key: cord-0940260-k1l16pmm
authors: Wen, M.; Chen, L.
title: Impacts of regional climate on the COVID-19 pandemic
date: 2020-06-13
journal: nan
DOI: 10.1101/2020.06.13.20130013
sha: 9441e5a9c5f2b43d8863dc450cec06544c792943
doc_id: 940260
cord_uid: k1l16pmm

The COVID-19 pandemic has led to six million confirmed cases by May 31, 2020. Impacts of regional weather and climate on epidemics have been investigated but need further study with new methods. We combined the number of monthly confirmed new cases and death with month, latitude, temperature, humidity, rainfall, and sunshine ultraviolet (UV) to explore the climate impact on epidemics in 116 countries and territories with at least 1000 confirmed cases. Correlation and regression analyses were performed with Stata. Humid subtropical climate regions had the most confirmed COVID-19 cases (24.4%). The case mortality in temperate marine regions was the highest (11.6%). Case-weighted means of the latitude, monthly maximum temperature, relative humidity, rainfall, and sunshine UV were 36.7 degrees, 20.5, 63%, 63mm, and 53.5, respectively. The case mortality was 7.44% in cold regions but only 4.68% in hot regions, 7.14% in rainy regions but only 3.86% in rainless regions, and 7.40% in cloudy regions but only 4.64% in sunny regions. Monthly confirmed cases increase as the temperature, rainfall, and sunshine UV rise in cold regions (r=0.34, 0.26, 0.26, respectively), but no correlation in hot regions. Every 1 increase in monthly maximum temperature leads to an increase in the natural logarithm of monthly confirmed new cases by 2.4% in cold regions. Monthly confirmed cases increase as the temperature, rainfall, and sunshine UV rise in arid regions (r=0.29, 0.28, 0.26, respectively), but no correlation in humid regions. Monthly confirmed new cases increase as the temperature and sunshine UV rise in rainy regions (r=0.30, 0.29), but no correlation in rainless regions. Monthly confirmed new deaths increase as the temperature and sunshine UV rise in cloudy regions (r=0.30, 0.30), but no correlation in sunny regions. It is wise to escape from an epicenter full of miasma to a hot sunny place in dry season without pollution. As peaking in the spring depends on the climate, the peak will go in the summer.

The COVID-19 pandemic emerged in Wuhan in December 2019 and has led to 6000000 confirmed cases and 367000 death by May 31, 2020. 1 Impacts of regional weather and climate on the pandemic have been investigated widely. For example, one study revealed that daily temperature, humidity, wind speed, and UV index were associated with a lower incidence of 5 COVID-19. 2 UV light was strongly associated with lower COVID-19 growth rates in the early phase before intervention (β = -0· 44). 3 The temperature and humidity could explain 18% of the variation in case doubling time. 4 Every 1°C increase in temperature was associated with a 3· 08% reduction in daily new cases and a 1· 19% reduction in daily new deaths. 5 The doubling time of COVID-19 transmission is expected to increase by 40-50% when the temperature rise from 5°C 10 to 25°C . 6 The growth rate was affected by precipitation seasonality and warming velocity rather than temperature. 7 COVID-19 exploded during the darkest January in Wuhan in over a decade, as daily irradiance correlated with case growth seven days later. 8 A one-degree increase in absolute latitude is associated with a 2· 6% increase in cases per million inhabitants. 9 Experimental studies revealed that novel coronavirus could not be quickly destroyed by hot 15 temperature but UV radiation in the summer. Ninety percent of infectious SARS-COV-2 virus was inactivated every 6· 8 minutes in simulated saliva and every 14· 3 minutes in culture media when exposed to simulated sunlight. 10 However, UV light may be meaningless for inactivating viruses in areas with high air pollution where UV light turns into heat. 11

We investigated the climate impacts on the COVID-19 pandemic in 116 countries and territories with more than 1000 cumulative number of confirmed COVID-19 cases from December 2019 to May 2020 (table) . We obtained province-level case data from the National Health Commission of China and its . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 13, 2020. . https://doi.org/10.1101/2020.06.13.20130013 doi: medRxiv preprint provincial branches. 12 There are 29 provinces and cities with more than 100 confirmed cases in China. US state-level data is from The New York Times, based on reports from state and local health agencies. 13 All 50 states and Washington DC had more than 100 confirmed cases in the United States. We collected case data of each month from 194 epidemic regions across 116 countries and territories. The number of monthly new cases and deaths were transformed by 5 natural logarithm. These 194 regions are distributed in 12 climate regions based on Köppen climate classification: Tropical rainforest, tropical grasslands, tropical monsoon, tropical desert, humid subtropical, Mediterranean climate, subtropical highland, subtropical desert, humid continental, temperate marine, temperate grasslands, temperate desert, subarctic.

We extracted latitude data from the Google map and climate data from Weather-atlas. 14 We did 10 not use the minimum temperature data but the maximum temperature in terms of collinearity. We calculated the means of absolute latitude, monthly maximum temperature, relative humidity, rainfall, and sunshine UV weighted by the number of monthly confirmed new cases. We created a new climate indicator, sunshine UV, the product of average sunshine time times average UV index. For example, the average sunshine is 7· 5 hours per day and the average UV index is 6 in 15 April, so the average sunshine UV is 45.

Correlation and nonlinear regression analyses were performed with Stata software to estimate the impact of the climate on the number of monthly confirmed new cases. In order to evaluate the impacts of climate, all epidemic regions were divided into two temperature-groups, two humidity-groups, two rain-groups, and two UV-groups: hot regions and cold regions (≥20°C , 20 <20°C ), humid regions and arid regions (≥60%, <60%), rainy regions and rainless regions (≥40mm, <40mm), sunny regions and cloudy regions (≥50, <50).

The seasonal pattern is inverted between the northern and southern hemispheres. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 13, 2020. . . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 13, 2020. . https://doi.org/10.1101/2020.06.13.20130013 doi: medRxiv preprint Figure 3 shows monthly climate data of those center cities of all COVID-19 epidemic regions.

The means of absolute latitude, monthly maximum temperature, relative humidity, rainfall, and sunshine UV weighted by the number of monthly confirmed new cases are 36· 7 degrees, 20· 5°C, 63%, 63mm, and 53· 5, respectively.

The case mortality was 7.44% in cold regions but only 4.68% in hot regions. Pearson correlation analysis shows that monthly confirmed cases increase as the temperature, rainfall, and sunshine UV rise in cold regions (r=0· 34, 0· 26, 0· 26, respectively), but no correlation in hot regions.

Absolute latitude was weakly correlated to the natural logarithm of monthly confirmed new 10 deaths in hot regions (r=0· 37).

A nonlinear regression analysis was performed to model the relationship between the number of monthly confirmed new cases (y) and monthly maximum temperature (x) in cold regions. The R-squared of this model is 0· 87, p<0· 0001. This regression model predicts that every 1°C increase in monthly maximum temperature leads to an increase in the natural logarithm of 15 monthly confirmed new cases by 2· 4% in cold regions. The model is as follows,

The number of confirmed cases in humid regions was 70% of the total. Pearson correlation analysis shows that monthly confirmed cases increase as the temperature, rainfall, and sunshine UV rise in arid regions (r=0· 29, 0· 28, 0· 26, respectively), but no correlation in humid regions.

Monthly confirmed cases decrease as the relative humidity rises in humid regions (r= -0· 21) but no correlation in arid regions.

The number of confirmed cases in rainy regions was 70% of the total. The case mortality was 7· 14% in rainy regions but only 3· 86% in rainless regions where the monthly average rainfall is . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 13, 2020. . https://doi.org/10.1101/2020.06.13.20130013 doi: medRxiv preprint less than 40mm. Pearson correlation analysis shows that monthly confirmed new cases increase as the temperature and sunshine UV rise in rainy regions (r=0· 30, 0· 29), but no correlation in rainless regions. Monthly confirmed new cases decrease as the relative humidity rise in rainy regions (r= -0· 26) but no correlation in rainless regions.

The case mortality was 7· 40% in cloudy regions but only 4· 64% in sunny regions where the 5 monthly average sunshine UV is less than 50. Pearson correlation analysis shows that monthly confirmed new deaths increase as the temperature and sunshine UV rise in cloudy regions (r=0· 30, 0· 30), but no correlation in sunny regions.

We found that the impacts of temperature, humidity, rainfall, and sunshine UV on the COVID-19 10 pandemic were consistent with several previous studies. Aerosol transmission of the virus by water microdroplets and pollution particles as carriers is plausible since the virus can remain viable and infectious in aerosols for hours and the UV radiation can destroy the virus in minutes.

Many epicenters are humid and cloudy with air pollution and all kinds of viruses, forming so-called miasma. As the mortality rate was 12% in Northern Italy but only 4· 5% in the rest of 15 the country by March 21, 2020, researchers found a correlation between the high level of COVID-19 lethality and the atmospheric pollution in Northern Italy. 15 An increase of only 1 μg/m 3 in PM2· 5 is associated with an 8% increase in the COVID-19 death rate in the United States up to April 22, 2020. 16 It is wise to escape from an epicenter full of miasma to a hot sunny place with clean air. On the 20 contrary, extremely locking down an epicenter to prevent escape is not the best policy. Wearing masks is necessary for epidemic prevention, especially in an epicenter full of miasma.

Authorities should control air pollution, increasing fossil energy taxes to promote clean energy.

UV lamps, air heaters, and oxygen bottles are necessary to improve the air quality in ICU and isolating rooms for patients. We suggest UV lamps should be compulsory in all service spaces, 25 including classrooms, dormitories, hospitals, hotels, bars, restaurants, offices, warehouses, and public transport.

Why SARS disappeared in 2003 but MERS has continued until now? SARS virus was occasionally from the civet cat to mankind but the MERS virus was continuously from camels.

The animal-human interspecies transmission of the COVID-19 virus is likely another occasional 30 event. Additionally, the infectivity, virulence, and lethality of any virus or bacteria is certain to decline after several months of reproduction. The discovery of this principle led to the development of the first vaccine in the history of mankind: A doctor in Sichuan, China, inoculated against smallpox in the tenth century. 18 We suppose a new pandemic of new virus mutants and interspecies jumping is more possible than another COVID-19 pandemic.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted June 13, 2020. . https://doi.org/10.1101/2020.06.13.20130013 doi: medRxiv preprint

MW proposed the project and wrote the paper. LC supported the project and revised the paper.

We declare no competing interests. 5 We thank Deepak Gupta and Jessica Carr helpful feedback.

All data needed to evaluate the conclusion in the paper are present in the paper or the supplementary materials.

. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 13, 2020. . https://doi.org/10.1101/2020.06.13.20130013 doi: medRxiv preprint 35 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 13, 2020. . https://doi.org/10.1101/2020.06.13.20130013 doi: medRxiv preprint Supplementary Materials: Figure S1 -S3 Tables S1-S4

External Databases S1 5 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted June 13, 2020. . https://doi.org/10.1101/2020.06.13.20130013 doi: medRxiv preprint

Coronavirus disease (COVID-2019) situation reports. 132

Temperature, humidity, and wind speed are associated with lower Covid-19 incidence. medRxiv

Seasonality and uncertainty in COVID-19 growth rates. medRxiv

Role of temperature and humidity in the modulation of the doubling time of COVID-19 cases

Effects of temperature and humidity on the daily new cases and new deaths of COVID-19 in 166 countries. Sci total 10 environ

Temperature dependence of COVID-19 transmission. arXiv

Global COVID-19 transmission rate is influenced by precipitation seasonality and the speed of climate temperature warming. medRxiv

Follow the Sun: Slower COVID-19 Morbidity and Mortality Growth at Higher Irradiances. SSRN

COVID-19 and climate: global evidence from 117 countries. medRxiv

Simulated sunlight rapidly inactivates SARS-COV-2 on surfaces

UV light influences covid-19 activity through big data: trade offs between northern subtropical, tropical, and southern subtropical countries. medRxiv

Covid-19) Data in the United States

Weather-atlas.com

Can atmospheric pollution be considered a co-factor in 30 extremely high level of SARS-CoV-2 lethality in Northern Italy? Environmental Pollution

Exposure to air pollution and COVID-19 mortality in the United States. medRxiv