key: cord-318043-1x3dp1vv
authors: Ahmadi, Mohsen; Sharifi, Abbas; Dorosti, Shadi; Ghoushchi, Saeid Jafarzadeh; Ghanbari, Negar
title: Investigation of effective climatology parameters on COVID-19 outbreak in Iran
date: 2020-04-17
journal: Sci Total Environ
DOI: 10.1016/j.scitotenv.2020.138705
sha: 
doc_id: 318043
cord_uid: 1x3dp1vv

Abstract SARS CoV-2 (COVID-19) Coronavirus cases are confirmed throughout the world and millions of people are being put into quarantine. A better understanding of the effective parameters in infection spreading can bring about a logical measurement toward COVID-19. The effect of climatic factors on spreading of COVID-19 can play an important role in the new Coronavirus outbreak. In this study, the main parameters, including the number of infected people with COVID-19, population density, intra-provincial movement, and infection days to end of the study period, average temperature, average precipitation, humidity, wind speed, and average solar radiation investigated to understand how can these parameters effects on COVID-19 spreading in Iran? The Partial correlation coefficient (PCC) and Sobol’-Jansen methods are used for analyzing the effect and correlation of variables with the COVID-19 spreading rate. The result of sensitivity analysis shows that the population density, intra-provincial movement have a direct relationship with the infection outbreak. Conversely, areas with low values of wind speed, humidity, and solar radiation exposure to a high rate of infection that support the virus's survival. The provinces such as Tehran, Mazandaran, Alborz, Gilan, and Qom are more susceptible to infection because of high population density, intra-provincial movements and high humidity rate in comparison with Southern provinces.

Since late December 2019, patients presenting with viral pneumonia due to an unidentified microbial agent were reported in Wuhan, China. It was an outbreak of the novel Coronavirus disease named 2019 novel coronavirus (COVID-19; previously known as 2019-nCoV). The disease has rapidly spread from Wuhan to other areas and affected 196 countries worldwide by March 25, 2020, which raised intense attention internationally (Chen et al., 2020; Lu et al., 2020; Phan et al., 2020; Xu et al., 2020) . It is really important to find out all the factors that play a role in COVID-19 spreading in Urban. The transmission of viruses can be affected by many factors, including climate conditions (such as temperature and humidity), population density and medical care quality (Wang et al., 2020) . Therefore, understanding the relationship between the geographical features of a country and the transmission of COVID-19 is key to making the best decision to control and prevent the pandemic. In other words, the discipline of geography is playing in the fight against the virus SARS-CoV-2, which causes coronavirus disease . Therefore, urban geography can be very helpful in that the spatial organization of the city determines the spatial pattern of the spread of the disease (Boulos and Geraghty, 2020; Wang et al., 2020) . While a set of health experts are banking on warmer weather conditions to slow down, if not completely halt, the Coronavirus, it is yet not clear whether the coming months will bring any respite to the world. Based on recent research, which has yet to be peer-reviewed, indicated that two factors include high temperature and humidity, directly correlated with the spreading of COVID-19 in a region (Wang et al., 2020) . (Araujo and Naimi, 2020) . However, the spreading of COVID-19

virus in hot and humid conditions will not stop entirely. Malaysia has confirmed more than 1,500 cases of the virus; more than 500 people are infected in Indonesia; and in Singapore, where the average temperature is around J o u r n a l P r e -p r o o f

year-round, despite rigorous detection methods and strict quarantine rules. The results of the research show that longer-term dramatic change of solar flux of ionizing radiation leads to provide an opportunity to create nono-metrical viruses like SARS and MERS (Qu and Wickramasinghe, 2017) . Also, high solar radiation prevents an outbreak with inactivating (or semi-infectious) coronaviruses (Gupta et al., 2015; Qu and Wickramasinghe, 2017) . Based on the results of (Qu and Wickramasinghe, 2017) identified. Finally, the discussion and conclusion presented.

The main variables in this study include geographical indicators with averaged data from February 19 to

March 22 from Weather Spark online web service ("The Typical Weather Anywhere on Earth -Weather Spark," n.d.). Variables including the number of infected people with COVID-19 reported by ("WHO | World Health

Organization," n.d.), population density, intra-provincial movement, infection days to end of the study period, average temperature( • C) (Yuan et al., 2006) , average precipitation (mm) (Araujo and Naimi, 2020) , humidity(%) (Wang et al., 2020) , wind speed (km/h) (Yuan et al., 2006) and average solar radiation (kWh/m 2 ) (Qu and Wickramasinghe, 2017) in the study period. The infection rate as a dependent variable defined as Eq.

(1).

This variable indicates the rate of infection or speed of the COVID-19 spreading. Table 1 shows the descriptive statistical and meteorological data of Iran from February 19 to March 22. 

(2)

Where P is average annual precipitation (mm), T is the average annual temperature ( • C) and I is De Martonne aridity coefficient. In this equation, the evaporation is indirectly considered. This method is more widely used in Iran for two reasons. The first availability of the factors and second the classification of this method can define diverse climates (Zareiee, 2014) . The De Martonne classification is shown in table 2. 

Iran can be divided into at least four different climate zones. Different clustering current climate regions in Iran shown in Fig. 1a . More than half of Iran's area is Arid and consists of lands with high average temperature shown in orange color. However, the semi-arid area includes high mountain and agricultural lands. This area is gray in Fig. 1a . Also, three provinces of Iran have the Mediterranean, wet and very wetlands consisting of jangles and grasslands.

The outbreak of COVID-19 based on the first case of observation is shown in the geographical map ( Fig.1b ). The first case of COVID-19 infection occurred in Qom and then spread to all provinces over the 13days period. Regarding the figure, Qom and Tehran can be described as the center of outbreaks that have spread throughout the south and north. In some provinces, this outbreak has occurred earlier, which has been spectacularly significant in terms of intra-provincial movement. In some provinces, this outbreak has occurred after 13 days, which can be attributed to a lack of awareness of the people. In order to get a better understanding of the COVID-19 outbreak in the Iranian provinces and the study of geographical factors, the meteorological data recorded in the study period are referred. The results of Pearson correlation analysis between the variables are shown in Table 3 . In this study, the infection rate was defined as the independent variable and its correlation with geographical variables analyzed. Do these results show whether the rate of disease outbreaks in the provinces depends on geographical factors or not? Using these results, we can infer the meteorological impact scenario on the outbreak of COVID-19. According to Table 3 

In this part of the research, sensitivity analysis between variables is discussed. Therefore, we use two methods of Partial correlation coefficient (PCC) and Sobol'-Jansen method to examine the importance of variables and prioritize it. In the PCC method, a linear regression must be made between the independent variables of the problem and the infection rate. On this basis, the basic geographical data should first be standardized.

Linear regression is created after changing data range between -1 and 1. Table 4 ). We are not satisfied with linear regression, to investigate these variables more consistently, we will continue to use one of the most powerful methods for identifying In this paper, we used the multi-layer perceptron (MLP) technique with 10 hidden neurons to create a more accurate model. We used 70% of the data for training, 15% testing, and 15% model validation, to obtain the best model with the highest accuracy (R = 92%). The results were recorded as shown in Fig. 2 .

By applying the MLP model on Sobol'-Jansen method, results are depicted in Fig.3 d. Figure Figure 3 shows the contour infection rate of COVID-19 disease concerning geographic variables (population density and movement). The higher value of the population density and intra-provincial movement leads to a higher value of infection rate so that in high-density provinces, the rate of disease growth reaches 150 people per day. Figure 3 b also shows the bubble plot of the average temperature and relative humidity over the study period. The size of bubbles also indicates the rate of infection, and the bubbles have different colors for different climates from Iran. Results show that in arid and semi-arid regions, disease rates are higher in areas with low humidity than in areas with high humidity.

However, two wetlands of Mazandaran and Gilan are places with a high rate of infection. Because these provinces are areas with very high population density. With the classification of the area to different climates, the average rate of disease spread in wetlands is higher than in other areas of Iran (Fig. 3 c) . Figure 4 illustrates the governing variables involved in the change of Coronavirus COVID-19 spreading rate. The maps are depicted and colored based on data quartiles. Figure 4 i shows the COVID-19 infection rate over the study period.

Observations show that the infection rate in the provinces of central Iran is higher than in the border regions, and these provinces are both high in population density and movement. A comparison of the virus infection rate charts and meteorological maps can justify the results obtained in the aforementioned methods.

According to Fig. 4 f, it can be seen that the humid area includes north, west, and northwest of Iran. However, the main part of the outbreak is in humid, very humid, and semi-arid (or mountainous) areas. The wind direction in most parts of Iran is from the west and in the northeast to west. Wind speeds are low in most parts of the country. Based on Fig. 4 h solar radiation in central Iran is lower than southeast with a low infection rate. We can conclude that where the infection rate is high, the population density and movement are high. Also, in this area humidity, wind speed and solar radiation have a reverse relationship with the rate of the disease spreading.

The 

The in that the spatial organization of the city determines the spatial pattern of the spread of the disease.

Iran is one of the countries with different types of climate. The infection rate in provinces is different; therefore, understanding the main variables that play an important role in this various spreading rate is crucial. In this study, the correlation of nine main variables includes the number of infected people, population density, intra-provincial movement, days of infection, average temperature, average rain, humidity, wind speed, and solar radiation with infection rate analyzed. The sensitivity analysis between variables determined by two methods Partial Correlation Coefficient (PCC) and Sobol'-Jansen to examine the importance of variables and prioritize it.

Based on the PCC method, the population density, intra-provincial movement, day of infection have a direct relation with infection outbreak. Conversely, wind speed, humidity, and solar radiation have an indirect correlation with the infection rate. However, in two humid regions of Iran, the rate of virus spreading is high.

The Sobol'-Jansen method also approved the aforementioned results and confirmed that the main variables that play an important role in the COVID-19 outbreak are population density, intra-provincial movement, wind 

Spread of SARS-CoV-2 Coronavirus likely to be constrained by climate. medRxiv

Geographical tracking and mapping of coronavirus disease COVID-19/severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) epidemic and associated events around the world: how 21st century GIS technologies are supporting the global fight against outbr

The effects of temperature and relative humidity on the viability of the SARS coronavirus

Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study

Application of gene expression programming and sensitivity analyses in analyzing effective parameters in gastric cancer tumor size and location

Do sunspot numbers cause global temperatures? Evidence from a frequency domain causality test

Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding

Importation and human-to-human transmission of a novel coronavirus in Vietnam

SARS, MERS and the sunspot cycle

Variance based sensitivity analysis of model output. Design and estimator for the total sensitivity index

The Typical Weather Anywhere on Earth -Weather Spark

High Temperature and High Humidity Reduce the Transmission of COVID-19

WHO | World Health Organization

Pathological findings of COVID-19 associated with acute respiratory distress syndrome

A climatologic investigation of the SARS-CoV outbreak in Beijing, China. Am

Evaluation of changes in different climates of Iran, using De Martonne index and Mann