above 32 F. while the lower layers of the air are still below the freezing point. As a result, the raindrop freezes during its fall, and reaches the ground as a hailstone. The quantity is usually small, although on rare occasions the ground may be covered to a depth of three or four inches by hail of this kind. The so-called soft hail consists of small white pellets of what looks like compacted snow'. It occurs usually in very small quantities during March and April, and occasionally during the autumn. It goft ^ falls nearly always from an overgrown cumulus cloud. It is supposed to be formed from frozen cloud particles mixed with rain- drops and compacted by a high wind. In the summer time hail never occurs except during a thundershower. The hailstones are usually large, in some cases several inches Summer in diameter, and they consist of concentric layers of compact hail. 242 METEOROLOGY snow and ice. Hail nearly always falls at the beginning of a shower, and sometimes great damage is done. There are records when the hail- stones during a single shower have covered the ground to a depth of more than a foot. The full consideration of the formation of this kind of hail will be deferred until the mechanism of a thundershower has been studied, but the structure of the hailstones would seem to show that they had been formed in a cloud whirling about a horizontal axis. The nucleus is carried up and coated with snow ; it then falls and is coated with water ; it is then carried up again, the water freezes, and it is once more coated with snow. This process continues, adding coat after coat, until the hailstone becomes too heavy to be longer sustained, and it falls to the ground. As will be seen later, it is in the squall cloud at the front of a thundershower that these conditions are actually realized. Section 119 of the instructions for preparing meteorological forms of the U. S. Weather Bureau says : " Care should be taken in deter- mining the character of precipitation when in the form of Bureau sleet or hail. Only the precipitation that occurs in the form definition of o f frozen rain should be called sleet. Hail is formed by accretions consisting of concentric layers of ice, or of alternate layers of ice and snow. It frequently happens that snow falls in the form of small, round pellets, which are opaque, having the same appear- ance as snow when packed. This should never be recorded as sleet." The above is perfectly definite, and in use at all Weather Bureau stations. According to it, the winter hail or frozen raindrops should be called sleet. The soft hail should be called snow, and only the summer hail should be recognized as hail. Unfortunately, in most dictionaries, books on meteorology, and the popular mind, snow mixed with rain is considered sleet, and several kinds of hail are recognized. 244. Ice storms. It sometimes happens that it rains very soon after a continued period of cold while the temperature of the ground and the layer of air next it is still considerably below the tenstics of freezing point. As a result, the rain freezes to everything storm ^ at ^ toucnes > and trees, shrubs, vines, and the ground itself become covered with a layer of ice. This is known as an ice storm, and considerable damage is sometimes done by breaking down trees and vines on account of the weight of the ice which forms on them. These storms are particularly prevalent in New England, and Fig. 106 illustrates such a storm. 1 1 See Monthly Weather Review, December, 1900. THE MOISTURE IN THE ATMOSPHERE 243 245. Rain-making. It has been a favorite popular belief that rain can be produced by cannonading and heavy artillery fire. The frequent occurrence of thundershowers on July 4 has L.^n used as Rain on an argument in favor of this belief, and also the fact that rain July 4 and has followed so many of the great battles. As regards after battles - July 4, statistics at many stations for many years show that there are no more thundershowers on the Fourth than on the third or fifth. In connection with rain following battles, it has been pointed out that the fact was mentioned long before gunpowder was invented. There are two possible reasons why the fact has been so often noted. In the first place, the discomfort and suffering brought about by the rain would surely cause it to be mentioned. And again, an army usually gets into position during good weather while the roads are good, so that by the time the battle begins a rain period would be due. Influenced, perhaps, by this popular belief, many attempts have been made by the so-called rain-makers to cause rain by artificial means, and in fact considerable money has been expended. The Rain- methods employed are either to cause violent explosions in makin S' the upper air or to liberate a large amount of gas in the upper air. The only way in which an explosion could cause rain would seem to be by furnishing through the smoke particles numerous nuclei of condensa- tions, or by causing cloud particles to coalesce, due to the wave motion caused by the explosion. The liberation of gas might cause convection. But all these effects would seem to be on too diminutive a scale to influ- ence the immense masses of air which must be affected in order to produce rain over a considerable area. As far as results The results. are concerned, it has never been proven that any ram has fallen which would not have fallen if the experiments had not been tried. 246. Cooling produced by precipitation. As soon as it begins to rain, particularly in summer, the air is usually much cooler. One reason for this is the fact that the raindrops are from 3 to 15 F. Raindrops colder than the air at the earth's surface. It is to be noted in are colder this connection that a raindrop in falling from a cloud is not compressed and heated by the compression, as air would be, but retains its temperature during the descent. There are other causes, Other however, of the cooling caused by precipitation, such as causes of the cutting off of insolation by the clouds, the coming c down of cooler air from above, and the cooling due to evaporation from the wet ground. 244 METEOROLOGY Front View. Vertical Section. Receive* THE DETERMINATION OF PRECIPITATION AND THE RESULTS OF OBSERVATION 247. The measurement of a rainfall. The only measurement made in connection with a rainfall is the amount, that is, the thickness of the layer of water which the rainfall would produce on a level 1 he amount is measured surface, provided none were lost. The instrument for deter- mining this is called a rain gauge, and consists in its simplest form of a vessel for catching the rain and a measuring rod for determining its depth. Since the rain gauge is so simple in con- struction and rain is one of the most important of the meteorological elements, it is no wonder that rain gauges of one form or another have been in use for nearly three hundred years. The U. S. Weather Bureau standard rain gauge, as illus- trated in Fig. 107, consists of a galvanized iron cylindrical can eight inches in diameter and about twenty inches high. It is provided with a funnel-shaped cover FIG. 107. -The U. S. Weather Bureau Rain Gauge. r receiver > with a beveled rim sharp on the inside and accurately circular in order to catch the amount of rain which falls on a definite area. The shape is that of a funnel with a small opening, in order to prevent evaporation. The funnel opens into an inside brass cylinder which has just one tenth the area of the outer can. The depth of the water in this cylinder is determined by inserting a measur- ing rod and noting the height to which it is wetted. The measure- ments are made to the tenth of an inch, and thus the amount of the precipitation determined to the hundredth. It is ordinarily stated that the rain gauge should be exposed in the Exposure of open from three to six feet above the ground, and at a con- a ram gauge, giderable distance from trees, buildings, or any obstruction. The disadvantage of this exposure is that there is no protection against fforitontal Section, JTJT Description of the U. S. Weather Bureau rain gauge. THE MOISTURE IN THE ATMOSPHERE 245 the wind, and the eddy caused by the wind passing over and around the rain gauge itself lessens the amount of precipitation which is received. If the rain gauge is exposed at a greater height, wind velocities are larger, and the loss would be still greater. A rain gauge should not be exposed on a roof, as it is impossible to know what effect the eddies caused by the building will have on the amount collected. Ordinarily a roof exposure increases the amount collected somewhat. The rain gauge of the U. S. Weather Bureau stations are usually located on the flat roofs of build- ings in large cities, and thus, on account of their location, the indica- tions of the gauges may differ 5 or even 10 per cent from the correct amount. Recording rain gauges have also been devised, and these work either on the float or tipping bucket principle. In the case of the float instru- ments, as the water rises the float is carried up and makes a Recording record on a revolving drum. In the case of the tipping rain gauges, bucket instruments, the bucket becomes filled whenever a hundredth of an inch of precipitation has fallen. It then tips over, brings another bucket into place, empties itself, and makes a mark on a revolving drum. In addition to the amount, the time of beginning and ending is also noted in connection with a rainfall. If the amount is too other ot>- small to measure, it is recorded as T (trace) in the records, stations. 248. The measurement of a snowfall. Two measurements are made in connection with a snowfall : one is the depth of the snow which has fallen, and the other is the water equivalent of the snow- Two meas _ fall. The depth is determined simply by measuring it urements with a measuring rod. The only difficulty is to find some E place where the depth has not been changed by drifting. For this reason it is customary to measure the depth at three or four different places which seem to be as free from drifting as possible, and take the average. The depth is ordinarily recorded to the tenth of an inch. In order to determine the water equivalent, the snow is sometimes caught in the outer can of the rain gauge when the funnel and inner cylinder have been removed. It is then melted down by How thg placing it in a warm room or by adding a known quantity water equiv- of warm water, which is deducted as soon as it has melted. J ent 1S . determined. The water is then poured into the inner cylinder and meas- ured, and the water equivalent thus determined. If it is thought that too much snow has been blown out of the rain gauge, another method 246 METEOROLOGY may be used. A sample may be taken by inverting the can and pressing it down in the soft snow in some place free from drifting. By passing a piece of tin or a shingle underneath, a sample may be secured and melted down as before. It would, of course, be more accurate to repeat this several times. It has been found that the number of inches of snow which corre- sponds to an inch of water is by no means constant. It requires all Ratio of ^ ne wa y fro m 6 to 30 inches of snow to make an inch of snow to water, depending on the lightness of the snow. The aver- age value, however, is about ten. 249. Observations of precipitation. The observations which are made of precipitation at both the regular and cooperative stations The obser- ^ ^ ne ^- - Weather Bureau are the same. These are : the vations kind of precipitation; the time of beginning and ending; the amount of a rainfall; the depth and water equivalent of a snowfall ; the amount of snow on the ground each day. There are also many special stations which observe practically noth- ing else except rain and snow. 250. Normal values and precipitation data. From these observa- tions of precipitation, three sets of normals may be computed. These The various are ^ ne norma l hourly, daily, monthly, and annual amount normals of of precipitation (rain and melted snow) ; the normal monthly on ' and annual amount of snowfall ; the normal monthly and annual number of days with precipitation. Since all of these quanti- ties are mere numbers, these normals are computed in the regular way. The graph which represents the daily variation in the amount of pre- The daily cipitation is determined by plotting to scale the normal variation. hourly precipitations. Its form is usually quite irregular, and is very different for different places and different times of the year. 251. The accompanying table gives the amount of the precipitation (rain and melted snow) for the various months and for the year for several years, and also the normal values for Albany, N.Y. THE MOISTURE IN THE ATMOSPHERE 247 THE AMOUNT OF PRECIPITATION FOR THE VARIOUS MONTHS AND FOR THE YEAR FOR SEVERAL YEARS, AND THE NORMAL VALUES, FOR ALBANY, N.Y. JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. OCT. Nov. DEC. YEAR 1874 3.61 2.90 1.97 4:97 2.32 4.71 6.78 1.94 4.01 1.77 2.19 .76 37.93 1875 2.14 1.65 3.27 3.63 2.57 3.98 2.46 6.55 2.63 5.97 2.29 1.11 38.25 1876 1.57 4.09 4.28 3.51 2.96 4.40 4.97 .53 5.17 1.64 2.65 2.42 38.19 1877 1.95 .36 3.33 1.42 2.77 4.60 4.00 4.57 1.82 7,86 2.70 .71 36.09 1878 4.45 4.12 2.18 3.99 3.65 4.54 5.52 3.76 3.20 3.37 4.43 6.16 49.37 1879 2.54 2.80 3.79 3.17 .89 4.62 5.10 4.25 3.47 1.24 2.56 4.23 38.66 1880 2.96 2.67 2.17 2.75 3.35 2.21 3.78 2.84 2.86 2.45 2.49 2.01 32.54 1881 2.86 2.50 3.80 1.34 3.90 3.76 2.22 2.07 2.38 3.19 3.44 4.88 36.34 1882 2.64 3.31 1.79 1.27 4.15 3.98 3.Q7 1.38 7.79 .27 .97 2.24 33.76 1883 2.43 3.00 1.77 2.65 3.20 6.30 5.96 3.69 3.19 3.49 1.14 2.55 39.37 1884 2.98 3.85 4.00 2.09 2.79 1.80 5.04 5.27 1.80 2.64 3.44 3.20 38.90 1885 3.09 1.38 .62 2.89 1.92 1.98 1.98 7.58 2.00 5.54 3.90 1.51 34.39 1886 3.66 1.40 2.73 3.67 3.40 3.19 2.56 .87 2.51 2.43 5.40 2.19 34.01 1887 3.02 2.86 2.90 2.49 2.27 2.99 4.61 4.61 1.94 2.22 4.36 5.43 39.70 1888 3.04 2.07 5.62 1.95 2.98 3.18 2.52 4.74 4.68 6.10 4.48 3.30 44.66 1889 2.83 1.81 1.76 1.25 3.32 6.43 4.19 3.63 3.68 3.48 5.00 2.14 39.51 1890 2.28 2.52 3.72 1.64 5.19 2.72 2.37 5.66 8.91 5.76 1.18 2.94 44.89 1891 6.12 4.14 3.12 2.27 1.69 2.65 6.11 5.88 1.94 2.13 2.40 3.23 41.68 1892 4.08 2.13 1.64 .56 5.30 4.41 4.22 6.70 2.08 .60 2.29 .82 34.83 1893 1.31 4.63 2.00 2.10 5.08 2.92 1.82 7.21 3.20 1.67 .91 2.54 35.39 1894 2.54 2.61 .85 2.02 4.64 3.29 2.96 2.26 4.18 4.62 1.96 3.18 35.11 1895 1.65 1.63 1.31 3.09 1.72 1.72 4.02 3.14 1.80 2.35 4.78 2.59 29.80 1896 .98 4.03 4.66 .98 1.55 2.49 3.57 2.25 3.31 1.53 1.80 .73 27.88 1897 1.62 2.05 1.85 3.12 4.69 4.45 6.67 4.43 1.87 1.01 4.65 4.38 40.79 1898 2.96 3.57 1.08 2.63 4.07 5.58 1.07 6.67 1.61 4.29 4.22 .02 38.77 1899 2.50 2.92 3.97 1.03 2.23 1.61 2.69 1.77 6.33 .85 1.47 .55 28.92 1900 2.33 2.84 4.62 1.31 1.36 3.54 3.41 2.83 .74 1.83 3.95 .80 30.56 1901 1.59 .56 4.14 4.66 4.79 3.14 4.26 4.51 2.96 2.48 2.02 .42 40.53 1902 .67 3.04 3.24 2.33 1.91 3.91 5.37 3.98 4.15 2.80 .95 .13 37.48 1903 1.83 2.05 3.55 .79 .15 6.44 3.51 5.14 1.30 6.09 1.65 .59 34.09 1904 2.51 1.17 1.94 2.87 2.16 5.48 2.96 2.78 3.88 3.09 .64 .78 31.26 1905 2.66 .80 2.43 2.12 .96 3.58 2.00 3.83 3.37 2.38 1.49 .36 26.98 1906 .97 2.09 2.54 2.20 3.90 5.80 3.91 2.49 1.57 2.62 2.46 1.96 32.51 1907 1.71 1.15 .87 2.33 3.21 3.29 4.14 .74 5.81 3.71 4.15 2.52 33.63 1908 1.36 2.77 1.43 2.62 4.26 2.32 5.33 3.63 .64 2.07 .40 1.58 28.41 Sums 87.43 87.47 94.94 83.71 105.30 132.01 136.05 134.18 112.78 105.54 94.81 90.96 1265.18 Normals 2.50 2.50 2.71 2.39 3.01 3.77 3.89 3.83 3.22 3.02 2.71 2.60 36.15 It will be seen that the various months may depart widely from normal, as the amount varies all the way from almost nothing to more than twice the normal amount. In the case of the annual The annual amount a departure of 20 per cent from normal is not variation in uncommon. A departure of three inches is common, of six inches is unusual, and of eleven inches is generally record breaking. The maximum occurs in July and the minimum in February. The reason for this is the large amount of rain which falls during the summer thunder showers. Similar data can be com- puted for every station. 248 METEOROLOGY 252. The accompanying tables give the normal monthly and annual precipitation for various stations in the United States and the world. NORMAL PRECIPITATION IN INCHES FOR THE VARIOUS MONTHS AND THE YEAR STATION - K | 1 i | h S 1 1 i 1 o fe Q ANNUAL Albany, N.Y 84 2.61 2.47 2.75 2.67 3.50 4.03 4.12 3.82 3.37 3.41 2.97 2.67 38.39 Atlanta Ga 44 478 5 05 5 57 407 3 37 3 96 4.30 4.47 3.52 2.31 3.43 4.64 49 47 Baltimore, Md. . . . 39 3.23 3.54 3.91 3.15 3.53 3.79 4.83 4.20 3iS6 2.96 43'.06 Bismarck, N. Dak. . . 34 0.53 0.53 1.06 1.79 2.43 3.35 2.19 1.95 1.17 1.04 0^67 0.59 17.50 Boise, Idaho .... 32 1.85 1.51 1.54 1.16 1.49 0.96 0.18 0.18 0.48 1.20 1.07 1.64 13.27 Boston, Mass 91 3.74 3.51 4.14 3.81 3.66 8.14 3.51 4.15 3.44 3.71 4.11 3.78 44.70 Buffalo NY ... 53 3.07 2.94 2.90 2.48 3.18 2 99 3.20 2 98 3 34 348 3 25 3 38 37.19 Charleston, S.C. . . . 120 3.08 3.08 3.31 2.40 3.41 5.28 6.22 6.65 5.22 3.90 2.70 3.34 48.59 Chattanooga, Tenn. 31 5.24 5.06 5.95 4.36 3.79 4.05 3.69 3.88 3.43 2.72 3.42 4.43 50.07 Cheyenne, Wyo. . . . 40 0.38 0.54 0.98 1.73 2.48 1.65 2.06 1.47 1.00 0.71 0.41 0.39 13.80 Chicago, 111 39 2.08 2.30 2.59 2.72 3.63 3.52 3.62 3.02 3.06 2.43 2.52 2.05 33.54 Cincinnati, Ohio . . . 75 3.27 3.18 3.77 3.20 3.96 4.20 3.77 3.61 2.79 2.59 3.19 3.38 40.91 Cleveland, Ohio . . . 38 2.53 2.68 2.84 2.25 3.27 3.58 3.64 2.94 3.28 2.73 2.59 2.57 34.90 Columbus, Ohio . . . 34 2.97 3.01 3.49 2.84 3.80 3.41 3.65 3.21 2.41 2.32 2.91 2.66 36.68 Davenport, Iowa . . 39 1.66 1.58 2.24 2.68 4.26 4.06 3.63 3.73 3.15 2.29 1.83 1.53 32.64 Denver, Col. . .. . . 37 0.48 0.47 0.96 2.09 2.57 1.46 1.64 1.35 0.90 0.97 0.56 0.60 14.05 El Paso, Tex. .... 55 0.42 0.47 0.29 0.19 0.28 0.59 1.67 1.88 1.64 0.87 0.57 0.44 9.31 Erie Pa 35 2 99 2 90 2 68 2 43 3.54 3.76 3 10 3.11 3 54 3.68 3.50 2.92 38.15 Flagstaff, Ariz. . . . 21 2.91 2.62 2.90 1.33 1.59 0.47 2.45 2.89 1.43 1.18 1.61 2'.49 Galveston, Tex. . . . 37 3.54 3.05 2.99 3.11 3.29 4.32 3.96 4.79 5.84 4.35 4.03 3.75 46.78 H&VTG IVIont/. 30 07S 50 54 89 2 04 ? 93 1 89 1 21 1 06 62 68 54 13.63 Helena, Mont. . . . 29 1.00 0.66 0.83 1.06 2.13 2.26 1.13 0.69 1.13 0.77 0.76 0.79 Indianapolis, Ind. . . 38 2.96 3.09 4.04 3.39 4.00 4.23 4.06 3.21 2.99 2.66 3.48 2.98 4L09 Jacksonville, Fla. . . 56 2.82 3.25 3.39 2.70 3.93 5.64 6.37 6.60 8.16 4.60 8.81 2.86 52.53 Kansas Citv, Mo. . . 38 1.23 1.68 2.40 3.13 4.59 4.93 4.47 4.31 3.71 3.06 1.96 1.36 36.94 Key West, Fla. . . . 72 2.02 1.58 1.63 1.80 3.06 4.65 3.56 4.89 6.49 5.11 2.13 1.94 38.26 Lincoln, Neb. . . . 34 0.66 0.95 1.23 2.60 4.49 4.61 4.31 3.48 2.53 2.08 0.79 0.70 28.43 Los Angeles, Cal. . . 32 3.03 3.00 3.05 1.02 0.48 0.08 0.01 0.03 0.10 0.75 1.33 2.85 15.75 Madison, Wis. . . . 51 1.63 1.50 2.08 2.54 3.66 4.01 3.80 3.15 3.08 2.32 1.76 1.72 31.25 Memphis, Tenn. . . 49 4.99 4.63 5.17 5.07 4.34 4.38 3.39 3.33 2.94 2.55 4.44 4.30 50.22 Milwaukee, Wis. . . 39 1.63 1.50 2.08 2.54 3.66 4.01 3.80 3.15 3.08 2.32 1.76 1.72 31.25 Minneapolis, Minn. 47 1.04 0.96 1.69 2.46 3.63 4.21 3.34 3.61 3.56 2.14 1.40 1.31 29.35 Mobile Ala . . 38 480 545 7.17 4.48 4.24 5.60 6 68 6.90 5.12 3. 15 3 65 456 61 80 New Orleans, La. . . 64 454 4?8 4 56 4 53 4 06 *> 39 6 53 5 65 449 1 95 381 454 55.63 New York, N.Y. . . 84 3.29 3.27 3.45 3.33 3.55 3.41 4.08 4.38 3.44 3.42 3.55 3.30 42.47 Omaha, Neb. . . . 51 0.68 0.81 1.36 2.96 4.35 5.17 4.43 3.45 2.95 2.47 1.00 0.89 30.46 Philadelphia, Pa. . . 39 3.23 3.35 3.43 2.92 3.30 3.27 4.14 4.69 3.36 3.01 3.11 3.07 40.88 Phoenix, Ariz. . . . 32 0.85 0.87 0.61 0.33 0.11 0.08 0.89 0.94 0.69 0.37 0.67 0.86 7.27 Portland, Me. . . . 44 3.76 3.45 3.89 3.23 3.56 3.27 3.47 3.55 3.31 3.70 3.67 3.77 42.63 Portland, Ore. . . . 60 6.32 4.96 4.74 2.94 2.38 1.77 0.75 4.58 1.69 3.11 6.00 7.21 42.45 St. Louis, Mo. . . . 75 2.27 2.64 3.61 3.61 4.53 4.83 3.69 3.48 3.00 2.83 2.99 2.62 40.10 St. Paul, Minn. . . . 73 0.89 0.79 1.44 2.41 3.44 4.10 3.53 3.47 3.39 1.99 1.38 0.97 27.80 Salt Lake City, Utah . 36 1.33 1.48 2.04 2.07 2.16 0.77 0.50 0.850.91 1.43 1.39 1.40 16.33 San Francisco, Cal. 61 4.82 3.63 3.32 1.68 0.74 0.02 0.02 0.O20.31 1.03 2.62 4.64 22.96 Seattle, Wash. . . . 20 4.42 3.97 3.19 2.66 2.22 1.56 0.68 0.49 1.91 2.70 5.94 5.94 35.68 Tampa, Fla 56 2.56 2.88 2.76 1.87 2.73 7.58 9.36 9.026.32 2.41 1.71 2.29 51.49 Topeka, Kan. . . . 22 1.21 1.50 2.15 2.53 5.09 4.78 4.79 4.57 3.33 2.01 1.27 0.84 34.07 Washington, D.C. . . 73 3.13 3.09. 3.47 3.27 3.71 3.74 4.34 4.08 3.25 3.12 2.59 3.01 40.80 Yuma, Ariz 38 0.42 0.53 0.35 0.08 0.03 T 0.15 0.49 0.17 0.21 0.29 0.41 3.13 The year 1908 is the last one included in these normals. T signifies trace less than one one hundredth of an inch. THE MOISTURE IN THE ATMOSPHERE 249 STATION 1 LENGTH OF RECORD 5 i a 5 S "-S 1 1 1 1 i 1 St. Petersburg 66 0.87 0.83 0.91 0.94 1.69 1.81 2.68 2.72 2.01 1.69 1.42 1.18 16.77 Stockholm . . 35 0.79 0.71 0.79 0.91 1.38 1.65 2.28 2.44 1.81 1.97 1.38 1.10 17.21 London . . . 40 2.01 1.61 1.69 1.65 1.93 2.24 2.40 2.40 2.40 2.72 2.28 2.13 25.47 Berlin . . . 30 1.54 1.46 1.85 1.38 1.73 2.48 2.72 2.24 1.65 2.01 1.85 1.93 22.84 Vienna . . . 30 1.34 1.46 2.01 1.97 2.83 2.80 2.64 2.68 1.65 2.01 1.81 1.89 25.08 Constantinople 48 3.43 2.72 2.44 1.65 1.18 1.34 1.06 1.65 2.05 2.52 4.02 4.80 28.86 Athens . . . 37 2.20 1.50 1.46 0.87 0.83 0.43 0.32 0.43 0.55 1.77 2.99 2.48 15.83 Jerusalem . . 32 6.38 5.08 3.54 1.73 0.28 0.00 0.00 0.00 0.04 0.39 2.28 5.51 25.24 Paris . . . 30 1.42 1.30 1.50 1.69 1.77 2 13 2.05 2 13 1 97 2.40 1.77 1.81 21 93 Rome . . . 55 2.87 2.32 2.48 2.32 2.17 1.50 0.63 1.10 2.72 4.09 4.46 3.27 29^92 Capetown . . 43 0.67 0.63 0.95 1.85 3.90 4.41 3.50 3.31 2.17 1.61 1.10 0.79 24.88 Adelaide . . 50 0.75 0.67 0.98 1.85 2.95 2.99 2.76 2.48 1.93 1.73 1.14 0.91 21.14 Peking . . . 37 0.12 0.20 0.24 0.63 1.42 3.03 9.45 6.34 2.56 0.63 0.28 0.08 24.96 Hong-kong 20 1.34 1.85 2.64 5.55 13.43 16.77 13.31 14.21 8.19 4.72 1.69 1.02 84.72 Tokyo . . . 2.17 2.95 4.37 5.04 5.91 6.54 5.16 4.29 7.99 7.28 4.29 *./68.11 Manila . . . 38 1.14 0.39 0.75 1.10 4.02 9.76 15.00 14.21 14.76 7.56 5.35 2.28 76.34 Bombay . . 84 0.12 o.oo 0.00 0.04 0.55 20.55 24.57 14.88 10.94 1.77 0.47 0.04 73.94 Havana . . . 30 2.72 2.28 1.81 2.83 4.49 7.17 5.04 6.02 6.69 7.40 3.07 2.17 51.69 Rio de Janeiro 40 4.69 4.33 5.39 4.57 3.62 1.85 1.61 1.85 2.28 3.07 4.25 5.43 42.95 Buenos Aires . 40 2.91 2.60 4.61 2.83 2.99 2.80 2.17 2.32 3.11 3.62 2.87 3.90 36.73 New York 42.47 Seattle 35.68 Chicago 33.54 Bismarck 17.50 n l. fc St. Louis 40.10 Washington 40.80 Denver 14.05 II San Francisco 22.96 FIG. 108. The Annual Variation in the Amount of Precipitation at 12 Stations in the U. S. 250 METEOROLOGY Rome 29.92 London 25.47 St.Petersburg 16.77 Jerusalem 25.24 ^11 jLflllM ? i sl Bombay 73.94 Manila 70.34 Peking 24.96 Tokio58.11 10 left! lils FIG. 109. The Annual Variation in the Amount of Precipitation at 8 Foreign Stations. In Figs. 108 and 109 some of these values are shown graphically. It will be seen at once that the annual variation is very different monthly and at different places. The chief factors which determine the annual pre- amO unt of precipitation and its distribution throughout the year are : the general wind system ; the temperature changes and also the contrast in temperature between the place in question and these regions from which the prevailing winds come ; eleva- which deter- tion and inclosure by mountains ; nearness to bodies of mine the water; the characteristics of the storms which occur. At distribution n . ..' . ofprecipita- San Francisco, for example, the precipitation occurs nearly a ^ ^ urm S tne winter, and almost no rain falls during the summer. San Francisco is located in the region of prevail- ing westerlies, and furthermore, a continent is warmer than the San Fran- adjoining ocean during the summer and colder in winter. cisco. Thus, during the winter we have the moisture-laden prevailing westerlies blowing from the warmer ocean over the colder land. Thus, condensation occurs, and the precipitation THE MOISTURE IN THE ATMOSPHERE 251 is copious. During the summer, the prevailing westerlies are blowing from a colder ocean over a warmer land. As a result, there is practi- cally no precipitation. At New York the precipitation is , , . . , , . New York. caused almost entirely by storms, extratropical cyclones, and thundershowers, and as a result, the distribution throughout the year is quite uniform. The maximum occurs during the summer when thundershowers are most prevalent. Thus by considering these dif- ferent factors, the amount of precipitation and its distribution through- out the year can be explained. 253. The accompanying table gives for Albany, N.Y., the amount of snowfall for the various months and for the year for several , Snowfall at years, and also the normal values. Albany. SNOWFALL FOR THE VARIOUS MONTHS AND FOR THE YEAR FOR SEVERAL YEARS, AND THE NORMAL VALUES, FOR ALBANY, N.Y. JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT. OCT. Nov. DEC. YEAR 1886 19.7 24 8.0 1.1 22.3 11.1 64.6 1887 20.5 15.4 22.2 3.5 1.8 27.2 90.6 1888 20.9 9.2 50.9 T T 9.5 3.0 93.5 1889 12.2 4.0 T 0.2 T 6.0 224 1890 5.0 5.2 11.0 25.9 47.1 1891 26.0 10.0 2.0 11.0 T T 1.0 1.0 51.0 1892 10.0 9.0 5.2 T T 1.0 25.2 1893 8.1 40.7 3.0 4.8 T 0.2 12.4 69.2 1894 14.9 24.7 1.0 2.5 T 8.6 24.4 76.1 1895 13.5 15.6 6.3 T T T 2.2 T 37.6 1696 7.1 15.3 23.8 3.2 0.1 5.0 54.5 1897 11.8 9.4 2.6 T 4.5 9.6 37.9 1898 19.8 23.5 2.0 1.2 10.3 7.3 64.1 1899 9.4 27.5 20.8 T T T '/.2 '59.9 1900 7.9 9.6 22.8 T T 14.4 1.5 56.2 1901 13.4 4.2 7.0 T 6.0 21.6 52.2 1902 2.1 16.8 9.1 T T T 1.8 33.5 63.3 1903 4.8 9.3 2.6 1.0 T 0.3 T 12.5 30.5 1904 20.8 8.2 11.1 5.0 T 6.2 12.5 63.8 1905 18.2 7.8 8.2 2.0 T 0.4 2.1 38.7 1906 2.5 16.6 15.8 6.7 T 10.4 6.7 58.7 1907 7.8 10.8 2.1 8.5 0.5 0.1 6.5 10.6 46.9 1908 5.1 20.6 5.6 0.4 T 1.7 8.7 42.1 Suma 281.5 315.8 243.1 51.1 0.5 0.4 107.9 245.8 1246.1 Normals 12.2 13.7 10.6 2.2 T T 4.7 10.7 54.2 It will be seen that the monthly amounts may depart widely from nor- mal, as the amount varies all the way from practically nothing Normal to more than twice the normal amount. In the case of the snowfa11 - annual amount, a departure of 40 per cent from normal is not unusual. 252 METEOROLOGY In the accompanying table are given the normal monthly and annual amount of snowfall for several stations in the United States. THE NORMAL AMOUNT OF SNOWFALL IN INCHES FOR THE VARIOUS MONTHS AND FOR THE YEAR STATION LENGTH OF RECORD fc n H K 5 i 1.2 2.5 2.2 2.7 3.8 0.0 0.8 1.3 9.9 0.0 0.0 2.0 5.8 1.0 0.0 0.9 0.2 0.3 0.0 4.0 T 0.8 3.6 2.5 0.0 0.8 0.4 (M a p >-s 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 T T 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 T 0.0 0.0 0.0 g 9 p 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 & J o u 54.7 33.8 24.5 45.4 71.3 T 36.6 22.7 50.4 2.4 0.7 37.0 54.7 23.3 0.5 37.0 23.9 24.0 0.6 74.3 15.2 21.1 37.1 50.2 0.1 19.2 23.4 * Albany, N.Y. . . . * Bismarck, N. Dak. . * Boise, Idaho .... Boston, Mass * Buffalo, N.Y * Charleston, S.C. . . * Chicago 111 ... 19 29 5 36 33 19 19 24 25 20 33 10 29 39 19 33 8 35 37 20 24 30 32 21 26 12.8 5.4 5.8 11.9 19.2 T 10.2 6.9 4.4 0.9 0.2 6.0 11.5 6.5 0.2 8.7 5.0 6.5 0.0 18.8 5.2 6.6 7.7 11.3 0.0 4.2 6.0 14.0 4.9 6.4 11.9 16.4 T 11.5 5.1 7.3 0.5 0.5 5.0 8.0 5.0 0.3 11.5 5.0 8.3 0.0 20.3 3.6 6.2 6.2 10.8 0.0 6.1 7.9 11.1 7.7 3.3 7.6 8.7 T 5.0 4.0 10.9 0.1 0.0 5.0 9.6 4.3 0.0 8.2 5.3 4.0 0.0 14.4 1.1 3.5 8.8 8.6 0.0 3.0 4.6 T 1.4 0.1 0.0 0.1 0.0 T T 2.0 0.0 0.0 5.0 1.4 1.0 0.0 T 0.0 0.0 0.0 T 0.0 0.0 0.2 0.5 0.0 0.1 T 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.8 0.0 0.0 T 0.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 T T 0.0 T 0.0 T 0.8 0.0 0.0 0.3 0.0 T T 3.8 0.0 0.0 3.0 3.1 T 0.0 T 0.6 T 0.0 T 0.0 T 0.2 1.0 0.0 0.1 T 4.7 6.3 2.2 2.5 6.5 0.0 2.5 1.6 4.5 0.3 0.0 7.0 6.6 1.4 0.0 1.8 2.4 0.9 0.0 4.8 0.8 0.8 4.7 5.8 0.0 1.3 1.0 10.0 4.8 4.5 8.8 16.3 T 6.6 3.8 6.8 0.6 T 4.0 8.1 5.0 T 5.9 5.4 4.0 0.6 12.0 4.5 3.2 5.7 9.7 0.1 3.6 3.5 Columbus, Ohio . . . Denver, Col * El Paso, Tex. . . . * Galveston, Tex. . . . Havre Mont .... Helena, Mont Indianapolis, Ind. . . New Orleans, La. . . * New York, N.Y. . . * Omaha, Neb. . . . Philadelphia, Pa. . . . * Phrenix, Ariz. . . . Portland Me Portland, Ore * St. Louis, Mo. . . . St. Paul, Minn. . . . * Salt Lake City, Utah . * San Francisco, Cal. Topeka Kan Washington, D.C. . . T indicates a trace ; less than one tenth of an inch. Ordinarily the last year included in the normals is 1908. If the station has a * the last year is 1903. 254. In the accompanying table will be found the normal number of days with precipitation for the various months and for Normal the year, for several stations in the United States. It will be days b with seen ^ na ^ ^ or ^ ne northeastern part of the country, precipita- precipitatioa tion occurs on nearly half of the days of a year. THE MOISTURE IN THE ATMOSPHERE 253 THE NORMAL NUMBER OF DATS WITH PRECIPITATION FOR THE VARIOUS MONTHS AND FOR THE YEAR h D j STATION w 5 8 JA fc 1 < 1 IH a 1 t> H^ o p o 5 ^ t> z B although it may make any angle Avith this button of the direction. The central pressure averages about 28.5 inches, elements. although pressures as low as 27 inches have been observed. 1 The belt of slightly higher pressure surrounding the cyclones and called 600 MILES FIG. 110. The Distribution of the Meteorological Elements about a Tropical Cyclone. the pericyclonic ring usually has a pressure of about 30.1 inches. It will be remembered that the occurrence of this high pressure was one of the first signs of the coming of the cyclone. The wind blows spirally inward, turning counterclockwise in the northern hemisphere. Its direction makes an angle of about 30 with 1 See Meteor ologische Zeitschrift, 1902, p. 474, for a barograph trace during a typhoon. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 269 the isobaric lines. The value of the angle is nearly the same all the way round, although somewhat greater, 35 to 40, in the northeast quad- rant and somewhat less, 20 to 25, in the southwe% quadrant. The wind velocity is small on the outside, nothing in the calm central eye, and attains its greatest velocity in the middle of the rain area. The maximum velocity nearly always reaches 100 miles an hour and some- times probably nearly 200. The arrows in the diagram by their direction and length show the direction and velocity of the wind. The cloud area is shaded in the diagram and is concentric with the isobars. On the outside, the cirrus radiates in all directions, while the FIG. 111. Temperature and Moisture Changes in a Tropical Cyclone. nimbus clouds occupy the inside portion. The transition clouds are usually cirro-stratus, perhaps cirro-cumulus. The calm central eye is sometimes almost cloudless. The precipitation is excessive, often several inches, and covers the same area as the nimbus clouds. It is sometimes accompanied by thunder and lightning, although this is most common at the end of the storm. It is said that the most violent cy- clones are never accompanied by thunder and lightning. The temperature and moisture are practically the same in all quad- rants, and thus would be represented in the diagram by ovals concentric with the isobars. As this would complicate the diagram too much, they are shown in section in Fig. 111. The temperature is high, 270 METEOROLOGY perhaps 80 F., before the coming of the cyclone. It drops to perhaps 70 when the nimbus cloud comes, due to the cooling caused by the precipitation. I^the calm central eye it rises nearly as high as before the coming of the cyclone and drops again when the precipitation begins afresh. The relative humidity is high at the beginning, is perhaps 95 per cent during the precipitation, and drops in the calm central eye to perhaps 70 per cent, depending on the rise in temperature. The whole formation is from 300 to 600 miles in diameter, while the central eye has a diameter from fifteen to twenty-five miles. The distribution of the six meteorological elements is thus completely portrayed by means of these two diagrams. The sequence of the changes which might be expected in the meteorological elements due to the passage of a cyclone centrally or obliquely pver a station can be determined by considering these diagrams as free and moving them centrally or obliquely over a given point. 270. Some special tropical cyclones. Hundreds, yes, more than a thousand, of these storms have been more or less carefully observed in different parts of the world since 1400 A.D. It is thus im- Possible to write up the life history of all of them or even the which have most important of them. Various books on meteorology served" often give partial lists of the most destructive of them and perhaps a more detailed description of one or two. For all the known facts about any one, the reader must be referred to the periodical literature of the subject. 1 (See Appendix IX.) 271. The last very destructive hurricane in the West Indies is the one which caused such appalling losses at Galveston, September 8, 1900. It is estimated that here above 6000 lives were lost and S'hSri- 68 " over $30,000,000 worth of property was destroyed. This cane on cyclone, unfortunately, is not very typical as regards its 8 ' behavior or the path which it followed. It appeared first as a small storm on the morning of September 1 southeast of the island of Hayti. It passed over Cuba on the 5th and reached south- western Florida on the 6th. Here it turned abruptly to the left and crossed the Gulf of Mexico, reaching Galveston on the evening of the 8th with its destructive energy at its maximum. It was then about 500 miles in diameter with a central pressure of less than 28.48 inches, and a wind velocity of more than a hundred miles an hour. The center passed south of Galveston but within less than 40 miles of the city. 1 For pictures of the wreckage caused by a hurricane see Monthly Weather Review, Sept., 1906. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 271 At Galveston, the lurid sunsets and the brick dust sky as heralds of an approaching hurricane were entirely lacking. The cirrus clouds made their appearance on the morning of the 7th, coming from the south- east, and the high tides and the heavy ocean swell appeared during the afternoon of the same day. The cirrus clouds became a mixture of cirrus, alto-stratus, and cumulus during the afternoon of the 7th, and later changed to strato-cumulus. On the 8th, in the early morning, the cloud forms were fracto-stratus and strato-cumulus with here and there a little blue sky, but this was soon followed by showery weather. The dense rain cloud came at noon and continued until midnight. The wind went to the northeast on the 7th and blew with constantly increas- ing velocity. On the afternoon of the 8th it was still northeast and blow- ing a gale. By 5 in the afternoon, it had reached full hurricane strength. Between 5 and 6.15 it blew at the rate of 84 miles an hour for several five-minute periods. At 6.15 the anemometer and other meteorological instruments blew away after a wind velocity of 100 miles per hour had been recorded for two minutes. By estimation, the wind velocity reached 120 miles per hour between 6.15 and 8.30 P.M. After 8.30 the wind still continued of hurricane strength, but shifted first to the east, then the southeast, and finally reached the south by 11 P.M. From that time on, the direction continued south and the velocity steadily decreased. The pressure dropped rapidly all during the 8th and reached its lowest, 28.48 inches, at 8.30 P.M. As the storm passed a little south of Galveston, the central pressure must have been a little less than this. After 8.30, the barometer rose rapidly. That all this would be the natural sequence of events can be seen by moving the diagram given as Fig. 110 from right to left below a given point. The destruction of life and property at Galveston was caused by a storm wave as well as by the excessive wind velocity. The water rose steadily all during the 8th, flooding a large part of the city, and at 7.30 P.M. there was a sudden rise of four feet in a few minutes. After passing Galveston, the storm turned to the right, passed inland and up the Mississippi Valley as far as Nebraska, where it then again turned, crossing Lake Michigan and Maine and passing out over Nova Scotia on the 12th. As soon as it went inland it lost its violence and many of the characteristics of a tropical cyclone and was gradually transformed into an extratropical cyclone. It also grew larger, and by the time it reached the St. Lawrence Valley it was more than a thou- sand miles in diameter. It retained, however, a low central pressure, rather high winds, and a compact form all through its life history. Figure 272 METEOROLOGY 112 shows its path in detail, and Chart XXVII reproduces the 8 A.M. weather map for September 8, 1900, when it was near Galveston. 272. Rules for mariners. The rules for mariners which are ordi- The rules narily given in connection with tropical cyclones are : first, for mariners. ^ o avo id running before the wind, particularly when the center of the cyclone is to the westward, as this would bring the vessel THE SECONDARY CIRCULATION OF THE ATMOSPHERE 273 towards the center and across the track of the coming cyclone ; secondly, to so direct the vessel as to avoid as far as possible the so-called " dan- gerous half " of the cyclone. To apply these rules necessitates the locating of the center of the cyclone, and this can be done from the wind direction. When cyclones were first studied, it was thought that the winds were truly L 0catin the circular, that is, followed the isobars ; and if this were the storm case, the application of Buys Ballot's famous law : " Stand centen with your back to the wind and the low pressure will be on your left hand," would give at once the direction of the storm center. But it is now known that the wind direction makes an angle with the isobars, and for this reason the storm center can be more readily located by means of a diagram. Let A (Fig. 113) represent the wind direction, then B must be the position of the isobaric line and the line C, at right angles to B, must give the direction toward the center of the cyclone. The direction of the center is thus known, and from observations of the sea and sky a close estimation of its distance can be made. The ISOBARIC, direction of the cirrus bands across the sky, the WIND location of the nimbus cloud on the horizon, and the direction of motion of the ocean waves are Center "of also guides as to the direction in which the Cyclone, storm center is located. The so-called " dangerous half " of a tropical cyclone is the north and northeast portion when the cyclone is moving northwest and the south and southeast portion when it is moving northeast. The Danger- All this applies to the northern hemisphere. These portions ous half of a are considered the more dangerous because the wind ve- cycc locities are somewhat larger. The reason for this is because in these portions the velocity of the wind about the center is combined with (added to) the velocity of the permanent winds in the region through which the cyclone is moving. On the other side of the storm, these two velocities are opposed. (See figure 115.) THE REGIONS AND TIME OF OCCURRENCE 273. Regions of occurrence. There are five regions of the earth where tropical cyclones occur, and these are shown in Fig. 114. 274 METEOROLOGY They are : (1) The West Indies, the Gulf of Mexico, and the coast of The five Florida ; (2) The China sea, Philippine Islands, and Japan ; regions of (3) each side of India in the Bay of Bengal and in the Arabian Sea; (4) east of Madagascar near the islands of Mauritius and Reunion ; (5) east of Australia, near Samoa. The WESTERN HEMISPHERE EASTERN HEMISPHERE FIG. 114. The Five Regions of Occurrence of Tropical Cyclones. tropical cyclones always occur on the west side of an ocean. This is true of the north Atlantic, the north Pacific, the south Pacific, and the south Indian oceans. Tropical cyclones never originate on land, and if they run ashore, they weaken, on the west lose their destructive violence, and are soon transformed ocean never * n * extratropical cyclones. A range of mountains 3000 on land or feet high is often sufficient to completely destroy a tropical Atlantic " 1 cyclone. It is also a very significant fact that tropical cyclones never occur in the south Atlantic Ocean. 274. Tracks of tropical cyclones. The tropical cyclones originate in the doldrums, not directly at^the equator, but from 8 to 12 from it on either side. Those in the northern hemisphere move the track of northwest through the trade wind belt-ivith a velocity of c clone* 1 ^ rom 6 to 12 mi ^ es an k ur< They curve to the right in about 30 north latitude, moving first due north and then northeast through the region of the prevailing westerlies. They grow somewhat larger, and the velocity of motion increases to 20, 30, or even 40 miles per hour. The path has somewhat the form of a parabola with THE SECONDARY CIRCULATION OF THE ATMOSPHERE 275 its vertex in 20 north latitude. In the southern hemisphere they move southwest at first, and then recurve in about 25 south latitude and move southeast through the region of the prevailing westerlies. Fig- ure 115 shows the direction of rotation of these storms, the dangerous half (shaded), and the form of the path for both hemispheres. S X* PREVAILING WESTERLIES X / NORTHEAST TRADE WINDS EQUATOR FIG. 115. The Direction of Rotation, Dangerous Half, and Path of Tropical Cyclones. The characteristics of the paths followed by the West Indies hurri- canes are well shown in Fig. 116, which gives the paths The tracks followed by all recorded hurricanes during September ^ies hurrL from 1878 to 1900. The mean track is also indicated. canes. 276 METEOROLOGY The tracks of the cy- In the case of the tropical cyclones which occur each side of I n ^ a m the Bay f Bengal and in the Arabian Sea, this characteristic parabolic path is lacking. Here the course fU we d is more irregular, and the paths are much shorter and more nearly straight lines. irregular FIG. 116. The Hurricanes of the West Indies during September, from 1878 to 1900. (After GARRIOTT, U. S. Weather Bureau.) 275. Frequency at different times of year. The frequency of occurrence at different times of year is exhibited in the accompanying table for the five regions of the earth where tropical cy- clones occur. This table gives the percentage frequency for the different months, the name of the investigator who collected the statistics, and in some cases the period cov- ered. The total number of these storms is also added, but these numbers are not comparable for the different regions, because the length of time is not the same and there was no common standard of intensity for determining how violent a storm must have been to be considered a tropical cyclone. It will be seen that in connection with the cyclones in the West Indies, and in the China Sea and Philippines, the greatest number occurs dur- The fre- quence at different times a year. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 277 ing the months of July, August, September, and October; that is, in the late summer.- The maxi _ In the case of the south mum num- Pacific and south In- dian oceans, the largest late i i summer. number occurs during December, January, February, and March, again in the late summer of that hemisphere. In the case of the cyclones of the Bay of Bengal and the they occur Arabian Sea there are between the monsoons. two maximums ; in April, May, and June, and again in October and November. Here they occur during the light, baffling breezes which prevail between the periods of the steadily blowing monsoons. THE ORIGIN OF TROPICAL CYCLONES 276. The convectional theory. The distribution of the meteoro- logical elements about . The convec- a tropical Cyclone has tional theory been fullv considered, wiiibepre- , . , " . f , sented from and the facts of ob- the deduc- tive stand- point. servation concerning the regions of occur- rence, the path followed, and the frequency at different times of year have been stated. It now remains to discuss the origin and development of these storms, and to account for all these facts of observation. The so-called con- vectional theory is the one usually 1010 ^00 O5CO 00 y CO Si COlO 1 (N it (NCO (NiO 00 - 1 1 TH COO o O (MCO TH C5CO i 1 1 I COrH o 1 | COCO OC5 i-H 00 Oi rH TH rH o 1 (NO (N(N 00 CO TH 10 TH CO COO Oi-H (NT* 00 00 ^ TH d rO W) a< M ce o> a? d QQ'S, |j f W !i d T Pacific Knipping t/2 t/3 8- si 3 d 1 i-H an d the angle made with the isobars larger than in distribution summer. The temperature difference between the west !?- me r- hat an d east side of the storm is usually much larger in winter ainerent in . summer than in summer. The only important difference between Sinter 1 winter and summer is in the nimbus cloud area. In winter, there is large, continuous nimbus cloud area mostly in the southeast quadrant, and precipitation falls steadily over this area. In summer, this nimbus cloud area is usually lacking. In its place are cirrus, cirro-stratus, and cirro-cumulus clouds, and, whenever local convection takes place, a thundershower is the result. Thus in winter over this area a continuous nimbus cloud with steadily falling precipitation is expected, while in summer cirriform clouds, sultry weather, and thundershowers are to be expected. The reason for this is the higher temperatures of summer, as the increased capacity of the air for water vapor permits it to be held as invisible water vapor, and it is not condensed into a nimbus cloud until convection takes place. The lows of spring and autumn are sometimes of one type and sometimes of 1 See Meteorologische Zeitschrift, Juli, 1903, pp.307, for the distribution about "Lows," at St. Louis, U.S.A., and for references to other articles. See also Annals of Harvard Col- lege Observatory, Vol. XXX. THE SECONDARY CIRCULATION OP THE ATMOSPHERE 287 the other. The winter type may also occur in summer, but the oppo- site is exceedingly rare. Each individual extratropical cyclone or low has its own peculiar characteristics and distribution of the meteorological elements, and it may differ much from the normal or type form. In order to gain anything like a full understanding of these storms, ^duafiow many individual lows as portrayed on the daily weather may differ maps must be critically studied. It would probably be better to defer beginning this study until the first part of Chapter VIII has been reached. At least twenty-five separate storms well distributed throughout the year should be carefully considered. In each case the actual distribution of the elements should be exactly noted and compared with the type form. If there is any discrepancy, it should be explained. There are three factors which cause a departure from the type form : (1) The surface topog- raphy of the country ; such facts as the location of bodies which cause of water, nearness to the ocean, the height, position, and ancies 50 "*" direction of mountain chains, etc. (2) The meteorological condition of the country ; such facts as to whether the ground is snow covered or not, whether the temperature and moisture are large or not, etc. (3) The neighboring meteorological formations ; that is, the posi- tion of the various lows and their antitheses, the highs, with reference to the low in question. All departures from a normal or type form in connection with the distribution of the meteorological elements can be explained on the basis of these three factors. 287. There is one modification of the distribution of the meteoro- logical elements aboulXan extratropical cyclone as just given, which is so marked, of so much practical importance, and of^such Manylows general occurrence that it is worthy of special considera- have a wind tion. In the southern quadrants of a low, a so-called wind s shift line often develops. This occurs in connection with about one low in seven in this country, and is much more common in Europe. The distribution of the meteorological elements about a low with a well- developed typical wind shift line is shown in Fig. 119. The oval isobars have a projection in the southern quadrants which appears like a pocket or a V-shaped bulge pointing usually towards the southwest. On the east side of this line, the wind continues from the south with small velocity ; the precipitation area usually disappears, and the nimbus cloud is replaced by the cirriform transition clouds; the temperature and moisture continue high. On the west side of this line, 288 METEOROLOGY there is a narrow belt of nimbus cloud with strong northwest winds. The temperature lines are also packed close together in this precipita- tion area. Apart from this modification in the southern quadrants, the distribution of the elements is the same as before. The weather map 1000 MILES FIG. 119. Distribution of the Meteorological Elements about a Law with a Typical Winci Shift Line. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 289 for 8 A.M., March 3, 1904, which is reproduced as Chart XXIX, shows an extratropical cyclone with a pronounced wind shift line. The weather maps for Jan. 23, 1906 ; Jan. 4, 1907 ; Jan. 21, 1908 ; May 6, 1909; Nov. 17, 1909; Nov. 23, 1909; Aug. 30, 1910, also show lows with typical well-marked wind shift lines. OF. O'F. 15 F. 15 F 30' F/ 45 F./ EARTH'S SURFACE 1 MILE CIRRUS CLOUD MOTION 4 MILES FIG. 120. The Structure of an Extratropical Cyclone at Various Levels. 288. The distribution of the meteorological elements about an extra- tropical cyclone, its characteristics and structure at the earth's surface, have been fully considered. It now remains, The struc _ in order to complete our knowledge of this formation, to ture of an consider its structure at various levels above the earth's c?cydane surface. This information has been gained from observa- at various tions made on mountains and by means of balloons and eves ' kites and by noting the direction and velocity of motion of clouds. The four small diagrams given as Fig. 120 will help to make clear the 290 METEOROLOGY structure of this storm at various levels. The first diagram shows again the well-known distribution of pressure and temperature about an extratropical cyclone at the earth's surface. The central pressure has been assumed to be 29.6 inches, and the temperature difference between the northwest and southeast parts has been considered 45 F. This difference is rather large, but not at all unusual. The second dia- gram has been obtained from the first by computation and shows the isobaric lines at the height of one mile above the earth's surface. It has been assumed in making the computation that the vertical decrease in temperature was everywhere the same and equal to the average vertical temperature gradient ; namely, 1 F. for 300 feet. Since the air in the southeast quadrant is warm, and thus light, the drop in pressure due to one mile of elevation will be less there than in the northeast quadrant, where the air is cold and heavy. As a result, the longer axis of the oval isobars now has a northwest-southeast direction, which is the direction of the greatest temperature contrast. The oval has also opened up so that the northwest portion is missing, and the center has also been dis- placed towards the northwest. This displacement of the center has been a matter of frequent observation. On the summit of Mt. Washington (elevation 6279 feet) the storm center passes on the average about three hours later than at the base of the mountain. This corresponds to a displacement of between one and two hundred miles in this small height. The third diagram was determined from the first in the same way as the second, and shows the isobaric lines at a height of about four miles. The oval form has now practically disappeared, and only a certain amount of distortion in the lines is now apparent. With increasing height this distortion grows less and less, and the isobaric lines become more nearly straight, running from west to east. At the earth's surface the air motion is a whirl about the area of low pressure, blowing spirally inward, making an angle with the oval iso- baric lines and turning counterclockwise in the northern hemisphere. With increasing elevation, the isobars become circular, even before the level of the clouds is reached, and the air moves nearly tangential to the isobars, rising slowly. By the time a height of one mile has been reached, as has been already shown, the isobars have again become oval, but with the axis now northwest-southeast, the center has been displaced, and the northwest portion of the storm is missing. An atmospheric whirl can now barely be said to exist, as the northwest portion is missing. The air motion approximates the form of two currents flowing side by side in opposite directions. On the right there is a warm current coming THE SECONDARY CIRCULATION OF THE ATMOSPHERE 291 from the southeast, and on the left a cold current coming from the northwest. With increasing height the isobars become less and less oval, and finally become distorted lines which gradually grow straighter and straighter, as was shown in the diagrams. While this is taking place, the whirling nature of the air motion becomes less and less apparent, and the two currents flowing in opposite directions become more pronounced. There is also a motion outward from the center, as the rising air must be injected into the general circulation aloft. With increasing elevation these air currents grow weaker, and by the time the level of the cirrus clouds has been reached (five miles on) only a slight distortion of the west to east motion of the outer layer of the atmosphere is apparent. The fourth diagram gives the motion of the cirrus clouds about an area of low pressure as observed at the Blue Hill Observatory near Boston, under the direction of Mr. A. L. Rotch. 1 The distortion caused by the currents is only slightly apparent. An extratropical cyclone is thus a formation of the lower part of the atmosphere. It is often stated that their influence ceases at five miles, but, considering the fact that some cross the Rocky Mountains, a greater height must be inferred in some cases at least. 289. A comparison of the tropical and extratropical cyclones. A tropical cyclone and an extratropical cyclone when considered super- ficially may seem very similar, but when the details in their structure are considered, they often stand in sharp contrast, ficiaiiy con-' Both are areas of low pressure, with spirally inflowing winds sidere . d thg y turning counterclockwise in the northern hemisphere. Both are attended by immense cloud areas and precipitation. They differ, however, in many ways : (1) In the case of an extratropical cyclone, the isobars are oval, and they lie closer together in the south- They differ west quadrant, while in the case of a tropical cyclone, the in many isobars are more nearly circular and equidistant from each other. The central pressure is also lower in the case of a tropical cyclone. (2) Wind velocities are much higher, and the angle made with the isobars less, in the case of a tropical cyclone. (3) In the case of a tropical cyclone, the temperature and moisture are the same in all quadrants ; that is, they are symmetrical with respect to the center. In the case of an extratropical cyclone, the values of temperature and moisture are much higher in the eastern part than in the western. (4) In the case of a tropical cyclone, the cirrus cloud extends out in all directions from 1 See Annals of the Harvard College Observatory, Vol. XXX., also Monthly Weather Review, March, 1907. 292 METEOROLOGY the center, and the transition clouds are the same in every quadrant. In the case of an extratropical cyclone, the cirrus cloud is found only on the east side, and the transition clouds are entirely different on the east and west sides. (5) The rain area is concentric with the low in the case of the tropical cyclone, while it lies mostly in the southeast quad- rant in the case of an extratropical cyclone. All these contrasts are well brought out if the two diagrams (Figs. 110 and 118) which represent the distribution of the meteorological elements about a tropical and an extratropical cyclone are carefully compared. The great difference in the two storms is noticed in connection with the central calm or eye. A tropical cyclone usually has a calm central area where the pressure is lowest, the clouds break through, difference is ^ ne temperature rises, and the relative humidity is much the presence less. An .extratropical cyclone probably never has such a centrafeye. cent ral calm or eye. Some have thought that if the extra- tropical cyclone is very violent, an approximation to a calm central eye exists. This is probably not the case, however, and the suggestion of an eye was given by the fact that the low in question had a pronounced wind shift line. If such a low passes a little to the north of an observer, the precipitation may stop, the clouds change from nimbus to a cirriform variety, the temperature and moisture may remain high, and the wind may continue to blow gently from the south. Soon after it may be again raining, with the wind blowing sharply from the north- west and the temperature dropping rapidly. This would give the impression of an eye, but it is really only the passage of a wind shift line. How this sequence of changes is possible can be readily seen from the diagram (Fig. 119), which gives the distribution of the meteoro- logical elements about a low with a typical wind shift line. 290. Description of the approach and passage of an extratropical cyclone. The first signs of the approach of an extratropical cyclone or low are usually these : The wind begins to blow gently quence'of from the east, the pressure decreases slightly, cirrus clouds weather make their appearance, and the temperature and moisture tothTap- 116 begin to increase. Next the barometer drops a little more, proach and the wind direction changes to the southeast and the velocity alow? 6 becomes a little greater, the cirrus clouds thicken to cirro- stratus or cirro-cumulus, and the temperature and moisture continue to rise. In the winter time, as the popular phrase goes, the weather has begun to moderate. In the summer time, it is the beginning of a period of sultriness. The pressure now drops still more, THE SECONDARY CIRCULATION OF THE ATMOSPHERE 293 the wind veers a little and blows harder, the cirriform clouds go through their regular transition into nimbus, and the temperature and moisture are high and increasing. Next comes a period of rain or snow, with the barometer still dropping and finally reaching its lowest. The wind, meantime, has slackened somewhat and veered a little and is perhaps now blowing from the south or southwest. The temperature and mois- ture still continue high. The wind now veers rather quickly into the southwest, then west, and finally northwest. The barometer begins to rise, the precipitation grows less, and the temperature and moisture decrease. Soon the nimbus clouds break up into fracto-nimbus, perhaps disclosing the upper cloud area. The fracto-nimbus then changes into strato-cumulus, and finally cumulus or fracto-cumulus, with a clear sky at night. In the meantime, the wind blows from the northwest with increasing velocity, the barometer is rising, and the temperature drops rapidly. The air also becomes much dryer. In the summer, the dry, cool, northwest wind has replaced the oppressive sultriness of a few days before. In the winter, the thaw or warm spell has been replaced by a cold snap. How often this sequence of weather changes has been noted by every one without realizing that it was but the approach and passage of an extratropical cyclone. This series of weather changes requires from two to four days in winter for its completion and nearly double the time in summer. The series of weather changes which has just been described applies to a nearly central passage of a low. If it passes to the north or south of an observer, the sequence would be different and depend veering and upon the distance of the center of the low. Just what the backing sequence would be, can be determined at once by consider- ing the diagram given as Fig. 118, movable and moving it from left to right centrally over or above or below a given point. It is worthy of mention that if the low passes north of the observer, the wind veers, while if it passes south of the observer, the wind backs. Since most lows pass north of the United States, veering winds are the rule and backing the unusual. If the low has a wind shift line, the sequence of weather changes can be determined in the same way by using the dia- gram given as Fig. 119. In the early part of the last century, when this sequence of weather changes was first noticed, it was explained, as was first done The cause by Dave, by assuming that there were two currents of air, of weather one warm and moisture laden from the south and the other cold and dry from the north. These were supposed to flow above each 294 METEOROLOGY other in the tropics and side by side in the temperate latitudes. It was to the ceaseless struggle of these opposing air currents and to the temporary victory of first one, then the other, that the weather changes were ascribed. This gave rise to the popular expression that the wind makes the weather. It is now known that the mechanism back of the ceaseless changes in our weather is the approach and passage of extra- tropical cyclones or lows. ANTICYCLONES 291. Just as two valleys are an impossibility without a hill or ridge of land between them, in the same way two areas of low pressure can- not exist without a region of high pressure between. These clones'or areas of high pressure stand in many ways in sharp con- highs are trast with the lows, and have many characteristics which of lows. 08 are exac tly the opposite. For this reason they are called anticyclones, or, sometimes, simply highs. An anticyclone or high can be best defined and described by stating its chief characteristics. It is an area of high barometric pressure with Outline spirally outflowing winds turning clockwise in the northern description, hemisphere and the opposite in the southern; the wind velocity is usually very moderate, and calms are frequent ; but few clouds are to be seen, and precipitation is usually lacking ; the changes in temperature and moisture are large and well marked. The whole formation is from several hundred to several thousand miles in diameter, and moves with moderate velocity, sometimes loitering for a day or two, from some westerly to some easterly quarter. 292. The distribution of the meteorological elements about a typical anticyclone is shown in Fig. 121. The isobars are often irregular in form, and in the center there may be several highs instead of a single peak of pressure. The more usual form, however, is the oval, and the longer axis generally ex tends northeast-southwest, north and south, or north- butionofthe west-southeast. The central pressure averages about 30.6 ^sdliemeSs mcnes > although this varies all the way from a little over about an 30 to more than 31 inches. or high 10 * The winds blow spirally outward from the center, turning clockwise in the northern hemisphere. The wind direction makes an angle with the isobars which is least in the northeast portion, where it averages about 20 or 30 and is greatest in the southwest por- tion, where the average is from 60 to 70. The wind velocity is THE SECONDARY CIRCULATION OF THE ATMOSPHERE 295 very moderate and decreases towards the center, where calms are fre- quent, particularly at night. There is a decided drop in temperature in the northeast portion, where the winds are from the north, and a decided rise in temperature \ \ \ FIG. 121. The Distribution of the Meteorological Elements about a Typical "^ Anticyclone or High. in the southwest and west portions, where the~winds are from the south. The moisture changes are not shown in the diagram, but they follow the temperature. In the northeast portion, both the absolute and the relative humidity are low, while in the southwest portion both are high. 296 METEOROLOGY On the east side the strato-cumulus clouds of the departing low may be still tisible. In the middle portion, convection-caused cumulus or fracto-cumulus clouds may be found These usually disappear at night and are most numerous on the east side of the center. In the winter, a thin alto-stratus cloud due to radiation may be formed during the night and be visible in the early morning. On the west side, the cirrus and cirro-stratus of the coming low may have already made their appear- ance. Precipitation very seldom falls in connection with an area of high pressure. Occasionally a cumulus cloud may become sufficiently overgrown to yield a snow flurry in winter or a few raindrops in summer. " The diameter of the whole formation averages perhaps 2000 miles and varies from several hundred to several thousand miles. The weather map for 8 A.M., April 23, 1906, which is given as Chart XXX, shows a very typical anticyclone or high with its center over Illinois. The weather maps for May 1, 1905 ; Oct. 28, 1907 ; Jan. 3, 1908 ; Jan. 30, 1909 ; March 22, 1909 ; April 10, 1909 ; June 18, 1909 ; Sept. 26, 1909; Oct. 29, 1909; March 14, 1910; Dec. 21, 1910, also show typical highs at various times of year. 293. It must be remembered that the distribution of the meteorologi- cal elements about an anticyclone or high which has just been so fully stated is the typical one, and applies to the eastern part of the United States. This typical distribution is slightly is different different in different parts of the world and at different times countries 11 * ^ vear > an( ^ the ac tual distribution in the case of any indi- and at dif- vidual high may depart widely from the type form. 11 ^ n Europe, the direction of the longer axis of the oval is more northeast-southwest than in the United States, and the quadrant in which the wind direction makes the smallest angle with the isobars is also different. In winter, the highs are larger and the central pressure is higher than in summer. Each individual high has its own peculiar characteristics, and distri- bution of the meteorological elements, and it may depart much from the type form. The only way to gain detailed knowledge about highs differ this formation is to study carefully many individual cases from the as they occur on the daily weather maps. In each case, note carefully the actual distribution of the elements, com- pare this with the type form, and then explain all discrepancies. The three causes of departure from type are the same as for lows ; namely, THE SECONDARY CIRCULATION OF THE ATMOSPHERE 297 X (1) the surface topography of the country, (2) the meteorological condi- tion of the country, (3) the neighboring meteorological formations. 294. The characteristics of a high at the earth's surface have now been fully stated, and it remains to consider its structure at various levels above the earth's surface. The change in form with elevation is very similar to that in the case of a low, and for ^ Jj^~ that reason it will only be sketched in outline here. At the anticyclone earth's surface the air motion is a whirl about an area of fevdJ 10118 high pressure. The air moves spirally outward, making an angle with the oval isobaric lines, and turning clockwise in the northern hemisphere. Since the northern portion is much colder than the south- ern, the pressure decrease with elevation will be greater in the northern portion than in the southern. By the time a height of one mile has been reached, the center has been displaced towards the cold area (that is, towards the north), the oval isobars have opened out, and the northern portion of the formation is lacking. An atmos- pheric whirl can now be hardly said to exist, and the air motion approximates the form of two currents flowing side by side in opposite directions. On the right there is a cold current coming from the north, and on the left a warm current* coming from the south. With increasing elevation, the oval isobars open out more and more, and finally become distorted lines which tend to become straighter. While this is taking place, the whirling nature of the air motion be- comes less apparent, and the two currents, flowing in opposite directions, become more pronounced. There is also a component of motion in towards the center, for the air which moves spirally outward at the earth's surface is replaced by a gentle descending air current which must be supplied from above. With ever increasing elevation, the isobars become straighter and straighter and the two air currents become less and less pronounced, until, by the time the level of the cirrus clouds have been reached (five miles on), only a slight distortion of the west to east motion of the outer layer of the atmosphere is apparent. The motion of the cirrus clouds about an area of high pressure as observed at Blue Hill is given in Fig. 122. This distortion caused by the air currents beneath is only slightly apparent. Thus the high, like the low, is a formation of the lower atmosphere. X FIG. 122. The Motion of the Cirrus Clouds about an Area of High Pressure. 298 METEOROLOGY 295. The sequence of weather changes brought about by the approach and passage of an anticyclone or high has been experienced many times by every one without knowing the cause of the changes. A quence'of ^ ow nas P r bably just passed; its rain area has gone by; weather the cloud form has changed to strato-cumulus ; the wind tothTap^ 116 nas g ne to the northwest; the temperature has dropped proachand markedly; the air has become much dryer; the barometer hfgh^ 6 ( * i g r i sm g- All this may be expressed meteorologically by saying that a low has passed and the weather con- trol is being transferred to a coming high. It may be expressed popularly by saying that it has cleared off and several days of good weather are in store. The wind continues in the northwest and de- creases in velocity. In fact the nights are almost perfectly calm and the wind blows only during the day. The barometric pressure con- tinues to rise, usually holding about steady during the daytime and rising quite a little during the night and early morning. The strato- cumulus clouds become cumulus or fracto-cumulus and decrease in number and size. Since these are convection-formed, they disappear at night and are visible only during the daytime. The temperature drops fairly low at night and rises rapidly during the day so that the daily range is excessive. Soon the center of the high is reached. There is an almost perfect calm both day and night. The pressure is at its highest. The sky is cloudless or at most shows a few cumulus clouds. The air is dry and the daily range in temperature continues large. Then the barometer begins to fall. The wind goes to the northeast or east and, blowing at first very gently, soon increases in velocity. The temperature begins to rise, and soon the cirrus cloud puts in its appear- ance. The barometer continues to fall; the wind blows harder and perhaps shifts to the southeast ; the temperature rises rapidly ; the air becomes more moist; the cirrus clouds thicken into cirro-stratus or cirro-cumulus ; and the weather control passes from the departing high to the coming low. This series of weather changes requires from two to four days in winter for its completion and nearly double the time in summer. The series of weather changes which has just been described applies to the nearly central passage of a high. If it passes to the north or south of an observer, the sequence would be somewhat different. Just what the sequence would be can be determined by moving the diagram given as Fig. 121 from left to right centrally over or above or below a given point. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 299 In summer, if a high loiters for several days, there is a series of days with rather high temperature during the day but with cool nights. If it loiters unduly long, a spell of dry weather is the result. In winter the low temperatures, sometimes many degrees below zero, occur during the first two days of a high. In autumn the morning fogs occur usually during anticyclonic weather, because it is then that the daily range of temperature is so large. In spring the destructive frosts usually occur just after the weather control has passed from a low to a strong coming high. THE TRACKS AND VELOCITY OF MOTION OF EXTRATROPICAL CYCLONES 296. Tracks in the northern hemisphere. If an observer far above the earth could look down upon the whole northern hemisphere, he would see a ceaseless procession of lows between the thirtieth and eightieth parallels of latitude, moving eastward and encir- eastward cling the poles. Their big cloud areas would gleam white and in the reflected sunlight. In Fig. 123 (after Loomis) the tracks of a number of lows are shown. Taking the northern hemisphere as a whole, the tracks followed by lows may be characterized by stating that they move eastward and slightly poleward. The reason for this eastward motion is because this is the prevailing direction of both the surface winds and the fast-moving upper air currents in these latitudes. The lows thus drift with the general wind system. Lows may originate anywhere. A good part of those which visit Europe originate over the Atlantic Ocean. In the United Theymay States, a favorite place of origin is the Mississippi Valley originate just east of the Rocky Mountains. The length of the path an y where - followed varies from a few hundred miles to, in a few rare cases, more than half of the circumference of the globe. The tropical cyclones come up from the tropics and join the extra- tropical ones in two places : over the West Indies and over the Philip- pines and Japan. As soon as a tropical cyclone enters the . extratropical region it loses its violence; the central eye cyclones disappears; the rain area becomes excentric; its velocity of motion is larger, and it soon assumes all the characteristics of an extratropical cyclone. In fact, in a few cases, it has been known to merge with an extratropical cyclone. 297. Tracks across Europe. If a smaller portion of the world is considered, as for example Europe, it is found that not all areas are 300 METEOROLOGY covered by the same number of lows. The lows seem to prefer to The Van travel along certain rather definite paths; and if the actual Bebber sys- paths followed by the lows for a number of years are general- tra'k&f f r * ze< ^ or summar i ze d, ^ * s found that a more or less definite Europe. system of storm tracks is the result. This has been done Fio. 123. The Tracks of a Number of Lows in the Northern Hemisphere. (After LOOMIS.) THE SECONDARY CIRCULATION OF THE ATMOSPHERE 301 Zugslrasse der Minima,. with particular skill by Van Bebber, the head of the German meteoro- logical service, and the resulting storm tracks for Europe are shown in Fig. 124. The width of the track is pro- portional to the fre- quence with which it is followed by lows. Track I is used more than any other, and chiefly in the autumn and winter. It is but little frequented in spring. Tracks II, III, and IV are traversed at all times of the year. Tracks II and III are used a little more during the cold portion of the year and IV in summer and autumn. Track Va is almost never traversed in summer, all times of year. 298. Tracks across the United States. The tracks of lows across the United States have been generalized by Bigelow, Russell, and Van Cleef and the results of all three investigators are presented here. In Fig. 125 the Bigelow system of tracks is shown. * t ^** { The main track follows the northern boundary of the United tracks States across the Great Lakes and out the St. Lawrence Valley, *.<* the This main track is joined by three others coming up from states, the south. One comes up from Colorado and Utah and joins it near Lake Superior. Another comes up from Texas and joins it near Lake Huron. The third comes up the Atlantic coast and joins it near Nova Scotia. There is a second main track across The Bige _ Texas and the Gulf States to the Atlantic coast, where it low system either turns northward or goes out over the ocean. The oftracks - broken lines show the average daily movement. This system of tracks is rather too highly generalized. It has been made simple by throwing out .too many tracks as erratic or exceptional. The percentage of lows which will follow these tracks has thus been reduced. FIG. 124. Storm Tracks for Europe. (From BERBER'S Lehrbuch der Meteorologie.) Tracks V6, Vc, and Vd are frequented at 302 METEOROLOGY FIG. 125. The Bigelow System of Storm Tracks across the United States. (The Tracks of Highs are also shown.) (U. S. Weather Bureau.) FIG. 126. The Russell System of Tracks across the United States. (From RUSSELL'S Meteorology.) THE SECONDARY CIRCULATION OF THE ATMOSPHERE 303 299. In Fig. 126 the Russell system of tracks is shown. Since there are eleven tracks, this system is not as highly generalized as The R ussel i the first one, and a larger per cent of the lows will follow system of some one of these tracks. tracks. 300. In Fig. 127 the Van Cleef system of tracks is shown. This is the least generalized of the three and thus accounts for the paths followed by the largest number of lows. In preparing The Van this diagram, the tracks of the lows from 1896 to 1905, Cleef system 1160 in number, were used. Of these 1160 only 57 oftracks - were erratic and did not follow some one of the tracks. The fre- quency with which any track is traversed is indicated by its width. FIG. 127. The Van Cleef System of Storm Tracks across the United States. (Twenty- seven Tracks are represented.) (U. S. Weather Bureau.) 301. If a low does not originate in the United States, it enters it from the northwest, west, or south. Those which enter north of the middle of the country have a tendency, in crossing the Missis- A smnmary sippi Valley, to move towards the south and then recurve of the three towards the northeast. In approaching the Atlantic coast, syst< all lows move towards the northeast. Although lows may originate within the country or enter it from various directions, they nearly all 304 METEOROLOGY W r OCOO^COO>0^ "- S TH f> ^^ S Is 8 J ? I I fc 2 > i I i-H TH T i i 1 rH si fc Q 2 S ^ t , "3 II g M 1 1 COCOOCOl>COCOQOI>I>t^T-l i-i CO g o h- 1 rH rH si 1 fi - COQO ^^^2c^ ss B EHH leave it by way of the St. Lawrence Valley and Newfoundland. The phrase " all roads lead to Rome " might be paraphrased in connec- tion with lows into "all storm tracks lead to New England." It must not be sup- posed that all lows fol- low one or the other of these many tracks with- out exception. A low may loiter in one place for a day or two, or turn sharply aside in one direction or another, or move erratically in almost any direction. These tracks simply represent normal be- havior. In order to gain familiarity with the paths followed by lows, many individual cases, as indicated on the daily weather maps, must be studied. 302. Velocity of mo- tion. Several impor- tant facts in The number and velocity Connection of motion w j th the of lows. velocity of motion of lows are brought out by the accompanying table, which gives the number and velocity of motion THE SECONDARY CIRCULATION OF THE ATMOSPHERE 305 of lows for the ten years (1882 to 1891), classified according t.o the Russell system of tracks. The average velocity for all tracks for all times of year is 31.7 miles per hour. The winter velocity is, however, nearly twice as large as the summer (January, 40; June and July, 24.9). The velocity of motion does not differ much in the different tracks and is greatest for the tracks which have the greatest curvature, the velocity Some tracks are much more frequented than others ; com- 01011 f pare, for example, I or V with IV or XI. Some tracks are also more frequented at certain times of year ; I, for example, with 3 in April and 25 in December, or VIII with 1 in July and 13 during both December and February. It is also a very important fact that the number of lows in winter is more than double the number in summer ; compare, for example, December, 71 with June, 33. Lows move faster across the United States than any other country. One authority, which gives the average velocity for the velocity in United States as 26 miles per hour, finds for Japan a veloc- various ity of 24; Russia, 21; the north Atlantic, 18; and Europe, countries - 16. That lows move faster in winter than in summer is found to be true for all parts of the world. It must not be supposed that the velocity of motion of individual lows always conforms to the normal. A low may loiter for a day or two, drifting, perhaps, but one or two hundred miles in that individual time. Another low may rush across the country, covering a lows are distance of sixteen or seventeen hundred miles in a single day. As has been so often stated, the only way to learn the character- istics of lows and highs, their tracks and velocity of motion, is to study them as they are portrayed from day to day on the weather maps. In making this study, the normal behavior should be held in mind, com- parisons made, and discrepancies explained. ' THE TRACKS AND VELOCITY OF MOTION OF ANTICYCLONES 303. Anticyclones or highs are more erratic than lows, both as regards the track followed and the velocity of motion. In Fig. 125, the Bigelow system of tracks for highs was shown. Highs usually enter the country from the extreme low system northwest or over California. To those entering from the ? hg Cks for extreme northwest, two courses are open. They may move eastward and slightly southward along the northern boundary of the 306 METEOROLOGY United States until the Atlantic coast is reached, where they turn to- wards the northeast and proceed in the direction of Iceland. The highs entering from the northwest may also move southeast over Kansas and the Gulf states to the Atlantic coast near Florida. The highs coming in over California usually move southeast and join this track just south of Kansas. After leaving the Atlantic coast, near Florida, the highs usually continue to move towards the southeast in the direction of Bermuda. Paths classified 741 Paths miscellaneous 98 Paths incomplete Total paths FIG. 128. The Van Cleef System of Tracks for Highs across the United States. (U. S. Weather Bureau.) The Van Cleef system of tracks for highs is shown in figure 128. This The Van system of tracks is based upon the highs from 1896 to 1905, Cleef sys- 928 in number. Only 98 of the 928 were eratic and did not tracks for follow one of the indicated tracks. The width of the track highs. i s proportional to the frequency with which it is traversed. 304. The velocity of motion of highs averages somewhat less than that of lows, but this is because a large high often remains almost sta- The veioc- tionary for a day or two and perhaps a week, and this brings ity of mo- down the average velocity of motion. Small highs, par- g s * ticularly those wedged in between lows, usually move with about the same velocity as the lows themselves. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 307 The following table, compiled by C. F. von Herrmann, is taken from the Bulletin of the Mount Weather Observatory, Vol. II, part 4, p. 196. It c6ntains the average, highest, and least velocity of motion of both highs and lows for the various months and for the year. The data summarized cover the twenty-seven years from 1878 to 1904. For the purpose of comparison, the values for lows de- termined by Loomis for the period 1872-1884 are also added. The truth of the generalizations which have been stated for the velocity of motion of both highs and lows is at once apparent from this table. 305. The four weather maps for 8 A.M., Jan. 30, Jan. 31, Feb. 1, and Feb. 2, 1908, which are Four illus- given as Charts XXXI- trative XXXIV, will show the actual distribution of the meteorological elements about highs and lows in individual cases, and will also illustrate the statements which have been made concerning the tracks followed and the velocity of motion on the part of both highs and lows. 1 The weather maps for Dec. 26 to 29, 1904; Jan. 2 to 5, 1906 ; Jan 14 to 17, 1906 ; Dec. 8 to 12, 1907; Dec. 21 to 24, 1907; Dec. 28 to 31, 1907; Jan. 10 to 13, 1908 ; Feb. 3 to 7, 1908 ; Feb. 13 to 1 For other sets of weather maps to illustrate typical lows, see : Monthly Weather Review 1904, end of volume ; Climatic charts of the United States ; Climatology of the United States (Bulletin Q of the U. S. Weather Bureau). CD CO OOCO Oi CO I>COCO cqco 1>OCO OOOO t>i-< CO COOO NiOrt COt>O OiOOO (NOOO OOOO CO to i I r-I |> CO CO' O i-H CO COtO (NtOi-H -, ^ eddy theory place in three layers. In the northern hemisphere, the of the origin U pp e r layer moves from the west, the middle layer from the northwest, and the surface layer from a little south of west. These layers also have different velocities of motion, so that it is easy to think of eddies as forming where these layers meet. If a stone is dropped into a fast-flowing stream, an eddy often results. Now the air which moves from the equator poleward in the outer layer of the atmosphere must drop down in extratropical regions to commence its journey back to the equator. This dropping down of air masses might readily cause eddies. If a stream flows over a rough, irregular bed, eddies are usually formed. The surface of the continents is very rough and irregular and the air moving over it might readily have eddies formed in it. There are thus at least three ways in which the formation of eddies in the atmosphere is easily thinkable. Each eddy would develop a small amount of centrifugal force, and this would cause a slightly lower barometric pressure and a low has thus originated. Since the circula- tion of the atmosphere is more vigorous in winter than in summer, eddies would form in greater numbers in winter than in summer. This objection at once suggests itself: the direction of rotation of an eddy would be entirely a matter of chance, so that as many would rotate An objec- one wa ^ as tne otner - Thus, when fully grown, the whirl tion and its about lows should be clockwise as well as counterclockwise, answer. Observations, however, show that the whirl about a low is al- ways counterclockwise in the northern hemisphere. A partial answer can be given to this objection. Due to the earth's rotation, the deviation is always to the right, and the whirl must be counterclockwise. If an eddy turned in the other direction, it would be immediately stopped and put out of existence, so that only those eddies turning in the right direc- tion could continue to exist and grow. 309. Another way of accounting for lows is to consider them simply the antitheses or results of highs. An area of high pressure is accom- LOWS may panied by descending air currents and winds blowing spirally be caused outward at the earth's surface. Unless the air is to congest by highs. ^ ^ e earth's surface, there must be some upward escape. At some point the air rises to relieve this congestion, and thus a low is formed. If this explanation of the origin of lows is accepted, the THE SECONDARY CIRCULATION OF THE ATMOSPHERE 311 question has simply changed from how lows originate to how highs originate. 310. During the last few years another very important theory as to the origin of lows has been developed. This is the work of Bigelow of the U. S. Weather Bureau and is sometimes called the counter current theory. According to this theory, the rent theory origin of a low is to be found in two great air currents of oppo- site direction and very different temperature. It was shown, when the general wind system of the world was being considered, that if the surface of the earth were level and the same every- where, the belts of high and low pressure would be uniform The cause and would extend all around the earth and the exchange of of the coun- air between the equator and pole in the extratropical regions ter currents< would take place in three regular layers, one above the other. The surface of the earth is, however, not level, and it is a diversified surface, being part land and part water. As a result, these belts of pressure break up into peaks and depressions, that is, permanent highs and lows, which change their intensity and position between summer and winter. The prevailing westerly air currents in the outer layer of the atmosphere seem to preserve their identity and regularity in spite of this breaking up of the belts of pressure, but the two lower layers are very much con- fused and mixed, so that here the exchange of air between the equatorial and polar regions takes the form of great jets or air currents which move now equatorward, now poleward. Those from the equator are, of course, warm and moisture-laden, while those from the polar regions are relatively cold and dry. The region of strongest air currents, and thus most vigorous exchange of air between equator and pole, seems to be at an elevation of about 1.5 miles above the earth's surface. Suppose that there are two great air currents at this level, flowing side by side, but counter cur- of opposite direction and of very different temperature, the jj ents pro " one on the right being warm and flowing poleward and the one on the left being cold and flowing equatorward. Due to this dif- ference of temperature, at the earth's surface there will exist an area of low pressure with spirally inflowing winds turning counterclockwise in the northern hemisphere. The truth of these statements becomes evident when the structure of a low above the earth's surface as given in section 288 is considered. It was there shown that the atmospheric whirl which exists at the earth's surface, degenerates at a height of a little more than a mile into two counter currents, flowing in different 312 METEOROLOGY directions with very different temperatures. The converse is equally true. Two currents in the upper atmosphere necessitate a low with inflowing winds at the earth's surface. A low, then, with its spirally inflowing winds may be considered simply as the surface effect of two powerful counter currents of very different temperature in the upper atmosphere. Since the exchange of air between equator and pole is more vigorous in winter than in summer, the number and intensity of lows should be greater in winter than in summer. This theory, then, seems to meet all the observed facts, to be free from objections, and to be plausible, but not so easy to picture as some of the other theories. 311. Highs and lows are very closely related, so that the origin of anticyclones or highs must be accounted for as well as the origin of Highs may extratropical cyclones or lows. One way of explaining the be caused origin of highs is to consider them the antitheses or results of lows. The air rises in areas of low pressure. This air, ejected from lows, must collect somewhere and increase the pressure or form a high. A high, then, may originate in the congestion of the air ejected from lows. If this explanation of the high is accepted, the real question has become the question as to the origin of the lows. 312. Another way of accounting for highs is to ascribe their origin to the congestion of air currents in the outer layer of the atmosphere due to the convergence of the meridians. In the outer layer of be caused the atmosphere the air moves spirally from the equator by the con- towards the pole. All meridians (that is, north and south the poleward lines) converge towards the pole so that the area steadily moving air - decreases with increasing; latitude. As a result the air must currents. congest and finally drop or be forced down to start equator- ward in the lower layer. This congesting of the air currents may give rise to areas of high pressure. 313. A high may also be caused by radiation. If there is an area particularly free from clouds and moisture, the radiation of heat from the ground and the lower air will go on with greater rapidity. Highs may .11 i i be caused particularly at night, than in surrounding regions. As a by radia- result, the air here will become particularly cold, dense, and heavy. It will contract somewhat and air will come in aloft to take its place. As a result, an area of high pressure has formed. Bi elow's 3I4 ' Bigelow's counter current theory may also be used to counter cur- account for highs. If at the height of a mile or so there are of'th^ori^ * wo coun ^ er currents, one on the right, cold and dry, flowing of highs. equatorward and one on the left, warm and moist, flowing THE SECONDARY CIRCULATION OF THE ATMOSPHERE 313 poleward, there will exist of necessity at the earth's surface an area of high pressure with spirally outflowing winds turning clockwise in the northern hemisphere. When the structure of a high above the earth's surface was being considered, it was found that it too degenerated from an atmospheric whirl at the earth's surface to two counter currents at an elevation of a little more than a mile. A high, then, can be considered simply as the surface effect of two powerful counter currents in the upper air. Four theories as to the origin of a low have been given and four for the origin of a high. These theories can be better contrasted and compared after the methods of growth, development, and maintenance of both highs and lows have been considered. 315. Growth of an extratropical cyclone. The crucial question in connection with the growth or continued existence of an extratropical cyclone is the source of the energy. A full-grown extra- How is the tropical cyclone has an activity of millions of horse power energy and does work at this rate for days or perhaps weeks. This gamed? immense amount of energy must be gained from some source, and the method by which this energy is gained will tell the story of the growth and development of an extratropical cyclone. There are three possible sources of this energy : the latent heat liberated by condensation of The ^f ee moisture to form cloud and precipitation ; the energy of sources of motion of the general wind system of the world ; the relative e displacement of masses of cold and warm air. If the energy is gained from the condensation of moisture, then what- ever may have been the origin of a low, it will build itself up in accord- ance with the convectional theory and will be essentially a How ^ convectional formation. The method of growth has been energy from treated in full in connection with the tropical cyclone. A ne at may* small area of low pressure has originated with rising air build up a currents. The rising air cools, reaches the dew point, forms cloud and precipitation, and liberates an immense amount of latent heat. This heats the rising air, maintains it at a higher tem- perature than its surroundings, causes further rise, and thus acts like forced draft in a chimney. A more violent indraft at the earth's surface now takes place. It becomes a whirl due to the earth's rotation. Centrifugal force is developed and the central pressure becomes lower. The formation has grown, due to the latent heat, and acquired more energy. It builds up and continues until the supply of energy lessens, when it gradually dies out. There have been violent lows with very 314 METEOROLOGY little cloud and no precipitation. It would thus seem that they must have some other source of energy. Furthermore lows are most violent in winter, when the temperatures are lowest and the air is dryest and thus the supply of available energy is least. It would thus seem that the latent heat due to condensation cannot be the only source of energy for lows. When the eddy theory for the origin of lows was being worked out, it was thought that these eddies might gain their energy from the general winds of the world, just as water eddies gain their energy nughTcome ^ rom the currents in which they are formed. If this were from the the case, the energy acquired by a low would be taken from system. W ^ ne general wind system and the circulation would thereby be hindered and slowed down. It seems incredible, however, that a low five miles thick and 3000 miles wide, thus resembling a piece of paper, and of a size comparable with the great air currents them- selves, could receive all its energy from these currents, just as water eddies receive their energy from the currents which cause them. If the Bigelow counter current theory of lows is accepted, then the . energy comes from the relative displacement of masses of might come cold and warm air. These two counter currents, due to their temperature difference, produce the area of low pressure ment of and the atmospheric whirl at the earth's surface. The more warm and vigorous these currents, the greater the temperature differ- ences and thus the more vigorous the effect, namely, the low, which they produce at the earth's surface. There is nothing to prevent all three of these sources of energy from contributing to the growth of a low at the same time and sources of such is probably the case. Careful observations and many energy o f them, particularly in connection with the beginning of tribute at lows and at various levels in the atmosphere, would permit the same ^he relative importance of these three sources of energy to be determined. The last one is sufficient in itself to account for lows, the first two are not. 316. Growth of an anticyclone. In connection with the growth of There are an anticyclone or high, the source of the energy is again the three chief thing to be considered. If a high is due to the accumu- energy S for lation of air from lows or to radiation, it is gravity which the growth supplies the energy ; and either of these processes which origi- nated the high may continue indefinitely and become more vigorous and thus account for the growth or continuance of the high. v- A \ . THE SECONDARY CIRCULATION OF THE ATMOSPHERE 315 If highs are due to the congestion of poleward moving air currents, then the energy comes from the general wind system, and this process can continue an indefinite time. If highs are due to counter currents in the upper air, then the source of energy is the relative displacement of masses of warm and cold air. The high at the earth's surface is simply the effect of these currents, and the more vigorous the currents the more vigorous the high. There is nothing to prevent all four of these processes from working simultaneously to build up and develop a high. Sometimes it would seem that one or the other of the processes stands out conspicuously as the origin of a particular high, but it is impossible to say that the others are not present. Any one of the four is sufficient by itself to produce a high. 317. The characteristics which lows and highs should have and the comparison with the observed facts. The various theories as to the origin of lows and highs and the sources of the energy which causes their growth and development have been fully con- inconnec- sidered. The general features of both the lows and highs ti nwit jj have been accounted for. There now remain many details highs must in connection with the distribution of the meteorological elements about typical lows and highs, their paths, and velocity of motion in connection with which the demands of theory on the one hand and observed facts on the other must be noted and a comparison made. On the east side of an area of low pressure the wind is from some southerly quarter, and this transports warm, moisture-laden air, raising the temperature and humidity. On the west side, the wind Tfa direction is from some northerly quarter and here the air is terfstics of cold and dry. The great difference in temperature and moisture between the two sides of a low is thus due to the wind direction. The oval form of the isobars and the direction of the oval is due to the temperature difference between the two sides. With increasing altitude this northeast-southwest oval first becomes circular in form, then again oval, but with the longer axis northwest-southeast. All this change in form is due simply to the temperature differences on the two sides. The cloud and rain area is located chiefly in the southeast quadrant because here the temperature and moisture have their largest values. The cirrus cloud is formed only on the east side because the air which rises in a low is injected into the rapidly moving westerly air currents of the outer layer of the atmosphere. This air, still containing 316 METEOROLOGY considerable moisture, is carried rapidly eastward, but makes very little headway towards the west against the air current. In connection with an area of high pressure, the cloud forms on the east side are the last reminders of a departing low, while the cloud forms on the west side are the first heralds of a coming low. teristics of Only convection-caused cumulus clouds or ^radiation clouds a high are m ^he early morning are found in the central part of a typical high. The northeast portion of a high is the coldest and the southwest the warmest, due to the transportation of air by the wind. Lows and highs move in general eastward. . This is because both the surface winds and the upper air currents move from a little south of west towards the east. In other words lows and highs drift The direc- . . _. . -11 tionandve- m the general wind system. If a small area is considered, locity of as f or example the United States or Europe, the path followed lows and by any individual low or high is determined by four factors : highs are (i) ^he topography of the country, (2) the meteorological condition of the country, (3) the surrounding highs and lows, (4) the distribution of the meteorological elements about the individual low or high itself. These four factors and the motion of individual lows and highs will be considered more fully in the chapter on weather prediction. Since these four factors are for the most part constant or subject to periodic variations, it is to be expected that lows and highs will follow more or less exactly a rather definite system of tracks. Lows and highs move faster in winter than in summer. This is be- cause the general winds of the world, which carry the lows and highs, move faster in winter than in summer. THE EFFECT OF PROGRESSION ON THE DIRECTION AND VELOCITY OF CYCLONIC WINDS 318. The conception which is often held of a low (or a high) is that . the air which constitutes the formation moves bodily forward companied a over the earth's surface. A low moves usually eastward and low, the slightly poleward. In the southern quadrants of a low the ity on the air motion with respect to the center is from the southwest to south side ^ e northeast. Here, then, the motion of progression and would be the motion of revolution about the center agree and the very dif- velocities would be added. In the northern quadrants of a low the air motion with respect to the center is from the northeast to the southwest. Here the motion of progression and the THE SECONDARY CIRCULATION OF THE ATMOSPHERE 317 motion of revolution are opposed and the velocities would be sub- tracted. To make the picture more definite, suppose that a low is moving towards the northeast with a velocity of thirty miles an hour. Suppose also that the velocity of revolution is thirty miles an hour. If the air moved with the formation, there would be then a wind velocity of sixty miles an hour in the southern quadrants and a dead calm in the northern quadrants. Now, no such difference in wind velocity as this exists between the northern and southern quadrants. The truth of the matter is that the low, the formation, goes forward, but the air which constitutes it at any moment does not. A low takes in air at the front, moves it a short distance, and abandons it. The low as a formation moves forward, but the air which tion moves ; constitutes it at different stages in its journey is entirely * he *"" does not. different. This is exactly analogous to the progress of an ocean wave in deep water. Each mass of water describes a little oval path and A wave comes back nearly to where it was before, but the wave anal gy- has gone on. The wave as a formation has gone on, but the water which constitutes it is entirely different. Over the ocean, in connection with tropical cyclones, and at consider- able altitudes, the velocity of the wind is found to be greater in the southern quadrants of lows than in the northern. This means that in these cases there is a slight carrying forward of the air which constitutes the low. Over the land, however, and near the earth's surface the friction is too great to permit the actual transportation of air by the low. 319. If a low progresses as a forma- tion, it now remains to consider what actually determines the direc- . . . . (. .-, . j The factors tion of motion of the air and which FIG. 129. Diagram Illustrating the Determination of the Wind Direc- tion and Velocity at any point near a Passing Low. its velocity at any point when mine the di- a low passes. Let US Consider, ^6^? for the sake of definiteness, and its ve- , , . , Tr j -IT- locity at any the two points X and Y in point Fig. 129, at the earth's surface on opposite sides of a passing low in the United States. These points are located in the region of the prevailing westerly winds and thus at 318 METEOROLOGY each point there would be the component A, which, in direction and length, is to represent these prevailing westerly winds in which the low drifts. Of course, this component marked A is really the com- bination of two things : the exchange of air between equator and pole and the deviation due to the earth's rotation. If the earth did not rotate, the surface wind direction in extratropical regions in the northern hemisphere would be south. This is deviated to the right by the earth's rotation and becomes the prevailing westerly winds with a direction from a little south of west. Since a low is near, there is a pressure gradient towards the center, and the air tends to move along the gradient at right angles to the isobaric lines, but it is deviated to the right by the earth's rotation. This cyclonic component is indicated in the figure by B in each case, and is the combination of two things : the pressure gradient towards the center, and the deviation due to the earth's rotation. On account of friction, the air has its greatest velocity of motion about a low, not at the earth's surface, but some distance above, where the isobars are nearly circular and the direction of the air motion is nearly tangential to the isobars. Due to fluid friction this more rapid motion of the upper air would drag around the surface air to a slight extent with it. If, for example, an iron ring were suspended in a tank of water and made to' revolve rapidly, the water would have a certain amount of motion imparted to it through friction. In the same way the rapidly revolving ring of air above the earth's surface will impart a certain amount of motion to the more slowly moving air above and below. This component is indicated by C in each case. If these three components are summed up, the resultant R in each case is obtained. It will be seen that the velocity is larger at X than at Y and that the direction is more nearly tangential to the isobars at X than at Y. If the dis- tribution of the meteorological elements about a low is considered, it will be seen that all this is in exact accord with the results of observation. If other points at different levels in connection with lows and highs are considered, two more components must at times be added. In the upper part of a low the rising air is forced out and injected into the surrounding air. There is here a component directed directly from the center. In the case of the upper part of a high, air is drawn in to take the place of the descending air current. Here there is a component directed directly in towards the center. The other component to be considered is due to the forward motion of the low or high. If no air were carried along with the formation, this component would not exist, THE SECONDARY CIRCULATION OF THE ATMOSPHERE 319 but as a small amount particularly at higher levels and over the ocean is carried along with the formation, there is this last component which has the direction of the low or high. If it is desired to determine the wind direction and velocity at any level and at any point in connection with either a low or a high, it is only necessary to determine these five components in direction and relative magnitude and find their resultant. THE CORRELATION OF THE METEOROLOGICAL ELEMENTS 320. Suppose, for example, the wind direction and the fact as to whether precipitation was falling or not have been recorded for the same hour each day for many years. A table of statistics can be prepared which would show for each wind direction J^icai^ie-" the number of times precipitation had occurred with that mentsare direction. It would be found for the northeastern part of fated" 6 " the United States that precipitation occurred with the great- est frequency when the wind was southeast. East, south, and south- west would also stand high. It would also be found that precipitation occurred the least number of times when the wind direction was north- west. West, north, and northeast would also stand low. There is thus a decided connection or correlation between the wind direction and the frequency of precipitation. The nature of the correlation is determined by summarizing statistics* The correlation of any observation of any element with any other can be worked out. Many of these correlations are of no practical value or importance. A few of those which are of some interest are the follow- ing : maximum temperature and wind direction ; minimum temperature and wind direction; pressure and temperature; temperature and precipitation. The reason why a correlation of the meteorological elements should exist must now be considered. To do this, the question as to what causes our weather must be raised and answered. Our jhe reason weather consists of two things : the typical weather which for the cor- would exist if there were no disturbances, and the influence of * the passing lows and highs. Now, the typical weather which would exist if there were no disturbances is indicated by the normal values of all the meteorological elements together with their diurnal and annual variations. Lows and highs are definite, well-known formations mov- ing along well-recognized tracks. A correlation of the meteorological 320 METEOROLOGY elements therefore exists because the weather is not haphazard or a matter of chance, but is the resultant of two factors, both of which are definite, well-known, and nearly always the same. C. THUNDERSHOWERS DEFINITION AND DESCRIPTION 321. Definition and chief characteristics. A shower is usually defined as a copious rainfall of short duration. The intermittent down- pours which occur during certain types of spring weather thunder- are, for example, spoken of as April showers. If a shower shower de- j s accompanied by thunder and lightning, it is considered a thundershower or thunderstorm. Some would reserve the word thunderstorm for a more than usually violent thundershower, but the distinction is seldom made. The presence of thunder and light- ning while it is raining does not necessarily mean that a and Hght- thundershower is in progress, for all violent rains which fall ningaccom- from thick clouds are accompanied by thunder and light- occurrences nm S- Thunder and lightning accompany tropical cyclonesV tornadoes, desert whirlwinds, and volcanic eruptions as wel, as thundershowers. It was formerly thought that thunder and light- ning were the essential things in connection with a thundershower; in short, they were considered the cause of the storm. It is and light- now known, however, that they are extremely secondary ning a result, an( j pj a y a verv unimportant part. In fact, they are a result and not a cause. They are the inevitable consequence and accompaniment of copious condensation in a thick cloud, and play no part whatever in the mechanism of a thundershower. A thundershower is a too well-known phenomenon to need a careful definition and a full description. It can be best defined and described ^ ,. j by stating its chief characteristics. A thundershower is Outline de- . . scription of an immense cumulo-nimbus cloud accompanied by copious a thunder- precipitation, a marked drop in temperature, and a peculiar, often violent, outrushing, squall wind which just precedes the rainfall. Thunder and lightning are always present and hail some- times falls. It is a violent, local storm, covering a comparatively small area and lasting but a short time. Damage is often caused by wind, hail, and lightning. 322. Description of the approach and passage of a thundershower. Thundershowers occur everywhere from the equator to the pole and THE SECONDARY CIRCULATION OF THE ATMOSPHERE 321 have thus been more generally observed than any other storm. It is the typical thundershower on a hot summer afternoon whose approach and passage is here described in outline. It has been a hot, sultry, oppressive day in summer. The air has been very quiet, perhaps alarmingly quiet, interrupted now and then by a gentle breeze from the south. The pressure has been i 11 -i r_ i i Description gradually growing less. The sky is hazy ; cirrus clouds are O f the ap- visible ; here and there they thicken to cirro-stratus or P roach and cirro-cumulus. The temperature has risen very high and atypical the absolute humidity is very large, but owing to the high th " nder temperature, the relative humidity has decreased some- what. The combination of high moisture and temperature and but little wind has made the day intensely sultry and oppressive. In the early hours of the afternoon, amid the horizon haze and cirro-stratus clouds in the west, the big cumulus clouds, the thunderheads, appear. Soon distant thunder is heard, the lightning flashes are visible, and the dark rain cloud beneath comes into view. As the thundershower approaches, the wind dies down or becomes a gentle breeze blowing directly towards the storm. The temperature perhaps drops a little as the sun is obscured by the clouds, but the sultriness and oppressive- ness remain as before. The thundershower comes nearer, and the big cumulus clouds with sharp outlines rise like domes and turrets one above the other. Perhaps the loftiest summits are capped with a fleecy, cirrus- like veil which extends out beyond them. If seen from the side, the familiar anvil form of the cloud mass is noticed. Just beneath the thunderheads is the narrow, turbulent, blue-drab squall cloud. The patches of cloud are now falling, now rising, now moving hither and thither as if in the greatest commotion. Beyond the squall cloud is the dark rain cloud, half hidden from view by the curtain of rain. The thunderheads and squall clouds are now just passing overhead. The lightning flashes, the thunder rolls, big, pattering raindrops begin to fall or perhaps, instead of these, damage-causing hailstones. The gentle breeze has changed to the violent outrushing squall wind, blowing directly from the storm, and the temperature is dropping as if by magic. Soon the rain descends in torrents, shutting out everything from view. After a time, the wind dies down but continues from the west or northwest ; the rain decreases in intensity; the lightning flashes follow each other at longer intervals. An hour or two has passed ; it is growing lighter in the west; the wind has died down; the rain has almost stopped. Soon the rain ceases entirely; the clouds break through and become 322 METEOROLOGY fractostratus or cirriform ; the temperature rises somewhat, but it is still cool and pleasant; the wind has become very light and has shifted back to the southwest or south. Now the domes and turrets of the retreating shower are visible in the east ; perhaps a rainbow spans the sky; the roll of the thunder becomes more distant; the storm has passed, and all nature is refreshed. 323. Distribution of the meteorological elements about a thunder- shower. The distribution of the meteorological elements about a thundershower is best shown by means of the accompanying The changes diagram, Fig. 130, which depicts the changes in the ele- in the mctc~ . oroiogical ments during a hot summer day with a typical thunder- elements shower between three and five in the afternoon. which occur onahotsum- The temperature rises unduly high during the day, and mer day with reac hes a maximum just before the coming of the storm, thunder- It drops slightly when the sun is obscured by the coming shower in clouds, but the large drop in temperature, which may amount noon. to from 6 to 20 F., occurs during the first twenty minutes of the thundershower. The temperature then continues to drop slowly until the end of the thundershower is reached. After the storm passes, the temperature usually rises again, but does not begin to attain the height reached just before its coming. The pressure usually sags somewhat during the day. When the squall wind begins to blow, there is a sudden increase in pressure of six or seven hundredths of an inch and the pressure oscillates up and down slightly all through the storm. At its end, the pressure is generally a little higher than just before the beginning. The pressure sometimes begins to fall after the shower has passed, but it generally rises, particularly if the shower has been a heavy one and no more are following in quick suc- cession. These oscillations of pressure are easily traceable in the indications of a barograph and from an interesting record of a thunder- shower. 1 The wind is light and from the south during the day. As the thunder- shower approaches, it shifts to the east and blows gently directly towards the approaching storm. This is suddenly replaced by the violent, sometimes damage-causing, squall wind which blows from the west or northwest directly from the thundershower. During the thundershower the wind holds its direction but steadily lessens in velocity. After the storm passes, it is light and often shifts back to the southwest or south. 1 For thundershower barograph curves, see Monthly Weather Review, 1898, Vol. XXVI, p. 592. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 323 The relative humidity is high, but drops during the day, due to the great rise in the temperature. When the thundershower commences and the temperature drops, it rises rapidly and attains a value of 85 or 90 per cent. After the storm passes it drops again slowly, due to the DURATION OF RAINFALL 7A.M. 9 68 F; 29.82 29.78 29.74 29.70 48 40 ESSUR 90^ 60V 30V 03T inches. 32J-S-H 24 16 nch.es =miles=p r=ho 3P.M. 3:30P.M. 4P.M. 4:30f.M. 5P.M. 7 9P.M. & MPS iunc / \ M ^H, TEMPERA1 WIND UR HUMID TY RA NF LL ecti n Ve ocitj 7A.M. 11A.M. 3P.M. 3.30P.M. 4P.M. 4.30P.M. 5P.M. 9PM. PIG. 130. The Change in the Meteorological Elements during a Hot Summer Day with a Typical Thundershower in the Afternoon. 324 METEOROLOGY rise in temperature. Haze and cirriform clouds are prevalent during the day. As the storm passes, the thunderheads, squall cloud, and nimbus clouds follow in succession, the whole being called a cumulo-nimbus cloud. After the thundershower passes, stratiform clouds exist for a while, but the cirriform clouds usually again make their appearance. The rain starts with a few large pattering drops, then falls in torrents, and then gradually lessens in intensity until the end of the thunder- shower is reached. The average duration of a thundershower is a little less than two hours. The first thunder is ordinarily heard about an hour, or a little more, before the coming of the rain, and the last thunder is heard about the same time after the cessation of the rain. 70 MILES FIG. 131. The Cross Section of a Typical Thundershower. It must always be held in mind that the changes in the meteorological elements which have been so fully described and illustrated in the dia- gram apply to the typical thundershower. Thundershowers are slightly different in different parts of the world, and the thundershowers which occur in winter and at night are somewhat different from those which occur on a hot summer day. Each individual thundershower may also differ widely from the type form. There is no more interesting or profitable study than to watch the progress of a thundershower with the type form always in mind. 324. Cross section of a thundershower. In the cross section of a typical thundershower, Fig. 131, the air circulation and the section of a distribution of the cloud masses are shown. The rising air typical thun- wn ich builds the storm, the descending and forward mov- ing air currents underneath the thundershower, and the whirl in the squall cloud with the peculiar outrushing squall wind are THE SECONDARY CIRCULATION OF THE ATMOSPHERE 325 all shown. The thunderheads with their cirruslike caps, the immense nimbus cloud, the stratiform clouds at the back, and the turbulent squall cloud are also depicted. The vertical and horizontal scales are not the same because a thundershower has an average length of perhaps seventy miles and an average height of only about four miles. The anvil-like form of the immense cumulo-nimbus cloud is here clearly seen. Two cumulo-nimbus clouds are represented among the illustrations of the cloud forms in connection with section 217. In one the thunder- head is especially prominent while in the other, which is a side view of a thundershower, the anvil-like form of the cloud is particularly notice- able. 325. The observations of thundershowers. Many special stations for a certain period of years have been maintained by several states in the United States and by several countries in Europe for the detailed study of thundershowers. Most of the information tions have which we have has been gained from the observations made be . en main - at such stations. Since a thundershower is such a small local formation, the regular and cooperative stations of the U. S. Weather Bureau are too far apart to make a detailed study of them possible. If instruments are not used, the observations ordinarily made are : the time of beginning and ending ; the direction of motion ; the violence of the thunder and lightning ; the presence of hail ; the in- tensity of the precipitation; the amount of the drop in vations temperature ; the violence of the wind. If instruments are wh *? h are at hand, the observations include a record of the changes in all the meteorological elements, together with the observation of the time of beginning and ending, direction of motion, thunder and light- ning, and hail. In determining the time of beginning and ending in the case of a thundershower, it is the time of occurrence of the first and last thunder that is usually taken. This is better than to use the time of occurrence of the first and last lightning, as the lightning can be seen so much farther at night than during the daytime. In fact, the so-called " heat lightning " usually indicates the presence of a thundershower, for it is probably the reflection from clouds or the hazy sky of lightning which is accompanying a shower which is below the horizon of the observer. 326 METEOROLOGY THE REGIONS AND TIME OF OCCURRENCE 326. Geographical distribution. Thundershowers occur in nearly every part of the world, but the number decreases rapidly from the d equator towards the pole. Within the tropics there are showers many places where there are nearly 200 days in the course of occur every- a vear w ith thundershowers. The number of days with thundershowers decreases rapidly with latitude, until in the polar regions but one or two thundershowers in the course of several years may be recorded. Fewer thundershowers occur over the ocean than over the land, and mountainous regions have far more than level country. In the United States the largest number occurs in the Gulf States, where there are on the average about sixty days in the course of a year with thundershowers. The number decreases both north and west. In New England the average is not much over fifteen. The accompany- ing table gives the normal number of days with thundershowers for several stations in the United States. 1 These normals are based on the ten years, 1901 to 1910, inclusive. Statistics as to the number of days with thundershowers or the number of thundershowers are very unreliable and cannot be compared. The Statistics reason is because there is no uniformity among observers as are unre- to what constitutes the presence of a thundershower. Some liable. count only the presence of rain, others the audibility of thunder or the visibility of lightning or the occurrence even of the so- called heat lightning. 327. Relation to extratropical cyclones and V-shaped depressions. Nearly all thundershowers which occur in extratropical regions are to Thunder- De f un d in the southern quadrants of a low. The thunder- showers showers which are due to purely local causes, such as a southern* 116 mountain or a peculiarity of the general wind system, are, quadrants of course, to be expected. It will be remembered that the great difference between a winter and summer low was the fact that the nimbus cloud area is unusually absent in summer and its place is taken by cirriform clouds, with sultry weather and thundershowers. Chart XXXV, which gives the daily weather map 1 For the number of thundershower days in the United States during 1903 see Monthly Weather Review, 1903, end of volume. For a chart of the normal number of thunder- shower days in the United States, see : Climatology of the United States, by A. J. HENRY (Bulletin Q of the U. S. Weather Bureau). For the number of thundershowers at Albany, N.Y., each year from 1884 to 1910, see section 28. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 327 X830HVI l> Oi (N Oi CX) tq i-H CO 00 I-H O ^ rH CO CQ CO CO CO T}H |> ' ' ' i-H (N co 06 d 06 ; i-H 00 O (N i-H CO 00 CO CO 00 O Oi ^ (N ^ ^ 00 '''' 10 1 (N t- CO t- ; 05 1 10 00 00 O CO O CO [ T-J 1C TH O 00 i-j * | C^J (N 1 1> IO 00 00 O O "* O t' t-' ^ ^jt t' ^i' t' 06 ,H t^! oo d o* t> 06 d 10 1-' d oo' od ^ji d oo* 10 co d I-H o> o iH rH iH r-l i I rH i-H (N O 00 co co to r- i- T- i^. ^ co i> > aay i^ O ^ O ^ CO C< OO r^ O O H O CO ^ W (N OO O i- W OO COW dddddi-nddddoidrHc *.NVf 328 METEOROLOGY for 8 A.M., Sept. 2, 1904, shows a low with several thundershowers in its southern quadrants. The presence of a thundershower is indicated by /~2. The weather maps for June 30, 1904; July 14, 1904; March 1, 1907; July 17, 1908 ; April 30, 1909 ; May 6, 1909 ; May 15, 1909, are also particularly good illustrations of this. Thundershowers also occur in great numbers in connection with V-shaped depressions. Whenever an isobaric line, instead of being V-shaped straight or uniformly curved, is bent so as to have the form depressions. o f a p OC k e t O r trough, it is spoken of as a V-shaped depres- sion. Whenever a low has a pronounced well-developed wind shift line, the isobars in the southern quadrants nearly always have this Thunder- peculiar V-shaped bulge. This has been fully described and showers are illustrated in section 287. If a low with a wind shift line wmTaiTv- an d this peculiar V-shaped bulge crosses the country during shaped de- the summer, it is nearly always attended by thundershowers ins ' particularly along the wind shift line. A V-shaped depres- sion sometimes indicates that a secondary low is forming. These secondary lows are very common in Europe, but not so frequent in this country. They sometimes develop all the characteristics of a regular low and are even more violent. They usually form in the southern quadrants of a low, and their motion is eastward and north- ward usually with a greater velocity than the parent low. As a result, they seem to circle about it in a counterclockwise direction. Whenever these secondaries occur in summer, they are nearly always attended by thundershowers. Sometimes thundershowers will occur in connection with a V-shaped depression when it is neither a wind shift line or the beginning of a secondary low. Thus whenever a weather map is being studied, all V-shaped depressions should be carefully noted and the characteristics of the meteorological elements about them critically examined. 328. Path across a country. When an overgrown cumulus cloud first develops into a cumulo-nimbus cloud FIG. 132. The Form of a an( ^ becomes a thundershower, it usually Typical Form thunder- covers but a small area. It is perhaps a of a Thunder- shower. f -i i j -i j shower. few miles long and a mile or two wide. As it moves across the country, it becomes constantly larger, so that at the end of six or seven hours, which is about the average life of a thundershower, it has a front some 150 or 200 miles long and is perhaps 40 miles wide. Its form is that shown in Fig. 132. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 329 FIG. 133. The Typical Path of a Thundershower. The typical path of a shower across the country is thus pear-shaped and is shown in Fig. 133. Much larger individual thundershowers have been noted. In some cases Form of they have lasted more than twelve path - hours and have covered a path more than 500 miles long and nearly as wide in its widest part. A large thundershower, which traversed Ger- many August 9, 1881, has been carefully charted by Koppen and is shown in Fig. 134. 329. Direction and velocity of motion. Thundershowers may move in any direction, but in extratropical regions the great majority of them move from west to east. Direction of This agrees with the direction of motion of the upper air motion - currents and also with the direction of motion of the surface winds in the southern quadrants of a low. Those due to purely local causes, as a mountain, sometimes remain stationary. The average velocity for the United States is between 30 and 40 miles per hour; in Velocity of Europe, it is between motion - 20 and 30 miles. The velocity of motion is greater over the ocean than on land and is greater in winter and at night than in sum- mer and during the daytime. Individual showers often depart widely from the type as regards direction and velocity of motion. 330. Time of day and season of occurrence. Thundershowers may occur any hour of the day or night, and any month in the year. The great majority, however, occur during the warmest months in the year, June, July, August, and during the and season hottest part of the day, 3 to 5 P.M. of occur ~ Most stations in the United States have a pro- nounced maximum between 2 and 6 in the afternoon, and a small secondary maximum between 3 and 5 in the early morning. The forenoon and the hours near midnight show the smallest number. FIG. 134. The Path of a Large Thunder- shower across Germany. 330 METEOROLOGY Over the ocean more occur during the night than during the day, and the same is true of Iceland and some coast stations. 331. Periodicity of thundershowers. In addition to the prominent well-marked diurnal and annual periodicity in the occurrence of thunder- showers, it has been thought that other faintly marked periods are also P resent - Some think that the moon influences the ficant occurrence of thundershowers ; the number being slightly thToccur!- larger during new moon and first quarter than during full rence of moon and last quarter. Some think that there is a tidal showers". influence, the number being greater during the high tide than during low tide. Others have thought that there was a 26-day period corresponding to the time of revolution of the sun and an 11-year period corresponding to the sun spot cycle. If there were uniform observations at many stations covering, say 100 years, it would be possible to answer the questions as to the existence of these periods at once. As it is, the observations are far from uniform and the records are often short and fragmentary. As a result, it seems impossible at present to be sure whether these periods exist or not. They are certainly not large or well marked. * THE ORIGIN AND GROWTH OF A THUNDERSHOWER 332. Three classes of thundershowers. All thundershowers can- not be ascribed to the same cause, and for this reason, in studying the The three or igi n f thundershowers, it is more convenient to divide classes of them into three classes ; namely, heat or convection thun- thunder- dershowers, cyclonic thundershowers, thundershowers due to local conditions. Each of these three classes will be considered separately. 41! thundershowers have this in common: All due to tnev are caused by moist rising airl This moisl; rising air the rising of cools due to expansion, reaches its dew point, and builds the immense cumulo-nimbus cloud with its copious precipitation. As a result of this copious condensation in a thick cloud, lightning and thunder occur. 333. Heat- or convection-caused thundershowers. Heat- or convec- The origin tion-caused thundershowers have their origin in masses of opment V of"a warm > m i st air. This air rises, due to convection, cools by convection- reason of expansion, reaches its dew point, and builds first thunder- a cumuuis cloud. If the mass of rising air is not too large, shower. or if it rises in different places in smaller amounts, only THE SECONDARY CIRCULATION OF THE ATMOSPHERE 331 cumulus clouds are the result. If the mass of rising air is large, the cumulus cloud becomes overgrown and the condensation of moisture is sufficiently vigorous to cause precipitation. The cumulus cloud has now become a cumulo-nimbus cloud and underneath it a de- scending air current forms. There are three causes of this descending air current. In the first place, the falling raindrops carry down air with them. In the second place, the air underneath the cloud becomes cooler and heavier because of the cold raindrops which pass through it and also because it is shielded from insolation by the cloud. In the third place, there is a certain amount of reaction against the rising air, for a large amount of air is injected into the upper atmosphere and this must cause an outward movement in all directions. Thus, underneath the cumulo-nimbus cloud, there is a descending air current which is brushed forward when it reaches the earth's surface, both as a result of the for- ward movement of the shower and to take the place of the air which is rising due to convection. In between this rising moist warm air in front of the thundershower and this descending air current underneath the cloud mass, a vigorous eddy or whirl about a horizontal The expla _ axis forms. This is seen in the turbulent squall cloud and nation of the the peculiar violent outrushing squall wind is a part of it. squaU * It might seem that the squall is very vigorous compared with the gentle air currents which build it, but it must be remembered that the squall cloud and wind are very small compared with the immense amount of rising and descending air. Furthermore, the squall cloud is near the axis of the whirl. All these air motions are shown well in Fig. 134. The air motion in a thundershower is very similar to what would be pro- duced by placing a card edgewise on a table, inclining it backward so as to make an angle of about 60 with the table, and then moving it slowly forward. The air would rise over the card, descend back of it to fill the vacant space as it was moved forward, and an eddy about the upper horizontal edge of the card would also form. The origin and growth of a heat- or convection-caused thundershower have now been fully stated and the reason for the structure of the storm and the changes in the meteorological elements during The exp i a _ its approach and passage should be apparent. The baro- nation of the metric pressure rises as soon as the squall wind arrives. The the'meteor- reason for this is the downward movement of the air masses ological eie- underneath a thundershower, and also the fact that the ments * air beneath the cloud mass is cooler and thus denser and heavier. There are three reasons why the temperature drops rapidly when the 332 METEOROLOGY squall wind comes. Underneath the cloud mass there are cold descend- ing air currents. Furthermore the air is cooled by the colder raindrops which fall through it, and it is shielded from insolation by the cloud mass. The reasons for the change in the wind, moisture, cloud, and precipitation are too evident to need explanation. This class of thundershowers ought to occur in the greatest number and with the most vigor when large masses of warm moist air are most Time and numerous. This would be during the hottest time of year place of and during the hottest time of day. Furthermore, it would Lce ' be when a place was located in the southern quadrants of a low, for it is then that the southerly winds are transporting the largest amount of warm, moist air. It will be seen that all these requirements of theory are in good accord with the observed facts. There are two or three observed facts in connection with thunder- showers which have been observed so often by professional meteorolo- Rivers bin- gists and others as well that an explanation of them should der thunder- be given. One is the fact that a thundershower often seems to be unable to cross a large river. It advances to one bank, remains stationary, and perhaps weakens and disappears or builds sidewise along the river. If it crosses the river, it is practically a- new shower which builds on the other bank. This has been observed too often to question the fact. It must be that the river is colder than the surrounding country and is thus the seat of a gently descending air current, which, however, has sufficient vigor to prevent the convec- tional rise of air which is the condition of life and advance on the part of the thundershower. In winter, when the river is warmer than the land, this hindrance to the advance of thundershowers should not exist. It also should be much less at night. Another fact, which has been often observed, is that it rains harder after each lightning flash. If this is true, it may be that the small rain- drops in the cloud are all charged with the same kind of harder after electricity and kept from uniting by electrical repulsion, a lightning AS soon as they are discharged by the lightning flash, they coalesce much more readily and build the larger drops which soon fall to the earth's surface. The interval of time between the lightning flash and the increase in the rainfall ought to give the time required for the raindrop to fall. It may also be that, for some reason, the small drop- lets suddenly unite to form large ones, and thus, as a result, the lightning flash occurs. In one case, the lightning flash is the cause; in the other, the result. Interesting observations in this connection could be made. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 333 334. Convection can be caused by cooling the top of a layer of air as well as by heating the bottom. Thus thundershowers might be caused by cooling the upper layers of the atmosphere as well as by heating unduly the surface layer. This probably method of often occurs. Excessive radiation during the night from beginning of the upper layers of the atmosphere, particularly if there is a caused thin cloud layer, may cool the upper air sufficiently to cause thunder- unstable equilibrium and thus convection and a thunder- shower. These thundershowers would occur chiefly in the early morning when the air was coldest. The secondary maximum in the frequency of thundershowers in the early morning which has been observed at most stations in the United States is probably due to thundershowers formed in this way. 335. Hail. Hail falls wherever thundershowers occur. It is esti- mated that from one half to one tenth of all thundershowers are accom- panied by hail. Hail practically never occurs except during The charac _ the hottest part of the year and the hottest time of day. teristics of In this respect, it shows a much more marked annual and a daily periodicity than do thundershowers. Hail never falls except dur- ing the beginning of a thundershower. The area covered by a fall of hail is very much smaller than the area covered by the thundershower which accompanies it. It is usually not more than 6 or 7 miles wide and perhaps 40 or 50 miles long. Thundershowers which are accompanied by hail usually have a very well-developed squall cloud and are more than usually violent. The structure and characteristics of this particu- lar kind of hail have already been fully stated in section 243. The hailstones are formed in the whirling squall cloud of a thunder- shower. The nucleus is carried up and coated with snow ; it then falls or is carried down, and is coated with water; it is then Theforma- carried up again; the process continues, adding coat after tionofthe coat, until the hailstone becomes too heavy to be longer sustained and it falls to the ground. It will be seen at once that all of the facts concerning the structure of a hailstone and concerning the time and place of occurrence of hail are in accord with this explanation. The possibility of the existence of temperatures sufficiently low to cause hail to be formed might be questioned. Suppose a thundershower is 4 miles thick, and that the cloud commences one mile The cause above the earth's surface, and suppose, furthermore, that the of the low temperature is that of the rising air. This air cools 1.6 F. for tem P erature - every 300 feet until the cloud level is reached. Then the rate of cooling 334 METEOROLOGY will be roughly one half of that until the freezing point is reached and then still less. A rough calculation will show that the top of the cloud ought to be from 75 to 100 F. colder than the temperatures at the surface. Thus, even if surface temperatures are well up towards 90 F., the top of a thundershower must be well below the freezing point and even in the neighborhood of zero and thus composed of snowflakes or ice crystals. There is no difficulty then in finding sufficiently low tem- peratures for the formation of hail. 336. Cyclonic thundershowers. Cyclonic thundershowers are due to the passing of a low and the coming of a high ; they are thus condi- tioned on the change in the weather control from an area of The transi- _ . _ tion from a low pressure to an area of high pressure. If a low goes by low to a north of a station so that its southern quadrants pass over Duu* the place in question, warm moist air is brought to the place by the southerly winds which accompany the low. A coming high is heralded by rather brisk, cold, dry jets of air from the northwest. The transition from one to the other is usually slow and gradual. There are times, however, when it is abrupt, and this is particularly the case when the low has developed a prominent wind shift clonic thun- line in its southern quadrants. The jets of cold dry air dershowers f ro m the northwest may overrun or underrun the warm are caused. moist air of the low. If the cold dry air overruns the warm moist air, then there is unstable equilibrium and all the conditions for convection are fulfilled, for there is warm air below and cold, dense air above. Convection will take place and, in summer, thundershowers will often form. These are really, in their growth and development, con- vection thundershowers, but their origin is cyclonic because it is condi- tioned upon the interaction of a high and a low. If the cold dry air of the coming high underruns the warm moist air of the departing low, this warm moist air will be raised bodily and forced to rise. This forced rise of warm moist air in summer, and even in winter, often causes thundershowers. Thundershowers formed in this way by the inter- action of a high and low often have a very long front. It may extend even the whole length of the wind shift line, and in England these showers are called " line showers." Cyclonic thundershowers ma3 r occur any hour of the day or night or at any time of year. How local 33 ^' Thundershowers due to local conditions. Thun- stationary dershowers due to local conditions are nearly always showers'are causec ^ kv the presence of a mountain and a sea breeze, or caused. a mountain breeze, or a wind caused by a passing high or THE SECONDARY CIRCULATION OF THE ATMOSPHERE 335 low. If a strong sea breeze blows against a mountain side near the shore and the air is forced to rise, a permanent cumulus cloud is formed over the mountains. If the sea breeze is particularly strong or moisture-laden, the cumulus cloud may become overgrown and develop into a thundershower. Such a thundershower would remain stationary near the mountain and disappear in the late afternoon with the dying down of the sea breeze. A mountain breeze blowing up a mountain side during the daytime could produce a stationary thundershower in the same way. Many of these stationary thundershowers over moun- tain peaks are reported in the Alps and in the regions near the Rocky Mountains. Air set in motion by a passing high or low and forced to rise by a mountain could also produce a thundershower. It will be seen in every case that the cause of the thundershower was air forced to rise by a barrier. A cumulus cloud is first formed, and this may become overgrown and develop into a thundershower. THUNDER AND LIGHTNING 338. Since lightning and thunder are only attendant phenomena caused by the copious condensation in a thick cloud and have These wm no part in the mechanism of a thundershower, they will be be consid- considered later in Chapter XI in connection with atmos- eredlat e r ' pheric electricity. D. TORNADOES DEFINITION AND DESCRIPTION 339. Definition and chief characteristics. The tornado is the most diminutive and yet the most violent and destructive Of all storms. It is peculiar to the United States, although in a slightly modified form, it at times occurs in other parts of the world. The charactens- name is derived from the Spanish and refers to the twisting tics * a 5 tornado. or rotating nature of the storm. It is always associated with a violent thundershower, which is usually accompanied by hail, a pronounced squall wind, and violent thunder and lightning. It occurs almost exclusively during the warmer months of the year and during the hottest part of the day. The term cyclone is still often used popularly, and even in the newspapers, in referring to these storms, but they should be called tornadoes, and the term cyclone should be reserved for the tropical cyclone or the extratropical cyclone. 336 METEOROLOGY The most distinctive thing about a tornado is the peculiar black funnel- shaped cloud which extends downward from the heavy cloud masses above, usually reaches the earth's surface, and causes complete devasta- tion wherever it touches. In the United States, nearly a hundred lives and several million dollars of property are lost annually by tornadoes, while a single violent one may cause four or five times this amount of loss. 340. Description of the approach and passage of a tornado. Since a tornado is always associated with a heavy thundershower, the charac- The funnel teristics of the day and the weather changes which precede cloud. ^h e com i n g o f a tornado are the same as those which herald the coming of a violent thundershower on a hot, sultry summer after- noon. Nearly all observers agree that just before" the formation of the funnel cloud, the clouds have an ominous greenish black appearance, and seem to rush together and start whirling with great violence. Then the black funnel cloud appears, which drops lower and lower until the surface of the ground is reached. Here it enlarges slightly, so that some have described the tornado cloud as having the form of an hourglass. It usually sways slightly from side to side and often writhes and twists. Sometimes the funnel cloud jumps a certain strip, only to touch the ground again farther on. The whole formation is usually less than 1000 feet in diameter and passes a given point in less than half a minute, but this is sufficient for complete destruction. The destruction wrought by a tornado seems to be caused both by excessive wind velocities and also by an explosive action. The explo- Two causes s ^ ve ac ^ion is probably due to the sudden decrease in the ofthede- barometric pressure. The barometric pressure is normally between fourteen and fifteen pounds per square inch. If the pressure were suddenly reduced one half, it would cause a pressure of over seven pounds per square inch on the inside surface of all objects containing air which could not quickly escape. The destruction wrought by a tornado is often weird and unusual as well as terrible. Due to tornadic action, large trees are stripped of their branches, broken off near the ground, or torn up by the roots ; heavy brick and The charac s * one buildings are crushed and destroyed as if they were teristics of card houses ; tin roofs are torn from buildings and carried the^destruc- manv m il e s through the air ; loaded cars and even locomo- tives have been blown from the track; heavy iron girders have been carried over the tops of buildings ; iron bridges have been moved from their foundations. Straws have been driven FIG. 13& Two Views of the same Tornado at Goddard, Kansas, May 26, 1903, about 4 P.M. (No barograph disturbance was noticed at Wichita about 18 miles distant.) FIG. 136. A Tornado at Oklahoma City, May 12, 1896. FIG. 137. Damage caused by a Tornado at Rochester, Minn., August 21, 1883. FIG. 138. Wreckage of Anchor Hall, Jefferson and Park Avenues, St. Louis, May 27, 1896. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 337 through boards, laths through trees, and small sticks of timber through iron plate. When the funnel cloud strikes a building, it often seems to explode. The roof is carried up and the side walls fly apart. Chests explode ; corks are drawn from empty bottles ; chickens are stripped of their feathers. Soot is often seen to rise in quantity from the chimneys of near-by houses and window panes often fly outward. The noise which accompanies a tornado is tremendous. It has been likened to the noise of a thousand express trains rush- The accom _ ing through tunnels or to the sound produced by thousands panying of wagons loaded with iron and moving rapidly over an E uneven pavement. The uproar is so great that the crash of individual buildings is seldom heard. The funnel cloud of a tornado is usually associated with the front of the violent thundershower. But little rain usually falls R e i at i nto before its coming, and hail usually follows it. The lightning the thunder- at times is almost incessant so that the funnel cloud has a reddish, lurid look. After a tornado passes, all the characteristics of a violent thundershower usually appear. In Figs. 135 and 136 photographs of distant tornadoes are repro- duced, and in Figs. 137 and 138 some of the damage caused by tor- nadoes is pictured. 1 341. Distribution of the meteorological elements about a tornado. The changes in the meteorological elements brought about by a thundershower have already been considered. Only the changes brought about by the funnel cloud, the tornado sure and proper, will be here stated. The barometric pressure in wind in a the center of the funnel cloud of a tornado has never been determined, as the instruments have always been smashed or exploded. Almost instantaneous drops in pressure of nearly an inch have been observed within a few hundred feet of a tornado, but the drop in the center of the 'funnel cloud is, without doubt, much larger than this. From its explosive effect, it has been estimated that the pressure per- haps drops to one half its value in the center. The wind velocity has never been measured. It certainly goes well over a hundred miles per hour, and may even reach 500 miles per hour, in violent tornadoes. The direction of the whirl about the center is always counterclockwise, and would thus seem to be determined by the rotation of the earth, or 1 For additional pictures of tornadoes and the damage caused by them see Monthly Weather Review, July, 1899 ; September, 1905, p. 400 ; June, 1906, p. 276 ; June, 1907, p. 258. 338 METEOROLOGY perhaps the rotation of air about the low in the southern quadrants of which the thundershower and tornado usually form. The only distinc- tive cloud form is the blue-black funnel cloud which extends down from the cloud masses above to the earth's surface. No noticeable changes in temperature, moisture, or precipitation occur. 342. Observation of tornadoes. Deliberate, careful observations are not often made during the passage of a tornado, not even if the path The obser- ^ destruction is several hundred feet from the observer, vations to be Such observations are, however, much needed. They should cover three things: (1) the appearance of the clouds, their color and motions, the direction and velocity of motion of the tornado, the characteristics of the thundershower which it accompanies; (2) the changes in the meteorological elements during the whole day, but more particularly during the passage of the thundershower and tornado ; (3) the kind and peculiarities of the destruction wrought. In making full notes of any meteorological occurrence, it is always better to record them in full on the spot and not trust at all to one's memory later. THE REGIONS AND TIME OF OCCURRENCE 343. Geographical distribution. Tornadoes occur almost exclu- sively in the United States, although in a slightly modified form they sometimes occur in the other countries where violent thunder- The place of occur- showers are common. They do not occur with the same tornadoes frequency in all parts of the United States, but visit chiefly the Mississippi Valley and certain of the southern states. They are practically unknown in the Rocky Mountain region and near the Appalachian system. They are most frequent over level country which is not heavily wooded and are practically unknown in mountainous or forest covered regions. In Fig. 139, the distribution of all recorded tornadoes from 1794 to 1881, as determined by Lieutenant J. F. Finley, is given. 1 The number of observed tornadoes is increasing each year, but this does not necessarily mean that the number of tornadoes is increasing, The number as the country is becoming more thickly settled and the of tornadoes, occurrence and details of tornadoes are more fully published. During the year 1877 to 1887 f on the average, 146 tornadoes occurred annually in the United States. 1 See C. ABBE, " Tornado Frequency per Unit Area," Monthly Weather Review, June, 1897, p. 250. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 339 FIG. 139. The Distribution of all Recorded Tornadoes from 1794 to 1881. (From GREELEY'S American Weather.) 344. Relation to extratropical cyclones. Tornadoes nearly always occur in the southern or southeastern portion of an extratropical cyclone and from 200 to 800 miles from its center. It will be remem- bered that it is here that violent thundershowers form in i ow which the largest numbers. It is sometimes said that the low p ve snseto tornadoes. which is attended by tornadoes has such definite and peculiar characteristics that the probability of a tornado can almost be predicted from the characteristics of the low. These peculiar characteristics are the following : The isobars are distinctly oval, and extend exactly north and south. The wind shift from south to northwest or west is very sharp, and often a V-shaped bulge in the isobars in the southern quadrants has developed. The temperature lines are packed very closely together along this wind shift line, and the bend in them is some- times so sharp that they bend backward on themselves. The tempera- ture difference between the east and west portion of the low is large and well-marked. Charts XXXVI and XXXVII which give the daily weather map for 3 P.M. March 11, 1884 and 8 A.M. April 25, 1906, repre- Tornado sent typical tornado lows. In the first case, there were prediction. 340 METEOROLOGY several distinct tornadoes in the southeast portion, and in the second case a tornado occurred in Texas. Unfortunately these charac- teristics are not always present when a low is attended by tornadoes. In fact, a tornado can occur when none of them are present. A careful study of, say, a hundred lows which have been attended by tornadoes will always leave the impression that it is impossible to do any predic- tion of the occurrence of a tornado from the characteristics of the low. 1 345. Path across a country. The path of a tornado is from a few feet to perhaps 2000 feet wide, and from a mile to sometimes 200 or 300 miles long. The direction of motion is easterly or south- easterly and the velocity of motion from 20 to 50 miles per hour. As a result a tornado passes a given point in less than half a minute. 346. Time of day and season of occurrence. Tornadoes are most frequent from 3 to 5 in the afternoon, and least frequent from 7 to 9 in the morning. The diurnal variation is very large and and season well-marked. Of those occurring in the evening, most have of occur- originated during the afternoon and have continued their rence. . & . existence into the evening. Tornadoes occur chiefly during May, June, July, and August, but, particularly in the Southern states, they are common at all times of the year. THE ORIGIN AND GROWTH OF A TORNADO 347. The origin of tornadoes. The method of formation of the various kinds of thundershowers has already been fully treated. In considering the origin of a tornado it is only necessary to explain why the tornado funnel cloud forms. underneath certain showers. The forma- tion of a tornado is usually ascribed to one of two causes : violent local convection at a given point or the existence of an energetic eddy. If a thundershower is viewed from the side, it is often noticed that one or two thunderheads rise much higher than the rest. This means Convectional a more violent convectional updraft at these points than origin of elsewhere. If, at some point, the convectional updraft were a tornado. especially energetic, there would develop a well-marked indraft at the earth's surface to replace the rising air. This moving air would be deviated to the right, due to the earth's rotation, and a vigorous atmospheric whirl would build itself up. The growth and 1 For a long series of weather maps when tornadoes occurred see, FINLET, "Tornado Studies for 1884," Prof. Papers of the Signal Service, No. XVI. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 341 development would be almost exactly the same as in the case of the tropical cyclone and would soon lead to a fully developed tornado. The direction of revolution would be here determined by the rotation of the earth. At the level of the lower clouds there are numerous air currents with slightly different directions and very different velocities. These could easily form an energetic eddy which would make itself felt Tornadoes at the earth's surface. Since a tornado is so small and the may be due path covered so short as compared with the thundershower * which it accompanies, it is not unreasonable to think of it simply as an eddy. It is a well-known principle that when one eddy forms inside another, the small eddy takes the direction of rotation of the larger. Now tornadoes form in the air which is moving counterclockwise about an area of low pressure and thus would rotate in the same direction. Many more exact observations of tornadoes are necessary before every step in the origin and development of a tornado can be fully stated. 348. Explanation of the facts of observation. The theories as to the origin of a tornado account for the existence of the whirl, its direction of rotation, and the high velocity. There are several other _ xpl fl.X13.tlO U facts of observation which need explanation. The rapid O f various whirl causes centrifugal force, and this is the cause of the facts f ob ~ . . servation, very low barometric pressure at the center, which in turn is particularly the cause of the tornado funnel cloud. The air has been the funnel - . cloud, relieved of perhaps one half of the pressure upon it ; it has expanded quickly, cooled below its dew point, and produced a cloud extending the whole length of the whirl. The funnel cloud has often been observed to touch the earth's surface for a certain distance, then rise above the earth, leaving a strip without destruction, and then descend again to the earth's surface and continue its work of destruction. This simply means that the whirl weakened for a certain distance. The weakening of the whirl accounts at once for the absence of destruction. It also would mean less centrifugal force, a higher barometric pressure at the center, and thus less cooling due to expansion, and perhaps ab- sence of the cloud as the dew point would not be reached. PROTECTION FROM TORNADOES 349. The subject of protection from tornadoes can be treated from two entirely different points of view ; first, the precautions to be taken 342 METEOROLOGY in order to secure safety if one is overtaken by a tornado and, secondly, tornado insurance again loss caused by tornadoes. If it is a frame building, the southwest corner of the cellar is probably the safest place. This is on the assumption that the building will be The places carried away by the first blast of wind. If the building is of of greatest stone or brick, or if one is caught out of doors, it is better to lie down in the open. Do not seek safety under a tree. In places often visited by tornadoes, so-called " cyclone cellars " are some- times constructed, and these are even provided by the more cautious with tools and certain supplies, in case everything is demolished, or the opening is covered with debris. Insurance against loss by tornadoes is now provided by several tornado insurance companies. Until recently they were not on a very Tornado satisfactory basis, but at present the distribution of tornadoes insurance. an( j t he amoun t o f damage caused by them is sufficiently well known to permit the probable risk to be computed. 1 E. WATERSPOUTS AND WHIRLWINDS WATERSPOUTS 350. Waterspouts are simply tornadoes which occur over bodies of water. This is abundantly proved by the fact that whenever a tornado Waterspouts crosses a river, pond, or body of water, it immediately becomes are simply a waterspout ; and waterspouts which run inland develop at once all the characteristics of tornadoes. Waterspouts are most common in the warmer and calmer seas, although they may occur wherever violent thundershowers are found. When the funnel cloud touches the surface of the water, it is greatly agitated, and the water has been observed to rise even eight or ten feet. This gives an idea of the diminution in the barometric pressure in the center of a waterspout. If it were a vacuum, the water would rise a little more than thirty feet. The spray is, of course, often carried up much higher, but the stories of large quantities of water being carried up from the sea into the clouds is pure myth. When a waterspout crosses a vessel, it has been found that the water in it is fresh. It is thus a condensation product, and has not been carried up from the sea. Waterspouts have usually been observed from considerable distances 1 See : Tornado Insurance, by HOWARD E. SIMPSON, Colby College, Waterville, Me., Insurance Monitor, February to September, 1883. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 343 and with no great care. The waterspout off Cottage City, Mass., August 19, 1896, occurred under circumstances remarkably advantage- ous for making observations and photographs, and has been A notable for this reason more carefully studied than any other. It water- has been critically studied and treated by Frank H. Bigelow spout< in four articles in Vol. XXXIV (1906) of the Monthly Weather Review. One of the photographs there given is reproduced as Fig. 140. This waterspout occurred in connection with a thundershower which was due to the overrunning of the surface air by a layer of colder air, as the weather control passed from an area of low to an area of high pressure. WHIRLWINDS 351. The violent sandstorms and the dusty whirlwinds of desert regions are probably only tornadoes of a slightly modified form and small intensity which are occurring under rather unusual Desert conditions. Over a dry desert region, the amount of mois- whirlwinds, ture in the air is so small that violent convection causes no precipitation or even cloud. The surface layer of air is heated to very high tempera- tures, as the sky is always cloudless, and energetic convection must often take place. If the sandstorm is a straight blow with rather long front, then it has all the characteristics of a thundershower, except that the cloud and precipitation are lacking. Whenever a dusty whirlwind with a small slender column appears, it is probably a tornado of small intensity, but without a funnel cloud, as there is not moisture enough to form it. In fact, even the little whirlwinds which spring up along a dusty roadway on a hot summer afternoon and move a short distance and rise to a height of twenty or thirty feet, remind one of a diminutive tornado. Even when carefully considered, as to origin and character- istics, they still have some things in common with tornadoes. F. CYCLONIC AND LOCAL WINDS INTRODUCTION 352. In the last part of Chapter IV the general winds of the world were classified and discussed. It remains in connection with this chapter to treat those winds which are due to the passing of Cyclonic cyclonic and anticyclonic disturbances. There are three of winds - these which are well known in the United States. They are : first, the 344 METEOROLOGY warm moist south wind often called by its Italian name, the sirocco ; second, the cold dry northwest wind which accompanies the coming There are of a cold wave, and may under certain circumstances be United 1 thC cons idered a blizzard ; third, the chinook, which is perhaps states. better known by its Swiss name, the foehn. THE CYCLONIC AND LOCAL WINDS OF THE UNITED STATES 353- Warm wave (sirocco). If a low is passing north of a sta- tion, warm moisture-laden air from the south is transported to the Cause of a place in question. If the low moves slowly or if the in- warm wave. d ra ft i s particularly energetic, a decided rise in tempera- ture may take place, and this is spoken of as a warm wave. The pass- ing of a cyclonic disturbance is particularly apt to cause a warm wave in the states east of the Mississippi Hiver, as the air transported from the Gulf States is especially warm and moist. In summer, the warm wave takes the form of several hot, sultry, oppres- sive days. The temperature and moisture are high, the wind blows from the south, and the sky is hazy and partly cloud-covered, usually with cirriform clouds. Relief comes when the center warm wave of the low approaches sufficiently near to cause rain, or when an^winter. ^ ne ^ ow P asses and the wind shifts to the northwest. In winter it takes the form of a thaw, and the sleighing may be spoiled or the ground freed from snow. Ice storms often result from the coming of this south wind, since the wind velocity is much greater at the height of a mile or so above the earth's surface. This transports warm moist air much more rapidly at this level, so that the upper air may be much warmer than the air at the earth's surface. This warm moist wind from the south which causes a warm wave is often called by its Italian name, the sirocco, but this name is ordinarily The sirocco em pl ve d to designate a particularly warm moist wind, and not to designate the south wind in front of any coming low. 354. Cold wave (blizzard). The meaning of the term cold wave has been made definite by the U. S. Weather Bureau. It is defined as Definition of a drop of a certain number of degrees in twenty-four hours a cold wave. w jth a minimum below a certain temperature. The amount of the drop and the minimum are different for different times of year, Cause of a and for different groups of states. A cold wave is brought cold wave, about by the cold, dry, northwest wind which marks the passing of a low and the coming of a high. In the popular mind, the THE SECONDARY CIRCULATION OF THE ATMOSPHERE 345 drop in temperature and the wind are associated together, although technically the cold wave has to do with the drop in temperature only. If it still continues to snow after the wind has gone to the northwest, and the temperature has dropped, or if the high northwest The wind drifts the light snow, a blizzard is said to be in progress. blizzard - There is no exact definition of a blizzard, but its characteristics are supposed to be high northwest wind, driving snow, and low and falling temperature. The origin of the cold air which constitutes a cold wave has been ascribed to several causes. Some have claimed that the cold air was transported from the far north, and that cold waves thus The origin have their place of beginning in the extreme northwest, of cold Others have claimed that the cold air descends from the v upper atmosphere, while still others look for the cause in the con- tinued radiation of heat to the clear sky night after night. That a cold wave is made on the spot, so to speak, seems more plausible than that the air is transported long distances. 355. Chinook (foehn). The chinook occurs chiefly on the eastern side of the Rocky Mountains, and is particularly common in the states of Wyoming and Montana. It is a hot dry wind coming from the _ J . -. . Description. west across the mountains. It usually makes its appearance suddenly, and the temperature may rise even 40 in fifteen minutes. The snow disappears as if by magic, for it evaporates, due to the dryness of the air, as well as melts. It often removes as much snow in a day as ordinary spring thawing would remove in two weeks. If it is common, it raises the normal annual temperature of a place by several degrees. This wind was first noted and studied in northern Switzerland, where it was called the foehn. It occurs chiefly on the north side of the Alps, although it does appear on the south side as well. In some The foehn valleys, it blows from 30 to 50 days during the months from and its char- November to March and raises the normal annual tempera- a ture considerably. If the foehn continues for several days, or if it is particularly vigorous, not only does all the snow melt, but everything becomes so dry that special precautions are taken to prevent a general conflagration in case of fire. Sometimes all fires are extinguished while the foehn blows. This wind is also found in Greenland and New Zea- land. In fact, it is found wherever a mountain chain and passing lows are associated together. This wind is usually called the chinook in the Western states, although it is better known as the foehn in other parts of the world, and this last name is fast becoming universal. 346 METEOROLOGY FIG. 141. Diagram Illustrating the Formation of the Foehn Wind. The cause of the foehn can best be stated in connection with a dia- gram (Fig. 141). The air is passing over a mountain towards a The cause cyclonic center and is forced to rise by the mountain. It of the foehn. exp ands and cools at the rate of 1.6 F. per 300 feet, soon reaches its dew point, becomes cloudy, and yields copious precipitation. The latent heat liberated by the condensation of the water vapor warms the rising air and prevents its cooling at so rapid a rate. In fact the rate is cut down from 1.6 F. per 300 feet to a little less than half this amount. After ^e top of the mountain is reached, the air begins the descent on the other side. It is now being compressed and grows warm at the rate of 1.6 F. per 300 feet. The precipitation ceases, the clouds disappear, and the air continues to grow warm at the same rate during the whole of the descent. As a result, it may reach the same level on the other side of the moun- tain 20 or even 40 warmer than before it started the ascent. It has also become very dry, as so much moisture was removed by precipitation. This explanation of the foehn was first given by Hann of Vienna. It was formerly thought that the Swiss foehn was due to hot dry air which came in some way from the desert of Sahara. A chinook should occur on the eastern side of the Rocky Mountains whenever a well-developed cyclonic storm passes across the northern The relation ^^^ ^ ^ ne United States. It ought to occur on the western of the side of the mountains when the storm center passes across the southern part of the country. As most cyclonic storms pass along the northern boundary of the United States, the chi- nook has been observed on the east side much more frequently than on the west. A cyclonic storm, central over Germany, ought to cause a well-marked foehn on the northern side of the Alps, and a foehn on the south side ought to be caused by a low, central over Italy or the Medi- terranean Sea. As a matter of fact, the foehn is much better developed and of much more frequent occurrence on the north side of the Alps than on the south side. The reason is because the air coming north from Italy is already warm and dry, and thus lends itself much more readily to the formation of a foehn wind. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 347 THE CYCLONIC AND LOCAL WINDS OF OTHER COUNTRIES 356. The various cyclonic and local winds which occur in different parts of the world may be divided in four groups according to cause, and each of these groups will be treated in order, four groups (I) The first group includes those which correspond to the warm, moist, south wind which is the cause of the warm waves in the United States, and is the well-known sirocco of Italy. The solano is a very hot, dusty, southeast wind which The sirocco occurs in the Mediterranean, especially on the eastern and similar coast of Spain. The leveche is a hot, dry, southwest wind, which also occurs in Spain. The leste is a very hot parching south wind which occurs in Madeira and northern Africa. All these winds have the same origin. They are due to the indraft of warm air from the south towards the center of an advancing low. They will be dry or moist, according as the region over which they ad- vance is dry or moist. In the southern hemisphere, the corresponding winds would come from the north. The brickfielder of southern Australia is a hot, dry, north wind. The zonda of the Argentine Republic is a hot northerly sirocco. (II) The second group includes those winds which correspond to the dry, cold, northwest wind which ushers in the cold waves in the United States, and is often a blizzard. The buran, or purga, is a very cold northeast wind, which zar d and occurs in Russia and central Asia. The snow is blown by the wind, and it is often blizzard-like. The pampero is a dry, cold ; southwest wind which is common in southern Brazil, Argentina, and Uruguay. The tormentos of Argentina is the same kind of a wind. All of these winds are due to the cold dry northwest wind (northern hemisphere) which marks the passing of a low and the coming of a high. (Ill) The winds of the third group are due to the bringing down of cold dry air from a plateau into a valley by a passing anticyclonic dis- turbance. Masses of air might be brought down from a The m i stra i plateau into a valley by a passing low as well as by a high, and similar but the air would not be cold and dry, as the presence of a high is necessary to give the air over the plateau these characteristics. The air is warmed somewhat by compression during the descent, but if 348 METEOROLOGY the descent is not too rapid, it still reaches the valley colder and dryer than the valley air. The mistral (Italian : magistrate = masterly) is a dry, cold, north- west wind found in the Rhone Valley, and due to masses of air coming into the valley from the plateau in the southeastern part of France. The bora (Greek : /Jo'peas = north wind) is a furious northerly wind, coming into the Adriatic Sea from the plateau of Carinthia. The tramontana is a searching northerly blast found on the Italian side of the Adriatic. The gregale of Malta is a cold, dry, unhealthy wind, and is due to the same cause. The williwaus of Terra del Fuego, in Patagonia, is perhaps of the same origin. This is a hurricane-like wind coming down from the moun- tains and blowing with great violence only 8 or 10 seconds. (IV) The fourth group includes those winds which get their peculiar characteristics from the topography and condition of the regions from The kham- which the winds come. The wind is, of course, caused to sin and sim- blow from these particular directions, due to the passing of aar winds. a h ; gh or low The harmattan is a hot, dusty, east wind, which is found in the west side of the desert of Sahara. It is most common during December, January, and February, and brings much sand with it. The khamsin (Arabic : khamsun = fifty) is found in Egypt and is a hot wind from the desert. Its direction is usually from the south or southeast, and it blows from 20 to 50 days in the course of a year. It will be seen that all of these are cyclonic or local winds, in the sense that they are due to the passing of highs or lows, and in some cases to the characteristics of the surrounding region. QUESTIONS (1) State the chief characteristics of the tropical cyclone. (2) What names are applied to them in different parts of the world? (3) Trace the historical rise of our present information about tropical cyclones.-" (4) Describe the approach and passage of a tropical cyclone. (5) Describe in detail the distri- bution of the meteorological elements about a tropical cyclone. (6) Illustrate the distribution by means of two diagrams. (7) Give the life history of some tropical cyclone. (8) What are the rules for mariners in connection with tropical cyclones? (9) How is the storm center located? (10) Which is the dangerous half of the tropical cyclone and why? (11) Where are the regions of occurrence of tropical cyclones? (12) Describe the form of the track followed. (13) At what times of year do they occur? (14) State in full the convectional theory of the origin of tropical cyclones. (15) Explain the formation of the calm central eye. (16) Explain the path followed and the time of occurrence of THE SECONDARY CIRCULATION OF THE ATMOSPHERE 349 the tropical cyclone. (17) Why do tropical cyclones occur where they do? (18) Compare the tropical cyclone and the circumpolar whirl in the general wind system. (19) State the chief characteristics of the extratropical cyclone. (20) Describe in detail the distribution of the meteorological elements about a low. (21) How does the distribution differ at different times of a year and in different parts of the world? (22) What are the factors which cause a departure from the type form? (23) Describe the distribution of the meteorological elements about a low with a wind shift line. (24) Illustrate the distribution by means of a diagram. (25) Describe in detail the structure of an extratropical cyclone at various levels above the earth's surface. (26) Contrast the tropical and extratropical cyclone. (27) Describe the approach and passage of a low. (28) Define a high and state the distribution of the meteorological elements about it. (29) Describe the structure of a high at various levels above the earth's surface. (30) Describe the sequence of weather changes as a high approaches and passes. (31) Describe the tracks followed by lows in the northern hemisphere. (32) Where do they originate? (33) Describe the system of tracks across Europe. (34) Describe the three systems of tracks across the United States. (35) Treat the velocity of motion of lows across the United States. (36) Describe the tracks followed and the velocity of motion of highs. (37) Treat in full the theories as to the origin of extratropical cyclones and anticyclones. (38) De- scribe the methods of growth of an extratropical cyclone. (39) Describe the growth of an anticyclone. (40) In connection with both high and lows explain the peculiarities in the distribution of the elements and the direction and velocity of motion. (41) Does the ah* accompany a high or low? (42) What factors determine the direction of the wind and its velocity at any point near a low or high? (43) What is meant by the correlation of the meteorological elements and why does one exist ? (44) Define a thundershower and state its chief char- acteristics. (45) Describe the approach and passage of a thundershower. (46) Describe and illustrate by means of a diagram the changes in the meteoro- logical elements as a thundershower passes. (47) Draw a cross section of a thundershower. (48) What observations are made of thundershowers ? (49) Where do thundershowers occur? (50) What is the relation of thundershowers to extratropical cyclones? (51) How are they related to V-shaped depressions? (52) Describe the form of a thundershower and the form of its path. (53) What is their direction and velocity of motion ? (54) At what time of day and season of the year do they occur? (55) Are thundershowers periodic in their occur- rence? (56) Slate the three classes of thundershowers. (57) Describe in detail the development of a heat- or convection-caused thundershower. (58) De- scribe the structure and formation of hail. (59) How are cyclonic thunder- showers caused ? (60) How are thundershowers due to local conditions formed ? (61) Define a tornado and state its chief characteristics. (62) Describe the approach and passage of a tornado. (63) State the changes in the meteoro- logical elements as the tornado passes. (64) When and where do tornadoes occur? (65) How are they related to extratropical cyclones? (66) Explain the origin and growth of a tornado. (67) Describe a waterspout and state its cause. (68) State the cause of the violent sand storms and the dusty whirl- winds of desert regions. (69) Name the three cyclonic winds peculiar to the United States. (70) Describe the wind which causes the warm wave and is often called the sirocco. (71) Describe the wind which causes the cold wave and may sometimes be considered a blizzard. (72) What is the chinook wind? (73) Where does the foehn occur and what are its characteristics? (74) State the cause of the foehn wind. (75) Name the four classes of cyclonic or local winds and describe some examples in each class. 350 METEOROLOGY TOPICS FOR INVESTIGATION (1) The complete life history and characteristics of some tropical cyclone. (2) The calm central eye of a tropical cyclone and the evidence of its exist- ence. (3) The extent to which a tropical cyclone retains peculiar characteristics after entering extratropical regions. (4) The difference in the distribution of the meteorological elements about a typical low in Europe and in America. (5) General laws derived from statistics on highs and lows. (6) The theories as to the origin of highs and lows. (7) The complete life history of some thundershower. (8) The periodicity of thundershowers. (9) The theories as to the formation of hail. (10) The size and structure of hailstones. (11) The complete life history and characteristics of some tornado. (12) The characteristics of the low which is accompanied by tornadoes. (13) Tornado insurance. (14) The complete life history and characteristics of some waterspout. (15) The characteristics of the blizzard. PRACTICAL EXERCISES (1) Note the distribution of the meteorological elements about several highs and lows and in each case explain in full any departure from the type form. (2) Note the path followed by several highs and lows and in each case explain any departure from the normal path. (3) From a 10-year file of the daily weather maps work up statistics on some point in connection with highs and lows. (4) Work out the correlation of the meteorological elements for one or more stations. (5) Determine from a file of weather maps the location of thundershowers with reference to the lows which they accompany. (6) Determine for one or more stations the normal number of days with thundershowers for the various months and for the year. (7) Work up statistics as to the time of day and season of occurrence of thundershowers . (8) If a thundershower with hail occurs, study critically the hail stones. REFERENCES TROPICAL CYCLONES ALEXANDER, WILLIAM H., Hurricanes. Bulletin 32, U. S. Weather Bureau, 1902. ALGUE, JOSE, The Cyclones of the Far East, 2d ed., 4, 283 pp., Manila, 1904. BERGHOLZ, PAUL, Orkane des fernen Ostens, xii + 260 pp., Bremen, 1900. DAVIS, WILLIAM M., Whirlwinds, Cyclones, and Tornadoes, 24, 90 pp., Boston, 1884. DOBERCK, W., The Law of Storms in the Eastern Seas, 4th ed., 8, 44 pp., Hongkong, 1904. ELIOT, SIR JOHN, Handbook of Cyclonic Storms in the Bay of Bengal for the Use of Sailors, 2d ed., 8, 2v., Calcutta, 1900-1901. THE SECONDARY CIRCULATION OF THE ATMOSPHERE 351 FISHER, ALFRED, Die Hurricanes oder Drehsturme Westindiens, 4, 70 pp., Gotha, 1908. GARRIOTT, E. B., West Indian Hurricanes, 4, 69 pp., 1900. Bulletin H, U. S. Weather Bureau. PIDDINGTON, HENRY, The sailors Horn-book for the Law of Storms, 6th ed., 8, 408 pp., London, 1876. For the life history of various tropical cyclones and other storms see the periodical- literature. EXTRATROPICAL CYCLONES AND ANTICYCLONES For the theories as to the origin and nature of extratropical cyclones and anti- cyclones, see : ABBE, CLEVELAND, Mechanics of the Earth's Atmosphere (2 vols. of collected papers). BIGELOW, FRANK H., Monthly Weather Review (various articles from 1902 on). HILDEBRANDSSON, H. H., ET TEissERENC DE BORT, Les bases de la meteorologie dynamique, Paris, 1900-1907. Report of the chief of the Weather Bureau, 1898-1899. (Report on the inter- national cloud observations.) STREIT, A., Das Wesen der Cyclonen, Wien, 1906. For the distribution of the meteorological elements about highs and lows, see : Annals of Harvard College Observatory, Vol. XXX. HANN, JULIUS, Lehrbuch der Meteorologie. HANZLIK, STANISLAV, Die rdumliche Verteilung der meteorologischen Elemente in den Antizyklonen, 94 pp., Wien, 1898. LOCKYER, WILLIAM J. S., Southern Hemisphere Surface Air Circulation, 4, 109 pp., 1910. Meteorologische Zeitschrift, p. 307, Juli, 1903. Monthly Weather Review, March, 1907. For the tracks followed by highs and lows, see : BIGELOW, FRANK H., Storms, Storm Tracks, and Weather Forecasting, 8, 87 pp., Washington, 1897. Bulletin 20, U. S. Weather Bureau. DUN WOODY, H. C., Summary, of International meteorological Observations, 1878-1887. Bulletin A, U. S. Weather Bureau. VAN CLEEF, Monthly Weather Review, Vol. 36, 1908. THUNDERSHOWERS CONGER, N. B., Report on the Forecasting of Thunderstorms during the Summer of 1892. Bulletin 9, U. S. Weather Bureau, 1893. GOCKEL, ALBERT, Das Gewitter, 246 pp., Koln, 1905. PLUMADON, J. R., Les orages et la grele, 8, 192 pp., Paris, 1901. TOMLINSON, CHARLES, The Thunderstorm, London, 1877. TORNADOES American Meteorological Journal, August, 1890. (Four prize essays on tornadoes.) FERREL, WILLIAM, Cyclones, Tornadoes, and Waterspouts. Professional Papers, U. S. Signal Service, No. XII, Washington, 1882. 352 METEOROLOGY FINLEY, J. P., Report of the Tornadoes of May 29 and 30, 1879, in Kansas, Nebraska, Missouri, and Iowa. Professional Papers. U. S. Signal Service No. IV, 1881. FINLEY, J. P., Characteristics of six hundred tornadoes. Professional papers, U. S. Signal Service, No. VII, 1884. FINLEY, J. P., Tornadoes, New York, 12, 196 pp., 1887. FINLEY, J. P., Tornado studies for 1884. Professional papers, U. S. Signal Service, No. XVI, 1885. HAZEN, H. A., The Tornado, New York, 12, 143 pp., 1890. Report of the Chief of the Weather Bureau, 1895-1896. (A Study of the Tor- nadoes, 1889-1896.) WATERSPOUTS BIGELOW, F. H., four articles in the Monthly Weather Review, 1896, Vol. 34. CHAPTER VII WEATHER BUREAUS AND THEIR WORK A BRIEF HISTORY OF THE U. S. WEATHER BUREAU, 357 THE PRESENT ORGANIZATION, 358, 359 THE STATION EQUIPMENT AND THE OBSERVATIONS TAKEN, 360, 361 THE DEVELOPMENT OF THE DAILY WEATHER MAP, 362 THE DAILY WEATHER SERVICE OF THE U. S. WEATHER BUREAU The taking and sending of the observations, 363. The charting of the observations, 364, 365. The construction of the weather map, 366. The distribution of the map, 367. Other methods of distributing the forecasts, data, and warnings, 368. The weather service in the evening and on Sundays and holidays, 369. OTHER WORK AND PUBLICATIONS OF THE U. S. WEATHER BUREAU, 370, 371 THE WEATHER BUREAUS OF OTHER COUNTRIES, 372 SOME SPECIAL METEOROLOGICAL OBSERVATORIES; THEIR EQUIPMENT AND WORK, 373 A BRIEF HISTORY OF THE U. S. WEATHER BUREAU 357. In 1747 Benjamin Franklin made a very important discovery in connection with storms. He had arranged with his brother in Boston to make some observations of an eclipse of the moon, while Benjamin he took simultaneous observations in Philadelphia. Shortly Franklin ob- before the occurrence of the eclipse, a strong northeast wind ^stonn^sa set in in Philadelphia, bringing with it clouds and rain, so moving for- that the observations could not be secured. Since the wind came from the northeast, he supposed, of course, that the observations had not been made in Boston. Great was his surprise when, several days later, he received word that the observations had been secured in Boston, and that a heavy storm had commenced the following morning. From further observations collected in connection with this storm, and from other observations as well, he came to the conclusion that a storm was a moving formation and that, although the wind usually commenced to blow from the east or northeast, its motion was from some westerly 2 A 353 .1 354 METEOROLOGY to some easterly quarter. As soon as the fact was fully recognized that a storm was a moving formation, the desire at once arose to keep track of it and herald its coming, but the means of communication were too slow and uncertain to make this practicable. After the invention of the electric telegraph, in 1837, the receiving of simultaneous observations from different parts of the country, and the heralding of storm, was put into partial operation, but this was brought to an end by the breaking out of the Civil War. The storms on the Great Lakes had attracted attention on account of their severity and the losses caused by them. In 1869 Professor Cleveland Abbe, who was then Director of the Observatory i ng e fJhe " at Cincinnati, was asked by the Board of Trade of that city, weather ser- to undertake the forecasting of these storms, and the Western Abbe. 7 Union Telegraph Co. transmitted the messages free of charge. The work was so successful and the results so satisfactory that the attention of the whole country was attracted to it. In 1870 a bill was introduced into Congress by Hon. H. E. Paine, from Wiscon- sin, perhaps at the suggestion of Professor I. A. Lapham of Milwaukee, to appropriate $20,000 to make the weather service national. The bill was passed, and the weather service of the United States was thus inaugurated. It was made part of the Signal Service and placed under the War Department, and thus General Albert J. Myer became its head, The first warning was issued November 1, 1870, and the first The first daily weather map was issued January 1, 1871. This weather made the United States the fourth country to issue daily weather maps, as the Netherlands, England, and France had already done so. General Myer continued to be the head of the weather service until his death, August 24, 1880. He was a strong man of great executive The differ- ^lity and under him the weather service developed rapidly, ent heads of He also enjoyed popular approval and confidence. He was succee( ^ e ^ ^y General William B. Hazen who continued in office until his death, January 16, 1887. His administra- tion was characterized by a fostering of the scientific side rather than by a development of the executive side. He was succeeded by General, then Captain, A. W. Greely, who continued in office until the weather service and signal service were separated. This separation took place July, 1891, and the weather service was then made a separate bureau and placed under the Department of Agriculture. Mark W. Harring- ton, who was then Professor of Astronomy and Director of the Observa- WEATHER BUREAUS AND THEIR WORK 355 tory at the University of Michigan, became the first chief of the U. S. Weather Bureau. He was succeeded July 4, 1895, by Willis L. Moore, who is at present the able head of the service. The Weather Bureau has always been characterized by steady ad- vance, development, and improvement, along both scientific and executive lines. In 1870 there were only 24 stations. Weather In 1893 there were nearly 136 stations, and the annual cost was about $900,000. At present there are nearly 200 sta- IzeTby* tions, and the annual cost is about $1,600,000. It is often stead y ad ~ state.d, even by foreign scientists and weather service officials, that the U. S. Weather Bureau is, in many respects, a model weather service. THE PRESENT ORGANIZATION 358. The central station of the U. S. Weather Bureau is, of course, located at Washington, and the rest of the country is divided into dis- tricts in two entirely different ways, for two different purposes. The ^i^ In the first place, there are twelve climatological districts, ciimatoiogi- conforming to the twelve principal drainage areas of the c United States. This scheme affords the best system of territorial units for the compilation and discussion of climatological data. These twelve districts are : (1) North Atlantic States, (2) South Atlantic and East Gulf States, (3) Ohio Valley, (4) Lake Region, (5) Upper Mississippi Valley, (6) Missouri Valley, (7) Lower Mississippi Valley, (8) Texas and Rio Grande Valley, (9) Colorado Valley, (10) Great Basin, (11) California, (12) Columbia Valley. The country is also divided into six (formerly eight) forecast dis- tricts for the purpose of forecasting the weather and pre- The six dieting storms. forecast Prior to 1909 there-were twenty-one climatic subdivisions for the purpose of preparing climatological data and statistics. The observations were summarized for each of these districts as a whole. These were New England Lower Lake Southern Slope Middle Atlantic States Upper Lake Southern Plateau South Atlantic States North Dakota Middle Plateau Florida Peninsula Upper Mississippi Valley Northern Plateau East Gulf States Missouri Valley North Pacific Coast Region West Gulf States Northern Slope Middle Pacific Coast Region Ohio Valley and Tennessee Middle Slope South Pacific Coast Region 356 METEOROLOGY The country was also divided into forty-five sections for the collection and dissemination of weather observations. Each state was usually The former a section, but Maine, New Hampshire, Vermont, Massa- method of chusetts, Connecticut, and Rhode Island were grouped subdivision. together as N ew England, and Maryland and Delaware, and Oklahoma and Indian Territory, were considered a single section in each case. Alaska, Hawaii, and Porto Rico were included in the forty- five. Up to 1909, each section, with the exception of Alaska, published a monthly climatological report, and a weather bulletin weekly during the summer. At present the section practically does not exist, al- though in the supervision of substations and in the collection of obser- vations the section directors still perform their old duties within their respective stations. Scattered throughout the country are Weather Bureau stations of three different kinds. There are, first, the regular stations of the Weather Bureau, of which there are nearly 200. There are, kinds of 66 m addition, nearly 3000_jcooperative stations, and about weather 500 special stations^ Observations are also received from dons** S1 many stations in Canada, Mexico, and the West Indies, and from some stations in Europe, Asia, and various islands. The accompanying map (Fig. 142) shows the 12 climatological districts and the 6 forecast sections into which the country is divided. The centers of the 12 districts are indicated by a while the centers of the 6 forecast sections are underlined. Only about half of the regular stations are shown. For a full list of the regular stations, see Appendix VII. 359. The chief of the U. S. Weather Bureau has his headquarters at Washington, and closely associated with him in executive and scientific The organi- WOI> k are the assistant chief, the chief clerk, the professors, zation at the heads of the divisions into which the work at the central station is divided, and the inspectors. The assistant chief acts in place of the chief when he so desires or is absent. The chief clerk has charge of the correspondence and all questions of personnel. The professors, of whom there are eight at present, are particularly well-trained men who are connected with the scientific, rather than the executive, work. The ten divisions into which the work at the cen- tral station at Washington is divided are : forecast, river and flood, instrument, publications, supplies, telegraph, accounts, marine, library, and climatological, and each is presided over by an efficient head. 1 It 1 The number of divisions, however, is a matter which is subject to frequent change. Very recently (1911) the forecast, marine, and river and flood divisions have been prac- tically consolidated into one, the division of observations and reports. WEATHER BUREAUS AND THEIR WORK 357 358 METEOROLOGY is the duty of the inspectors to visit the regular stations from time to time, both to make investigations in case anything has gone wrong, and also to make suggestions as to how the efficiency and usefulness of the station may be increased. At the regular Weather Bureau stations outside of Washington from one to fourteen men are employed. The head of each station receives the title of official in charge, and the one next in authority to him is usually called the first assistant. At the central station at Washington about 200 are employed, and at all the regular stations outside of Washington about 500, so that there Number of are, all told, about 700 commissioned employees in the service employees. o f the Weather Bureau. In addition, nearly 100 display men and observers receive compensation for their services. The salary of the chief is $6000, and that of the assistant chief and chief clerk, $3000 each. At the regular weather bureau stations outside of Washington the salaries run from $3500 down. The observers at cooperative sta- tions receive no compensation for their services, except the publications of the bureau. THE STATION EQUIPMENT AND THE OBSERVATIONS TAKEN 360. The central station of the U. S. Weather Bureau is pictured in Fig. 143, which is used as the frontispiece of this book, and consists of The central a mam building and several adjoining buildings located at station at 24th and M streets in Washington. On the ground floor of the main building are located the rooms for the weather forecasting and the library. The library now contains nearly 30,000 volumes. It is the best meteorological library in this country, and one of the best in the world. On this same floor are located the accounts, telegraph, and river and flood divisions. On the second floor are located the rooms for the marine work, and the offices of the chief, assistant chief, and chief clerk. The instrument division, the printing outfit, the climatological division, and the division of supplies are located in the adjoining buildings. The observations taken and the instruments used are essentially the same as those at all the regular stations of the Weather Bureau. There are, however, several special instruments in use at Washington. The regular stations of the U. S. Weather Bureau are usually located in the larger cities, as the weather maps and weather forecasts can be more quickly distributed and reach more people. From three to WEATHER BUREAUS AND THEIR WORK 359 eight or ten rooms are usually occupied by a Weather Bureau station. In some cases the Weather Bureau owns the building, but more often the rooms are simply rented. They are generally located on the top floor of some high office building, and a flat mien*!!?!?" roof must also be available for exposing the instruments. reguiarWea- In New York the rooms were on the 20th floor of 100 %%"** Broadway, 1 and the instruments were on the 23d floor. In Boston, the Weather Bureau is located in the main tower of the post office building. On the roof will be found a thermometer shelter, containing a maximum and a minimum thermometer, wet and dry bulb thermometers which can be whirled, a Richard Freres thermo- graph, and sometimes a recording hygrometer. A tipping bucket rain gauge, a Robinson cup anemometer, an electric contact sunshine recorder, and a contact-making wind vane will also be found suitably exposed. If the station has but three rooms, one would probably con- tain the desks of the official in charge, or local forecaster official, and his first assistant, the library, and perhaps files and supplies. A second room would contain the apparatus in actual use and additional apparatus for demonstration purposes and to replace anything which might become broken. Here would, at least, be found the " triple register " for recording the wind velocity, wind direction, sunshine, and precipitation ; a Richard Freres barograph ; and two good mer- cury barometers. The third room would be a work room for prepar- ing the weather map and repairing apparatus. Here would be found the printing presses and the addressograph for addressing the daily weather maps and other publications. At the section centers and at stations located in large cities more room is necessary, as there is a larger number of employees, but the general arrangement is the same. At a cooperative station, only a maximum and minimum thermometer in a thermometer shelter and a rain gauge are necessary so that the equipment is very simple. The thermometer shelter is usually located in the open over sod, and some five or six ment of a feet above it. It may be fastened to the north side of some cooperative unheated building, leaving ample space for ventilation be- tween it and the building. The rain gauge is usually located in the open, a few feet above the ground. 361. At a regular station of the U. S. Weather Bureau the following observations are taken : 1 May 1, 1911, the station was moved to the Whitehall Building, 17 Battery Place. The office rooms are now on the 29th floor of a 31-story building. 360 METEOROLOGY Continuous Pressure with barograph.* Wind direction! Wind velocity I ( j^ ecor( j e( j on the same revolving drum ; instruments in the open.) Precipitation J Moisture with hygrometer, f Temperature with thermograph. f 8 A.M. and 8 P.M. Maximum and minimum temperature. f Pressure.* Temperature.! Moisture (wet bulb thermometer).! Clouds (upper and lower). Precipitation. Miscellaneous Fog, frost, hail, thunder, halo, aurora, sunset and sunrise colors, smoke, haze. * Instrument in the office. t Instrument in the thermometer shelter. Form No. 1083 Met'l. U. S. DEPARTMENT OF AGRICULTURE, WEATHER BUREAU. (Station} (Date) , 19 (75th Meridian Time.) 8 A.M. 8 P.M. Dry thermometer . . Wet thermometer Maximum thermometer Minimum thermometer Precipitation. {Upper . Lower Barometer Attached thermometer Observed reading Total correction Station Reduced Compared with Form No. 1001 Met'l. NOTES: WEATHER BUREAUS AND THEIR WORK . 361 The observations at 8 A.M. and 8 P.M. are taken down on form 1083, which is here reproduced. The data on this form and values obtained from the sheets taken from the recording instruments are copied into forms 1001 and 1014. Form 1001 and 1083 to- gether with the original triple register, thermograph, and barograph sheets are sent monthly to Washington. The tionand necessary data are also copied into the " Climatological Record," a book retained by the local station. A monthly summary is also printed by each regular station. A pamphlet entitled " Instructions for Preparing Meteorological Forms " and the " Station Regulations of the Weather Bureau " give detailed information as to the taking and recording of the observations. The Weather Bureau also publishes a large number of blank forms, and form 4024 is a check list of the various forms. Samples of these various forms may be se- cured from the regular Weather Bureau stations and are very valuable illustrative material of the work of a station. At a cooperative station, the following observations are required : Maximum temperature. Snow on ground. Minimum temperature. Duration of precipitation. Set maximum. Prevailing wind. Precipitation. Cloudiness. Snowfall. Miscellaneous phenomena. Only a monthly report is required, and this is made out in triplicate. One copy is retained by the observer and two are sent to the section center. One of these is eventually sent to Washington for filing. taken at a At a special station, those observations are taken for station? which the station was founded. * THE DEVELOPMENT OF THE DAILY WEATHER MAP 362. In 1780 the Meteorological Society of the Palatinate, with its center at Manheim, began the collection and publication of detailed meteorological observations for Europe. A few reports The first from America were also included and the work continued weather map was for about thirteen years. Based upon these observations, constructed the first synoptic charts or weather maps were made by by Brandes - H. W. Brandes in 1820. He constructed them for the whole of the year 1783, and described the results, but did not publish the maps. From 1821 until his death in 1857, William Redfield of New York 362 METEOROLOGY collected meteorological observations and studied storms by means of weather maps. He seems to be the first one in this country, and Redfieid's the first one after Brandes, to have charted simultaneous work - observations in order to study storms and atmospheric move- ments and changes. In this country, E. Loomis and J. Espy also made many weather maps during the first half of the last century. In Eng- land, Mr. James Glaisher began constructing weather maps in 1848, and during the London World's Fair in 1851 the first weather maps based upon observations received by telegraph were produced. After 1851, however, the work was discontinued. In 1863 Leverrier of the The first ^ ar i s Observatory began the publication of the series of regular daily weather maps for Europe, based upon telegraphic daily 8 observations, which has continued unbroken until the present weather time. The series of daily weather maps for the United States was begun in 1871, which makes the United States the fourth country to issue regular daily maps, as the Netherlands, England, and France had already done so. At first the maps were issued in this country three times a day, at 7 A.M., 2 P.M., and 9 P.M. The change was made from three maps to two maps, issued 8 A.M. and 8 P.M., January 1, 1889. Since September 30, 1895, only one daily map, the 8 A.M. map, has been issued. In 1868 Leverrier began the publication of a daily weather map for the whole globe, but the work was discontinued after several years. Weather From 1875 to 1895 > the U ' S ' Weather Bureau made a maps for weather map for the whole northern hemisphere, and the ma P s were published for the first ten years and form a very valuable collection. This map is still made in manuscript at Washington, and one of them is reproduced as Chart L. At present, the Deutsche Seewarte, at Hamburg, and the Danish Meteorological Office, at Copenhagen, publish jointly a daily map of the North Atlantic. This together with the maps for Europe and the United States cover a good part of the northern hemisphere. THE DAILY WEATHER SERVICE OF THE U. S. WEATHER BUREAU 363. The taking and sending of the observations. At all the regular The obser- stations of the U. S. Weather Bureau, and at the Canadian vations are and Mexican stations as well, the observations are taken at eight! East- 8 A>M - an( * $ P - M - Eastern Standard Time, that is, five hours era stand- from Greenwich. This means, for example, that they are ard une. taken in the morning at 7 A.M. in Chicago, at 6 A.M. in WEATHER BUREAUS AND THEIR WORK 363 Denver, and at 5 A.M. in San Francisco. In fact, the observations are commenced a little before eight, so that the work may be finished and the telegram, which is to convey them to other stations, prepared by 8 o'clock. The observations are not telegraphed as figures, but they are reduced by means of an elaborate, yet very ingenious, complete, and A code is satisfactory code, to a series of words. According to the ^k trans code book, the data for transmission by telegraph may con- mission, sist of fifteen words and they are to be enciphered in theVollowing order:, 1. Name of station (or telegraphic designation therefor, as furnished from the central office) . 2. Pressure and temperature (corrected readings) . 3. Precipitation. 4. Direction of wind, state of weather, and maximum temperature. 5. Current wind velocity and minimum temperature, or current wind velocity and maximum temperature. 6. Minimum or maximum temperature (in twenty-four hours) which oc- curred more than twelve hours previous to observation. 7. Marked rise or fall of pressure (from stations specially designated). 8. Report on the river observation (from stations specially designated and in certain authorized reports) . 9. Frost (light, heavy, or killing). 10. Thunderstorms. 11. Fog, haze, or smoke. 12. Upper clouds. 13. Lower clouds (when sent). 14. Maximum wind velocity and direction (in accordance with special in- structions) . 15. Special monthly or weekly reports. The code is of such a character that one familiar with it is able to trans- late it without using a code book or memorizing words. The figures are conveyed, in general, by the first vowel and the consonant which pre- cedes it. The following illustrations of the second word in the telegram will make this clear: PRESSURE TEMPERATURE WORD 30.24 50 Demulsion 30.24 58 De mo crat 30.24 88 Desolate 30.54 50 Mel my 30.54 58 Memory 30.54 88 Me so type 30.56 88 Misogamy 364 METEOROLOGY Not all of the 1 fifteen words are usually sent. Numbers 1, 2, 3, 4, 5, 14, are nearly always sent; number 6, 9, 10, are generally sent; and the others only occasionally. The purpose of using a code instead of figures is primarily to secure brevity and thus save expense. As it is, the telegraphic tolls of the The purpose Weather Bureau amount to more than $100,000 annually of the code. anc j constitute one of the largest bills of the bureau. The use of a code also prevents the observations from being tampered with or read in transit, and at the same time greater accuracy is secured, as words are always transmitted with fewer mistakes than figures. The weather telegram containing from six to fifteen words is thus filed at the local telegraph office by each regular station at eight o'clock. HOW the These messages have precedence over all others, so that telegrams usually before 9.30 o'clock Washington, New York, and other large stations are in possession of the observations made at all regular stations, and all stations which produce weather maps have received a sufficient number of reports to construct an accurate weather map. This exchange of telegrams is accomplished by a circuit system. The telegrams are not sent individually from each station to every other station, but are collected at some center and then sent as a unit through a series of stations connected up as a continuous circuit. It is not customary to wait until all the telegrams have been collected and then form the circuit. As soon as a sufficient number has been collected, they are sent around the circuit. At map-producing stations, the ob- servations thus ordinarily came in in three or four installments. The last ones usually get in before 10 o'clock. The system of circuits is changed from time to time, and is worked out with the greatest care, so as to make the amount of telegraphing, and thus the expense, a minimum. At both 8 in the morning and 8 at night, the observations are sent out as just described. 364. The charting of the observations. The central station at Washington and many of the regular stations of the U. S. Weather Bureau issue daily weather maps which give, both in the form of a chart and in tables, the observations taken at 8 in the morning. A weather map, based upon the 8 P.M. observations, is prepared in manuscript at Washington and a few other stations, but it is never published except by the newspapers. The weather map (form DD) was used, prior to 1910, at nearly all of the map-producing stations, and is still used at a few of them. It is 16 X 11 inches and contains an outline map of the United States WEATHER BUREAUS AND THEIR WORK 365 10 X 6^ inches. This outline map is printed in brown and contains the names of the regular stations. On this map are charted the pres- sure, temperature, wind direction, and state of the weather. How the The pressure is indicated by a series of solid lines called iso- observations bars, drawn through places having the same pressure. The f^the 'daily lines are drawn for every tenth of an inch difference in pressure weather and the pressure is indicated at each end of the line. These map * lines inclose areas of high and low pressure, and the areas of highest and lowest pressure are marked with the words " High " and " Low." These are the extratropical cyclones and anticyclones which have been studied in a previous chapter. Dotted lines, called isotherms, connect places having the same temperature. These lines are drawn for every ten degrees, and the temperature of each line is indicated at both ends of it. The wind direction is indicated by an arrow which flies with the wind. The state of the weather is indicated by a set of symbols. O indicates a clear sky, 3 partly cloudy, cloudy, indicates rain, snow, that the report is missing. /? indicates that a thundershower occurred at the station during the preceding twelve hours. These explanations are stated briefly in the " Explanatory Notes " which are printed in one corner of the map and are here reproduced. EXPLANATORY NOTES Observations taken at 8 A.M., seventy-fifth meridian time. Air pressure reduced to sea level. ISOBARS, or continuous lines, pass through points of equal air pressure. ISOTHERMS, or dotted lines, pass through points of equal temperature. SYMBOLS indicate state of weather : O clear ; partly cloudy ; cloudy ; R rain ; S snow ; M report missing ; f& thunderstorm. ARROWS fly with the wind. i Shaded areas when used show regions of precipitation during past 24 hours. "T" in table, indicates amount too small to measure. The various weather maps which have been reproduced in the chapter on storms will illustrate the method of charting the observations. Further information is also conveyed by means of statistics in tabular form in the lower right-hand part of the map. Here are given tempera- ture at 8 A.M., 75th meridian time, temperature change in 24 hours, maximum temperature in last 24 hours, minimum temperature in last 12 hours, wind velocity 8 A.M., precipitation in last 24 hours. On the lower left-hand side of the map are given the weather predictions for 366 METEOROLOGY the particular region where the station is located and a general discus- sion of the weather conditions. Since 1910 most of the map-producing stations issue the so-called " commercial weather map," which is just the same as the maps which are published in the newspapers. The size is the same as commercial before, only the color is blue instead of brown. The isobars, weather ^ ne highs and lows, the method of indicating the direction filflP. of the wind, and the method of indicating the state of the weather are the same as before. Only four isotherms are drawn; for zero, freezing, 90 F., and 100 F. In addition, the temperature at 8 A.M., the amount of the precipitation, and the wind velocity are printed on the face of the map near each station. This, of course, ne- cessitates a change in the statistical material given in tabular form. The following legend is placed on these maps and makes the slight differences clear: Observations taken at 8 A.M., seventy-fifth meridian time. ISOBARS, or continuous lines, pass through points of equal air pressure. ISOTHERMS, or dotted lines, pass through point of equal temperature; they will be drawn only for zero, freezing, 90, and 100. SYMBOLS indicate state of weather : Q clear ; factories, large office buildings, in fact, where- butionofthe ever they will be exposed so that the information which they contain may reach quickly a large number of people. The other publications of the Weather Bureau are sent free of charge to libraries and institutions of learning. They may all be obtained by any one on payment of prices which are not excessive. WEATHER BUREAUS AND THEIR WORK 373 THE WEATHER BUREAUS OF OTHER COUNTRIES 372. Nearly all the large civilized nations of the world now maintain a national weather service and publish daily weather maps The coun _ and forecasts. This is true of all the larger countries of tries which Europe, Japan, China, Canada, India, Australia, New Zealand, ^01^ * Argentine Republic, and Algeria. weather The construction and appearance of the daily weather map is much the same in all countries. In the details, however, there is considerable difference. In all cases, the observations are taken at the same time at many stations. These are transmitted by The weather telegraph, using a code, to one or more central stations where maps of the daily weather maps are produced. The weather map other . j.i- L* i_ i. , j countries. always contains an outline map on which is charted some 01 the observational material. More of it is presented in tabular forms and forecasts and a general discussion of the weather conditions is always included. The observations are taken at different hours in different countries, but always in the early morning. The code used generally consists of a series of numbers, and does not make use of words. Wind velocity is often charted on the map as well as wind direction. This is usually done by placing a different number of barbs on the arrows. Sometimes temperature and pressure are charted on separate maps. Although the details may be different, one familiar with weather maps would find no difficulty in interpreting the weather map of any other country. The weather maps of Great Britain, France, Germany, Japan, and India are particularly interesting. Samples of the weather maps issued by the various countries may occasionally be secured from the U. S. Weather Bureau at Washington, by any one who will make good use of the illustrative material. The English weather service is managed by the Meteorological Office at South Kensington, London, S. W. There are upwards of 200 stations scattered over the British Isles and divided into The English five classes. In addition, the Royal Meteorological Society weather and the Scottish Meteorological Society have covered the country with a network of climatological and phenological stations. The weather map is issued daily, and is based on the observations received by telegraph from twenty-nine home and forty-four foreign stations. The home observations are taken at 7 A.M. and 6 P.M. West European time. The foreign observations are all taken in the early morning, but not at the same time. The telegram which conveys 374 METEOROLOGY the observations consists of six numbers of five figures each. The daily weather report consists of four quarto pages. Pages one and four contain data in tabular form, while pages two and three contain charts of the various meteorological elements. The Meteorological Office also issues a weekly and a monthly weather report. The central office of the weather service for the Dominion of Canada is located at Toronto. Here are received twice daily, by telegraph, the observations from forty-three Canadian stations and dian sixty-four stations in the United States. From these, the weather daily weather forecasts and bulletins are prepared and distributed from various centers. The preparation of the daily weather map, particularly in Europe, requires international cooperation. This has lead to International Inter _ Meteorological Congresses and to uniformity in nomen- nationai co- clature, time of observation, and the like. The adoption of the international system of cloud classification in 1891 is an example of this. There is also a set of international meteorological symbols for the various things observed. 1 It has also been proposed to have a set of international storm warnings. The system shown in Fig. 144 will probably be universally adopted. Day warnings International Storm Warnings. Cones)/ /m. For a gale commencing with wind in the NW. quadrant. ^J For a gale commencing with wind in the SW. quadrant. j^ 1 For a gale commencing with wind in the NE. quadrant. ^T For a gale commencing with wind in the SE. quadrant. For a hurricane. FIG. 144. International Storm Warnings. See Monthly Weather Review, December, 1905, p. 524. WEATHER BUREAUS AND THEIR WORK 375 SOME SPECIAL METEOROLOGICAL OBSERVATORIES; THEIR EQUIP- MENT AND WORK 373. The two most important meteorological observatories in this country which are not directly connected with the routine work of a Weather Bureau station are the Blue Hill Meteorological Blue Hill Observatory near Boston, and the Mount Weather Research and Mount Observatory at Mount Weather, in Virginia. The Blue Weather - Hill Observatory is a private observatory, and was founded by Mr. A. L. Rotch in 1885. A few years later it became associated with the Astronomical Observatory of Harvard College and its records and results are printed in the Annals of that observatory. Particularly noteworthy has been the work of this observatory in the use of kites to explore the upper air and in cloud observations. The Mount Weather Research Observatory was founded by the U. S. Weather Bureau, and most of the time is to be devoted to balloon and kite work, a study of the solar heat, and the electrical condition of the atmosphere. Other observatories, some of them connected with institutions of learning, are making valuable series of observations and publishing results, but they are undertaking no elaborate pieces of research work. Among these may be mentioned the Experiment Station of the Massa- chusetts Agricultural College at Amjierst, Mass. The two most famous mountain observatories in this country were those on Pike's Peak in Colorado, and on Mt. Washington in New Hampshire. Both have now been discontinued, but valuable observations were secured. In Europe, the most famous meteorological observatories are, perhaps, those at Pottsdam and Lindenburg near Berlin ; Pavlovsk near St. Petersburg; Trappes, Mont Souris, and Pare St. m eteoroiogi- Maur in France ; and Kew and Greenwich in England, cai obser- The most famous mountain observatories are probably of Europe. Sonnblick and Hoch Obir in Austria, Santis in Switzerland, Pic du Midi and Puy de Dome in France, Wendelstein and Brocken in Germany, Etna in Italy, and Ben Nevis in Great Britain. There are now so many observatories in addition to the regular stations of the weather service of the various countries, that only the more important ones can be named. To attempt to describe their equipment and work in small space is out of the question. 376 METEOROLOGY QUESTIONS (1) What important discovery did Benjamin Franklin make in connection with storms? (2) State in outline the history f the U. S. Weather Bureau. (3) Name the chiefs of the Weather Bureau and t6e characteristics of the admin- istration of each. (4) What has been the cost ""of the Weather Bureau? (5) Describe the three different ways in which the United States is divided into districts by the Weather Bureau. (6) How many kinds of Weather Bureau stations are there? (7) Describe the organization of the Weather Bureau at Washington. (8) How is the work at the central station at Washington sub- divided ? (9) How many are employed by the U. S. Weather Bureau and with what salaries? (10) Describe the Weather Bureau station at Washington. (11) Describe a typical Weather Bureau station. (12) Describe the instru- ment equipment of a Weather Bureau station. (13) What observations are taken at a regular Weather Bureau station ? (14) What observations are required at a cooperative station? (15) Describe the development of the daily weather map. (16) Describe the development of the weather maps for the northern hemisphere. (17) Describe the taking and sending of the observations for the daily weather map. (18) Describe the code for sending them. (19) How are the observations charted on a weather map? (20) How does the Washington weather map differ from those issued at regular stations? (21) Describe the construction of the weather map at Washington. (22) Describe the chalk plate process of producing the weather map. (23) Describe the different ways in which weather forecasts, data, and warnings are distributed. (24) De- scribe the weather service in the evenings and on Sundays and holidays. (25) Describe the educational and research work of the Weather Bureau officials. (26) What are the periodical publications of the Weather Bureau? (27) What publications appear from time to time? (28) What differences are found in the weather maps of other countries? (29) Name and describe some special meteorological observatories in this country and in Europe. TOPICS FOR INVESTIGATION (1) The history of the weather map to 1850. (2) The various codes used to transmit weather observations. (3) The processes used in constructing a weather map. (4) The history and present organization of the weather service of some country. PRACTICAL EXERCISES (1) Construct a weather map for the United States. (The necessary observa- tions can probably be secured from any map-producing station of the U. S. Weather Bureau. Material for six maps can be found in WARD'S Practical Exercises in Elementary Meteorology.) REFERENCES For the history of the U. S. Weather Bureau and its present organization, consult : MOORE, JOHN W., Meteorology, 2d ed., London, 1910. (Chapter V, pp. 44-68.) WEATHER BUREAUS AND THEIR WORK 377 MOORE, WILLIS L., "Forecasting the' Weather and Storms/' The National Geographic Magazine, June, 1905. POLIS, DR. P., Der Wetter dienst und die Meteorologie in den Vereinigten Staaten von Amerika und in Canada, Berlin, 1908. Two pamphlets published at Washington and each entitled " The Weather Bureau." The annual " Reports of the Chief of the Weather Bureau." For a description of the station equipment and the methods of taking and record- ing the observations, see : For the United States : Instructions for Cooperative Observers. Station Regulations. Instructions for Preparing Meteorological Forms. (All published by the U. S. Weather Bureau at Washington, D.C.) For Great Britain : SCOTT, R. H., The Observer's Handbook. For France : ANGOT, ALFRED, Instructions Meteor ologiques. See also Appendix IX. For the equipment and work of special meteorological observatories, see : WALDO, FRANK, Modern Meteorology, London, 1893 (pp. 160-203). For a description of the weather services of other countries, see : CAMPBELL, BAYARD, " Government Meteorological Organizations in Various Parts of the World," Quart. Jour. R. Met. Soc., April, 1899. BEBBER, W. J. VAN, Lehrbuch der Meteorologie, Stuttgart, 1890 (pp. 359-384). BORNSTEIN, R., Leitfaden der Wetterkunde, Braunschweig, 1906 (pp. 178-200). MOORE, JOHN W., Meteorology, 2d ed., London, 1910 (pp. 26-43, 69-74). PERNTER, J. M., Wetterprognose in Osterreich, Wien, 1907. Monthly Weather Review, August, 1907, p. 364. CHAPTER VIII WEATHER PREDICTIONS INTRODUCTION The general method of weather forecasting, 374, 375. The work of the U. S. Weather Bureau, 376-378. WEATHER PREDICTION CONSIDERING THE Low AS THE DOMINATING FORMATION Locating the storm center twenty-four hours ahead, 379, 380. Determining the distribution of the meteorological elements, 381. The prediction, 382, 383. WEATHER PREDICTION WHEN V-SHAPED DEPRESSIONS AND OTHER SECONDARY ISOBARIC FORMS ARE PRESENT, 384 WEATHER PREDICTION CONSIDERING THE HIGH AS THE DOMINATING FORMATION, 385 WEATHER PREDICTION BY SIMILARITY WITH PREVIOUS MAPS, 386 THE PREDICTION OF PARTICULAR AND DANGEROUS OCCURRENCES The prediction of frost, 387. The prediction of cold waves, 388. The prediction of tornadoes, 389. The prediction of destructive wind velocities, 390. The prediction of floods, 391. THE ACCURACY AND VERIFICATION OF PREDICTIONS The terms used in official predictions, 392. The system of verification, 393. The accuracy attained, 394. The popular idea of the accuracy of forecasts, 395. LONG RANGE PREDICTIONS Prediction from station normals, 396. Weather cycles, 397. Tendency of a weather type to continue, 398. Popular superstitions and credulity, 399. FORECASTS FROM LOCAL OBSERVATIONS AND APPEARANCE OF SKY Prediction from the readings of instruments and appearance of the sky, 400. Weather proverbs and weather rules, 401, 402. 378 WEATHER PREDICTIONS 379 INTRODUCTION 374. The general method of weather forecasting. Weather has been defined as the condition of the atmosphere at any particular time and place. The condition of the atmosphere is determined by the six so-called meteorological or weather elements ; and weather namely, temperature, pressure, wind, moisture, cloud, and precipitation. The best way, then, to describe the weather or to depict exactly the condition of the atmosphere is to state the numerical values -for the meteorological elements. To forecast or predict the weather is thus to foretell the values which the meteoro- logical elements are expected to have. There are three factors which determine the condition of the atmos- phere. The weather may thus be looked upon as the composite or resultant of three things : (1) the typical or normal condi- The weather tion of the atmosphere which would exist if there were no is a corn- disturbances or local influences : (2) the disturbances caused p site * . three things. by such passing meteorological formations as extratropical cyclones, anticyclones, thundershowers, tornadoes, and the like ; (3) local influences such as land and sea breeze, the presence of large bodies of warm water, mountains, etc. The typical or normal weather which would exist if there were no disturbances due to local causes or the passing of meteorological forma- tions would be different for different parts of the country. Nonnal or For the northeastern part of the United States, it would be typical something like this. The average daily temperature would wea be highest the last part of July, and then decrease steadily, and grad- ually, day by day, until the last of January, when it would be least. It would then begin to steadily and gradually increase. Each day there would be the regular daily oscillation of temperature with its maximum in the early afternoon and its minimum at sunrise. The pressure would be very constant, slightly higher in winter than in summer. There would also be the small daily oscillation with its chief maximum at 10 A.M. and its chief minimum at 4 P.M. The wind would be always moderate in velocity, blowing harder in the winter than in summer, and by day than by night. The wind direction would shift from north- west in winter to southwest in summer and back again. The absolute and relative humidity would also show a regular daily and annual variation. There would be very few clouds ; perhaps now and then a cumulus due to convection. Precipitation would be entirely lacking. 380 METEOROLOGY In short, the regular daily and annual variations in the meteorological elements would be present, but there would be no irregular fluctuations. It would be difficult to give numerical values to this theoretical state of things. They would be different from the normals found from obser- vation for any station, as cloud and precipitation are lacking. How- ever, in practical forecasting, one is not concerned with this theoretical state of things, but with the normals as derived from observation, since cloud and precipitation are actually present. The various normals, then, in connection with the meteorological elements, may be considered to represent normal or typical weather. The passing meteorological formations which exert the chief influence in the United States are the lows and highs (the extratropical cyclones and anticyclones). V-shaped depressions and other second- The passing J . ' * K meteoroiogi- ary forms of isobars should perhaps also be mentioned. cai forma- Occasionally, a tropical cyclone visits the Gulf States. Thundershowers and tornadoes should also be included, but thundershowers always accompany a low or V-shaped depression, and tornadoes are always associated with thundershowers. Thus, extra- tropical cyclones and anticyclones are by far the most important forma- tions to consider. In fact, New England is crossed by such a ceaseless procession of these two formations, that the weather is nearly always dominated by one or the other, and can hardly ever be said to be typical. All of these formations have already been critically studied in the chapter on storms. Local influences are not usually very numerous or of great importance. If a place is located on the seacoast, then land and sea breeze must be Local in- taken into account. In forecasting the weather for New fluences. York State, the presence of the Great Lakes play an impor- tant part, particularly in the late autumn and early winter, when they are still much warmer than the land and are putting large quantities of moisture into the atmosphere. A large river also sometimes influences weather conditions. 375. The various normal values, when once determined for the different meteorological elements, hold for all time. Thus normal The lows weather is known for weeks or years ahead. The local and highs influences are usually unimportant, and can be estimated portant ** with fair precision. The chief difficulty, and practically the things in only difficulty, is thus to determine the influence of the tmg< passing meteorological formation. If the lows and highs were even always typical, weather forecasting would be an easy matter. WEATHER PREDICTIONS 381 That is, if they all followed typical courses with known velocities and were typical as regards the distribution of the meteorological elements about them, their influences could be readily estimated. As it is, weather forecasting is by no means an easy matter, as the lows and highs are seldom typical as regards path, velocity, or characteristics. Weather forecasting, then, really turns on this one thing, the estimation of the influence of the passing meteorological formations. The general method to be followed in predicting the weather is thus apparent. The first thing to do is to estimate the influence which the passing meteoro- logical formation is going to exert one or two days in the future. The local influences must next be estimated. When these two have been combined with normal weather, the resultant is the expected or pre- dicted weather. The details in this process will be considered a little later. 376. The work of the U. S. Weather Bureau. For the purpose of weather forecasting, the United States is divided by the U. S. Weather Bureau into six forecast districts with centers at Washing- where the ton, Chicago, Denver, San Francisco, Portland, and New forecasts of Orleans. These districts are shown on the map in figure 142. Based on the 8 A.M. observations, the officials in charge Bureau are of each forecast district issue weather forecasts and warnings for the respective districts. As soon as made, these are forwarded by telegraph to the Central Office at Washington, and forecasts for all districts appear on the Washington weather map. Based on the 8 P.M. observations, forecasts and warnings are issued from Portland and San Francisco only for the respective districts. The forecasts and warnings are issued from Washington for all the other districts. A local forecast official at a map-producing station issues forecasts for the immediate vicinity of the station only. On the weather map, this forecast appears, and also the forecasts for the state, and perhaps adjoining states, which have come from the respective forecast centers. The prediction based on the 8 A.M. observations are for 36 hours, while those based on the 8 P.M. observations are for the following 48 hours. The forecasts consist of predictions of temperature, wind, and state of the weather. This last includes the amount of cloud and the kind and quantity of precipitation. Pressure and moisture are Of what the never forecasted. Cold waves, frosts, and high wind ve- locities (storms) are also predicted and the appropriate warnings sent out. A local forecast official at a map-producing station forecasts temperature, wind, and state of the weather only. All fore- 382 METEOROLOGY casts and warnings in connection with high wind velocities (storms), cold waves, and frosts come from the forecast centers. These forecasts and warnings of the Weather Bureau are not only printed on the daily weather map, but they receive a wide distribution by means of newspapers, cards, flags, and special messengers. These various methods of distribution have already been discussed in section 368. 377. Long training is required on the part of the Weather Bureau officials before they are allowed to issue forecasts. An official in charge The training ^ a station where forecasts are not issued and first assistants of a fore- at large stations are permitted to make practice forecasts. After this has been continued with fair success for a year, the authority to issue forecasts may be given. Each official who is authorized to make local forecasts is still required to make practice forecasts for the state in which the station is located. He must also make practice storm, cold wave, and frost forecasts for the local station where he is. All of these practice forecasts are immediately mailed to Washington for verification. Ability to forecast well depends upon characteristics of mind as well as careful training. Greely in his American Weather says : " The Th skill of a weather predictor arises largely from his alert teristics of a comprehensiveness of mind, accurate and retentive memory, successful phlegmatic . but confident temperament, and long experi- ence." A weather forecaster must take in many details at a glance ; he must recollect past occurrences ; he must not lose confi- dence in himself or his ability, even if an occasional prediction goes wrong. 378. The general method of weather prediction in other countries is the same as in this ; although the details may be quite different. Most The countries issue but one set of forecasts and warnings each methods of day. This is often in the early afternoon, instead of the forecastlfin mornm g- Some countries get supplementary telegrams other coun- from a few stations just before the forecasts are issued. The forecasts are usually issued from a central station for the whole country, although this is not true for the larger countries. The signals used and the methods of distributing the forecasts are often quite different. WEATHER PREDICTIONS 383 WEATHER PREDICTION CONSIDERING THE Low AS THE DOMINATING FORMATION 379. Locating the storm center twenty-four hours ahead. Since the estimation of the disturbances caused by passing me- teorological formations is the thing of chief importance in forecasting the coming weather, weather prediction may be tions must considered under three heads : (1) when a passing low is ^ered the dominating formation; (2) when a V-shaped depression or some secondary isobaric form is exerting the chief influence ; (3) when a passing high is exerting the dominating influence. If it is a low which is going to dominate the weather for the immediate future, the first thing to do is to determine where the center is expected to be twenty-four and forty-eight hours ahead. A fairly exact estimation of where the center will be twenty-four hours J t ^ e [ i ocat _ ahead must be made ; the position forty-eight hours ahead ing a low need be only roughly estimated. The general rule is to hours ahead determine the track which the low seems to be following, and then put it ahead on this track the distance normally covered in twenty- four and forty-eight hours. If a low has just formed, or has just come into the field of observation, the past is no guide as to the track. The safest rule in these cases is to assume that the low probably will follow the most frequented track which passes through the locality where it appeared. For this the Van Cleef system of tracks (see section 300) will probably be the most satisfactory. If the forecasting is being done for the northeastern part of the United States, the lows have usually been observed and charted on the weather map a day or two before they become the dominating influence. If this is the case, note which track the low has been following and put it ahead on this track. Either the Bigelow, Russell, or Van Cleef system of tracks may be used. It probably would be better to use that system which contains the track which the low seems to be following most exactly. The normal velocity of motion is about 900 miles a day in winter and about 600 in summer and intermediate values in other seasons. (See section 302.) After it has been determined where the center of the low would be twenty-four and forty-eight hours ahead, if it followed its track with normal velocity, it is customary for the skilled forecaster to modify this estimation by taking account of other considerations. This simply means that as a result of his long experience, he has formed certain empirical rules for his own guidance. This shows the necessity of long 384 METEOROLOGY experience before reliable forecasts can be made and also the difficulty in explaining exactly how a weather forecast is made. Some of these Someempir- rlues are the following: (1) Lows near each other tend to ical rules coalesce. This is particularly true if one low is in the ing the 7 " northern Mississippi Valley and the other in the southern, normal posi- or if one low is coming up the Atlantic Coast and the other is coming eastward across the Great Lakes. They usually coalesce at an intermediate point and become more intense than either component. Often the appearance of coalescing may be given by the fading away of one low, while the other becomes more intense and dom- inant. (2) Lows and highs repel each other. If a high is directly on the track of a coming low, the low is likely to be retarded or to be de- flected from its normal track. (3) Lows tend to follow the record of previous days. This means that if a low is a slow-moving one, it tends to retain that characteristic all through its life history. (4) Lows tend to move toward the areas of greatest rainfall during the preceding twenty-four hours. (5) Lows tend to move toward the areas of highest dewpoint. (6) Lows tend to move toward areas of least wind velocity. (7) Lows tend to grow more intense as they approach bodies of water as the Great Lakes or Atlantic coast. (8) Lows tend to move faster as the pressure grows less. (9) Lows with above normal winds weaken, while lows with below normal winds develop lower pressure. (10) Lows on a curved track tend to move faster after they begin to move northeastward. This set of rules makes no pretense at being complete. Only the well-recognized and more important ones have been given. Each forecaster, as he acquires experience, will prefer to formulate his own experience. They will be found very useful, however, by a be- ginner. The method, then, of determining where the center of a low is expected to be twenty-four and forty-eight hours ahead is to put the low forward The method on ^ ne track which it seems to be following the normal dis- briefly tance, and then to modify this estimation by taking account of certain general principles which have been learned from experience and perhaps formulated as rules. As a result of this process, the forecaster will come to a very definite conclusion as to where the center of the low is expected to be twenty-four hours ahead, and he will also have a general idea of where he expects it to be forty-eight hours ahead. 380. In this connection, the question can be well raised as to what really determines the track which a low is to follow and its velocity WEATHER PREDICTIONS 385 of motion. The path and velocity are probably completely determined by the general drift of the atmosphere, the characteristics of the dis- tribution of the meteorological elements about the formation itself, the location and characteristics of the surrounding W hf c h deter- meteorological formations, the surface topography of the mine the country, and the meteorological condition of the country. It velocity of will be seen at once that the factors which enter in are large motion f * in number and very complex. If the value and relative im- portance of all of those factors could be determined from a study of many weather maps, it then would be possible in any particular case to determine the amount of motion, and the direction of motion, which each factor would cause, and the resultant for all the factors would give the direction of motion and the velocity of motion of the low. This can, however, probably never be done with precision. These factors are, however, of very different importance. Some are of prime importance, and many have such insignificant the problem influence that they can almost be neglected. The general 2bk P S " drift of the atmosphere and the distribution of pressure about the low itself are probably the most important factors in determining the path and velocity of motion of a low. The distribution of the tempera- ture over a country is probably the next most important factor, while the distribution of the winds about the low as regards direction and velocity, and whether the low is accompanied by unusually high or low wind velocities, are also very important considerations. The location of the rain area during the preceding twenty-four hours and the direction and intensity of the surrounding highs and lows perhaps stand next in importance. As an approximation to the solution of the general problem, the path and velocity of motion of a low might be worked out, using the two most important factors only. This has been done by Mr. Edward H. Bowie, who was then local forecast official at method of St. Louis, Mo., and has been recently transferred to Wash- locating a ington to be forecaster there. His investigation will be found in the Monthly Weather Review for February, 1906 (Vol. XXXIV, p. 61), and the reader must be referred to this article for a full discussion of his method. He first determines, for the various months in the year, the twenty-four hour drift of the atmosphere for all parts of the United States. Then, in the case of the low in question, the pressure gradient toward the center from the north, northeast, east, southeast, etc., is found from the weather map. The resultant of these eight pressure 2c 386 METEOROLOGY gradients is then found, and this represents the direction and magnitude of the unbalanced pressure which is forcing the low to move. The com- bination of this unbalanced pressure with the general drift of the atmos- phere gives the direction of motion and the velocity of motion of the low. Only two of the general factors are taken account of in this method, but the high degree of accuracy which Mr. Bowie has attained in forecasting, by the use of his methods, attests the prime importance of these two factors and the value of his method. Even when the Bowie system is not used in full, the underlying principle is of great value to the fore- caster. One can notice very easily in which quadrant of a low the iso- bars are most closely packed together. Here the pressure gradient will be largest, and the low will, in general, move away from tlnVarea'. Trabert 1 has recently called attention to the fact that the distribution of temperature over a country may determine the direction of motion of a low. Usually the temperature increases from the north toward the south. In this case, the winds on the eastern side of a low are bringing warmer, moisture-laden air from the south, while the winds on the western side are bringing cold, dry air from the north. The pressure will thus be falling on the eastern side and rising on the western side, and the low will thus move towards the east. If the temperature dis- tribution over a country were different, a different direction of motion might result. The motion, in general, would be at right angles to the temperature gradient. In 1905 the Belgian Astronomical Society instituted an international competition in forecasting at Liege, and the first prize was awarded to The Guil- Gabriel Guilbert of Caen. Later (1909) Guilbert published bert rules. hj s me thod of forecasting in book form under the title " Nouvelle methode de prevision du temps." His system is really based upon three rules, all of which relate to the winds which surround the low. These rules may be summarized and stated as follows : 2 (1) Every depression that gives birth to a wind stronger than the normal will fill up more or less rapidly. On the other hand, every de- pression that forms without giving rise to winds of corresponding forqe will deepen, and often depressions that are apparently feeble will be transformed into true storms. (2) When a depression is surrounded by winds having varying degrees of excess or deficiency, as compared with the normal wind, it moves 1 See : "Die Zugrichtung der Depressionen," by WILH. TKABERT, in Das Wetter, February, 1911. 2 See Monthly Weather Review, May, 1907, p. 210. WEATHER PREDICTIONS 387 towards the region of least resistance. These favorable areas are made up of regions in which the winds are relatively light, and especially of such as have divergent winds with respect to the center of the depression. (3) The rise of pressure takes place along a direction normal to the wind that is relatively too high, and it proceeds from right to left ; an excessive wind causes a rise of pressure on its left. Here, again, we have a whole system of forecasting built upon one of the more important of the many factors which determine the path and velocity of motion of a low. The wind direction and its velocity and the pressure gradient are very closely related, so that, in a way, the JBowie system and Guilbert's rules rest on the same foundation. These rules are of great value to the forecaster, as they can be easily applied, and help to guide one's estimation as to the probable direction and velocity of motion of a low. It will also be noticed that all of the empirical rules which were stated in section 379 for locating the center of a low twenty-four hours ahead depend on the influence of one or several of the various factors which determine the path of a low. 381. Determining the distribution of the meteorological elements. The method of determining the probable location of the center of the dominating low twenty-four and forty-eight hours ahead First study has just been fully discussecf. The next step is to determine critically the the probable distribution of the meteorological elements J^^^eie^ about the low in its predicted location. In order to do this, ments about first study critically the distribution of the elements about * the low as indicated on the weather map which is serving as the basis for the forecast. Notice in detail just what the values of temperature, pressure, wind, moisture, cloud, and precipitation are in different parts of the area covered by the low. If there is any departure from the normal distribution about a low, try to explain each departure from normal. It might be well in this connection to emphasize the neces- sity of noting carefully whether the low has a wind shift line, or is accompanied by the winter or summer type of cloud area. Next decide whether the predicted location of the center of the low twenty- Next esti . four hours ahead is going to change the previous distribution mate the in any way. Nearness to a body of water or a range of causedlby mountains and the meteorological condition of the country the new probably exert the greatest modifying influence. Finally assume that the distribution of the elements about the low, twenty- four hours ahead, will be the same as on the previous weather map with 388 METEOROLOGY the exception of any changes which the new location of the low may be expected to cause. The rule, then, briefly expressed, is to assume The rule that the distribution of the elements will be what it was, briefly ex- with the exception of the changes caused by the new sur- pressed. roun dingS. 382. The prediction. The various steps which must be taken be- fore a weather forecast can be made have been discussed individually and in detail. It has been seen that the weather is the corn- factors must posite or resultant of three factors : normal weather, local be summed influence, and the disturbances caused by passing meteoro- logical formations. Normal weather is known for days and months in advance. The local influences are usually insignificant. No rules can be laid down for these. They must be learned from experience for each individual place. In the case of a low, the method of locating its center and determining the distribution of the elements about it, and thus its influence twenty-four or forty-eight hours ahead, has been explained. It now remains simply to sum up these three factors in order to make a weather prediction. It is the best of practice for a beginner, or even for one who has ac- quired considerable skill in forecasting, to force oneself to form an exact The making picture of what the weather is expected to be. This exact- of exact ness can well be carried, in some Cases, to the point of estimat- gives a vaiu- m & ^ ne numerical value which each of the elements is expected able train- to have. It must not be expected, however, that such a fore- mg * cast will verify in every detail. Something will almost always be wrong, but fortunately the forecaster is not left to helplessly wonder what went wrong. As soon as the weather has occurred, and the next weather map has been received, it is always possible to see exactly what was overestimated or underestimated, and these very mistakes become stepping stones for the acquirement of more skill, ac- curacy, and confidence in forecasting. It will bear repetition that these exact forecasts have great value as a training. If such an exact forecast is to be made, the best order in which to predict the values of the mete- orological elements is perhaps this : (1) pressure ; (2) wind direction and velocity ; (3) clouds amount and kind ; (4) moisture ; (5) tem- perature ; (6) precipitation kind and amount. The forecasts of temperature, wind, and state of the weather issued by the U. S. Weather Bureau are definite and explicit, but do not usually contain estimates of the numerical values that the elements are ex- pected to have. The various phrases which may be used, and their WEATHER PREDICTIONS 389 exact meaning, are laid down in the " Station Regulations " of the U. S. Weather Bureau. Some of these regulations are here quoted. " Forecasts based on the 8 A.M. reports will be for the night The fore _ of the current day, 8 P.M. to 8 A.M., and for the following casts as day, 8 A.M. to 8 P.M. They may cover the afternoon of the & current day only when marked changes are expected. Am- Weather biguous expressions will be avoided in the preparation of Bureau - forecasts. When conditions are so uncertain as to make the accuracy of a forecast doubtful, the word ' probably ' or ' possibly ' may" be used. " Forecasts of temperature changes will- be made when the 24-hour changes at 8 A.M. and 8 P.M. of the following day are- expected to equal or exceed 6 in the months, June, July, August, and September ; 8 in April, May, October and November; and 10 in December, January, February, and March. The terms ' warmer, colder, decidedly warmer, decidedly colder ' will be used in describing the corresponding changes. Forecasts of fair, partly cloudy, or cloudy, will be made when pre- cipitation to the amount of 0.01 inch or more is not expected. When precipitation of 0.01 inch or more is expected, its character will be indicated by the use of such terms as ' rain, snow, local rain, showers, local snow, snow flurries, thundershowers, thunder- storms/ etc." 383. The following example will serve to illustrate the method of making an exact prediction and also a more general one, such as would be issued by a local forecast official. The weather maps for 8 A.M., Sunday, Dec. 8, 1907, and for 8 A.M., Monday, Dec. 9, are reproduced as Charts XXXVIII and XXXIX. which the The first table contains various observations made at many j^based. regular weather bureau stations at 8 A.M. on Monday, while the second table contains the detailed observations made at Albany, N.Y., at 8 A.M. and 8 P.M. on Sunday and at 8 A.M. on Monday, The problem is, on the basis of this material, to form a definite picture of the weather changes expected at Albany during the next thirty-six hours, and then to formulate such a forecast as would be issued by a local forecast official. The forecast must be made separately for Monday night (8 P.M. Monday to 8 A.M. Tuesday) and for Tuesday (8 A.M. to 8 P.M. of that day), thus covering a period of thirty-six hours from the time of the last observations which can be utilized. Similar observational material and the weather maps are always available whenever a fore- cast is made. 390 METEOROLOGY OBSERVATIONS 8 A.M. MONDAY, DEC. 9, 1907, AT VARIOUS WEATHER BUREAU STATIONS a o TEMPERATURE 1 * g TEMPERATURE P 1 1 S i OH O Q hJ W w ^ O O h-J O S ^ w M fc *5 ~~ fc STATIONS 00 ^ a S O hH STATIONS P 00 esi o 1 " 1 FH Min. in Max. in H < & Iss Min. in Max. in ! ^3 |g | Last 12 Last 24 ^ K M S ^ Last 12 Last 24 * 02 ^ j 5 o Hours Hours a g z * h-J r Van Cleef system of tracks may be used. Each of the low contains a well-marked track which corresponds with the tours^ahead. m tion of the low. One would expect it to move up the Mississippi Valley, across the Great Lakes, and down the St. Lawrence Valley. The rule for locating the center twenty-four hours ahead is to put it along on the track which it seems to be following the normal amount, and then to modify the estimation by other considera- tions. For December the normal amount is something less than 900 miles in twenty-four hours. 1 One would thus expect the center to be just north of lake Erie at 8 A.M., Tuesday. Now, are there any modify- ing circumstances ? Near-by lows tend to coalesce and lows and highs repel each other. There is no near-by low with which to coalesce, and the Atlantic high is too weak and too far away to retard it. The west- ern high is pushing in vigorously, and might tend to push the low east- ward a little. Lows tend to follow the record of previous days. The low has hardly covered the 900 miles during the previous twenty-four hours, so that it seems to be one moving with a little less than normal speed . Lows tend to move faster after they recurve, and they also tend to move faster and grow a little deeper as they approach the Atlantic Ocean. It will thus be seen that the slow motion of the previous twenty-four hours has been explained, and one would expect greater speed during the next twenty-four hours. If the Bowie system of forecasting is used, it will be seen at once that the pressure lines are crowded together in the north- western and western portion. This means that the formation would be pushed eastward. It will also be noticed that the region of high winds is between the low and the following high. If the Guilbert rules are held in mind, one would expect again that the formation would move more 1 For the numerical data about highs and lows, see Chapter VI, part B. WEATHER PREDICTIONS 393 than the normal amount eastward. As a result of all these modifying considerations, one would put the center a little farther along its course and somewhat eastward. As a final conclusion, then, Lake Ontario might be chosen as the probable location of the center of the low, 8 A.M. on Tuesday. The Atlantic high will have moved far out over the ocean, and the western high will follow eastward behind the low; but neither of these formations will exert any influence on the weather able motion at Albany during the time interval in 'question. of the other The next question is this. What will be the probable dis- tribution of the meteorological elements about the low on this Tuesday morning, when its center is over Lake Ontario. To answer this one must first study critically the present distribution and then esti- The ^ stri , mate the probable changes caused by the new location. The bution of central pressure is 29.6 inches. This is just about normal JJogSaiTie- for a fully developed typical low, and the form of the isobars ments about and the direction of the longer axis of the oval are also very e low * typical. The wind directions form a very perfect counterclockwise spiral about the area of low pressure. At only a very few stations there are un- usual or unexpected wind directions. In the northern, eastern, and south- ern quadrants the wind velocity is light or moderate. In the belt, at the west, between the low and the high, the wind velocities are large, reach- ing the velocity of a storm wind, forty or more miles per hour, at a few stations. The temperature lines are much distorted. The low is char- acterized by a very marked rise of temperature in front of it, and a sharp drop on the western side of the center. The cloud area extends far out to the east, reaching to the coast. On the western side, the nimbus cloud area quickly gives place to a clear sky. The precipitation has taken the form of rain, and has been general in the northern, eastern, and southern portions of the storm. The quantity has not been large, as it varies from a mere trace to not more than half an inch at any sta- tion. What changes may be expected in this distribution of the elements due to the new location of the center? As a low crosses the Great Lakes and nears the Atlantic coast, it comes into regions of greater moisture and usually becomes a little deeper. One would thus expect, Tuesday morning, to find the central pressure a little less, and the amount of precipitation somewhat larger, but in other respects without change. During Tuesday, one would expect the low to pass down the St. Lawrence Valley and past the Albany station. 394 METEOROLOGY The definite picture of the expected changes in each element may now be formed. Pressure. At 8 A.M., Monday, at Albany, the pressure was 30.19, the wind direction was north, and its velocity was two miles per hour. The definite ^.11 this means that the departing high was just on the point picture of of passing over the weather control to the coming low. weather* 1 ^ ne wou ld thus expect the pressure to drop steadily during changes at Monday and Tuesday. If the estimated central pressure of the low at 8 A.M. -on Tuesday is placed at 29.4, then one would expect the pressure at Albany to be about 29.5 or 29.6 at this time. It should continue to fall during Tuesday, reaching its lowest, perhaps 29.4 or even a little lower, Tuesday afternoon or night. After the center of the low had passed nearest to Albany, the pressure would, of course, begin to rise. Wind. At 8 A.M., Monday, the wind was north with a velocity of only two miles per hour. It will be remembered that the wind velocities on the eastern side of the low were light or moderate. One would thus expect only light or moderate winds until the center of the low had passed. The direction should change to the east or southeast by Mon- day night. It would then shift more to the south, and should continue from some southerly quarter during most of Tuesday. Later on Tues- day the wind should shift to the southwest and west, and eventually northwest. The Hudson River Valley, however, runs north and south. Thus, due to a local influence, one would expect the easterly and westerly components to be lessened and the winds to be mostly southerly or northerly. Cloud.. At 8 A.M., Monday, it was foggy almost to the point of rain, but no rain had fallen. One would expect it to remain totally cloudy during the whole of Monday and Tuesday. The nimbus cloud area ought to be reached before Monday night, and it ought to continue during the whole of Tuesday. If it should rain intermittently, the cloud would, of course, be called some form of stratus, instead of nimbus when there was no precipitation. Moisture. At 8 A.M., Monday, it was foggy with a relative humidity of 100 per cent. With the rising temperature, the relative humidity would grow somewhat less, but it should remain high all during the storm. Temperature. At 8 A.M., Monday, the temperature was 28.0. The maximum during the previous day had been 44. The coming storm was characterized by a marked rise in temperature. One would WEATHER PREDICTIONS 395 thus expect a maximum on Monday well above 44, say 50 ; but little drop during the night ; a maximum on Tuesday even higher than on Monday, say 55 to 60. By Tuesday night, the temperature should begin its rather sharp drop. Precipitation. Up to 8 A.M., Monday, there had been no precipi- tation. With the rapidly rising temperature, the kind of precipitation would, without doubt, be rain. The amount at various stations had been moderate, a trace to - a half inch. The quantity should grow larger. One would thus expect some during Monday night and a half inch or more during Tuesday. The definite picture of the expected weather changes during Monday night and Tuesday has now been given. It remains to formulate the forecast such as a local forecast official would issue. There The official are no predictions of a cold wave, frosts, or destructive winds forecast - to be made. The forecast will thus be concerned with temperature and state of the weather. A forecast in accord with the definite picture which has just been formed would be : For Albany and vicinity, rain and warmer to-night ; Tuesday, rain with continued high tempera- tures. The weather map for 8 A.M., Tuesday, Dec. 10, 1907, is repro- duced as Chart XL, and the accompanying table contains the detailed observations at Albany at 8 P.M. Monday, and at 8 A.M. Theverifica- and 8 P.M. Tuesday. By studying this map and the table tion - it becomes evident at once to what extent the predictions verify. The illustration which has just been given will serve to elucidate the method of forecastings. There is no royal road to becoming a skillful forecaster. Practice is the essential thing. At the begin- Experience ning many mistakes will be made. Influences will be over- is essential - estimated or underestimated, or overlooked. When the next map is issued and the weather occurs, the forecaster is not left to wonder help- lessly what went wrong. It is nearly always evident what was incor- rectly estimated, and these very mistakes are the most valuable things in the acquirement of skill and experience. 396 METEOROLOGY OBSERVATIONS AT ALBANY DEC. 9, 1907 DEC. 10, 1907 Temp. (F.) 8 A.M 45.5 51.0 27.0 53.0 53.0 45.5 48.2 55.0 48.2 Max. previous 12 hours. Min. previous 12 hours 8 P.M Max. previous 12 hours Min. previous 12 hours. . . . Pressure reduced to sea level 8 A.M 29.95 30.19 29.95 29.52 29.95 29.52 29.40 29.52 29.40 Max. previous 12 hours Min. previous 12 hours 8 P.M. Max. previous 12 hours. Min. previous 12 hours. Wind [Direction 8 A.M S. 15 S. . N. 13 3 Direction 8 P.M Velocity 8 A.M Velocity 8 P.M Moisture f Rel. humid. 8 A.M 70 90 92 (Rel. humid. 8 P.M Clouds IKind and direction of motion 8 A.M. Cloudiness 8 A.M Stratus from S. Cloudy (10) Nimbus from S. Cloudy (10) STimbus from N . Cloudy (10) light fog Kind and direction of motion 8 P.M. Cloudiness 8 P.M Precipi- tation in inches Snow on g Snow on g Amount at 8 A.M. during previous 12 hours. Trace Rain Trace .10 .58 Rain Trace Trace Amount at 8 P.M. during previous 12 hours Kind round 8 A.M round 8 P.M. . WEATHER PREDICTION WHEN V-SHAPED DEPRESSIONS AND OTHER SECONDARY ISOBARIC FORMS ARE PRESENT 384. Whenever a weather map is to serve as the basis of a weather forecast, the form of the isobaric lines must be critically studied. If these All second- ^ nes nave anv otner ^ orm tnan tnat of ova ^ s surrounding ary isobaric areas of low or high pressure, they must be given particular beotcd USt atten ti n > anc * this is especially the case if the general east- ward motion of all meteorological formations will bring them near enough to the station in question to exert an influence during the WEATHER PREDICTIONS 397 following thirty-six or forty-eight hours. Two of these isobaric forms, other than the oval lows and highs, have received special names and merit consideration. These are V-shaped depressions, and secondary lows. Other isobaric forms have been named, and have, perhaps, individual characteristics, but their influence on weather is so slight that they do not merit consideration. A V-shaped depression is a pocket-like bulge or projection in an iso- baric line, and may have several different meanings. If two areas of low pressure are quite near together, the isobaric lines in The various crossing the ridge of higher pressure between them, must kinds of v- have this V-shaped bend. The same is true of the saddle pre-Sons* 5 " which separates two areas of high pressure which are near and what together. In these two cases, the V-shaped bulge has no they Slgmfy> particular significance or importance. If the isobaric lines in the southern quadrant of a low all have this V-shaped bulge, it will prob- ably be found that the low has a well-marked wind shift line. Lows with this peculiarity have already been fully discussed, and the significance of this V-shaped bulge, as far as coming weather is concerned, is evident. If an isobaric line surrounding an extensive high shows such a bulge, it often means that an area of low pressure is about to form, that is, it marks the origin of an extratropical cyclone. This is particularly liable to be the case if, in addition to this bulge, the wind is blowing from dif- ferent directions at near-by stations and a small cloud area has formed. Sometimes a V-shaped depression which does not become a definite low will, however, cau^e thundershowers in summer or a snow flurry in winter. A V-shaped bulge in connection with an area of low pressure may mean that a secondary low is about to form. The general rule for weather forecasting, in connection with V-shaped depressions is, then, to note carefully if they exist and, if so, to remember that they may signify simply the dividing line between near-by lows or highs, the presence of a wind-shift line in connection with a low, the formation of a new area of low pressure, or the formation of a secondary low. A secondary low is an area of low pressure of much smaller extent but sometimes of greater violence, which may exist in the southern quad- rants of a much larger low. They are not very common in secondary this country, but quite common in Europe. They usually lows - move eastward and northward, faster than the big low, so that they appear to circle around it in a counterclockwise direction. Sometimes the big low fades away and the secondary assumesN^t^f importance. 398 METEOROLOGY In weather prediction, it is best to consider the secondary low simply as another low, and forecast accordingly. The explanation of the popular expression that " the weather, instead of clearing off as it gave promise, has had a relapse " is often to be found in the existence of these second- ary lows. Charts XLI and XLII, which represent the daily weather maps for Dec. 27 and 28, 1904, illustrate well the formation of a second- ary low from a V-shaped bulge. On the 27th only a slight V-shaped bulge is in evidence, while on the 28th the secondary low is already fully formed. The weather maps for Jan. 27 and 28, 1908 ; May 20 and 21, 1908; April 30 and May 1, 1909; March 6 and 7, 1910; Oct. 27 and 28, 1910 ; Nov. 27 and 28, 1910, also illustrate well the forma- tion of secondaries. WEATHER PREDICTION CONSIDERING THE HIGH AS THE DOMINATING FORMATION 385. If it is a passing high instead of a low which is going to dominate the weather for the coming twenty-four or forty-eight hours, the general method of procedure, in making a forecast, is just the same oAhe as *f ** were a low. Determine first where the center of the method of high will be twenty-four and forty-eight hours ahead. This predictions. * s done by noting the path which it seems to be following, and then putting it ahead on this path the normal distance covered in the given time. This estimated position is then modified by taking into account other considerations, which may be simply the general result of experience or may have taken the form of definite rules. Next assume that the distribution of the meteorological elements about it will be what it was, with the exception of any changes which its new location might cause. It is important in this connection to hold in mind the two chief ways in which an area of high pressure builds and becomes more intense. These are through the discharge of air from lows and through growing colder due chiefly to radiation to a clear sky. After the probable location of the high and the distribution of the elements have been determined, the method of arriving at the fore- cast is just the same as that for the low, which has been treated in detail. WEATHER PREDICTION BY SIMILARITY WITH PREVIOUS MAPS 386. Prediction by similarity with previous maps is an entirely dif- ferent method Ji^weather forecasting from that which has just been de- WEATHER PREDICTIONS 399 scribed in detail. It might well be called the mechanical or auto- matic method of forecasting, because individual judgment plays no part. 1 The first step is to find, in the series of weather This is the maps, that one which is the most similar to the map which mechanical is to be the basis of the forecast. Weather maps have method - been issued regularly in this country since 1871, so that The method the file covers nearly forty years. In order to use this stated * method of forecasting, it presupposes that the maps have been classi- fied and indexed so that the similar map can be found. In this long series, there probably would always be at least one map which would be fairly similar to the map under consideration. The method of making the forecast is simply to assume that the weather changes which took place in the previous instance will again occur in the present case. The six daily weather maps which are reproduced as Charts XLIII- XLVIII illustrate well this method of forecasting. The three maps for Thursday, Jan. 9, 1908, for Tuesday, Jan. 14, 1908, and for An aiustra _ Monday, Jan. 27, 1908, are extremely similar. In each tionofthe case, there is a high with a very moderate central pressure method - (30.1-30.3) with its center over the Gulf States, and this high, in each case, has a lip or projection extending over the Great Lakes. In each case, there is a low of considerable intensity which is departing by way of the lower St. Lawrence Valley, and another low of considerable inten- sity which* is pushing in from the extreme northwest. The high is accompanied in each case by a decided drop in temperature over the Great Lakes. Further points of similarity can be detected by studying critically the distribution of the elements in the three cases. The three maps which follow are extremely similar. In each case, the high drifted eastward the same amount and the extension over the Great Lakes became more pronounced and was accompanied by a still further drop in temperature. In each case the coming low moved eastward about the same amount and developed a double center, becoming really a trough of low pressure extending from the Great Lakes to the Gulf. Weather prediction for the Middle Atlantic and New England states based upon the principle of similarity would have received very com- plete verification. THE PREDICTION OF PECULIAR AND DANGEROUS OCCURRENCES 387. The prediction of frost. The method of predicting the damage- causing frosts of the late spring and early autumn was described in 400 f METEOROLOGY Chapter V, B, (sections 206-210). A frost is predicted by the officials of the U. S. Weather Bureau in exactly the same way as any other The method temperature. On the basis of the weather map the fore- of frost caster must estimate the probable minimum temperature on ' (real air temperature in the thermometer shelter) on the fol- lowing morning. In making this estimation he will be guided largely by the probable clearness of the sky during the night and the prob- able wind velocity. It will be remembered that a clear sky and ab- sence of wind are essential for a large drop in temperature and a con- sequent frost. If, after the probable minimum temperature in the thermometer shelter has been estimated, it is desired to determine the probable temperature of low-growing vegetation in the open at various The ca "^ s points in a limited area surrounding the station in question, ence be- three things must be taken into account : (1) Plant tempera- tween the tures go below the real air temperature because they are not in a ther- sheltered and are free to radiate their heat. (2) Vegetation mometer j s i oca ted near the ground and not at the height of the ther- shelter and of vegetation mometer shelter. (3) The variation in temperature over a under dif- limited area is often considerable. Thus the temperature ferentcondi- . tions. of vegetation in the open, near the ground, and in the coldest part of a limited area may be expected to be from 5 to even, in extreme cases, 20 lower than the estimated minimum in the ther- mometer shelter. These facts are of vital importance and must not be overlooked. Frost predictions and warnings are not issued by local forecast officials, but are issued only by the district forecasters from the respective centers. What con- Warnings of light and heavy frost will be verified by the stitutes a occurrence of light and heavy frost respectively ; and also Weather* 6 ^ v a re Prted minimum temperature of 40 and 32 respec- Bureau tively, accompanied by clear or partly cloudy weather and light winds or calm during the period for which frost is fore- cast. Thus what constitutes a frost has, in a certain sense, been defined by the Weather Bureau. 388. The prediction of cold waves. The meaning of the term What con- " co ^ wa ve " has been made definite by the U. S. Weather stitutes a Bureau. Nos. 299-304 of the " Station Regulations," which vave * are reproduced here, contain the definition. " Cold wave warnings will be ordered when it is expected that a 24- hour fall in temperature, equalling or exceeding that specified for the WEATHER PREDICTIONS 401 district, will occur within the 36 hours following the observation upon which the order is based, accompanied by a minimum temperature of the required degree, or lower. The districts and the respective temper- ature falls and minimum temperatures required to verify cold wave warnings during the different seasons are as follows : " In northern Maine, northern New Hampshire, northern Vermont, northeastern New York, western Wisconsin, western Iowa, Minnesota, North Dakota, South Dakota, Nebraska, Montana, Wyoming, Idaho, eastern Washington, and eastern Oregon : a 24-hour fall of 20, with a minimum of zero in December, January, and February, and a minimum of 16 from March to November inclusive. "In southern Maine, southern New Hampshire, southern Vermont, Massachusetts, Rhode Island, Connecticut, New York, (except north- eastern part), northern New Jersey, Pennsylvania, Ohio, Indiana, Michigan, Illinois (except Cairo), western Maryland, West Virginia northern Kentucky, Missouri, eastern Iowa, eastern Wisconsin, Kansas, Colorado, the Texas panhandle, northern New Mexico, northern Ari- zona, Utah, and Nevada; a 24-hour fall of 20 with a minimum of 10 in December, January, and February, and a minimum of 24 from March to November, inclusive. " In southern New Jersey, Delaware, eastern Maryland, the District of Columbia, Virginia, western North Carolina, the northwestern quar- ter of South Carolina, northern Georgia, northern Alabama, northern Mississippi, Tenessee, Cairo, southern Kentucky, Arkansas, Oklahoma, Indian Territory, northern Texas (except the panhandle), southern New Mexico, western Washington, and western Oregon : a 24-hour fall of 20, with a minimum of 20 in December, January, and February, and a minimum of 28 from March to November, inclusive. " In eastern North Carolina, central South Carolina, central Georgia, central Alabama, central Mississippi, northern Louisiana, and central Texas : a 24-hour fall of 18 with a minimum of 25 in December, January, and February, and a minimum of 32 from March to November, inclusive. "In the coast region of South Carolina and Georgia, extreme southern Georgia, Florida, extreme southern Mississippi, southern Louisiana, the Texas coast, California, and southern Arizona : a 24-hour fall of 15, with a minimum of 32 in December, January, and February, and a minimum of 36 from March to November, inclusive." The best precept to follow in making a cold wave forecast is perhaps the following, which is a slight modification of the one advocated by 2D 402 METEOROLOGY The area for which a cold wave is to be pre- dicted. LOW the U. S. Weather Bureau. Through the coming area of high pres- sure (a cold wave is always caused by a coming high) draw two axes at right angles to each other extending north and south, and east and west. The area for which to predict a cold wave will be an oval area located entirely in the southeast quad- rant. A cold wave is particularly sure to occur if a passing low is located to the east of the high, and the high has an oval form with the larger axis extending northeast-southwest. The ac- companying figure makes clear the area for which the cold wave is to be predicted. The brief rule is then to determine first if the prob- able drop in temperature will be sufficient to cause a cold wave, and then to predict it for an area located as stated above. Great caution should be used in predict- ing it for any other area, as many failures have resulted from so doing. Cold wave forecasts and warn- ings are issued only by district forecasters, and not by local fore- cast officials. On the Washing- ton weather map, these forecasts are indicated by the letters CW printed near each station for which a cold wave is predicted, ^reliction . Chart XLIX, which reproduces the daily weather map for indicated. Jan. 6, 1909, illustrates well the area for which a cold wave is predicted, and the method of indicating the prediction on the weather map. Bulletin P (W. B. publication 355) of the U. S. Weather Bureau, by Edward B. Garriot, entitled " Cold Waves and Frosts in the United illustrations States," contains 328 charts which reproduce the daily of cold weather maps and exhibit the meteorological conditions which attended the principal cold waves from 1888 to 1902 inclusive. This invaluable publication also contains a brief account of the origin and cause of cold waves, and presents a chronological account of the historic cold periods in the United States prior to 1888. No one wishing to make cold wave forecasts could do better than study criti- cally this exhaustive treatise. FIG. 145. Diagram Illustrating the Area for which to Predict a Cold Wave. WEATHER PREDICTIONS 403 389. The prediction of tornadoes. Since a tornado usually covers an area only a few hundred feet wide and a few miles logf it is decidedly undesirable to alarm a whole state or several states witli the . forecast that a tornado could occur. For this reason, tor- O f tornadoes nadoes are never predicted by the U. S. Weather Bureau. are not No. 244 in the " Station Regulations " covers the point. " Forecasts of tornadoes are prohibited. When conditions are favor- able for the occurrence of destructive local storms, the term severe thunderstorms ' or ' severe local storms ' may be used by district fore- casters. The phrase ' conditions are favorable for the destructive local storms ' will be used only by the Chief of Bureau, or, in his absence, from the Central Office, by the chief of the Forecast Division." In a previous chapter (sections 343 to 346) there was given a full discussion of the portions of the country most frequented by tornadoes, the time of day and season of occurrence, the type of low wh entor _ which most frequently gives rise to the violent thunder- nadoes are showers which are accompanied by tornadoes, and the loca- posslble - tion of the tornado with reference to the low. In forming one's own opinion from a weather map as to whether a tornado might occur, there are three things to note : First, is the type of low one which is suggestive of tornadoes, secondly, is it the season of the year when they are likely to occur, thirdly, is the place where they could be expected to occur one much frequented by tornadoes. 390. The prediction of destructive wind velocities (storms). Storm warnings are displayed when the wind is expected to attain a velocity for a period of five minutes, equaling or exceeding the verify- ^^ con _ ing velocity within the 24 hours following the time that stitutes a the warning is ordered hoisted. The above may serve as a s definition of what is meant by a destructive or storm wind. There are at present nearly 60 Weather Bureau stations on the Atlantic coast and Gulf coast, on the Great Lakes, and on the Pacific coast in addition to the many adjacent storm warning display stations, where these warn- ings are displayed. The verifying velocity is different for The veru y_ each station, and varies from roughly 20 to 60 miles per ingveioci- hour. For example, at New York it is 44 miles per hour, * at Boston 32, at Portland, Me., 28, at Norfolk 26, at Buffalo 26, at Chicago 46, at Duluth 40, at Seattle 35, and at San Francisco 36. The probable occurrence of a verifying velocity is pre- dicted in the same way as any other wind velocity. A storm warning should contain the location of the storm center, tion. 404 METEOROLOGY and the probable direction in which it will move, with a forecast of the force, direction, and shifts of the wind. These warnings are issued only by a district forecaster, and the flags used to announce them are described in section 348. Bulletin K (W. B. publication 288) of the U. S. Weather Bureau, by E. B. Garriott, entitled " Storms of the Great Lakes/' contains 952 . charts, which reproduce the daily weather maps and exhibit the meteorological conditions which attended the storms which were accompanied by a wind velocity of the verifying amount from 1876 to 1900 inclusive. This exhaustive treatise on the storms of the Great Lakes must be studied critically by any one who would predict wind velocities for this region. 391. The prediction of floods. The method of predicting floods will be given in Chapter X, which treats of floods and river stages. The reguia- The regulations concerning river and flood forecasts are cerrin C river the followm g : and flood " Flood warnings and forecasts of river stages will be issued forecasts. from forecast centers that are river centers, and from specially designated river centers. Copies of all river forecasts and warnings will be promptly transmitted to the Central Office ; regular forecasts on the Daily River Forecast card (Form No. 1086 Met'l), and special fore- casts on the Special River Forecast card (form No. 1087 Met'l). Flood warnings will be immediately transmitted by telegraph to the Central Office." THE ACCURACY AND VERIFICATION OF PREDICTIONS 392. The terms used in official predictions. The terms which may be used by the officials of the U. S. Weather Bureau in making predictions are all prescribed and each has a definite mean- useVin^ffi- m &- This, to a certain extent, hampers individuality cial predic- and a free expression of opinion as to the coming prescribed weather, but it is necessary in order to prevent am- ty the biguity and hedging, and to make possible a systematic Bureau* verification of the predictions. The terms which may be used, and the meaning of these terms in connection with forecasts of temperature and state of the weather, have been given in section 382. What constitutes a dangerous wind velocity (storm), a cold wave, and a frost has been stated in sections 390, 388, 387, respectively. WEATHER PREDICTIONS 405 393. The system of verification. All forecasts and warnings are sent to the Central Office of the Weather Bureau for verification, and the only official verification takes place there. In the case All forecasts of the forecast districts, the forecasts and warnings are sent and warn- each day to Washington by the forecasters for the district. ^w^shing- All local forecasts of temperature and state of the weather ton for veri- made at regular stations, and all practice forecasts in connec- cation - tion with temperature, state of the weather, storm winds, cold waves, and frosts are entered on the appropriate form (No. 1069 Met'l) and forwarded to Washington. At the central office of the U. S. Weather Bureau at Washington, these forecasts and warnings are verified in accordance with a definite set of rules, and the percentage of The ^ accuracy of each forecaster can be determined. The fore- verified on casts and warnings are not verified as a whole, but on five five differ - ent counts. different counts that is, the forecasts of temperature, state of the weather, storm winds, cold waves, and frosts are all verified separately. These rules for verifying can best be given by quoting from the " Station Regulations." " Forecasts and warnings and practice forecasts and warnings will be verified by the following rules : " Rain or snow forecasts and all modifications thereof, The rules containing the terms l possibly/ 'probably/ etc., will be veri- jngpre- 7 " fied by the occurrence of 0.01 inch or more of precipita- dictions, tion. " Precipitation to the amount of 0.01 inch or more which is not fore- cast, or which is forecast and does not occur, will be charged as a failure. " A forecast of precipitation will be considered to be neither a success nor a failure when a trace of precipitation is recorded within the period specified. " In determining the percentage of verification, the number of fore- casts verified will constitute the dividend; the divisor will be the sum of the number of forecasts verified, the number of forecasts not verified, and the number of occurrences of 0.01 inch of precipitation or more which were not forecast. " Forecasts of temperature change, and all modifications thereof con- taining the terms ' possibly/ ' probably/ etc., will be verified by the occurrence of a temperature change, of the kind forecast, equaling or exceeding the stationary limit for the season. " Temperature changes equaling or exceeding the stationary limit 406 METEOROLOGY which are not forecast, or which are forecast and do not occur, will be charged as failures. " In determining the percentage of verification, the number of forecasts verified will constitute the dividend ; the divisor will be the sum of the number of forecasts verified, the number of forecasts not verified, and the number of occurrences of temperature changes equaling or exceed- ing the stationary limit not forecast. " A storm warning will be verified by the occurrence of a verifying velocity for a period of 5 minutes at the station where the warning is displayed, or by the occurrence of a verifying velocity within 150 miles of the station, within 24 hours after the warning was ordered hoisted, and without regard to changes in the original order that may be made during the period. " The continuance of a verifying velocity for a period of 20 minutes without a warning will be considered a storm without a warning. " Not more than one verification, or one storm without warning, will be counted for any one station during any 24-hour period. " In determining the percentage of verification, the number of warn- ings verified will constitute the dividend ; and the sum of the number of warnings verified, the number of warnings not verified, and the number of storms without warnings will constitute the divisor " Cold wave warnings will be verified by the occurrence of the re- quired 24-hour fall in temperature, accompanied by the required mini- mum within the 36 hours following the regular observation on which the warning is based. " The occurrence within a 24-hour period of the temperature change and minimum temperature required to verify a cold wave warning, for which a warning was not issued, will be counted a cold wave without warning. " Not more than one verification, or one cold wave without warning, will be counted at any one station during any 24-hour period. " In determining the percentage of verification, the number of cold wave warnings verified will constitute the dividend ; and the sum of the number of warnings verified, the number of warnings not verified, and the number of cold waves without warnings will constitute the divisor." " In determining the percentage of verification, the number of frost warnings verified will constitute the dividend; and the sum of the number of frost warnings verified, the number of warnings not verified, and the number of frosts and verifying temperatures without frost for which warnings were not issued will constitute the divisor." WEATHER PREDICTIONS 407 394. The accuracy attained. The statement is usually made that the accuracy attained by the official forecasters of the U. S. Weather Bureau is between 80 and 85 per cent. This is, however, a general average for all forecasters, all sections of the coun- try, and all five of they lines along which forecasts and warn- is above 8o ings are verified. The accuracy which can be attained in different parts of the country is very different, (it is probably easier to forecast for California than any other part of the country, and the hardest part is perhaps New England.^ The forecasts of cold waves probably verify least often, as the accuracy is J^cy dc- U " only about 70 per cent. Some forecasters are more skillful pendsonthe than others, but the accuracy of different forecasters pre- ^ e ^ er e dieting for the same locality would probably not differ as and the much as 10 per cent. An accuracy of 85 per cent means that Jcte/' 6 " on the average, one day in seven will see a complete failure of the weather predictions. It must be held in mind in this con- nection, however, that a mere guess ought to yield the accuracy of but 50 per cent. The accuracy attained in Europe is about the same as in this country, although the terms used in forecasting, and the method of verification, are quite different in different countries. Weather forecasting, as regards accuracy, may be considered on about the same level as the practice of medicine. The forecaster can diagnose the present condition of the atmosphere with as much pre- weather cision as a physician can determine the bodily condition of forecasting a patient^ He can predict the coming weather certainly with certainty as a physician can predict the exact course! practice of and outcome of any well-known disease. Weather fore- casting can probably never hope to attain the accuracy of astronomy in predicting celestial occurrences. The increase in accuracy during the last fifteen years has been small, and this would seem to indicate that present methods can yield no greater accuracy. This certainly does not mean that no elab- The accu orate investigations have been undertaken to determine if racy has not a new system could be devised or the present methods im- increased proved. The weather map for the whole northern hemi- sphere has been constructed daily, and the permanent highs and lows have been critically studied in order to determine if changes in them exert an influence on the passing highs and lows and thus determine their path, velocity, and characteristics. The weather maps for differ- 408 METEOROLOGY ent levels above the earth's surface have been constructed in order to determine if the passing meteorological formations, as portrayed at these . levels, are less erratic in their behavior than at the earth's vestigations surface. The areas of greatest and least pressure and tern- have been perature change have been studied critically to determine undertaken f*. . J to improve if these are any more regular in their behavior tha^n highs or lows. The variations in the energy received from the sun have been determined and studied in order to ascertain if these influence the highs and lows. Of these, the study of the per- manent highs and lows has yielded the greatest results, but none of these investigations has revolutionized the methods of forecasting or added much to the accuracy. 395. The popular idea of the accuracy of weather forecasts. The newspapers and periodicals are essentially just, fair, and considerate in their attitude towards the Weather Bureau, and the papers and failure of some predictions to verify. It is very seldom that periodicals a hostile editorial is seen or a vindictive, sarcastic attitude are fair. or comment is met with. There are, of course, many jokes at the expense of the Weather Bureau and its failures. There are also many jokes about the repairs necessary to an automobile, but that does not prove that the automobile is not both a useful and pleasure-giving vehicle. There are a very few who sneer at the Weather Bureau and The declare that it should be done away with. Those who have Weather this attitude are usually those who really know very little very profit- a bout meteorology, the work of the Weather Bureau, or the able in- value of its predictions. The Weather Bureau certainly saves to this country annually at least three and probably ten times its cost. It thus gives very handsome returns for what is expended, and any one who sneers at its work is purely a useless, destructive critic, unless he can propose some plan which would lead to greater returns. It has often been said that it would be very desirable if the weather map could receive such wide distribution that it would come to the notice of every one, and at the same time the public could be educated to interpret the map. Every person would then make his own predic- tions, and the attitude of a person towards what he does himself, even if imperfect, is always very different from what it is towards the same thing done, perhaps somewhat better, by some one else. WEATHER PREDICTIONS 409 LONG-RANGE PREDICTIONS 396. Prediction from station normals. The definite predictions made by the officials of the U. S. Weather Bureau on the basis of the daily weather map are for only thirty-six or forty-eight hours ahead, and it is impossible to extend the period with meant by any certainty to more than three days, even in the case of ^ng-Taxige the most typical maps. Recently the attempt has been f made to issue a general prediction for a whole week, and the basis of this will be discussed later. The popular desire is, however, not for predic- tions for a few hours ahead or at most a week ahead, but for predictions for a month, or several months, or a whole year in advance. These are called long-range predictions. The attempt has sometimes been made to use the station normals for the various meteorological elements as the basis of long-range pre- dictions. For example, for the northeastern part of New No Definite York State, it could be stated with certainty on the basis of predictions these various normals that the temperature would fall 6 made below zero at least once during the winter, that there would station be at least one thundershower during the summer, that at E least two feet of snow would fall during the winter, etc. These state- ments are, however, not really long-range predictions, but merely a description of the climate of the locality. Furthermore, they give no indication as to what the weather is going to be on any definite date, or even whether the season is going to be too hot, or too cold, too wet, or too dry. The attempt has also been made to use departures from normal as the basis of long-range predictions. There is a strong popular belief that, for example, if a certain month is too hot or too dry, Departure the next month will be too cold or too wet, in order to com- from no * mai cannot be pensate. Now these departures from normal, particularly used as a in the case of temperature and precipitation, have been {^ sis r ^ studied critically for many stations and for many years, 1 weather and the conclusion drawn from such research work has always predictions. been that, from the departure from normal on the part of any particular month, no conclusion whatever can be drawn as to whether the follow- ing month will be too hot or too cold, too wet or too dry. Much research work has also been done on the probable character- istics of one season as determined by the departure from normal of a i See Bulletin U of the U. S. Weather Bureau. 410 METEOROLOGY previous season. Koppen has found, for the middle of Europe, that Seasonal the probability of a departure from normal in the other di- sequence. rection in the case of temperature, from one season to an- other, is as follows : winter to spring 0.49 spring to summer 0.45 winter to summer 0.44 spring to autumn 0.40 summer to autumn 0.38 autumn to winter 0.45 summer to winter 0.50 autumn to spring 0.52 This means, for example, that the probability of a cold autumn follow- ing a warm summer is 38 out of 100. It will be seen at once that the preponderance of probability is too slight to make this of much value in forecasting. At any rate, no definite forecasts can be made. Hellmann has found, for Berlin, that there is the greatest probability of the follow- ing successions, but the preponderance of probability was always small : moderately mild winter cool summer very mild winter warm summer moderately cold winter cool summer very cold winter very cool summer moderately warm summer moderately mild winter very warm summer . cold winter Similar probable sequences could be worked out for any station on the basis of the observations, and the preponderance of probability could be determined for each sequence. It is of too little value in forecasting, however, to make it worth while. If it has been done, one might say, for example, after a cold winter had passed, that the probability of a warm summer was perhaps 46 out of the 100. This would be very far from a satisfying long-range prediction. 397. Weather cycles. The daily and annual change in the values of the meteorological elements is, of course, very marked. There is always a pronounced diurnal and annual variation, and the charac- cycies have teristics and magnitude of these variations in the case of each been de- element have been fully discussed in previous chapters. Inves- tigators have also thought they have found several minor periods or cycles which underlie the changes in the values of the elements. Cycles of 3 days, 5.5 days, 11 years, and 35 years without doubt exist. 1 Cycles of 26.7 days, corresponding to the sun's rotation, 7 years, A criterion 19 years, corresponding to the nutation period, and many of value. others have been announced by different investigators. In fact, the number of cycles has become so large that it is a serious ques- 1 See Monthly Weather Review, April, 1899, p. 156. WEATHER PREDICTIONS 411 tion if some criterion should not be used to determine if a cycle is worthy of consideration. It could well be said that a cycle is not worth con- sidering, unless (1) it is very well marked, if it occurs at only a few sta- tions ; (2) it occurs at many stations, if it is of small magnitude ; (3) it corresponds to something else which has the same period and could be its cause. The three-day cycle is probably the average interval between the passing of a low and the coming of the following high. The 5.5-day cycle is, in a certain sense, a double three-day cycle. It is The s.s-day probably the average time interval between the passage of cycle - well-marked lows near any given station. It is evident to any one who follows the daily weather maps, day after day, that these two cycles may be said to exist, but they are very far from definite and could not be projected into the future more than two weeks at the very most, and only then with great uncertainty. The eleven-year cycle corresponds to the sunspot period, and is very pronounced in terrestrial magnetism and atmospheric electricity. It is easily found in certain of the meteorological elements, The eieven- but the magnitude of the variation is extremely small, and vear c y cle - the observations must be carefully averaged at many stations fora long time, to detect it with certainty. Its presence has been detected chiefly in connection with the number of tropical cyclones, the amount of precipitation, and the tracks followed by lows. 1 The cycle of 35 years is due to Bruckner, and is particularly noticeable in connection with temperature and precipitation. Not only have direct observations been used in the discussion, but such The thirty- indirect data as the dates of harvests, the opening and clos- five-year ing of navigation, the height of inclosed seas, the severity of cycle * winters, etc. In fact, even the size of the circle of annual growth in the case of very old trees has been used. European observations show that during the last two centuries 1746-1755, 1791-1805, 1821-1835 ; 1851-1870 were relatively warm. 1731-1745, 1756-1790, 1806-1820, 1836-1850, 1871-1885, were rela- tively cold. 1756-1770, 1781-1805, 1826-1840, 1856-1870 were relatively dry. 1736-1755, 1771-1780, 1806-1825, 1841-1855, 1871-1885 were relatively wet. The amount of the oscillation in the case of the precipitation is very dif- ferent in different countries, and is usually greater in the interior than near 1 See W. J. HUMPHREYS, Astrophysical Journal, September, 1910. 412 METEOROLOGY the coast. On the average, it amounts to about 20 per cent of the total amount. The period is not strictly 35, but varies from 34 to 36 years. It will be seen at once that all of these cycles are too indefinite and of too little importance to play any part in weather fore- are of no casting or serve as a basis for long-range predictions, forecastin 398 ' Tendenc y of a weather type to continue. If the ceaseless changes in the values of the meteorological elements are carefully studied, it will be apparent that the number of changes in the weather from one day to the next is smaller than the number of There is a continuations of the same kind of weather. That is, after a decided ten- cold day, there is a greater probability of another cold day fhTexisting tnan ^ a warm day, and after a rainy day there is a greater weather to probability of another rainy day than of a fair day. There is thus a decided tendency for the existing weather to continue. At Brussels, for example, the probability of a change after a certain number of days with a certain character as regards tempera- ture and rain is given in the following table : After a continuance of 1 2 3 4 5 6 7 10 15 days Temperature .25 .24 .22 .21 .17 .17 .15 .15 .13 Rain .37 .32 .30 .26 .27 .24 .25 .23 .23 This means that if there have been four rainy days in succession, the probability of the next day being fair is only 26 out of 100. A similar table could be prepared for any station, and the same decided tendency of existing weather to continue would be found. This table also shows that if a weather forecaster in Brussels should each day systematically predict that the next day would be the same as regards temperature and rainfall, he would attain an accuracy in forecasting of about 75 per cent. If longer periods of time are considered, it will be found that, as soon as a definite type of weather has become established, there is a decided tendency for this type to continue several days or perhaps weather also even several weeks. A warm rainy period is apt to continue tends to f O r a week or two, and cold dry weather usually lasts equally long. During the types, the highs and lows tend to follow the same tracks and have the same characteristics. This is a matter This is due of great practical importance in forecasting, and more and manen? er " more attention is being paid to it. This tendency of a areas of weather type to continue is probably the result of what Teis- iow h pres- serene de Bort calls the " centers of activity " of the atmos- sure. phere. If the chart which exhibits the isobaric lines for January is studied, it will be seen that in the northern hemisphere there WEATHER PREDICTIONS 413 are four permanent highs, an immense one over Asia, a smaller one central over North America, and two still smaller ones over the Atlantic and Pacific oceans. There are also two permanent areas of low pressure with their centers over the north Atlantic near Iceland and over the north Pacific south of Alaska. These areas are the so-called centers of activity and remain practically fixed. They are the permanent landmarks of pressure, and stand in sharp contrast to the moving areas of low and high pressure. If any change occurs in the location or intensity of these permanent areas, the moving areas of high and low pressure have dif- ferent characteristics and move over different tracks, and thus give rise to a different type of weather. Since these permanent areas of pressure change their characteristics slowly and reciprocally, there is a tendency for a given weather type to continue for several days, or even 'several weeks. The general forecasts which have lately been attempted at Washington for a week ahead, are based upon a study of these perma- nent areas of pressure and this tendency of a weather type to continue. ' These centers of action change with the time of year. A series of iso- baric charts, exhibiting the normal pressure for every month in the year, would be necessary for the full discussion of the subject. The exact type of weather which exists for given characteristics on the part of these areas has not yet been fully determined. In fact, not much more has been done than to recognize the importance of these areas and the fact that the type of weather probably does depend, to a large extent on their characteristics. 1 The pressure distribution over the whole northern hemisphere for January 28, 1910, is given as chart L. This is an illus- tration of the daily pressure map as made at Washington for the whole northern hemisphere (see section 362). The Siberian high has an un- usual intensity. The low over Europe is the strong persistent one which caused the excessive precipitation over France, which resulted in the great floods at Paris in January, 1910. 399. Popular superstitions and credulity. In the foregoing para- graphs (sections 396, 397, 398), the science of meteorology has presented all that it has to offer in connection with long- range weather predictions. On the basis of the daily weather what can be map, exact predictions for the coming 36 or 48 hours can be JJ?^ made, and these verify in about 80 to 85 per cent of the along the cases. Based on the tendency of a weather type to con- ^thtr tinue, and on a study of the centers of activity in the prediction. 1 For a short bibliography of articles on this subject see Bulletin of the Mount Weather Observatory, vol. Ill, part 4, p. 237. 414 METEOROLOGY northern hemisphere, it is possible to make a general forecast for per- haps a week ahead. It was seen that long-range weather predictions could not be made with any desirable definiteness or certainty based on station normals, departures from normal, or weather cycles. Since this is the state of the case, the U. S. Weather Bureau and scientific meteorologists are frank and honest enough to admit that, at present, long-range weather predictions are an impossibility. Long-range weather predictions are very desirable, and every means of making them is being thoroughly investigated ; but simple honesty demands that the admission be made that, at present, there is no way known of making them. The popular desire, however, is for long-range weather predictions, and, this being the case, it is not remarkable that various attempts are The various ma de to satisfy this desire in one way or another. Some things which try to make long-range predictions for themselves and basis for a friends, using as a basis the action of some animal, or the long-range weather on some definite date, or something connected with recasts. ^ moon Others, bolder and usually with a desire for financial gain, publish these forecasts in almanacs and the like. Some newspapers even publish these predictions, and worse than this, they actually pay for the privilege of publishing these worse than useless pre- dictions. Now if these long-range predictions are not mere guesses, the various things which serve as a basis may be grouped under three heads : (1) the influence of the moon, the sun, or the planets in short, astronomical control ; (2) the actions of animals, birds, and plants ; (3) the weather during certain days, months, or season. Now, in the case of the sun and moon, as was stated in connection with weather cycles, there may be a very slight influence on some phases of the weather, but the influence is so slight that it is almost impossible to find it with certainty. It is certainly so small that it can play absolutely no part in forecasting. As for the rest, there is, in the first place, absolutely no scientific reason why they should have any connection with the coming weather ; and, in the second place, the observations at many stations have been averaged for many years to determine if they do have any influence, and the result has always been to find that they do not. Thus all long- range forecasts built upon such things are mere superstitions and have no foundation whatever. They are no better than mere guesses, and it should be remembered that a mere guess should be correct half the time, so that there should be no surprise at some chance verifications. It might perhaps be agreed that those who make these long-range forecasts WEATHER PREDICTIONS 415 simply for themselves and friends, should be allowed the possible pleas- ure of this probably harmless self-delusion. It is a very different matter, however, in the case of those who publish their forecasts and particularly those who publish them for financial gain. They are not only making gain at the expense of the superstition and credulity of the public, but they are making the public believe that something can be done which cannot be done. These published forecasts are of all degrees of definiteness. Some- times, in some almanacs, the statement " Snow may be expected at this time " will be found on the margin, and extending over a A typ j ca i third of a winter month. Since this would be verified by long-range snow occurring anywhere in the United States at any time f during the period, it is needless to add that it is certain of verification. Sometimes the more elaborate forecasts run something like this : Febru- ary 21 to the end of the month, constitutes a storm period of marked energy. There will be snow in the northwest, gales over the Great Lakes, and freezing weather on the Atlantic coast. A storm of marked energy will move from west to east across the country. Its coming will cause warmer weather with rain or snow. It will be followed by north- west winds and colder with snow flurries in New England. There will be thundershowers in the Southern states and the month which closes with this period will average too warm. The above is copied from no source, has no basis whatever, and can be applied to any year. It is simply a mixture of meteorological information, meteorological statis- tics, glittering generalities, and plain guesses : and any forecaster of the U.S. Weather Bureau, and any scientific meteorologist could write predic- tions like the above by the page and volume if he were willing to cheapen and degrade his science to that extent. There is never an in- terval of seven days at the end of February without a low of analysis of fair intensity crossing the United States somewhere. Lows such a always move from west to east, and are preceded by south winds and warmer weather and are followed by northwest winds and colder weather. All this is simply meteorological information. Now, as regards the gales on the Great Lakes, snow flurries in New England, and thundershowers in the Southern states, if a series of weather maps covering these seven days for the last twenty years were carefully studied, it would be seen that these occurrences took place about 18 out of the 20 years. The chance is thus ten to one in favor of these things happen- ing. It is therefore an extremely safe prediction considering statistics. Freezing weather on the Atlantic coast is a glittering generality, as the 416 METEOROLOGY 32 F. line always intersects the Atlantic coast somewhere. That the month would average too warm is a plain guess, and thus stands an even chance of being right for any station. Furthermore, there are certain to be many localities where it will verify. The widespread belief in the existence of an equinoctial storm and Indian summer comes, to a certain extent, under the head of popular . superstitions. If the equinoctial storm is defined as a rain- noctiai storm storm, lasting at least three days, and occurring within two and Indian or three days of the 21st of September, then there is seldom a summer. year when one occurs. If, however, the equinoctial storm is defined as any rainstorm lasting, say a day or longer, and occurring within two weeks of the 21st of September, then there is very seldom a year when several equinoctials do not occur. The reason for the belief in an equinoctial storm is probably the fact that about this time of year the first storms of the winter type, with steadily falling precipitation, make their appearance. They stand in sharp contrast to the summer type with the sultry weather and thundershowers. Storms of the winter type can occur, however, during any month of the summer. The amount of the precipitation near the 21st has been shown by averag- ing the observations at many stations to be no greater than before or after that date. 1 The case is similar with Indian summer. If Indian summer is defined as a spell of peculiar weather in the autumn, characterized by great warmth, smokiness, and haziness, and lasting for several weeks, then Indian summer seldom occurs. If, however, Indian summer is defined as a few days of slightly greater warmth and haziness, which only serve to emphasize our otherwise delightful autumn weather, then Indian summer nearly always occurs. 2 FORECASTS FROM LOCAL OBSERVATIONS AND APPEARANCES OF SKY 400. Prediction from the readings of instruments and the appear- ance of the sky. The question is often raised, if it is possible to predict A general the weather from the sky appearance and the indications cornin* the ^ me teorological instruments without using the daily weather weather can maps. It can readily be shown that a good general idea of rrom^sk 6 * 1 ^ e comm S weather can probably be formed in this way, appearance but that exact and definite prediction requires the use of the 1 See Monthly Weather Review, November, 1901, p. 508. 2 See Monthly Weather Review, January, 1902, p. 19. WEATHER PREDICTIONS 417 daily weather map. Suppose, for example, that it is January, and the wind has just changed to the southeast. Suppose, observa- furthermore, that the temperature is rapidly rising, that tions - the moisture is increasing, that the barometer is falling, that the sky is hazy, and the cirrus clouds are visible, perhaps thickening to cirro-stratus or cirro-cumulus. It is evident that an area of low pressure is about to dominate the weather. The normal sequence of weather changes during the next two days would be precipitation probably in the form of snow, winds shifting to the northwest, and then clearing and colder. A good general idea of the coming weather has thus been formed. If such questions as the probable amount of snowfall, the probable dura- tion of the snowfall, the probable rise in temperature, the possibility of the snow turning to rain, the probable drop in temperature after the storm passes, are to be answered, the daily weather maps are indispens- able. As a second illustration, suppose it is again January, and the wind has just gone to the northwest. Suppose, furthermore, that the barometer is rising, the moisture is decreasing, the temperature is falling rapidly, the clouds are breaking up into strato-cumulus and cumulus, and the air is becoming clear. It is evident that the weather control is passing to a coming high. The normal weather sequence would be two or more days with northwest winds, low temperature at night, and plenty of sunshine. The sky appearance and the indications of local instruments indicate whether a coming or departing low, a coming or departing high, is domi- nating the weather. As soon as this is determined, the probable sequence of weather changes for the next day or two is at once ap- parent. To make a definite prediction, however, weather maps must be used. The readings of local instruments are of such value in determining the location and direction of motion of a low, that the following is printed on the face of all weather maps issued by the U. S. Weather Bureau: " When the wind sets in from points between south and southeast and the barometer falls steadily, a storm is approaching from the west or northwest, and its center will pass near or to the north of the observer within twelve to twenty-four hours, with winds shifting to northwest by way of southwest and west. When the wind sets in from points between east and northeast, and the barometer falls steadily, a storm is approach- ing from the south or southwest, and its center will pass near or to the south or east of the observer within twelve to twenty-four hours, with winds shifting to northwest by way of north. The rapidity of the 2E 418 METEOROLOGY storm's approach and its intensity will be indicated by the rate and amount of the fall in the barometer." 401. Weather proverbs and weather rules. Weather proverbs are as old as written language. The generalization of weather experience into proverbs and weather rules seems to have been one of many the first acts of civilized man. Weather proverbs dating *overbs ^ ac ^ to at least 4000 B.C. have been found on the clay tablets of Babylonia ; there are many of them scattered through the oldest manuscripts; their number has increased through centuries; and at present there are hundreds of them. The following collection, Dr. jenner's in versified form, usually ascribed to Dr. Jenner, the dis- coiiection. coverer of vaccination, is probably the most famous and interesting : "The hollow winds begin to blow, The clouds look black, the glass is low, The soot falls down, the spaniels sleep, And spiders from their cobwebs creep; Last night the Sun went pale to bed, The Moon in halos hid her head, The boding shepherd heaves a sigh, For see ! a rainbow spans the sky ; The walls are damp, the ditches smell, Closed is the pink-eyed pimpernel ; Hark how the chairs and tables crack ! % Old Betty's joints are on the rack; Her corns with shooting pains torment her And to her bed untimely sent her ; Loud quack the ducks, the peacocks cry, The distant hills are looking nigh ; How restless are the snorting swine ! The busy flies disturb the kine, Low o'er the grass the swallow wings ; The cricket, too, how sharp he sings ! Puss on the hearth, with velvet paws, Sits wiping o'er her whiskered jaws ; The smoke from chimneys right ascends, Then spreading back to earth it bends ; The wind unsteady veers around, Or setting in the South is found ; Through the clear stream the fishes rise, And nimbly catch th 7 incautious flies ; The glowworms, num'rous, clear, and bright, WEATHER PREDICTIONS 419 Illumed the dewy dell last night ; At dusk the squalid toad was seen Hopping and crawling o'er the green ; The whirling dust the wind obeys, And in the rapid eddy plays ; The frog has changed his yellow vest, And in a russet coat is dressed ; The sky is green, the air is still, The merry blackbird's voice is shrill, The dog, so altered is his taste, Quits mutton bones on grass to feast ; And see yon rooks, how odd their flight ! They imitate the gliding kite, And seem precipitate to fall, As if they felt the piercing ball. The tender colts on back do lie, Nor heed the traveler passing by. In fiery red the Sun doth rise, Then wades through clouds to mount the skies. 'Twill surely rain, I see with sorrow Our jaunt must be put off to-morrow." Several fairly complete compilations of weather proverbs have been made, and in the references to the literature given at the end of this chapter some of them are mentioned. Various countries have different weather proverbs and sometimes those well known in one country will be unknown or will not apply at all in another. Weather proverbs, particularly those referring to the sky appearance and the meteorological elements, are often called weather rules, or prognostics. 402. By far the most important question in connection with weather proverbs is whether they have any foundation or not ; that is, whether they are mere superstitions or whether they have a basis Thefive in fact. 1 In this regard, weather proverbs may be divided classes of into five classes. The first two are fairly well founded, weath f . . proverbs. while the last three are mere superstitions. The first class includes those which infer an impending weather change from the sky appearance and something connected with the meteoro- logical elements for example, " Rainbow in the morning, sailor take warning; rainbow at night, sailor's delight." Now a rainbow in the morning, means that the sun is shining classes. 1 See " Some Weather Proverbs and their Justification," by W. J. HUMPHREYS, Popular Science Monthly, 1911. 420 METEOROLOGY in the east, which is clear, and that it is raining in the west. Since thundershowers and storms move, in general, from west to east, this means that rainy weather is impending. Similarly a rainbow at night means that the sun is setting clear in the west while it is raining to east- ward. This indicates a departing storm and that a period of good weather is at hand. A second example of this same class of proverbs is the following : " Mackerel sky and mares' tails make lofty ships carry low sails." Now mackerel sky and mares' tails, as popularly expressed, mean technically that cirro-cumulus clouds are present in the sky. This is a transition cloud form from cirrus to the coming nimbus. The storm cloud is usually accompanied over the ocean by high winds, and this will cause a vessel to carry but little canvass. It is both inter- esting and instructive in connection with weather proverbs to try to trace out the scientific basis for them. The second class of weather proverbs are those which infer the coming weather from the behavior of animals, plants, and inanimate things. The coming of a low with its rain area and high shifting winds is usually heralded by an increase of temperature and moisture and a decrease of pressure. Drains are said to smell before rain. This may simply mean that the lower pressure causes some of the air to escape, thus causing the odor to become noticeable. The increase in temperature and mois- ture often causes a change in the behavior of animals. Their cries and actions are different, but this does not mean in any sense that the animals are endowed with prophetic vision. They are simply reacting to a changed present condition which is the forerunner of the rain period. The three classes of weather proverbs which have no scientific basis whatever, are (l) those which infer the future weather at some distant date from the actions of animals or plants ; (2) those which infer the future weather from the weather at some previous time; (3) those which infer the weather from some astronomical body. No credence whatever is to be placed in these sayings, because there is no reason why they should be true ; and statistics show that they are not true. The following will serve as examples of these : Squirrels gather more nuts before a hard winter ; If it rains St. Swithin's day, it will rain forty days ; The moon and the weather change together. A running commentary on Dr. Jenner's doggerel verses, quoted above, will illustrate the scientific basis for many of these prognos- A commen- tics. These are all rain prognostics, and belong to the first T'nner'? 1 ^ wo c ^ asses ^ wea ^her proverbs. Since rain is expected, it collection. can be inferred at once that a low is approaching. This WEATHER PREDICTIONS 421 means that the temperature and moisture are increasing, the pressure is lessening, the wind is in the southeast, and haze and cirriform clouds are prevalent. The glass is low refers, of course, to the falling barometer. The red color of the sun at sunrise, the presence of clouds, the halo around the moon, the pale appearance of the sun, all indicate the hazy condition of the atmosphere and the presence of cirriform clouds. The shifting, rising wind indicates that the storm is coming steadily nearer. The falling of soot in chimneys and the dampness of walls simply indi- cate that the moisture is larger. To the same cause, joined with higher temperatures, may be attributed the closing of sensitive flowers, the rheumatic pains, the shooting of corns, the low flight of insects and of birds in search of them, the restlessness of animals, and the other changes in their cries and movements. QUESTIONS (1) Define weather and weather prediction. (2) Of what three things is weather the composite? (3) Describe the normal or typical weather for the northeastern part of the United States. (4) Name the passing meteorological formations which exert the chief influence on the weather. (5) State what is meant by local influences. (6) State in outline the general method of weather forecasting. (7) How is the United States divided for the purpose of fore- casting? (8) What forecasts are issued by a local forecast official? (9) Of. what does a forecast consist? (10) What training do the weather bureau fore- casters receive? (11) What are the characteristics of a successful forecaster? (12) How do the methods of making weather forecasts differ in other countries? (13) Describe in full the method of locating the center of an area of low pressure 24 hours ahead. (14) What are some of the rules for modifying the estimated position ? (15) What really determines the track of a low ? (16) Which are the more impor- tant factors in determining the track of a low ? (17) Describe the Bowie method of locating a low. (18) Describe in detail the method of determining the distri- bution of the meteorological elements about a low 24 hours ahead. (19) What are the advantages of making an exact forecast? (20) What terms may be used in the forecasts made by the U. S. Weather Bureau? (21) What is meant by secondary isobaric forms? (22) Describe in detail the various kinds of V-shaped depressions and their significance. (23) What are secondary lows? (24) Describe in outline the method of making a weather prediction when a high is the dominating formation. (25) Describe the method of weather prediction by similarity with previous maps. (26) How are frost predictions made ? (27) How is the temperature of low-growing vegetation itself determined? (28) What constitutes a frost in the weather bureau sense? (29) What is meant by a cold wave ? (30) How is the area for which a cold wave is predicted to be determined? (31) How is the cold wave prediction indicated on the weather map? (32) Why are tornadoes not predicted? (33) When is a tornado likely to occur? (34) What is meant by a destructive wind velocity or storm? (35) What are the regulations concerning river and flood forecasts ? (36) What terms may be used in the official weather bureau predictions? (37) How are weather predictions verified? (38) On what counts are weather predictions verified? (39) What accuracy is attained by official forecasters? (40) What 422 METEOROLOGY investigations have been undertaken to improve the present methods of forecasting? (41) What is the attitude of newspapers and periodicals towards weather forecasts? (42) What is meant by a long-range prediction? (43) To what extent can the station normals be used in the making of long-range predictions? (44) To what extent can departure from normal be used as a basis for long-range predictions? (45) What are weather cycles? (46) To what extent can weather cycles be used in forecasting ? (47) Describe in detail the tendency of a weather type to continue. (48) To what is this tendency due? (49) What can be accomplished along the line of scientific weather prediction? (50) What are the various things which serve as a basis for long-range weather forecasts ? (51) Describe the typical long-range weather forecast. (52) To what extent can predictions from the readings of instruments and the appearance of the sky be made? (53) Describe weather proverbs. (54) Into what five classes may the weather proverbs be divided ? (55) What basis have weather proverbs ? TOPICS FOR INVESTIGATION (1) The Bowie method of locating the center of a low. (2) The rules for locating a low 24 hours ahead. (3) V-shaped depressions. (4) Secondary lows and their influence on the weather. (5) Local influences in forecasting. (6) The area for which to predict a cold wave. (7) The accuracy attained in weather forecasting. (8) The terms used in expressing weather predictions and the system of verification in other countries. (9) The methods used by the weather bureaus of other countries in making and distributing predictions. (10) Centers of action. (11) Long-range predictions. (12) Weather cycles. (13) Weather proverbs. PRACTICAL EXERCISES (1) Make several exact weather predictions and also in the form issued by the officials of the U. S. weather Bureau, when a low is dominant, when a high is dominant, and when secondary isobaric forms are present. In each case verify the forecast according to the regular rules. (For this work a file of weather maps will be very useful, although predicting for the coming day is always much more interesting.) (2) In several cases locate the center of a low 24 hours ahead, using the Bowie method. (3) From a file of weather maps select those which illustrate well the rules for modifying the estimated location of the center of a low. (4) From a file of weather maps select those which illustrate well forecasting when the secondary isobaric forms are of importance. (5) From a file of weather maps select those which illustrate well the predic- tion of daijgerous or damage-causing occurrences. (6) From a long series of observations (those at a regular Weather Bureau station would be necessary) compile statistics and determine if certain weather proverbs or weather rules or things used as the basis of long-range predictions have any foundation or value. Determine in each case the preponderence of probability. WEATHER PREDICTIONS 423 REFERENCES The following are the books and pamphlets which, as a whole or in part, deal with weather and weather forecasting in general : ABBE, CLEVELAND, Preparatory Studies for Deductive Methods in Storm and Weather Predictions, 165 pp., Washington, 1890. (Annual Report of the Chief Signal Officer for 1889 ; Appendix 15.) ABBE, CLEVELAND, The Aims and Methods of Meteorological Work, Balti- more, 1899. (Special publication of the Maryland Weather Service.) ABERCROMBY, RALPH, Principles of Forecasting by Means of Weather Charts, 122 pp., London, 1885. ABERCROMBY, RALPH, Weather, London, 1887. BEBBER, W. J. VAN, Handbuch der ausubenden Witterungskunde, 8, 2 parts, Stuttgart, 1885-6. BEBBER, W. J. VAN, Beurteilung des Welters auf mehrere Tage voraus, Stutt- gart, 1896. BEBBER, W. J. VAN, Die Wettervorhersage, 2d. ed., 219 pp., Stuttgart, 1898. BIGELOW, FRANK H., Storms, Storm Tracks, and Weather Forecasting. U. S. Weather Bureau, Bulletin No. 20 (W. B. No. 114). CHAMBERS, GEORGE F., The Story of the Weather, London, 1897. DALLET, G., La prevision du temps. FREYBE, OTTO, Praktische Wetterkunde, 173 pp., Berlin, 1906. GRANGER, FRANCIS S., Weather Forecasting, 121 pp., Nottingham, 1909. GUILBERT, GABRIEL, Nouvelle methode de prevision du temps, 343 pp., Paris, 1909. KLEIN, H., Wettervorhersage fur jedermann, Stuttgart. KUHLENBAUMER, TH., Unser Wetter und seine Vorherbestimmung, 164 pp., Minister, 1909. MOORE, WILLIS L., Weather Forecasting. U. S. Weather Bureau, Bulletin No. 25 (W. B. No. 191). MOORE, WILLIS L., " Forecasting the Weather and Storms," The National Geographic Magazine, June, 1905, Vol. XVI, No. 6. PERNTER, J. M., Wetterprognose in Osterreich, 61 pp., Wien, 1907. PERNTER, J. M., " Methods of Forecasting the Weather," Monthly Weather Review, December, 1903. SCOTT, ROBERT H., Weather Charts and Storm Warnings, 229 pp., London, 1887. SCOTT, ROBERT H., Notes on Meteorology and Weather Forecasting, 40 pp., London, 1909. SHAW, W. N., Forecasting Weather, 8, xxvii + 380 pp., London, 1911. Station Regulations of the U. S. Weather Bureau, Washington, 1905. WARD, ROBERT DE C., Practical Exercises in Elementary Meteorology, Ginn& Co., 1899. For the prediction of particular and damage-causing occurrences such as storm winds, cold waves, frost, floods, etc., see : GARRIOTT, EDWARD B., Cold Waves and Frosts in the United States. U. S. Weather Bureau, Bulletin P (W. B. No. 355). GARRIOTT, EDWARD B., Storms of the Great Lakes. U. S. Weather Bureau, Bulletin K (W. B. No. 288). In connection with long-range weather predictions, consult : GARRIOTT, EDWARD B., Long-range Weather Forecasts. U. S. Weather Bureau, BuUetin No. 35 (W. B. No. 322). 424 METEOROLOGY WALZ, F. J., "Fake Weather Forecasts," Popular Science Monthly, pp. 503-513, Vol. 67, 1905. For collections of weather proverbs, weather rules, weather folklore, and weather prognostics, see : CHAMBERS, GEORGE F., The Story of the Weather, London, 1897. DUN WOODY, H. H. C., Weather Proverbs. Signal Service Notes, No. IX, Washington, 1883. GARRIOTT, EDWARD B., Weather Folklore. U. S. Weather Bureau, Bulletin No. 33 (W. B. No. 294). HORNER, D. W., Observing and Forecasting the Weather, 46 pp., London, 1907. INWARDS, RICHARD, Weather Lore, 233 pp., London, 1898. MICHELSON, W. A., Wetterregeln, 17 pp., Braunschweig, 1906. SWAINSON, CHARLES, A Handbook of Weather Folklore, 275 pp, 1873. PART II PART I consists of eight chapters which treat in succession the atmos- phere, the heating and cooling of the atmosphere, temperature, pressure and wind, moisture in all its forms, the various storms, weather bureaus and their work, and weather predictions. These chapters are pre- sented in full, and the various topics are treated at length. This mate- rial is treated in full in practically all textbooks on meteorology, and makes a complete treatise in itself. Part II consists of five chapters which are treated in full in some books and passed over with a few words or pages in others. If meteorology is taken in the largest sense, as the science of all atmospheric phenomena, and should include all the work of weather bureaus, then these chapters should be included in a treatise on the subject. These chapters treat climate, floods and river stages, atmospheric electricity, atmospheric optics, and atmospheric acoustics. It was at first intended to consider these five chapters as an appendix and present simply the syllabus of each chapter, and the references to the literature. If this book is used as a textbook, the student has had sufficient material presented to him in a predigested form, and it would be best for him to work up these various topics for himself. The general reader, however, would prob- ably prefer to know what would be covered in these chapters without working up the subject from the literature. For that reason, most of the topics are sketched in outline, but no attempt is made at the full- ness or completeness of Part I. 425 CHAPTER IX CLIMATE DISTINCTION BETWEEN WEATHER AND CLIMATE, 403 CLIMATIC DATA AND CHARTS, 404 THE FACTORS WHICH DETERMINE CLIMATE, 405 THE CLIMATIC SUBDIVISIONS OF THE WORLD Subdivision on the basis of latitude, 406. Subdivision on the basis of temperature, 407. Subdivision on the basis of the general wind system, 408. Subdivision on the basis of surface topography, 409. " Other climatic subdivisions of the world, 410, 411. THE CLIMATE OF THE UNITED STATES Introduction, 412. Subdivision, 413. Detailed treatment of certain subdivisions, 414. THE CLIMATE OF OTHER COUNTRIES AND PLACES, 415 THE CONSTANCY OF CLIMATE, 416 THE SNOW LINE, 417 DISTINCTION BETWEEN WEATHER AND CLIMATE 403. Weather is the condition of the atmosphere at any particular time and place, and is best described by stating the numerical values of Weather * ne va us meteorological elements. Climate is generalized and climate weather, and has to do with a larger area and longer time. Climate is also variously defined as the totality of the weather or the customary course of the weather. Weather changes from moment to moment, but climate remains the same. Weather has to do with the particular values of the meteorological elements, while climate is concerned more with the normal values. 426 CLIMATE 427 Climatology is the science of climate and includes, not only a descrip- tion of the climate and a statement of the causes of its climatology characteristics, but also its effect on animal and vegetable defined - life and its relation to the occupations and activities of man. CLIMATIC DATA AND CHARTS 404. The material necessary for a description of the climate of a locality may be presented as tables of statistics or by means of graphs and charts. It should include the normal hourly, daily, monthly, and yearly values of all the meteorological elements of Jjjj e and all the tables of data which could be worked up in connec- which tion with the various elements. It should also include tables ^en. of data concerning the composition of the atmosphere, the amount of evaporation, the temperature of the ground, solar radiation, thunderstorms, fogs, the electrical condition of the atmosphere, the haziness of the sky, and the like. It might also include tables of data in connection with the dates of harvest, the freezing over of rivers, the arrival and departure of birds, and the like. Hann, in his Lehrbuch der Klimatologie, mentions thirty-six sum- maries of data which should always be included in a climatological study, and Abbe, in his Aims and Methods in Meteorological Work adds six more to the list. They are the following : (1) The monthly and annual mean temperature of the air. (2) The extent of the mean diurnal range of temperature for each month. (3) The mean temperature at two specific hours, namely, the early morning and midafternoon. (4) The extreme limits, or total secular range, of the mean temperatures of the individual months. (5) The mean of the monthly and annual extreme temperatures, and the result- ing non-periodic range. (6) The absolute highest and lowest temperatures that occur within a long interval of time. (7) The mean variability of the temperature as expressed by the differences of consecutive daily means. (8) Mean limit, or date, of frosts in spring and fall, and the number of con- secutive days free from frosts. (9) The elements of solar radiation as measured by optical, chemical, and thermal effects. (10) The elements of terrestrial radiation as measured by radiation ther- mometers. 428 METEOROLOGY (11) The temperature of the ground at the surface, and to a depth of one or two yards. (12) The monthly means of the absolute quantity of moisture in the atmos- phere. (13) The monthly means of the relative humidity of the air. (14) The total precipitation, as rain, snow, hail, dew, and frost, by monthly and annual sums. (15) The maximum precipitation per day and per hour. (16) The number of days having 0.01 inch or more precipitation, including dew or frost. (17) The percentage of rainy days in each month or the probability of a rainy day. (18) The number of days of snow, with the depth and duration of the snow covering. (19) The dates of first and last snowfall. (20) Similar data for the dates of hail. (21) Similar data for the dates of thunderstorms. (22) The amount of cloudy sky, expressed in decimals of the whole celestial hemisphere. (23) The percentage of cloudiness by monthly means, for three or more specific hours of observation. (24) The thickness of the cloud layer, or the amount of strong sunshine as shown by Campbell's sunshine recorder. (25) The number of foggy days, or the total number of hours of fog. (26) The number of nights with dew ; also the quantity of dew. (27) The monthly means or total of wind velocity, or estimated wind force. (28) The frequency of winds from the eight principal points of the compass, and the frequency of calms. (29) The frequency of winds for each hour of observation and the diurnal changes in the winds. (30) The meteorological peculiarities of each wind direction, or the respective wind roses for temperature, moisture, cloudiness, and rainfall. (31) The mean annual barometric pressure. (32) The total evaporation, daily, and monthly, or some equivalent factor, such as the depression of the dewpoint combined with the velocity of the wind. (33) Variations in the gases contained in the atmosphere, provided they are suspected to be of importance. (34) Impurities in the atmosphere, such as the number of dust particles, and especially the number of spores or germs of organic life. (35) The porportions of ozone, the peroxide of hydrogen, and nitric acid. (36) The electrical condition of the atmosphere, if there is any method of obtaining it. To these Abbe adds the following : CLIMATE 429 (37) The sensation experienced by the observer, such as mild, balmy, invigorat- ing, depressing, and other terms used to express the effect of the weather upon mankind. (38) The number of storm centers that pass over a given locality, or the storm frequency, monthly and annual. (39) Frequency of severe local storms. (40) The duration of twilight. (41) The blueness or haziness of the sky. (42) The number and extent of the sudden change from warm to cold, or moist to dry weather, and vice versa. Even the above list is by no means complete, as many more summaries could be added. Most of this material can be presented by means of graphs and charts, as well as by tables of statistics. THE FACTORS WHICH DETERMINE CLIMATE 405. The climate of different parts of the world is very different, and the chief factor which determines the climate is the latitude of the locality, for with the latitude varies the amount of insola- . . tion received from the sun. In fact, the word climate comes the chief from the Greek and means inclination. Primarily, then, it climati c was the varying inclination of the sun's rays at different latitudes which was considered. If latitude were the only factor in determining climate, all places with the same latitude would have the same climate. The climate which would exist if it depended on latitude alone is often called the solar climate. There are, however, several other climatic factors and the climate is not the same at all places hav- ing the same latitude. The solar climate as modified by these other factors is often called the physical climate, four other These modifying factors are : the relative distribution of land climatic and water, the altitude above sea level, mountain ranges, and the topography of the locality. Under this last is included the nature of the soil, and whether the country is vegetation-covered or forested. It is sometimes stated that the type of storm, the amount of rainfall, the direction of the prevailing winds, etc., are the factors which deter- mine climate. This would seem to be a mistake, as it is these Things things which constitute climate. The climate is different which in two localities, and this means that these things are differ- ^eTons^d- ent in the two localities. The climatic factors are the ered climatic factors which determine why these differences exist, and the one major factor and the four minor factors have just been stated. 430 METEOROLOGY THE CLIMATIC SUBDIVISIONS OF THE WORLD 406. Subdivision on the basis of latitude. The division of the world into geographical or climatic zones on the basis of latitude goes -back to the time of the Greek philosophers. The division of the WOI> ld into the five zones used at present is generally ascribed the present to Parmenides, who flourished about 450 B.C. These five fnto^ones 11 zones are the torrid zone, the two temperate zones, and the two frigid zones. The torrid zone is divided into two equal parts by the equator, and is bounded on the north and south by the tropics of Cancer and Capricorn. The width of the zone is thus 47, and the sun reaches the zenith on at least one day during the temperate,' year at every place within this zone. The two frigid zones and frigid ij e w holly within the Arctic and Antarctic circles and sur- round the two poles. The sun never rises at least one day in the year at all places within these zones. The two temperate zones lie between the frigid zones and the torrid zone, and each is 43 wide. The torrid zone covers 40 per cent, the two temperate zones 52 per cent, and the two frigid zones 8 per cent of the earth's surface. It will thus be seen that they are of very unequal size. The names of the zones are unfortunate, as they would seem to suggest a temperature basis for the subdivision, while in reality the subdivision is on the basis of latitude only. The torrid zone is more appropriately called the tropical zone, and the frigid zones the polar zones. This system of Parmenides which has come down to us was used by Aristotle (about 384 B.C.). Various other methods of dividing the world into zones were proposed by different Greek philosophers. Eudoxus of Cnidus, who lived about 366 B.C., divided the quadrant of the earth into 15 parts, of terns^!?^- wn i c ^ 4 were assigned to the torrid zone, 5 to the temperate, division on and 6 to the frigid. The tropics and polar circles were thus latitude! 8 f fixed at 24 and 54 of latitude respectively. Claudius Ptolemy, the great astronomer and geographer, who flourished at Alexandria in about 150 A.D., proposed a different system. Near the equator, the width of a zone was determined by a difference of 15 minutes in the length of the longest day. In higher latitudes, differences of half an hour, an hour, and finally a month, were used. Until within the last century or two all systems of subdivision were based on latitude only. This means that latitude was practically the only climatic factor which was recognized, although a few mentioned the importance of other things. CLIMATE 431 407. Subdivision on the basis of temperature. Temperature is the most important of the meteorological elements in its influence on plant and animal life and the occupations and habits of life of man. Isothermal lines do not follow parallels of latitude, and it has thus been proposed to use them instead follow of parallels of latitude for bounding the various zones. latitude 8 According to Supan, the equatorial or torrid zone should be limited on either side by the normal annual isotherm of 68 F. This would cross the United States from the southern part of California to the northern part of Florida. The intermediate or temper- supan's ate zone has for its poleward limit the isotherms of 50 F. subdivision, for the warmest month. This would cross North America from Alaska to a little north of Newfoundland. In this system of subdivision there would thus be five zones as before, only they would be bounded by iso- thermal lines instead of parallels of latitude. Koppen has proposed a system of subdivision, based upon tempera- ture, into nine belts or zones, a central zone and four on each side of it. Two of these are further subdivided into three parts. These Koppen's belts are as follows : system of (1) Tropical belt : All months hot, that is, with a normal subdivision - temperature of over 68 F. during all months. It would extend roughly from 20 N. to 16 S. latitude. (2) Subtropical belts : 4 to 11 months hot, that is, over 68 F. ; 1 to 8 months temperate (50 F. to 68 F.). (3) Temperate belts : 4 to 12 months between 50 F. and 68 F. (4) Cold belts : 1 to 4 months temperate ; the rest cold, that is with a normal monthly temperature below 50 F. (5) Polar belts : all months below 50 F. The two temperate belts are subdivided into three parts with these characteristics : (a) constant temperature during the year, (6) hot summers, (c) moderate summers and cold winters. 408. Subdivision on the basis of the general wind system. The importance of the general wind system in determining the geographical distribution of the various meteorological elements has been brought out in previous chapters. It was shown that it was zone sub _ the chief factor in determining the distribution of precipita- division tion, and the absolute and relative humidity were both the genera i closely correlated with it. It would thus seem that the wind world might be subdivided into climatic zones on the basis sys of the permanent wind system. When subdivided on this basis, 432 METEOROLOGY nine zones are usually recognized, a central zone and four on each side of it. The central zone is called the subequatorial zone, and here would be experienced the calms of the doldrums and trade winds blowing in opposite directions at different times of year. Next to this central zone, on either side, would be the two trade wind zones. Next to these would come the two subtropical belts or zones. Here would be found the calms of the horse latitudes and winds blowing from opposite directions dur- ing different parts of the year. Beyond these zones would come the zones of the prevailing westerlies, but as they cover such a very large part of the earth's surface, it has seemed best to divide each of these zones into two by the polar circles. There would thus be nine zones, corresponding to the terrestrial wind system. 409. Subdivision on the basis of surface topography. The climate of different localities in the same zone is by no means the same. This is true no matter if latitude, temperature, or the general A zone may be divided wind system has been made the basis of subdivision into mt S1 usin zones. It is thus necessary to subdivide the zones on the surface basis of the characteristics of the surface. Six subdivisions topography are usua iiy made : ocean, west coast, east coast, plain, plateau, and mountain. The climate on the east coast of an ocean is usually somewhat different from that on the west coast. For this reason, two coast or littoral climates must be recognized. It is also not sufficient to say a climate is continental as distinguished from marine or littoral. The low-lying areas, the plateaus, and the mountain regions must be treated separately. Thus, if the world is first divided into zones on the basis of latitude, or temperature, or the general wind system, and then if these zones are subdivided on the basis of nature of the surface, the resulting subdivi- sion will be small enough that a general description of the climate of each one can be given, and what is said about the climate of one locality will hold for all other places in the same subdivision. 410. Other climatic subdivisions of the world. There are four other climatic subdivisions of the world, each with a different basis, which deserve a brief consideration. These are the subdivisions of Supan, Koppen, Ravenstein, and Herbertson. Supan arbitrarily divides the world into 35 so-called climatic provinces. The attempt is here made to group together those near-by places which Supan's have the same surface topography and where the meteoro- ciimatic logical elements have the same characteristics. The number provinces. o f these provinces, namely 35, and their boundaries are CLIMATE 433 purely arbitrary. Provinces No. 25, 26, 27, and 28 include the United States, and their characteristics are as follows : No. 25, Californian province. Here it is relatively cool, especially in summer, and there is a marked subtropical rainy season. No. 26, North American mountain and plateau province. Here there are great daily and yearly ranges in the values of the meteorological elements, and it is also dry. No. 27, Atlantic province. Here the chief characteristics are contrast in temperature between north and south in winter, extreme climate even on the coast, a plentiful, evenly distributed rainfall, rapid changes. No. 28, West Indian province. Here there is an equable temperature and rain at all seasons, but with a well-marked summer maximum. Koppen's climatic subdivision of the world has a botanical basis. Five kinds of plants are recognized : (megotherms) those which need a continuously high temperature and abundant moisture; ^oppen's (xerophytes) those which like high temperature and dryness ; botanical (mesotherms) those which require moderate temperatures syst< and a moderate amount of moisture ; (mikrotherms) those which need lower temperatures ; (hekistotherms) those which will live in low tem- peratures. The five main divisions are further subdivided until the whole number reaches twenty-four. According to Ravenstein, the world is subdivided into sixteen cli- matic types, and the basis of the classification is tempera- R aven _ ture and relative humidity. stein's According to Herbertson, the world is subdivided into syst< six natural geographical regions, and these are subdivided until the whole number of recognized climates reaches fifteen. His Herbert- basis of classification is a combination of temperature, rain- son's fall, topography, and vegetation. system. 411. It will thus be seen that there are many methods of subdividing the world into smaller areas for the purpose of discussing the climate, and many different bases have been used for the various The purpose classifications. The purpose in each case has been to form of sub- a sufficient number of climatic provinces so that the climate lvlslon> at different localities in the same province would be essentially the same. In general, the larger the number of provinces, the more nearly alike will the climate be at different places in the same province. 434 METEOROLOGY THE CLIMATE OF THE UNITED STATES 412. Introduction. The United States is a country of such vast extent and with such a diversified surface that practically no statements The climate can ^ e ma ^ e about the climate of the country as a whole, in different The statements which would be true for one part of the country is 6 country would be entirely incorrect for another. In one very dif- part of the country, New England for example, the weather is dominated by an almost unbroken procession of passing highs and lows. As a result, great irregular changes in the values of the meteorological elements follow each other in quick succession. The precipitation is all storm-caused and quite evenly distributed throughout the year. The climate is thus extreme, very variable, and with copious, evenly distributed precipitation. In another part of the country, California for example, there is a well-marked rainy season. The pre- cipitation here is caused mostly by the general wind system of the world. The irregular changes in the meteorological elements are few, so that during any one season, one day is much like another. The cli- mate is thus very uniform with a well-marked wet and dry season. Subdivision The first essential, then, in discussing the climate of the essential. United States is to subdivide the country into smaller areas so that the climate of all places in the same district will be practically the same. 413. Subdivision. The United States might be subdivided into smaller areas for the purpose of discussing the climate, using any one of The three ^ ne systems of subdivision and classification which have just systems of been treated. Since, however, the U. S. Weather Bureau usedTyThe nas c H ec ted practically all of the climatological data, it Weather would be better to adopt the systems of subdivision which have been followed there. Formerly the country was divided into twenty-one climatic divisions (see section 357), and the climatological data and statistics were summarized for each of these districts separately. At present, these districts are still used to a slight extent in summarizing data. When the form of the Monthy Weather Review was changed in July, 1909, the country was subdivided into twelve climatological districts (see section 357), and the observations are now summarized for these districts as a whole. These districts are also adhered to as far as practicable in matters of administration. In 1910 was commenced the publication of the summaries of the climatological data of the United States, by sections, and for this purpose the country was divided into CLIMATE 435 106 districts. Thus in discussing the climate of the United States, it would be best to consider the country divided into 21, 12, or 106 districts, following the subdivisions of the U. S. Weather Bureau. 414. Detailed treatment of certain subdivisions. The scope of this book prevents the complete discussion of the climate of even one place or district. The reader who is interested in the climate of any particular locality must be referred to the literature of the subject for information. The general question, however, could well be raised as to what would constitute a full and complete discussion of the climate of a place. Such a complete treatise might conveniently consist of the following nine parts : (I) A map of the locality and surround- full, com- ing regions is usually presented first. This map should con- P. lete trea ~ tain the usual geographical features and the elevations. A climate of a brief description of the cities, mountains, etc., usually place , would accompanies the map. (II) Next a full description of the surface topography might be given. This would include the char- acteristics of the rivers, the nature of the soil, whether forested or not, and the like. (Ill) Next the climatological data might be presented as tables of statistics or as charts and graphs. The source of the data, the length of the records, and their probable accuracy might also be discussed. (IV) Next would come a discussion of the data, perhaps considering first the meteorological elements in order (temperature, pressure, wind, moisture, cloud, precipitation), and then the other meteorological and phenological occurrences, such as frost, fog, evapora- tion, time of harvest, migration of birds, etc. (V) Next might come a discussion of the types of storms, their prevalence and severity. (VI) Next the influence of the climate on certain diseases and the general healthfulness of the climate might be discussed. (VII) Next the influ- ence of the climate on agriculture and vegetation might be discussed. Here would be considered such questions as the kind of crops which could be best grown, the possibility of fruit growing, the kind of forest trees which would be most common, etc. (VIII) Next the general influence of the climate on the industries and habits of life of the people might be discussed. (IX) A bibliography might be added. These nine subdi- visions will serve to indicate what one might expect to find in a full, complete discussion of the climate of a locality. 436 METEOROLOGY THE CLIMATE OF OTHER COUNTRIES AND PLACES 415. It is entirely impossible to present, in a limited space, the com- plete discussion of the climate of even one country or place. This will not be attempted, but the chief characteristics of the three great zones, the torrid, temperate, and frigid, should perhaps be stated. The zones are, however, of such large extent that there are but few characteristics which are common to an entire zone. The climate of the torrid zone is characterized by great unifprmity, small irregular changes in the meteorological elements, high tempera- bar ^ ure > an d a small yearly variation in temperature. Nowhere teristics of else in the world is the weather so nearly the same day after the torrid jay. 'pjjjg means a verv uniform climate, and the reason zone* for it is the type of storms. Tropical cyclones and thunder- showers are the only storms. Tropical cyclones are few in number, occur at certain times of year, and cover a very small area. Thundershowers are very prevalent, occurring at many places almost daily. Large, irregu- lar changes in the meteorological elements following each other in quick succession are thus almost unknown. The noonday sun always stands high in the sky and the change in temperature during the year is small. Monsoons are well marked in many countries in the torrid zone, and the rainfall occurs either during the rainy season, caused by a monsoon, or almost daily, usually with a thundershower in the afternoon, when con- vection is most powerful. The seasons thus depend more on the general wind system and the rainfall than on changes in temperature. The temperate zone is characterized by a very variable climate and large changes in temperature between summer and winter. The tem- The temper- peratures are, of course, lower than in the torrid zone. The ate zone. temperate zone is constantly being traversed by passing highs and lows, and, as a result, the changes in the meteorological ele- ments are abrupt and large. This makes the climate very variable. The rainfall results both from the general wind system and the passing storms. The frigid zone is characterized by greater uniformity and much lower temperature than the temperate zone. The changes in the meteoro- The frigid logical elements, particularly temperature, between summer zone. anc j winter, are large. The daily changes at times are very small. Large, irregular changes are also present. CLIMATE 437 THE CONSTANCY OF CLIMATE 416. The question of the constancy of climate must be discussed for three different time intervals. First, has the climate remained constant during the recent past, say the last hundred years ? Secondly, has the climate remained constant during historic times, say the last 7000 years ? Thirdly, has the climate remained constant during recent geologic ages, say the last 10,000,000 years? There are many stations where meteorological observations have been made for more than a hundred years. In fact, a few records cover more than three hundred years. Based upon these observa- tions, the statement can confidently be made that the cli- has re- mate is essentially the same now as it was many years, or mained un- even a hundred years ago. This is largely contrary to popu- during the lar belief. It means that, taking one year with another, the last hun ~ snowfall is just as large now as then. It means that sleigh- ing lasts just as long now as then. It means that the winters are no milder now than then. It means that our summers are no hotter now than then. The constant statements by the older people, that the climate is different now than it used to be when they were much younger, are due to the tendency to magnify and remember the unusual while the ordinary is forgotten. Thus, in time, it is only the unusual snowfall or the extremely low temperatures that are well remembered, and un- consciously the abnormal has thus been substituted for the normal. These statements are also due to the fact that the attitude towards life, the amount of energy, the daily occupations, and perhaps the place of residence of the older people are very different now than when they were much younger. In discussing possible changes in climate during the last 7000 years, inference must be drawn from such recorded facts as the dates of har- vest, the kind and amount of crops raised, the kind of cloth- ing worn by the people, the habits of life of the people, the has re- existence of certain wild animals and forest trees, the size of mail ^ ed es ~ rivers, the height of lakes and inclosed seas, etc. From same during evidence of this kind, the conclusion has been drawn that h . istoric there have been no marked changes in climate during his- toric times. It has been often thought that certain climatic cycles have been detected. The 11-year and 35-year cycles have been well investi- gated, while cycles of much longer duration have also been suspected. The 11 -year cycle corresponds to the sun spot cycle and is very poorly 438 METEOROLOGY marked. Bruckner's 35-year cycle is on the contrary fairly well marked in Europe, both in temperature and precipitation. None of these cycles, with the possible exception of the 35-year cycle, are at all regular or well marked. (See section 397.) There can be no doubt th&t the climate has changed greatly during recent geologic ages. Almost tropical vegetation has existed in Green- Great land, and glaciation has extended many times far towards changes the equator. Various explanations of these changes have togkdages keen advanced. Among them are a change in the location and the pos- of the earth's axis, a change in the eccentricity of the earth's sibie causes. OJ fo^ the precession of the equinoxes which brings the long cold winter to the northern (land) hemisphere every 25,000 years, a change in the energy emitted by the sun, a change in the composition of the earth's atmosphere, a change in the elevation of the place, a change in the distribution of land and water, and thus the ocean currents. THE SNOW LINE 417. The temperature in general grows less with increasing altitude, and thus there are regions on high mountains even in the torrid zone Definition of near ^ e e q uator where the snow which falls during one the snow winter has not sufficient time to melt entirely during the line * following summer before the advent of the snows of the next winter. These are regions of perpetual snow, and the lower boundary of these regions is called the snow line. On the equatorial Andes, the height of the snow line is about 5000 meters, roughly three miles. With increasing south latitude, the height its height in ^ ^ ne snow ^ me increases somewhat due to scantier precipi- South tation, and reaches a maximum height of about 6000 meters America. in ^ soui ^ latitude. The height of the snow line then de- creases rapidly with increasing latitude. Its height is about 3500 meters in latitude 32 ; 1600 meters in latitude 42 ; 800 meters in lati- tude 50 ; 400 meters in latitude 55 ; and according to the observa- tions made by certain Antarctic expeditions, it reaches sea level in from 67 to 70 south latitude. In the northern hemisphere, the height of the snow line is somewhat its height in S rea ter than in South America. In the Alps in Europe (lati- other parts tude about 46) it is about 2900 meters. In Norway (lati- of the world. tude about 61 o) it ig about 1500 meters In the Himalaya Mountains the height is about 5000 meters. In North America, at CLIMATE 439 latitude 19 N., in Mexico, the height of the snow line is about 4600 meters. On Mt. Shasta the height is 2400 meters; in the Cascade Mountains, at the northern boundary of the United States, it is 2000 meters ; on Vancouver Island it is 1700 meters ; on Mt. St. Elias (lat. 60 N.) it is 800 meters. Even in polar regions it does not everywhere absolutely reach sea level. The height of the snow line does not depend on latitude alone, and thus is often somewhat different on two mountains in the same latitude. The other factors which influence the height of the line are The factors the amount of snow during the winter, the change in tern- ei perature between winter and summer, the steepness of the height. mountains, the exposure of its slopes, and the general wind system. TOPICS FOR INVESTIGATION (1) What a full list of climatic data would consist of. (2) The factors which determine climate. (3) Some system of climatic subdivision and its basis. (4) The contrast between marine and littoral climate in the same latitude. (5) The climatic characteristics of the continents. (6) Climatic changes during historic times. (7) Geological changes in climate. (8) The change in the height of the snow line between winter and summer. (9) The temperature of the snow line. (10) Effect of climate on the mental characteristics of the inhabitants. PRACTICAL EXERCISES (1) Prepare tables of climatic data. (2) Express these tables graphically. (3) Summarize the observations of some station with a long record to show the constancy of climate. (4) Treat fully the climate of some place. REFERENCES The following books and pamphlets deal with climate or climatology in gen- eral: BEBBER, W. J. VAN., Hygienische Meteorologie, Stuttgart, 1895. BONACINA, L. C. W., Climatic Control, viii + 167 pp., London, 1911. CULLIMORE, D. H., The Book of Climates for all Lands, London, 1890. HANN, JULIUS, Handbuch der Klimatologie, 3 vols., 2d ed., Stuttgart, 1897; 3d ed. of Vol. I, 1908; of Vol. II, 1910; of Vol. Ill, 1911. HERBERTSON, A. J. and F. D., Man and his Work, London, 1899. KOPPEN, Klimakunde, 2d ed., Leipsig, 1906. MEYER, HUGO, Anleitung zur Bearbeitung meteorologischer Beobachtungen fur die Klimatologie, 8, viii + 187 pp., Berlin, 1891. 440 METEOROLOGY MOORE, WILLIS L., Climate. Bulletin No. 34 of U. S. Weather Bureau (W. B. publication No. 311). RATZEL, Anthropogeographie, 2d ed., Stuttgart, 1899. STOCKMAN, WILLIAM B., Invariability of our Winter Climate, (W. B. publica- tion No. 312). SUPAN, A., Grundzuge der physischen Erkunde, 4th ed., Leipzig, 1908. WARD, R. DEC., Climate, G. P. Putnam's Sons, New York, 1908. WARD, R. DEC., Harm's Handbook of Climatology (translation of Vol. I), The Macmillan Co., 1903. WOEIKOF, A., Die Klimate der Erde, Jena, 1887. (Of the books just mentioned, the two by Ward are by far the most complete and readable in English. Taken together they cover the subject of climate in a most complete and interesting way. The three-volume work by Hann in German is by far the most complete and useful treatise on the subject. It is a veritable mine of information.) For a description of the climate of the United States as a whole and of various places in the country, see : BLODGET, Louis, Climatology of the United States, xvi -f- 536 pp., Philadelphia 1857. FASSIG, OLIVER L., The Climate and Weather of Baltimore, Baltimore, 1907. GREELY, A. W., Report on the Climate of Colorado and Utah, Washington, 1891 HAZEN, HENRY A., The Climate of Chicago. Bulletin No. 10 of U. S. Weather Bureau. HENRY, ALFRED J., Climatology of the United States, 1012 pp., Washington, 1906. Weather Bureau, Bulletin Q. (The standard descriptive and statistical work on this subject.) M'ADIE, ALEX. G., and WILLSON, GEORGE H., The Climate of San Francisco Bulletin No. 28 of the U. S. Weather Bureau, (W. B. publication No. 211). M'ADIE, ALEX. G., Climatology of California Bulletin L of U. S. Weather Bu- reau, (W. B. publication No. 292). SMOCK, JOHN C., Climate of New Jersey, Trenton, 1888. (Part of the final report of the state geologist.) Summary of the Climatological Data for the United States, by Sections (106 are to be issued, covering the whole country). WALDO, FRANK, Elementary Meteorology (Chapter 13). For a description of the climate of various countries and place, outside of the United States, see : ABBE, CLEVELAND, JR., The Climate of Alaska, Washington, 1906. (Extract from Professional Paper No. 45, U. S. Geological Survey.) ABBOT, HENRY L., Climatology of the Isthmus of Panama (W. B. publication No. 201). ALEXANDER, WILLIAM H., Climatology of Porto Rico, Monthly Weather Review, July, 1906. ALGUE, JOSE, The Climate of the Philippines, 103 pp., Manila, 1904. BEHRE, OTTO, Das Klima von Berlin, 8, 158 pp., Berlin, 1908. BLANFORD, Climates and Weather of India, 8, 382 pp., London, 1889. DAVIS, WALTER G., Climate of the Argentine Republic, vi + 154 pp., Buenos Aires, 1902. ELIOT, SIR JOHN, Climatological Atlas of India, Edinburgh, 1906. KNOX, ALEXANDER, The Climate of the Continent of Africa, 8, xii + 552 pp., Cambridge, 1911. CLIMATE 441 MOORE, JOHN W., Meteorology, 2d ed. London, 1894. (Climate of the British Islands, Chapters XXV and XXVI.) NAKAMURA, K., The Climate of Japan, 109 pp., Tokio, 1893. PHILLIPS, W. F. R., Climate of Cuba, Bulletin 22 of U. S. Weather Bureau, (W. B. publication No. 163). QUETELET, A., Sur le climat de la Belgique, 4, 2 vols., Bruxelles, 1849. Rizzo, G. R., II Clima di Torino. ROSTER, GIORGIO, Climatologia delV Italia nelle sue attineze con Vigiene e con agricoltura, 8, xxix + 1040 pp., Torino, 1909. For a treatment of the variations in climate, see : BUCKNER, EDWARD, Klimaschwankungen seit 1700, viii + 324 pp., Wien, 1890. ECKARDT, WILHELM R., Das Klimaproblem der geologischen Vergangenheil und historischen Gegenwart, 8, vi + 183 pp., Braunschweig, 1909. Die Veranderungen des Klimas seit dem Maximum der letzten Eiszeit (Pub. by llth International Geological Congress), 4, Iviii + 459 pp., Stock- holm, 1910. For climatic charts see the references in connection with the charts of the meteor- ological elements at the end of the various chapters. CHAPTER X FLOODS AND RIVER STAGES DEFINITION OF THE TERMS USED IN CONNECTION WITH RIVERS, 418. THE MEASUREMENTS MADE IN CONNECTION WITH RIVERS Velocity of flow, 419. River stage, 420. Cross section of a river, 421. River discharge 422. RIVER DATA, 423 THE DIFFERENT KINDS OF FLOODS, 424 THE CHARACTERISTICS OF INDIVIDUAL RIVERS, 425 THE PREDICTION OF RIVER STAGES AND FLOODS, 426, 427 SUDDEN RISES OF OCEANS AND LAKES, 428 DEFINITION OF THE TERMS USED IN CONNECTION WITH RIVERS 418. There are several terms used in connection with rivers and floods which require at the outset exact definition and brief consideration. Drainage area. By the drainage area of a river is meant the tract of country from which the water drains into the river. This is some- Drainage times called the catchment basin or the watershed. The area - area drained by some of the rivers of the world is enormous, that by the Amazon probably being the largest. The drainage area of the Mississippi River and its tributaries probably stands next in size. The Missouri River drains about 527,150 square miles ; the Ohio River, about 201,700 ; the Arkansas River, about 186,300 ; the Red River, about 90,000; so that the Mississippi River and its tributaries drain about 1,240,000 square miles. Even the smallest brook with a name usually has a drainage area of a good many square miles. A complete descrip- tion of the drainage area of a river would include an account of its topography, meteorology, and climate. Inland drainage area. If a tract of country is landlocked, that is, surrounded on all sides by land of greater elevation, the rivers will flow 442 FLOODS AND RIVER STAGES 443 to the lowest point and form a lake or sea without an outlet to the ocean. Such an area is called an inland drainage area, and the evapora- tion must here equal the precipitation. The sea or lake is Inland often below sea level and is usually salt. The reason is drainage because the rivers always carry down a small amount of salt and other minerals, and these are left behind when evaporation takes place. The Caspian Sea, with an area of 180,000 square miles, is the largest sea of this kind in the world. Great Salt Lake, Utah, contains 17 per cent of salt and other minerals, and is the best known inland lake in the United States. In Australia, 52 per cent of the whole country consists of inland drainage areas; in Africa, 31 per cent; in Europe and Asia, 28 per cent; in South America, 7.2 per cent; in North America, 3.2 per cent. Run-off. By run-off is meant the percentage of the precipitation which eventually drains into a river, and this varies all the way from a few per cent to nearly 90 per cent in some extreme cases. The run-off depends chiefly upon the characteristics of the drainage area, but also upon the amount and rapidity of the rainfall. Ground is ordinarily classified as permeable or impermeable. If rock or a stratum of material impervious to water is near the surface, the ground is im- permeable, and the run-off, in this case, will be large and occur very shortly after the precipitation. If the ground is permeable, the run-off is much smaller and much more gradual and regular. Most watersheds have also a decided seasonal change in characteristics. At one time of year vegetation may be luxuriant. At another time of year the ground may be frozen hard or covered with a layer of snow and ice. All these seasonal changes make a tremendous difference in the run-off. The run-off also depends, to a large extent, on the amount and rapidity of the rainfall. It increases both with the amount of the rainfall and the rapidity with which it falls. One of the hardest problems in connection with rivers is to try to estimate the run-off when, over a drainage area in a certain condition, a certain amount of rain falls in a certain time interval. The average run-off for all watersheds in the world is between 20 and 30 per cent. Thalweg. The term thalweg is sometimes used to designate the valley bottom through which a river runs. Thalweg. Regimen. The term regimen is used to designate the characteristics of a river. Its normal height, its greatest and least height, its normal discharge of water, its normal velocity of flow, its cross section all these things go to make up its regimen. 444 METEOROLOGY River stage. By river stage is meant the height of the surface of a river above some arbitrarily chosen zero point. The zero point may be mean sea level, or the lowest point reached by the river or the normal height of the river, or any arbitrarily chosen point. Thus, if a river stage is stated as 36 feet, it simply means that the surface of the river is 36 feet above the definite zero point from which all heights are reckoned. Flood line. The flood line is some definite river stage so chosen because a greater height than this can be considered a flood. Thus, if the flood line is 40 feet, it means that a river stage above 40 feet would result in an overflow and a damage-causing flood. River slope. By river slope is meant the change in elevation of the river surface with distance. It is usually expressed as so many inches per mile. In the case of great rivers, it is never more than a few inches per mile. In rapidly flowing streams it is much more and may amount to several feet. Wetted perimeter. ^ The length of a line from one side of a river Wetted to the other measured along the bottom is called the wetted perimeter perimeter. Mean hydraulic depth. The area of the cross section of a river -, . divided by the wetted perimeter is called the mean hy- Meanhy- J draulic depth, draulic depth. THE MEASUREMENTS MADE IN CONNECTION WITH RIVERS 419. Velocity of flow. The velocity of flow of a river is determined ordinarily by means of a current meter, which consists essentially of a A current propeller wheel which revolves faster the greater the velocity meter. o f ^ ne curre nt. The Price current meter as made by W. and L. E. Gurley of Troy, N.Y., is pictured in Fig. 146. This instrument is used by the U. S. Coast and Geodetic Survey and by many hydraulic engineers in different parts of the country. It consists essentially of five conical buckets so arranged that they turn easily with the slightest current. They are provided with a rudder consisting of four light metal wings or vanes in order to keep the wheel in line with the current. A heavy weight (about sixty pounds) with a wooden rudder is attached for deep-water work or where the current is particularly swift. The instrument is so constructed that electrical contact is made after every revolution of the wheel, and the number of revolutions is counted auto- FLOODS AND RIVER STAGES 445 matically by an electric register. A reduction table is furnished with the instrument for finding the velocity which corresponds to a given FIG. 146. The Price Current Meter. number of revolutions per second. It would be a little more accurate to determine the reduction table for each instrument separately. This 446 METEOROLOGY is done by dragging it at a known rate of speed through still water. Sometimes, instead of an electric register for counting the number of revolutions, the wheel is so constructed that a hammer strikes against a diaphragm after every ten revolutions, and the sound is conveyed to the ear of the observer. These are called acoustic current meters. In making a complete determination of the velocity of flow of a river, observations must be made at different depths and in different parts of a river. This practically amounts to determining the velocity of flow at all points in the cross section. If only the surface velocity is desired, it can be determined roughly by watching some floating object and determining the time required for it to be carried a known distance. The velocity of flow can also be determined by computation from the slope of the river and the mean hydraulic depth. In order to determine the slope, the difference in level of the river at methods of P omts several miles apart must be accurately determined determin- by surveying methods. Only approximate results can, onflow?"* 7 however, be obtained, as the constants in the formula for computing velocity from slope and mean hydraulic depth are too uncertain and depend upon too many things. 420. River stage. The river stage or the height of the river surface above some arbitrarily chosen point is determined by means of a river The river gauge. This consists ordinarily of a heavy plank, 8 or 10 gauge. inches wide and a couple of inches thick, and of sufficient length to cover the greatest possible fluctuations in the height of the surface of the river. The gauge is ordinarily divided into feet, possibly inches or tenths of a foot, and the foot marks are usually numbered. The gauge is placed vertically and securely fastened to a bridge pier, the end of a wharf, or the like. The height of the surface of the river can thus be readily read off. Sometimes the gauge is not placed vertically, but is inclined to follow the river bank. It should always be graduated to show vertical heights, however. Every river gauge should be provided with a permanent bench mark near by on shore. A bench mark is simply a very stable, permanent The bench point whose elevation above sea level is not supposed to mark. change. A copper bolt in the stone foundation of a building, the water table of a firmly placed building, the surface of some large stone in a building, all serve well as bench marks. By surveying methods, the height of the bench mark above mean sea level should be determined, and also the difference in level between the bench mark and the zero of FLOODS AND RIVER STAGES 447 the river gauge. If a river gauge is then repaired or carried away by a flood, the new one can be placed with its zero mark at exactly the same level as before. The zero of a river gauge may be placed anywhere on the gauge, but it is customary to put it so low that the lowest water will never reach it. Negative values of river stages are thus avoided. The following descriptions of the river gauges at Albany, N.Y., and New Orleans, La., taken from W. B. publication No. 227, will serve as illustrations. Albany, New York " Albany, N.Y., is on the Hudson River, 150 miles from its mouth. " The gauge is a self -registering tide gauge, patterned after those used by the United States Coast and Geodetic Survey in former years. It is the property of the United States Engineer Corps, and is located on the east side of the State Street Bridge. " The bench mark was established in 1896 by the United States Engi- neer Corps, is on the southeast corner of the east basement window on the south or State Street front of the United States Government build- ing near Dean Street. It is 18 feet above the zero of the gauge and 18.2 feet above mean sea level. " The highest water was 21.4 feet on February 9, 1857. It was due to back water. On October 4 and 5, 1869, the water reached a stage of 18.5 feet, the highest stage due to rainfall alone. The lowest water was -1.2 feet on Sept. 30, 1867." New Orleans, Louisiana " New Orleans, La., is on the Mississippi River, 108 miles above the Gulf. The river is 2400 feet wide. The drainage area above the sta- tion is 1 T 235,500 square miles. " The river gauge is the property of the city and is situated at the foot of Canal Street among a cluster of piles in rear of ferry wharf. It is made of cypress, and is painted white with markings in black. " Bench mark at corner of Common and Delta Streets, on iron cornice, 6 inches above sidewalk at E. Conery's store, is 16.5 feet above zero of gauge, and 14 feet above mean sea level. Curbstone under third window of customhouse from Decatur Street, and on Customhouse Street, is 11 feet above zero of gauge, and 8.5 feet above mean sea level. " Graduation is from zero to 17 feet above. Highest water was 19.5 feet on May 13, 1897 ; lowest, -0.2 feet on December 27, 1872. Dan- ger line is at 16 feet." 448 . METEOROLOGY 421. Cross section of a river. The determination of the cross section of a river belongs to hydrographic surveying. It is necessary to deter- The cross mine by sounding the depth of the water every few feet all section of a the way across the river. These observations can then be plotted to scale and a cross section of the river determined. 422. River discharge. River discharge may be found in two ways: by means of a weir or dam, and by computation from the cross section and the velocity of flow. we * r me ^hod is by far the most accurate, but can be termining applied only to small streams. It consists in forcing the charge!* water to flow over a weir or dam. By measuring the width of the stream and the depth of the water flowing over the dam, fairly exact values of the discharge can be found by computation. If the cross section of a river is known and the average velocity of flow has been determined, the discharge can be computed. The product of the area of the cross section in square feet by the velocity in feet per second gives the discharge in cubic feet per second. RIVER DATA 423. A complete description of a river and its characteristics would include material and data concerning both the watershed and the river What a itself. In connection with the watershed or drainage area, complete a full treatment would include an account of its topography, ofTriver* 1 meteorology, and climate. The most important items are would a map of the watershed showing the elevations and the char- acter of the surface, and normal values and data concerning the precipitation, snowfall, temperature, evaporation, and run-off. All other facts in connection with the topography, meteorology, or climate would be of interest and value, but of secondary importance. A treatment of the river itself would include a detailed description of the course of the river and data in connection with the flow of water. The cross section of the river at various points, its length, the height of its banks, and the area which would be covered by a rise of a given amount should all be known. In connection with the flow of water, normal values and data for the river stages, the velocity of flow at different stages, the river discharge at various stages, the causes of floods, and the velocity of progression of a flood wave should all be known. Complete data for a river and its drainage area are probably available for very few rivers, but the important items are known for most rivers. FLOODS AND RIVER STAGES 449 THE DIFFERENT KINDS OF FLOODS 424. Floods in rivers may be caused in a variety of ways. (1) Floods may be caused by the breaking of a dam, by the breaking of levees, or by a sudden change of course by a river. (2) Floods may . be caused by the temporary blocking of a river by an ava- kinds of lanche, landslide, or glacier. This choking of the river channel fl . oods in would cause a temporary lake back of the obstruction, and the giving way of this barrier might cause disastrous floods along the lower course of the river. (3) Floods can be caused by the luxuriant growth of vegetation which may choke the river channel and thus cause a rise of water. (4) Floods may be caused by the formation of ice dams when the ice breaks up in the spring. These cause floods both above and later below the dam. (5) Floods are often caused by rather sudden melting of the snow and ice over a considerable portion of a watershed. (6) Floods due to excessive precipitation over the water- shed are the most common of all floods. Of these various kinds of floods, those caused in the first two ways are of unusual occurrence, and are far from common. Those caused by vegetation and ice dams are fairly common in certain rivers. Those caused by the melting of snow and excessive precipitation are the commonest of all floods. The best-known flood due to the breaking of a dam is the one which occurred at Johnstown, Pa., June 1, 1889. A reservoir about 3 miles long and 1 mile wide and perhaps 100 feet deep was held by a iu ustr ations dam 1000 feet wide. The sudden breaking of this dam pre- of the va- cipitated the water upon Johnstown, 18 miles below, and nous kinds * caused the loss of nearly 3000 lives. The breaking of levees along the lower Mississippi River, particularly when the river is in flood, is of fairly common occurrence and often floods large areas. The sudden changing of the course of a river usually occurs in moun- tainous regions where the river is small and swift, or at the delta where a very large river flows into the ocean. In the first instance, very little damage is usually done. Disastrous results ordinarily follow when a large river changes its course. It is said that the Hwangho river in China is particularly prone to change its channel, and that the place where it empties into the ocean has varied as much as 300 miles during the last 4000 years. These great changes have caused the loss of many millions of lives. The temporary blocking of a river by an avalanche or landslide or by 2a 450 METEOROLOGY the forward movement of a glacier occurs only in mountainous regions near the source of a river. This blocking of the channel may cause, however, a fairly large lake ; and if the barrier breaks suddenly, disas- trous floods are sure to occur along the course of the river. Floods due to vegetation are said to be common along the Upper Nile and in the Parana River in South America. Ice dam floods are common in all rivers which freeze over to a con- siderable depth in winter. When the ice breaks up in the spring, it is apt to form a dam at a narrow part of the channel or where it is ob- structed by a bridge. Such a dam causes floods above it, and, if it breaks suddenly, floods below it are likely to occur. Floods due to the sudden melting of the snow and ice on the watershed or to excessive precipitation over the watershed are common in all parts of the world. These two causes are also likely to occur together during the late winter and early spring, in which case the rise of the rivers will be particularly large. Of these six kinds of floods, the last two are the only ones which can be predicted. THE CHARACTERISTICS OF INDIVIDUAL RIVERS 425. Limited space does not permit the detailed treatment of the characteristics of even one particular river. The reader must be referred to the literature on the subject for the characteristics of any river in which he may be especially interested. What would be included in such a complete treatise has already been stated in connection with river data. A few facts about the more important rivers of the world may be of interest. In the Nile the lowest water occurs ordinarily in June and the highest in September or October. At Cairo, Egypt, the about indi- rise is from 15 to 30 feet, and the country is inundated annually to a considerable distance on each side of the river. The fertility of Lower Egypt is due to the water gained in this way, and to the alluvium brought down by the river. The rise in the Nile is due to large rainfall in Abyssinia of the monsoon type. In the Yangtse-Kiang, the Amur, and the Hwangho rivers, the rise occurs during the late summer. This is due to monsoon rains over the interior of Asia and shows to what extent the monsoon penetrates the continent. Floods during the summer are particularly damage-causing FLOODS AND RIVER STAGES 451 as they occur during the time of growing crops. A summer rise due to monsoon rains is also true of the Congo, the Ganges, and the Brahma- putra. The Amazon River changes but little during the year, as the rainfall is more uniform, and when the tributaries on one side are in flood, those on the other side are usually not. In the Mississippi and Ohio rivers, and in the Rhine, Seine, and Elbe in Europe, the floods occur during the winter and spring. They are caused by large rainfall over the watershed while the ground is frozen, or by the sudden melting of large quantities of snow. Very often the two causes operate together. The rivers of New England and near-by states also have floods during the winter and spring for the same reasons. In these rivers ice dams often form when the ice breaks up in the spring. THE PREDICTION OF RIVER STAGES AND FLOODS 426. The prediction of river stages and the height and time of occur- rence of a flood crest may be made in two different ways. One is on the basis of what has occurred on the watershed, and the other -is on the basis of river stages which have been ob- methods of served on gauges higher up the river and on its various P redi cti n g , ., . river stages. tributaries. If the first method is followed, the condition of the watershed must be known ; that is, its extent, whether the ground is frozen or not, the amount of snow which may rest on it, etc. A measured int^^st amount of rain at a known temperature has fallen on the method oc- whole or a part of the watershed in a certain time. The S l JJSl on problem is to determine the resulting rise in the river due to shed are the run-off. It might seem that the problem could be solved used< theoretically; that is, that it could be determined what the run-off would be for the watershed in a known condition due to the rainfall in question. As a matter of fact, the problem is too complex to be treated with any accuracy in this way. If, however, records for a con- siderable time are available, it is possible to determine from past occur- rences about what the resulting rise in a river will be for certain happen- ings on the watershed. But even when done in this way, the results are so uncertain that this method is never used by the U. S. Weather- Bureau in forecasting. The second method of predicting floods and river stages is by means of gauges placed higher up the river and on its various tributaries. 452 METEOROLOGY When a flood occurs the water rises, attains its greatest height, and then falls. A flood may thus be considered as a wave which progresses down a river at a certain speed. By means of the river gauges omTmethod P lace d higher up the river and on its tributaries the location gauge read- and height of the different flood waves may be determined, used"' Based on past experience, the height of the flood wave at the station in question and its time of occurrence can be predicted. The whole prediction of floods rests upon a series of rules or tables derived by studying critically the records of previous floods. This method can, of course, be applied only to the lower part of the course of a river and not to its sources. 427. The prediction of river stages and floods is part of the regular work of the U. S. Weather Bureau, and belongs to the river and flood The work of sec ti n - It was formerly under one man, but the predictions the Weather are now made by many near the various rivers. Daily river gauge readings are made at 8 A.M. at many stations located on the various rivers. These observations are telegraphed to 32 centers, and the preparation of forecasts and warnings is, in most cases, intrusted to the officials of the Weather Bureau at these centers, under the super- vision of the forecast official at Washington. The regulations concern- ing the issuing of forecasts have already been stated in section 391. SUDDEN RISES OF OCEANS AND LAKES 428. Sudden rises of the ocean are caused either by earthquakes or tropical cyclones. There are several notable examples where tidal Sudden rise waves more than 50 feet high have been caused by earth- of the ocean. q ua kes and great loss of life has resulted. The rise of water caused by the Galveston cyclone (see section 271) is a good example of the second kind. A sudden increase in the height of the surface of a lake at a certain point is called a seiche^ and these are common in the Great Lakes of the United States and in some lakes in Switzerland, and other parts of the world. They are probably caused by thunder- showers or a sudden change in wind direction or velocity. They some- times amount to several feet. It is said that the surface of the water on the south side of Lake Erie is always several feet higher when the north wind blows than when the south wind blows. Thus a sudden change in wind direction might readily cause a seiche. After the surface of a lake had once been changed from a level surface by the action of the wind, if FLOODS AND RIVER STAGES 453 the wind should die down to a calm, the surface would then oscillate about a nodal line near the center of the lake until finally brought to rest by friction. The continuance of seiches after the cessation of wind can thus be explained. TOPICS FOR INVESTIGATION (1) The run-off under various conditions. (2) Inland drainage basins. (3) The various kinds of current meters. (4) The complete description of some river. (5) The seiche. PRACTICAL EXERCISES (1) Investigate critically some small brook, determining the characteristics of the watershed and the discharge and behavior of the brook under all condi- tions. (2) Determine the cross section of some river and the velocity at all points in it. REFERENCES FRANKENFIELD, H. C., The Floods of the Spring of 1903 in the Mississippi Watershed. Bulletin M ; W. B. publication No. 303. HENRY, ALFRED J., Wind Velocity and the Fluctuations of Water Level on Lake Erie. Bulletin J ; W. B. publication No. 262. MOORE, WILLIS L., The Influence of Forests on Climate and on Floods. MORRILL, PARK, Floods of the Mississippi River. Bulletin E ; W. B. publica- tion No. 143. RUSSEL'L, THOMAS, Meteorology, New York, 1895. (Chapters IX and X cover rivers and floods and river stage predictions.) Daily River Stages at River Gauge Stations on the principal Rivers of the United States (6 parts have been issued. The Publication was com- menced by the Signal Service and continued by the Weather Bureau.) Monthly Weather Review. (The condition of the rivers and the occurrence of floods are here summarized monthly. The hydrographs of the seven prin- cipal rivers are also given.) CHAPTER XI ATMOSPHERIC ELECTRICITY INTRODUCTION HISTORY, 429 THE ELECTRICITY OF THE EARTH, AIR, CLOUDS, AND RAINDROPS The measurement of potential differences due to the earth's electric field, 430. The electric field of the earth, 431. The conductivity of the atmosphere, 432. The source of the charged ions and the earth's negative charge, 433, 434. The source of the electricity of clouds and raindrops, 435. Air currents and earth currents, 436, 437. THE NATURE AND KINDS OF LIGHTNING The cause and kinds of lightning, 438. Zigzag lightning, 439. Other kinds of lightning, 440. DANGER FROM LIGHTNING AND PROTECTION FROM LIGHTNING Loss of life and property due to lightning, 441. Protection from lightning, 442. OTHER MANIFESTATIONS OF ATMOSPHERIC ELECTRICITY, 443. INTRODUCTION HISTORY 429. Lightning is an electric spark on a tremendous scale. Vague, indefinite opinions that this might be the case were expressed by various Early opin- scientists from 1600 on, but the first definite assertion was ions and ex- made by J. H. Winkler in Leipzig, in 1746, and he attempted lts * to prove it by analogy. Benjamin Franklin proposed experi- mental proofs in 1749, and in 1752 sent a communication to the Royal Society in London, recommending the use of rods with points as light- ning conductors. D'Alibard, at Marly sur Ville, near Paris, through translating Franklin's communication, received the incentive to carry out a series of experiments. An iron rod some forty feet long and pro- vided with a point was attached to an insulated support. When a thundershower approached, sparks an inch or more in length could be drawn from the rod. Franklin's famous kite experiment was performed a little later. In a letter dated October 19, 1752, he describes it as follows: 454 ATMOSPHERIC ELECTRICITY 455 " Make a small cross of light sticks of cedar, the arms so long as to reach to the four corners of a large, thin silk handkerchief when extended. Tie the corners of the handkerchief to the extremities of the f Tan j / ^ n y a cross, so you have the body of a kite which, being properly kite ex- accommodated with a tail, loop, and string, will rise in the penment - air like those made of paper, but being made of silk is better fitted to bear the wet and wind of a thunder gust without tearing. To the top of the upright stick of the cross is to be fixed a very sharp-pointed wire, rising a foot or more above the wood. To the end of the twine next the hand is to be tied a silk ribbon, and where the silk and twine join a key may be fastened. This kite is to be raised when a thunder gust appears to be coming on, and the person who holds the string must stand within a door or window, or under some cover, so that the silk ribbon may not be wet ; and care must be taken that the twine does not touch the frame of the door or window. As soon as the thunder clouds come over the kite, the pointed wire will draw the electric fire from them, and the kite, with all the twine, will be electrified, and stand out every way and be attracted by an approaching finger. And when the rain has wet the kite and twine, you will find the electric fire stream out plentifully from the key on the approach of your knuckle." De Ramas, in France, a few years later (1757) was able to get sparks ten to twelve feet long by means of a kite. A few days after D'Ali- bard's experiment, Le Monnier was able, by means of better The basis constructed and insulated apparatus, to prove that there was of the old a difference of potential between a point at a given height in conce P tion - the air and the earth at all times, even when the sky was clear. From this time on, long series of observations were made by many investigators in different parts of the world. The digest of all this experimental and observational material led to the formation of what we may now call the old conceptions concerning atmospheric electricity. This old con- ception has been considerably modified during the last twenty years, but still deserves careful consideration. The earth itself is a conductor, and is surrounded by a non-conducting medium, the atmosphere. The earth is highly charged with negative electricity and in the non-conducting atmosphere charged around it, there is thus a field of force. There would thus be rthsur - rounded by a difference of potential between a point at a given height in a non-con- the atmosphere and the earth, and this potential difference would be greater, the greater the elevation of the point in question. Since the earth has a charge of negative electricity, lines of 456 METEOROLOGY force must start from the earth and extend outward. The great ques- tion was, where these lines of force ended, in other words, where the corresponding charges of positive electricity were located. Some said in the clouds, others in the outer regions of the atmosphere, or on the heavenly bodies, or in the remote depths of space. The origin of the earth's negative charge was never satisfactorily explained, it being generally supposed that it had had its charge from the beginning. Floating in the atmosphere and carried from one point to another, are innumerable dust particles and minute water drops. These are con- The charged ductors and charged with either negative or positive elec- particies. tricity. Various ways in which these particles may have become charged have been suggested. (1) They may have become electrified by friction. Carried rapidly by the wind, these particles might strike against material objects or each other. Ice crystals and snowflakes in the upper air may strike against each other. (2) At the moment of evaporation, a particle may have become charged nega- tively by induction. (3) If a cloud forms, the lower side would have a positive charge induced on it, and the upper side would have a negative charge. If such a cloud should be suddenly broken through horizontally by the wind, the two portions would be charged and with opposite kinds of electricity. (4) Every particle floating in the atmosphere would have positive electricity on its lower side and negative electricity on its upper side due to induction by the charged earth. Now ultra-violet light readily discharges negative electricity. Thus the negative electricity on these particles might be discharged and scattered, which would leave the particle charged with positive electricity. There are thus at least four different ways in which these little conducting particles might be- come charged with negative or positive electricity. When a cloud forms, these particles become nuclei of condensation, and thus the cloud particles become highly electrified. When these cloud particles further unite to form raindrops, the quantity of electricity steadily increases, until finally there is a lightning flash between two clouds charged dif- ferently or between a cloud and the earth. This whole conception, then, briefly summarized is as follows. The earth is a nearly spherical conductor charged with negative electricity and surrounded by a non-conducting medium. It will be surrounded by equipotential surfaces and there will be lines of force going out from it. In the non-conducting surrounding medium there are conducting particles which become positively or negatively electrified. These serve as nuclei of condensation when a cloud forms, ATMOSPHERIC ELECTRICITY 457 and thus cloud particles and raindrops collect electricity until a light- ning flash occurs. The first difficulty with this explanation was encountered when it was found that the atmosphere was not a non-conductor, but a poorly con- ducting medium. It was found that a charged body in the atmosphere was slowly discharged. This was at first laid cuities with to poor insulation of the body, or the presence in the atmos- the ? ld con " phere of dust and water particles. It was soon found that this discharging of a charged body coulcl not be accounted for in this way, as the rate of discharge was slower when the moisture was high or even when it was foggy. Since the atmosphere conducted the elec- tricity more like an electrolyte than a metallic conductor it was soon assumed that there were present in the atmosphere, numerous small particles or portions of molecules, called gas ions, which were charged some positively, some negatively. The actual existence of these ions has since been experimentally demonstrated. The presence of the charged ions at once raised new questions. What was the origin of these ions, and how had they become charged, some positively, some nega- tively ? How did the earth retain its negative charge ; in other words, why was not the earth discharged? It was also soon found that these ions, particularly the negatively charged ones, also served as Read j ust _ nuclei of condensation. Another source of the electrifica- mentneces-] tion of the cloud particles and raindrops had thus been sary * found. A readjustment of the old conception was thus necessary, and this will be shortly given. THE ELECTRICITY OF THE EARTH, AIR, CLOUDS, AND RAINDROPS 430. The measurement of potential differences due to the earth's electric field. Since the earth is charged with negative electricity, it must be surrounded by an electric field of force and a poten- Two pieces tial difference must exist between any point in the atmos- of apparatus phere and the charged earth. In order to measure the B Jssary - potential differences between the earth and a given point in the atmos- phere, an electrometer and a " collector " are necessary. In the early determinations, a simple crude electroscope was used as the electrom- eter. It consisted of a glass globe or case, containing a metal rod to which was attached two pith balls or two light straws, or The eiec- two leaves of thin gold foil. By the divergence of these light frometer. objects the charge and thus the potential difference could be judged. 458 METEOROLOGY In Fig. 147, two simple electroscopes which may serve as electrometers are shown. One is a gold leaf electroscope which has been made more sensitive and accurate and has been provided with a scale. The other is Braun's electrometer. Here a light aluminum pointer moves over a scale and indicates the potential difference. For more precise measure- ments, some form of quadrant electrometer must be used. For the description, theory, and use of these well-known pieces of electrical apparatus, the reader must be referred to text-books on physics. FIG. 147. Two Simple Electroscopes. (Froni GOCKEL'S Die Luftelektrizitat.) The so-called collector is placed at the point in the atmosphere for which the potential difference as regards the earth is to be determined. The In the early experiments, it consisted of an insulated point collector. or p O i n t s connected by means of a wire with the electrometer. It would take up the potential of the point where it was located, and the difference of potential between this point and the earth would thus be indicated by the electrometer. Later, the flame of a lamp placed on an insulated support or some slowly burning substance was used. Still later, the water-dropping collector was devised. This consists simply of a vessel of water on an insulated support, from which the water is allowed ATMOSPHERIC ELECTRICITY 459 to fall drop by drop. This last collector has probably been the most widely used of any. In still more recent experiments, a small plate or rod covered with a radioactive substance has been used. The method of determining the potential difference is thus to connect one part of the electrometer with the earth, and the other part with the collector which is placed at the point, and the of h determin d electrometer reading indicates the potential difference. ingpoten- 431. The electric field of the earth. Since the earth is a conductor highly charged with negative electricity, it must be surrounded by an electric field of force, extending out indefi- nitely through the atmosphere, and an equipotential surface could be drawn through any point in this field of force. By an equi- potential surface is meant a surface containing all points tial surfaces which have the same potential difference as regards the and thcir earth. If the earth were a perfectly smooth conductor, the equipotential surfaces would be parallel to the earth's surface, that is, concentric with the earth. As a matter of fact, the surface of the earth is far from smooth and level, and the irregularities greatly distort the equipotential surface. Numerous investigations have been made to determine the effect of a hill, moun- tain, tree, building, or the like, on the equipotential surfaces. It has been found that the general effect of projections is to warp the equipotential surfaces upward and cause them to be closer together. This is represented roughly in Fig. 148. It will be seen from this that the change in potential with elevation would be very small beside a building or hill. On the other hand, above a tree or hill it would be particularly large. This must be held in mind in choosing a point for which to determine the potential difference as regards the earth. The most typical and usual values would The geo _ be found by choosing a point over a level plain. graphical, The change in potential with elevation amounts ordinarily n ^f f '^ to about 100 volts per yard, and a point in the atmosphere irregular va- is positive as compared with the earth. This change in ^ pote n_ potential with elevation is by no means a constant. It tial differ- grows rapidly less with altitude. It is very different in different parts of the world. It has a periodic daily and annual varia- FIG. 148. The Equipotential Surfaces over an Irregular Surface. (From HANN'S Lehrbuch der Meteorologie.) 460 METEOROLOGY tion and very large irregular fluctuations, which are closely correlated with the meteorological elements and storms. Near the earth's surface, as just stated, the change in potential amounts to about 100 volts for an ascent of one yard. At a height of two or three miles, the change per yard has decreased to nearly one half its value at the earth's surface and there is some evidence that, at the height of five miles or more, a change with elevation practically ceases to exist. This proves that the earth is not the only charged conductor, but that there are charges of electricity in the atmosphere itself. If the earth alone were charged, the change in potential with elevation would not cease to exist. The values found at various places on the earth's surface are very different. This may be, in a large part, due to irregularities in the surface, but the larger values seem to be found in middle latitudes. Smaller values seem to be found for cold or dry places. The daily variation is very complicated, and seems to show two maxima and two minima very similar to the daily variation in barometric pressure. The maxima occur in the middle of the morning and in the early evening. The minima occur in the early afternoon and before sunrise. The graph which represents the daily variation is very different for different places and sometimes has . only one maximum and minimum. The annual variation shows a maximum in winter and a minimum in summer. The values of potential difference grow less with higher temperatures. Under long-continued bright sunshine, the values are usually less. The values also grow less with increasing cloudiness. With increasing dampness and during foggy weather, the values are usually larger. The effect of wind and pressure is extremely small and has never been defi- nitely determined. During a snowstorm or thundershower tremendous irregular fluctuations occur. The positive potential difference often becomes negative and may attain values as high as 10,000 volts per yard. If under normal fair weather conditions, the change in potential per yard is taken as 100 volts, and if it is furthermore assumed that this state The charge ^ things is the same over the whole earth, the negative of the potential to which the earth must be charged can be com- puted. The value would be about 600,000,000 volts. 432. The conductivity of the atmosphere. Until recent times, the atmosphere was always considered a non-conductor. The Se C c