key: cord-0949964-ymuqe5eg authors: Yassin, Mohamed F.; Aldashti, Hassan A. title: Stochastic analysis of the relationship between atmospheric variables and coronavirus disease (COVID‐19) in a hot, arid climate date: 2021-08-18 journal: Integr Environ Assess Manag DOI: 10.1002/ieam.4481 sha: 0e121954eb93bb4b85ee3fa30a34557a9fc65f78 doc_id: 949964 cord_uid: ymuqe5eg The rapid outbreak of the coronavirus disease (COVID‐19) has affected millions of people all over the world and killed hundreds of thousands. Atmospheric conditions can play a fundamental role in the transmission of a virus. The relationship between several atmospheric variables and the transmission of the severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) are therefore investigated in this study, in which the State of Kuwait, which has a hot, arid climate, is considered during free movement (without restriction), partial lockdown (partial restrictions), and full lockdown (full restriction). The relationship between the infection rate, growth rate, and doubling time for SARS‐CoV‐2 and atmospheric variables are also investigated in this study. Daily data describing the number of COVID‐19 cases and atmospheric variables, such as temperature, relative humidity, wind speed, visibility, and solar radiation, were collected for the period February 24 to May 30, 2020. Stochastic models were employed to analyze how atmospheric variables can affect the transmission of SARS‐CoV‐2. The normal and lognormal probability and cumulative density functions (PDF and CDF) were applied to analyze the relationship between atmospheric variables and COVID‐19 cases. The Spearman's rank correlation test and multiple regression model were used to investigate the correlation of the studied variables with the transmission of SARS‐CoV‐2 and to confirm the findings obtained from the stochastic models. The results indicate that relative humidity had a significant negative correlation with the number of COVID‐19 cases, whereas positive correlations were observed for cases of infection and temperature, wind speed, and visibility. The infection rate for SARS‐CoV‐2 is directly proportional to the air temperature, wind speed, and visibility, whereas inversely related to the humidity. The lowest growth rate and longest doubling time of the COVID‐19 infection occurred during the full lockdown period. The results in this study may help the World Health Organization (WHO) make specific recommendations about the outbreak of COVID‐19 for decision‐makers around the world. Integr Environ Assess Manag 2021;00:1–17. © 2021 SETAC The coronavirus disease (COVID-19) is a viral infectious disease that is seriously threatening public health and the economy worldwide. COVID-19 was first identified in Wuhan, China, at the end of December 2019, and is caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The clinical symptoms of infection caused by the virus, such as dry cough, fever, pneumonia, dyspnea, and respiratory disorders, can lead to progressive respiratory failure and death in susceptible individuals. At the end of January 2020, the World Health Organization (WHO) declared the outbreak of this virus a public health emergency of international concern, after which it was declared a global pandemic on March 11, 2020. COVID-19 has spread across the globe, with more than 126 million confirmed cases of infection and 2.77 million deaths worldwide as of March 27, 2021 (World Health Organization, 2021 . Kuwait is considered a developing country and has been severely affected by this global health crisis. The Ministry of Health for the State of Kuwait declared the first case of SARS-CoV-2 infection on February 24, 2020. More than 226,000 confirmed COVID-19 cases and 1270 deaths occurred in Kuwait as of March 27, 2021 (World Health Organization, 2021 . Several vaccines became widely available during the review and post-acceptance of this paper, which has since greatly slowed transmission rates and case numbers where vaccination rates are high (World Health Organization, 2021) . Fear became a worldwide phenomenon during the period in which the transmission of the virus peaked because of the limited understanding surrounding its route of transmission. It is therefore important that parameters that may play a role in the transmission of SARS-CoV-2 and its spread in the outdoor environment are better understood. Transmission may be affected by various factors including atmospheric variables (such as temperature, humidity, wind speed, visibility, and solar radiation). Thus, understanding the relationship between atmospheric variables and the transmission of SARS-CoV-2 and its spread could be vital in making the best decisions that can minimize or prevent outbreaks of the virus; however, evidence of any possible relationship is still limited. Previous studies have indicated a relationship between the weather (short-term variation in meteorological conditions) or climate (long-term averages) and the outbreak and transmission of viruses (Casanova et al., 2010; K. Lin et al., 2006; Tamerius et al., 2013; Tan et al., 2005; Van Doremalen et al., 2013; J. Yuan et al., 2006; Lowen and Steel, 2014; Jaakkola et al., 2014) . Chan et al. (2011) and Park et al. (2020) investigated the effects of temperature, humidity, and the diurnal temperature range on both the SARS coronavirus and influenza. Their results indicated that high temperatures had a synergistic influence on the inactivation of the SARS coronavirus under high humidity, whereas the incidence of influenza increased under low temperatures and low/high humidity in temperate regions. Another study by Bhimireddy and Bhaganagar (2018) investigated the transport and dispersion of plume in the convective boundary layer conditions for Pure Convective, Shear-Buoyancy, and Near-Neutral regimes. The study's findings demonstrate that the horizontal transport of the plume scales with the near-surface releases and initial wind conditions for the surface, and the vertical transport scales with atmospheric stability parameters. The SARS-CoV-2 outbreak transmission is affected by several environmental variables, for example, weather, climate and atmospheric variables, virus infectivity, host defense mechanism, host behavior, and so forth. Atmospheric variables are essential parameters that are a global primary concern caused by the prevalence of COVID-19 around the world. Studies conducted in 2020 investigated the correlation between the weather or climate variables (e.g., temperature, humidity, wind speed, visibility, rain, and solar radiation) and the number of COVID-19 cases. Several of these studies have concentrated on different cities and countries around the world, for example, the United States and Europe (Bukhari & Jameel, 2020; Gharoie Ahangar et al., 2020) , Africa (Adekunle et al., 2020) , Russia (Lasisi and Eluwole, 2021) , Wuhan, China , 122 cities in China (Xie & Zhu, 2020) , India (Awasthi et al., 2020; Borah et al., 2020; Gupta, Pradhan, et al., 2020) , Jakarta, Indonesia (Tosepu et al., 2020) , Pakistan (Aslam et al., 2020) , Brazil (Aule et al., 2020; Pequeno et al., 2020) , Iran (Ahmadi et al., 2020) . Preliminary studies that found no relationship between COVID-19 transmission and atmospheric variables (e.g., temperature) were performed by Xie and Zhu (2020) , Gupta, Raghuwanshi, et al. (2020), and Yao et al. (2020) . Still, some studies found a positive and negative relationship between COVID-19 transmission and atmospheric variables. For example, Zhu, and Liu, Huang, et al. (2020) investigated the meteorological effects on COVID-19 transmission in South America using a multiple linear regression model. The results demonstrated that absolute humidity is negatively correlated with COVID-19 transmission in some regions. Other studies conducted in the USA and other tropical regions by Sobral et al. (2020) , J. , Bashir et al. (2020) , Melo et al. (2020) and Méndez-Arriaga (2020) assessed the nexus analysis between and the spread of COVID-19 climate variables using the regression models. The aforesaid authors indicated that the humidity and temperature often affect the spread of COVID-19. Eslami and Jalili (2020) discussed the role and impact of different environmental factors (e.g., humidity, temperature, wind speed, and solar radiation) in the spread of COVID-19. The results demonstrated that the increased temperature and sunlight could facilitate the destruction of COVID-19 and its stability on surfaces. Another study, conducted in 100 countries by Tzampoglou and Loukidis (2020) , investigated the impact of climatic factors on the intensity of COVID-19 transmission using statistical models for several countries. The results indicate a significant correlation between atmospheric temperature and COVID-19 transmission and death rates. Another study in Bangladesh by Islam, Hasanuzzaman, et al. (2021) discussed the meteorological impacts of the COVID-19 outbreak using a multiple linear regression model, a Monte Carlo method, a compound Poisson generalized linear model, and a random forest model. Results indicated that high humidity and temperature reduce COVID-19 outbreaks. Asyary and Veruswati (2020) demonstrated that exposure to sunlight was significantly correlated with patients' recovery from COVID-19 in Jakarta, Indonesia. Some studies in India by Kumar and Kumar (2020) , Mousavi et al. (2020) , Liu et al. (2020 , and Ladha et al. (2020) investigated the association between climatic conditions and COVID-19 transmission in different states of India (e.g., Mumbai, Delhi, Tamil Nadu, Maharashtra, and Gujarat), using the artificial neural network technique, Spearman's rank correlation model, Mann-Kendall method, Karl Pearson's correlation analysis, and Sen's slope estimator. The results found that relative humidity and pressure parameters are more active than the other parameters on the COVID-19 pandemic. The effects of meteorological indicators on COVID-19 cases and infection rates in Pakistan were investigated by Raza et al. (2020) using generalized Poisson regression. This study implies that COVID-19 cases increased with a temperature rise. Another study in Singapore was carried out by Pani et al. (2020) to investigate the relationship of the COVID-19 outbreak with several meteorological parameters using Spearman's and Kendall's rank correlation models. The results revealed that temperature, humidity, dew point, and water vapor have a positive correlation with the COVID-19 pandemic. Yang et al. (2021) and Pramanik et al. (2020) explored whether the relationship of COVID-19 transmission and meteorological factors with the season and geographical locations using the random forest algorithm, multiple stepwise regression, and Pearson correlations. The results revealed that the lognormal distribution model had the best fit for the changes of COVID-19 infected cases. Fu et al. (2021) studied the influence of governmental responses and meteorological factors on COVID-19 transmission in four European countries (UK, Italy, Germany, and Spain) using the distributed lag nonlinear model. Their results demonstrated that both the dry and cold environments are likely to facilitate COVID-19 transmission. Cacho et al. (2020) and Briz-Redón and Serrano-Aroca (2020a) evaluated the effect of ultraviolet radiation, humidity, and temperature on COVID-19 incidence and the severity across the Spanish cities. Ultraviolet radiation was significantly related to COVID-19 incidence and severity. Neto and Melo (2020) analyzed the correlation between the cases of COVID-19 and population size, and the weather in Brazil's cities. The analysis revealed a significant positive correlation between humidity and instances of COVID-19. A study conducted by Islam, Bukhari, et al. (2021) discussed the relationship between COVID-19 cases and humidity, temperature, and UV index globally. They found a negative correlation for COVID-19 patients with wind speed. Using data from different outbreak countries (e.g., Italy, UK, France, Sweden, Iran, the United States, Korea, and Australia), Rouen et al. (2020) studied the association between the case growth rate COVID-19 and temperature. They found a negative correlation between temperature variation and daily cases. Ş ahin (2020) examined the correlation between weather conditions and COVID-19 in Turkey. The results indicated that temperature and wind speed have a positive relationship with the number of COVID-19 cases. A new challenge in sustainable development in Italy by Pirouz et al. (2020) discussed the effect of urban and climate parameters in the cases of COVID-19 using multivariate linear regression and trend analysis. The analyses indicate that climate factors could affect the daily positive cases of COVID-19. The impact of the seasonal climate on the distribution of the COVID-19 pandemic was investigated by Scafetta (2020) . The weather of winter 2020 in the region of Wuhan (Central China, Hubei) Haque was similar to that of the northern Italian provinces of Milan, Brescia, and Bergamo. Several researchers investigated the impact of atmospheric temperature and humidity on the spread and transmission of COVID-19, for example, Demongeot et al. (2020) , Goswami et al. (2020) Xie and Zhu (2020) , and Yuan et al. (2020) . Their results demonstrated that temperature and humidity were negatively related to cases of COVID-19 and deaths. Bukhari, Massaro, D'agostino, et al. (2020) and Ogaugwu et al. (2020) explored the influence of weather on the spread of COVID-19. Results indicated a weak negative association between the spread of COVID-19 and temperature. The correlation between meteorological parameters and the outbreak of COVID-19 in the top 20 countries was tested by Sarkodie and Owusu (2020) using novel panel estimation techniques. The top 20 countries namely the US, Spain, Italy, France, United Kingdom, Germany, Turkey, Russia, Iran, China, Brazil, Canada, Belgium, Netherlands, India, Switzerland, Peru, Portugal, Ecuador, and Saudi Arabia. They demonstrated that rising temperatures and relative humidity reduce the survival, stability, viability, and transmission of COVID-19. In contrast, another study by S. in China, Hong Kong, and Singapore found that high temperatures reduce COVID-19 transmission, and high relative humidity increases disease transmission. Briz-Redón and Serrano-Aroca (2020) investigated the effect of temperature on the number of COVID-19 cases in Spain's provinces. They found no consistent evidence of a correlation between the number of COVID-19 cases and temperature at the provincial level in Spain. Another relevant study published by Prata et al. (2020) reported that temperature has a negative linear relationship with the number of confirmed COVID-19 cases. However, Menebo (2020) demonstrated that temperature is positively correlated with SARS-CoV-2 infection, whereas precipitation correlates negatively. Zhu, Liu, Huang, et al. (2020) also investigated the relationship between daily temperature and the number of confirmed SARS-CoV-2 infections in 122 cities in China. They found that daily temperature has a positive linear relationship with the number of COVID-19 cases, with a threshold of 3°C. Studies by Ma et al. (2020) and Wu et al. (2020) examined the impacts of temperature, diurnal temperature range, and humidity on the mortality of COVID-19 in the Chinese population. They concluded that the mortality of COVID-19 correlates positively with the diurnal temperature range but negatively with humidity. In addition, they demonstrated that the COVID-19 pandemic might be partially suppressed by an increase in the daily temperature and relative humidity. The effects of meteorological conditions, such as temperature, humidity, and rainfall, on the spread of SARS-CoV-2 in a tropical climate were modeled by Auler et al. (2020) . They revealed that higher temperatures and humidity favor the transmission of SARS-CoV-2, which differs from the results reported in colder countries. Furthermore, the impacts of atmospheric thermal stability, based on wind speed and air pollution, on the spread of the COVID-19 pandemic were investigated by Coccia (2020) using statistical evidence that supports the hypothesis. Results indicated that high atmospheric stability, with low wind speed, may accelerate the diffusion of SARS-CoV-2 in the atmosphere. A new study by Bhaganagar and Bhimireddy (2020) analyzed the local atmospheric parameters (e.g., atmospheric stability, wind shear, wind speed and direction, air turbulence, air temperature, and moisture content) that enhanced air-borne dispersion of coronavirus during March 9-April 6, 2020, in New York City, USA, using Weather-Research-Forecast model coupled with the Lagrangian Hybrid Single-Particle Lagrangian Integrated Trajectory (WRF-HYSPLIT) model. Their study concluded that atmospheric stability regimes, the transport direction, and the extent of blob spread correlate with the wind shear and temperature gradient, and also with the scales of the horizontal rolls and thermal updrafts and downdrafts. In hot, arid climates, some studies performed in Iran, Algeria, Pakistan, Mexico, Kuwait, United Arab Emirates, Oman, Ghana, India, Saudia Arabia, and Iraqi (Kurdistan region) by Meo et al. (2020) , Sasikumar et al. (2020) , Alkhowailed et al. (2020) , and Amin and Amin (2020) examined the impact of the climate variables on the COVID-19 outbreak. They demonstrated that positive COVID-19 cases increase as humidity, temperature, and wind speed decrease. Another study on a global scale by explored the connection between regional climate parameters and the spread of COVID-19. The results indicated that countries in the colder temperature zone display a faster rise in COVID-19 cases than countries in warmer temperature zones. In contrast, a review of the effect of environmental factors on the spread of COVID-19 was summarized by Azuma et al. (2020) , Briz-Redón and Serrano-Aroca (2020b), Kumar et al. (2021) , Mecenas et al. (2020) , McClymont and Hu (2021) , and Paraskevis et al. (2020) . Some of these studies demonstrated a negative correlation between COVID-19 cases with temperature and a positive correlation with humidity. Another study between Switzerland and the United Arab Emirates by Daneshvar et al. (2021) indicated distinct differences or similarities in the climate impacts on COVID-19 transmission using correlation modeling. The results demonstrated that the impacts of climate on COVID-19 transmission play different roles in the various countries based on climatic regions. Considering the studies above, conducted by researchers in different cities and countries worldwide using various statistical and modeling techniques, it is clear that much debate remains about the association between the spread and transmission of the COVID-19 pandemic and atmospheric variables. Some studies found no association of the COVID-19 outbreak with atmospheric variables, whereas other studies demonstrated strong and weak, positive and negative associations between the spread COVID-19 pandemic and atmospheric variables. Therefore, the association of the COVID-19 outbreak with atmospheric variables is inconsistent and needs more study and time to better understand and evaluate it. The WHO reported that, currently, there is no conclusive evidence that either meteorological or climate conditions have a strong effect on the transmission of SARS-CoV-2 (World Health Organization, 2020). Thus, proving a relationship between atmospheric variables and COVID-19 outbreaks could be helpful in predicting the pattern of infection over the coming months. However, various parameters may influence the progression of the COVID-19 pandemic. The main objective of this study is, therefore, to investigate the relationship between atmospheric variables (temperature, relative humidity, wind speed, visibility, and solar radiation) and the outbreak of COVID-19 in the State of Kuwait, with a hot, arid climate, using stochastic models. The results will have implications for policy-makers and the public. In this work, stochastic models with a probability density function (PDF) and cumulative density function (CDF) for normal and lognormal distributions were used to model the atmospheric variables and COVID-19 cases. Spearman's rank correlation and the multiple regression model were used to investigate the correlation between atmospheric variables and the transmission of SARS-CoV-2. The State of Kuwait is situated in the area between 29°30′N and 30°05′N and 46°03′E and 48°35′E in the northeast of the Arabian Peninsula and covers approximately 17,818 km 2 (Al- Rashidi et al., 2018; Yassin et al., 2018b,c) . The climate of Kuwait is characterized as arid with extreme heat and large quantities of airborne dust, low humidity, and little precipitation (Al-Dabbous and Kumar, 2014; Al-Dousari et al., 2019) . Kuwait has two primary seasons, namely, summer (during which the temperature ranges between 48 and 54°C), extending from May to September, and winter, which extends from December to February. In addition, there are two transient seasonsspring, which occurs during March and April, and autumn, which occurs during October and November (Al-Dousari et al., 2020; Yassin et al., 2018a; Yassin et al., 2021) . The present study gathered data during spring and summer. Daily data concerning COVID-19 cases and atmospheric variables were collected from February 24 to May 30, 2020, during periods of free movement, partial lockdown, and full lockdown. The period of free movement, during which no restrictions were applied to movement, fell between February 24 and March 21, 2020. Partial lockdown, in which movement was partially restricted, was implemented between March 22 and May 9, 2020, and the full lockdown was enforced between May 10 and May 31, 2020. The daily data for COVID-19 cases in Kuwait were extracted and collected from daily situation reports published by the Ministry of Health for the State of Kuwait (https://corona.e.gov.kw/en). Data describing the daily atmospheric variables were obtained from weather stations operated by the Directorate General of Civil Aviation (DGCA), Kuwait. Data describing the daily atmospheric variables were collected from weather stations operated by the DGCA, Kuwait. Variables included daily temperature, relative humidity, wind speed, visibility, and solar radiation from February to May 2020. These data are maintained by the meteorology department's quality control and assured process under the supervision of DGCA. The data processing system uses a nationwide network of fully automatic data collection units (DCU). Each DCU has a meteorological, agro-meteorological, and oceanographic sensor suite tailored to its specific application. The DCU units read the sensors in real time and partially process the raw values into meaningful, quality-checked, 1-min values. These values are then communicated back to a central data acquisition system, where the values are further processed into hourly, daily, and monthly statistics. These processed values are displayed on a graphical user interface at the meteorological and display working places. They are also used to generate standard World Meteorological Organization messages, such as SYNOP, SHIP, BUOY, and CLIMAT. The quality monitoring system (QMS) tests, analyses, edits, and monitors these data in the climate database management system (CDMS). CDMS is a relational database that stores meteorological and related data necessary to meet climate data's present and future needs. The data stored in CDMS form the basis of the climate record. It should be considered that there is a period of latency between the day on which a person becomes infected and the day on which the infection is confirmed; the average duration of this period is 3 days. This 3-day delay was therefore taken into account, meaning that the atmospheric data used described the conditions 3 days before infection was confirmed. A stochastic model is a tool used to assess the probability distribution of potential outcomes by allowing for random variation in one or more inputs over time. The intensity of COVID-19 cases varied randomly because of changes in the atmospheric conditions. Either the PDF or the CDF was applied as a continuous random variable in the stochastic models, with various applications. As a continuous random variable, the PDF is the density of the probability that a variable has value x. Because the probability of continuous distribution at a single point is zero, this variable is often expressed in terms of integrating two points. x is a random variable if the PDF is used, where f(x) is −∞ ≤ x ≤ ∞ such that the following is true (Al-Harbi et al., 2012; Yassin & Elmi, 2011) : The equation describing a continuous random variable X, which in this case is the CDF or F(a), is (Al-Harbi et al., 2012; Yassin & Elmi, 2011) : The relationship between the PDF and CDF for a continuous random variable was thus determined. The PDF (f(x)) of a continuous random variable x is defined as the derivative of the CDF F(x): It is sometimes useful to consider the CDF F(x) in terms of PDF f(x): The PDF for a normal distribution is: The CDF for a normal distribution is: The lognormal distribution of the PDF is: The lognormal distribution of the CDF is: where μ is the mean of the random variables and σ is the standard deviation. The random variables (i.e., the intensity of SARS-CoV-2 infection and atmospheric variables = √(x-mean(x)) were normalized by their corresponding mean (x/µ), where x and µ represent the intensity and mean of the data, respectively. The variable x/µ was included in the stochastic model. The infection rate (IR), growth rate (GR), and doubling time (DT) were used to analyze the number of COVID-19 cases during the free movement, partial lockdown, and full lockdown periods. The IR indicates the speed of the spread and transmission of COVID-19. The GR indicates the infection cases during the three periods. In contrast, the DT indicates the time taken for a count to double the infection cases for every period. The IR is defined in Equation (12) The GR and DT are defined in Equations (13 and 14) (Gupta, Pradhan, et al., 2020; Patel & Patel, 2020) : where Nf/Ne is the number of infection cases on the first and last day, and n is the number of infection days. Descriptive statistical analyses of the number of COVID-19 cases, IR, and atmospheric variables were performed using minimum, maximum, mean, standard deviation, skewness, and kurtosis distributions. As the data used in this study were not normally distributed, Spearman's rank correlation (r s ) was utilized to examine the correlation between the atmospheric variables and COVID-19 cases. The formula describing the correlation coefficient is defined in Equation (15): where n is the number of days and d i is a value representing the difference in the number of COVID-19 cases and the atmospheric variables over the period studied. In addition, a multiple regression model was used to evaluate the relationship between the atmospheric parameters and COVID-19 cases, with the atmospheric parameters as independent variables and COVID-19 cases as the dependent variable. Where multiple regression models are appropriate for the different scenarios in which multiple variables influence a single result. COVID-19 cases (Ŷ i ) were modeled as a function of five atmospheric variables (X i ): temperature (x 1 ), relative humidity (x 2 ), wind speed (x 3 ), visibility (x 4 ), and solar radiation (x 5 ) with their corresponding coefficients (β i ). The multiple regression model equation can be expressed as: A two-tailed p-value of <0.10 was considered statistically significant for all statistical analyses. The statistical analyses of daily SARS-CoV-2 infections and atmospheric variables were carried out using MATLAB version 2020b. The predictive performances of the stochastic models described in the previous section: "Stochastic Models" were first estimated. To investigate how infection with SARS-CoV-2 may change with differing atmospheric variables, descriptive statistics, stochastic analyses, and correlation analyses were applied to evaluate the relationship between infection and temperature, humidity, wind speed, visibility, and solar radiation during the three periods of free movement (February 24-March 21, 2020), partial lockdown (March 22-May 9, 2020), and full lockdown (May 10-30, 2020). Table 1 details the number of COVID-19 cases, the number of deaths, and the GR and DT during the periods of free movement, partial lockdown, and full lockdown, provided by the Ministry of Health for the State of Kuwait. In all, SARS-CoV-2 infections numbered 176, 8512, and 17,504 during free movement, partial lockdown, and full lockdown, respectively, and the number of deaths was 0, 58, and 147, respectively. In contrast, the GR was 4.4, 6.5, and 0.26, with the DT at 15. 26, 9.59, and 264.62, respectively. This table clearly shows that the GR of infection was 4.4 during the free movement period, but it moves up during the partial lockdown period and then drops during the full lockdown period. Simultaneously, the DT of infection was 15.29 during the free movement period, but it dropped during the partial lockdown period and quickly moved up during the full lockdown period. The fewest COVID-19 cases were reported during the period in which free movement was still allowed, estimated at 176.00. The fewest deaths were also recorded in this period, with a value of 0.00. The lowest GR of infection occurred when the full lockdown was a full restriction, estimated at 0.26. The lowest DT of infection was during the partial lockdown period, with a value of 9.59. Meanwhile, the most cases, the most deaths, the highest GR, and the highest DT of infection of 17,504.00, 147.00, 6.5, and 264.62, respectively, occurred during the full lockdown, except the GR in partial lockdown. Therefore, it is essential to investigate the factors that most influenced SARS-CoV-2 infection in Kuwait during free movement, partial lockdown, and full lockdown periods. Table 2 provides statistics describing the number of SARS-CoV-2 infections occurring daily, atmospheric variables, and IRs over the three periods under investigation. The relationship between atmospheric variables and COVID-19 cases was statistically evaluated for the three different periods to establish if any correlation between the different variables could be found. The relationships between each atmospheric variable and SARS-CoV-2 infection are as follows. The descriptive analysis listed in Table 2 demonstrates mean temperatures of 19.83, 25.73, and 32.93°C, respectively, during the periods of free movement, partial lockdown, and full lockdown. The corresponding mean number of infections with SARS-CoV-2 was 6.5, 152.0, and 884.0, respectively. These results suggest that the temperature was correlated with the number of infections occurring during the three periods studied. The number of infection cases and rates increased when the temperature increased. Accordingly, it is possible Integr Environ Assess Manag 2021:1-17 © 2021 SETAC wileyonlinelibrary.com/journal/ieam that temperature has a significant impact on SARS-CoV-2 infection. Therefore, air temperature is an environmental factor that affects the spread and transmission of the COVID-19 outbreak. The skewness (asymmetry of the probability distribution) and kurtosis (tailedness of the probability distribution) also support this positive relationship. Excess kurtosis was calculated for normal distribution, which is equal to zero. The skewness and kurtosis that were related to temperature decreased during the periods of free movement and partial lockdown, whereas the skewness and kurtosis associated with infection increased. However, the opposite was observed for full lockdown, during which the skewness and kurtosis associated with temperature rose, and the skewness and kurtosis of infection decreased. The descriptive analysis in Table 2 indicates the presence of a relationship between relative humidity and SARS-CoV-2 infection. The relative humidity during the periods of free movement, partial lockdown, and full lockdown was 51.28%, 41.30%, and 21.81%, respectively, indicating that a decrease in the relative humidity was related to an increase in infection cases and rates. There appears to be an inverse relationship between relative humidity with COVID-19 cases and infected rates during the three periods. Furthermore, as the skewness and kurtosis of relative humidity increased during the three periods studied, the skewness and kurtosis of COVID-19 cases increased during the periods of free movement and partial lockdown, whereas it decreased during full lockdown. The descriptive statistics in Table 2 suggest a relationship between solar radiation and SARS-CoV-2 infection. The amount of solar radiation attainment on the surface during free movement, partial lockdown, and full lockdown was 1230.10, 1437.00, and 1648.10 J/cm 2 , respectively. This indicates the possibility of a direct positive relationship between solar radiation and infection cases and rates during the periods of study. Thus, solar radiation had a significant effect on SARS-CoV-2 infection. Moreover, the skewness and kurtosis of solar radiation decreased during the periods of free movement and partial lockdown, whereas the parameters relating to the number of COVID-19 cases increased. The skewness and kurtosis of solar radiation increased during full lockdown, whereas the skewness and kurtosis of infection decreased. According to the descriptive statistics results in Table 2 , there was no direct relationship between COVID-19 cases and variations in wind speed or visibility. The average wind speed during the three periods was 3.94, 4.06, and 4.05 m/s, respectively, and the visibility was 3244.4, 4605.7, and 4478.6 m, respectively. This indicates that there was an increase in the wind speed and visibility during the periods of free movement and partial lockdown alongside the growth in the number of COVID-19 cases, whereas both factors decreased during full lockdown. These contradictions indicate that there is no consistent association between SARS-CoV-2 infection and wind speed or visibility. The above descriptive analyses revealed the possibility of a relationship between COVID-19 outbreaks and atmospheric variables. The IR for SARS-CoV-2 is directly proportional to the air temperature, wind speed, and visibility but inversely related to the humidity. The importance of each atmospheric variable during each period was therefore further investigated, and the results are discussed in Section: Stochastic Model Analysis. The intensity of COVID-19 cases and the atmospheric variables were normalized by the mean (x/mu) in this study. This was applied to clarify the variation in the PDF and CDF shapes that relate the number of infections to changing atmospheric variables. Figures 1-3 show the normal and lognormal PDF and CDF of SARS-CoV-2 infection and atmospheric variables during the three periods from February 24 to May 30, 2020. The characteristics of the normal and lognormal PDF and CDF for SARS-CoV-2 infection and atmospheric variables during the periods of free movement, partial lockdown, and full lockdown are as follows. (a) Free movement (without restriction) Figure 1 shows the normal and lognormal probability density functions (PDFs) and cumulative density functions (CDFs) of both SARS-CoV-2 infection and atmospheric variables during the period in which free movement was allowed. During this period, the normal PDFs of temperature and relative humidity exhibited symmetrical, sharp, and narrow curves, whereas the normal PDFs of infections, wind speed, visibility, and solar radiation exhibited asymmetrical, wide curves. The highest value of the normal PDF was 0.75 for temperature, and the lowest was 0.5, also for temperature. However, the highest and lowest values of the lognormal PDF were associated with temperature and relative humidity, respectively. Figure 1 shows that the normal and lognormal PDFs describing the number of COVID-19 cases were very similar to those of temperature and relative humidity in terms of shape and value. These similarities suggest that temperature and relative humidity had a strong relationship with the SARS-CoV-2 infection during the period of free movement. Figure 1 indicates that the normal PDF and CDF of COVID-19 cases were considerably different from those associated with wind speed, visibility, and solar radiation. This means that wind speed, visibility, and solar radiation had a weak relationship with infection during the period of free movement. One feature to note is that the value and shape of the normal and lognormal CDFs of COVID-19 cases were also similar to those of temperature and relative humidity, which is consistent with previous observations ( Figure 1) . Moreover, the normal CDFs of infections, temperature, and relative humidity (Figure 1 ) appear to be higher than those of wind speed, visibility, and solar radiation at 0.5 < CDF < 1.0. Conversely, the lognormal CDFs of infection, temperature, and relative Integr Environ Assess Manag 2021:1-17 © 2021 SETAC DOI: 10.1002/ieam.4481 FIGURE 2 Normal and lognormal PDFs and CDFs of of SARS-CoV-2 infection and atmospheric variables during the partial lockdown period. CDFs, cumulative density functions; PDFs, probability density functions humidity were lower than those of wind speed, visibility, and solar radiation at 0.0 < CDF < 0.5. Clearly, the CDF of visibility is lower than that of the other variables as 0 0.1); thus, no significant correlation is assumed between these three parameters and the number of COVID-19 cases during the period of free movement. The mathematical model of Ŷ for partial lockdown is: The regression analysis for this period reveals that F(5,43) is equal to 11.89 with a p-value <0.01. Meanwhile, the R 2 of the model is equal to 0.76 with an adjusted R 2 value of 0.53 and a standard error of 102.307. SARS-CoV-2 infection was significantly correlated with temperature and relative humidity during partial lockdown (p < 0.1). The temperature coefficient was greater than zero, whereas that of relative humidity was smaller than zero; thus, SARS-CoV-2 infection was positively correlated with temperature and negatively correlated with relative humidity during this period. The results in the present study are consistent with the results of the correlation analysis. The mathematical model of Ŷ during full lockdown is: Statistical analysis of the regression during full lockdown reveals that F(5,20) was equal to 2.628 with a p-value <0.10, but the R 2 of the model was equal to 0.68 with an adjusted R 2 value of 0.289 and a standard error of 128.849. During full lockdown, SARS-CoV-2 infection was significantly correlated with relative humidity, wind speed, and visibility (p < 0.1). In this period, SARS-CoV-2 infection was negatively correlated with relative humidity and positively correlated with wind speed and solar radiation during this period. The results are congruent with the results of the correlation analysis. The multiple regression model results indicated that the effect of atmospheric variables on COVID-19 cases differs in the free movement, partial lockdown, and full lockdown periods. The temperature was the most affecting factor in COVID-19 transmission in partial lockdown. Its contribution in infection cases was more significant than the free movement and full lockdown periods. This may be caused by the seasonal change from winter to spring. The humidity significantly affected the infection cases in the free movement period compared with the other periods. Its contribution was more significant than during the partial and full lockdown periods. This indicates that the decrease in humidity in the hot, arid region might boost COVID-19 transmission. The infection cases' wind speed significantly positively affected the infection cases in the full lockdown period, with a negative effect in the other periods. This may be the result of the full lockdown period being entirely restricted for people and vehicle movement and frequent air convection in spring. The feature of wind speed increases the diffusion rate and spread distance of SARS-CoV-2. In addition, higher wind speed is also helpful for the generation of turbulence within the urban area, which results in the accumulation of novel coronavirus near the surface ground, thus increasing the risk of infection (Yang et al., 2021) . The visibility was negatively affected by the COVID-19 infection in the free movement period and positively affected in the other periods because there is no dust during these periods. In contrast, the solar radiation effects were weak during all periods because the sun is mild in these periods. Compared with these periods, temperature and humidity were most directly related to COVID-19 transmission than other atmospheric variables, but their influence varied during the periods. Based on the multiple regression model analysis, ANOVA revealed the significance to be less than 0.10 for the equations of these periods. The R 2 increased in the partial lockdown and full lockdown periods. The R 2 ranged from 0.59 to 0.68, respectively. P-values ranged from 0.01 to 0.10, respectively. This indicates that the ranged from 56% to 76% approximately of the variations in COVID-19 infections can be affected by the atmospheric variables. The present study used data describing the conditions in the State of Kuwait during summer and spring 2020, which are characterized by hot and arid weather. The trend of changes in the atmospheric variables provides a better picture of the COVID-19 outbreak in Kuwait. Analysis of the findings in previous sections (Descriptive Statistics Analysis, Stochastic Model Analysis, Correlation Statistical Analysis, Multiple Regression Model Analysis) indicates a significant association between the number of COVID-19 cases and temperature, relative humidity, wind speed, visibility, and solar radiation during the three periods of free movement, partial lockdown, and full lockdown. The findings obtained were compared with those of previous and recent studies that have reported that the coronavirus outbreak is significantly associated with temperature and are consistent with the results of these studies. For example, both Sundell et al. (2016) and Park et al. (2020) reported that influenza is correlated with temperature. Chan et al. (2011) suggested that the coronavirus is correlated with high temperatures, and Casanova et al. (2010) found an association with low temperatures. Recently, Wu et al. (2020) and Sarkodie and Owusu (2020) indicated that the COVID-19 pandemic was suppressed as the temperature increased. Meanwhile, other studies conducted by Menebo (2020), Shahzad et al. (2020) , Tosepu et al. (2020) , Bashir et al. (2020) , Zhu, Liu, Huang, et al. (2020) , and Şahin (2020) clearly demonstrated that temperature is positively correlated with outbreaks of COVID-19. The results of both previous and recent studies investigating the link between the coronavirus and temperature are consistent with the present study's findings, which demonstrated a positive relationship during the periods of free movement and partial lockdown. In contrast, Prata et al. (2020) , Bukhari, Massaro, D'agostino, et al. (2020) , Ogaugwu et al. (2020) , Islam, Bukhari, et al. (2021) , Rouen et al. (2020) , and S. found that temperature is negatively correlated with SARS-CoV-2 infection. Relative humidity was also found to be influential in the increasing number of COVID-19 cases during the periods of free movement, partial lockdown, and full lockdown. Studies conducted by Barreca and Shimshack (2012) , Soebiyanto et al. (2015) , and Tamerius et al. (2013) revealed that influenza is negatively correlated with relative humidity. However, Urashima et al. (2003) indicated that influenza is positively correlated with relative humidity. Recently, Wu et al. (2020) , Ş ahin (2020), Sarkodie and Owusu (2020) , Meo et al. (2020) , Sasikumar et al. (2020) , Alkhowailed et al. (2020) , Amin et al. (2020) , Zhu, Liu, Huang, et al. (2020) , Islam, Bukhari, et al. (2021), and Ahmdi et al. (2020) indicated that the COVID-19 pandemic is inversely associated with relative humidity. This study indicates that SARS-CoV-2 infection was positively correlated with relative humidity during the period of free movement and negatively correlated with relative humidity during the partial and full lockdown. This result in the present study is congruent with other studies by Kumar and Kumar (2020) , Mousavi et al. (2020) , Liu et al. (2020 ), Sing et al. (2020 , Neto and Melo (2020) , and Ladha et al. (2020) . Moreover, Ş ahin (2020) reported that wind speed is positively correlated with the number of COVID-19 cases, which was consistent with our results for full lockdown. However, Bhaganagar and Bhimireddy (2020) and Coccia (2020) recently revealed that stable atmospheric conditions result in low wind speeds, low-level turbulence, and cool, moist ground conditions that favor the transmission of the SARS-CoV-2 infection. Their results are consistent with the present study's findings, which demonstrated a negative relationship during free movement and partial lockdown periods. In addition, the study by Bhaganagar and Bhimireddy (2020) revealed a significant contribution of the local meteorological conditions in spreading and transmitting the virus infection. The virus can spread and transfer e via 30 min in the atmosphere with a radius of 0.2 km and move a distance of 1.0-2.0 km from the source. Solar radiation was also found to be positively associated with SARS-CoV-2 infection during the full lockdown, which is consistent with the recent findings of Asyary and Veruswati (2020). To our knowledge, this study is one of few that have investigated the association between the daily number of confirmed COVID-19 cases and temperature, relative humidity, wind speed, visibility, and solar radiation in a hot, arid climate. Comprehensive analysis of the data was carried out with stochastic models, Spearman's correlation, and a multiple regression model. The significant findings of this study indicate a strong relationship between relative humidity and SARS-CoV-2 infection. The number of COVID-19 cases rose as the temperature, wind speed, and visibility increased and decreased as the relative humidity increased. The IR for SARS-CoV-2 is directly proportional to the air temperature, wind speed, and visibility, but it is inversely related to the humidity. The lowest GR and longest DT of the COVID-19 infection were observed in the full lockdown period. The main results for the three periods studied are summarized as follows. (a) During the period in which free movement was allowed, SARS-CoV-2 infection exhibited a weakly significant positive correlation with temperature and a significant positive correlation with relative humidity. (b) During the partial lockdown, a strongly significant positive correlation was found between temperature and SARS-CoV-2 infection, whereas a significant negative correlation was observed between SARS-CoV-2 infection and relative humidity. There was also a weak significant positive correlation between COVID-19 cases and wind speed, and a weak significant positive correlation between SARS-CoV-2 infection and solar radiation. A negative correlation was observed between infection with SARS-CoV-2 and visibility. (c) During full lockdown, there was no significant correlation between COVID-19 cases and temperature. Moreover, a significant negative correlation was obtained between SARS-CoV-2 infection and relative humidity, and a positive correlation was found between SARS-CoV-2 infection and visibility. Spatiotemporal analysis of meteorological factors in abating the spread of COVID-19 in Investigation of effective climatology parameters on COVID-19 outbreak in Iran Number and size distribution of airborne nanoparticles during summertime in Kuwait: First observations from the Middle East Cost and effect of native vegetation change on aeolian sand, dust, microclimate and sustainable energy in Kuwait Environmental and economic importance of native plants and green belts in controlling mobile sand and dust hazards Stochastic modeling of the impact of meteorological conditions on road traffic accidents Impact of meteorological parameters on COVID-19 pandemic: A comprehensive study from Saudi Arabia Gaseous air pollution background estimation in urban, suburban, and rural environments Climate analysis to predict potential spread and seasonality for global (COVID-19) in Iraqi Kurdistan region A correlation study between weather and atmosphere with COVID-19 pandemic in Islamabad Sunlight exposure increased Covid-19 recovery rates: A study in the central pandemic area of Indonesia Evidence that high temperatures and intermediate relative humidity might favor the spread of COVID-19 in tropical climate: A case study for the most affected Brazilian cities Statistical interpretation of environmental influencing parameters on COVID-19 during the lockdown in Delhi Environmental factors involved in SARS-CoV-2 transmission: Effect and role of indoor environmental quality in the strategy for COVID-19 infection control Absolute humidity, temperature, and influenza mortality: 30 years of county-level evidence from the United States Correlation between climate indicators and COVID-19 pandemic Local atmospheric factors that enhance air-borne dispersion of coronavirus-high-fidelity numerical simulation of COVID19 case study in real-time Short-term passive tracer plume dispersion in convective boundary layer using a high-resolution WRF-ARW model Examining the correlation between the weather conditions and COVID-19 pandemic in India: A mathematical evidence A spatio-temporal analysis for exploring the effect of temperature on COVID-19 early evolution in Spain The effect of climate on the spread of the COVID-19 pandemic: A review of findings, and statistical and modelling techniques Will coronavirus pandemic diminish by summer? Available at SSRN 3556998 Effects of weather on coronavirus pandemic Can climatic factors explain the differences in COVID-19 incidence and severity across the Spanish regions?: An ecological study. Environmental Health Effects of air temperature and relative humidity on coronavirus survival on surfaces The effects of temperature and relative humidity on the viability of the SARS coronavirus The effects of atmospheric stability with low wind speed and of air pollution on the accelerated transmission dynamics of COVID-19 Climate effects on the COVID-19 outbreak: A comparative analysis between the UAE and Switzerland Temperature decreases spread parameters of the new COVID-19 case dynamics The role of environmental factors to transmission of SARS-CoV-2 (COVID-19) Estimation and demographic analysis of COVID-19 infections with respect to weather factors in Europe Projections for COVID-19 pandemic in India and effect of temperature and humidity Estimating the impact of daily weather on the temporal pattern of COVID-19 outbreak in India Effect of weather on COVID-19 spread in the US: A prediction model for India in 2020 Association between temperature, humidity, and COVID-19 outbreaks in Bangladesh Are meteorological factors enhancing COVID-19 transmission in Bangladesh? Novel findings from a compound Poisson generalized linear modeling approach COVID-19 and climatic factors: A global analysis Decline in temperature and humidity increases the occurrence of influenza in cold climate A correlation study between meteorological parameters and COVID-19 pandemic in Mumbai Current understanding of the influence of environmental factors on SARS-CoV-2 transmission, persistence, and infectivity Association of environmental parameters with COVID-19 in Delhi Is the weather-induced COVID-19 spread hypothesis a myth or reality? Evidence from the Russian Federation Containing the spread of coronavirus disease 2019 (COVID-19): Meteorological factors and control strategies Environmental factors on the SARS epidemic: Air temperature, passage of time and multiplicative effect of hospital infection Discovering Correlations between the COVID-19 Epidemic Spread and Climate Impact of meteorological factors on the COVID-19 transmission: A multi-city study in China Roles of humidity and temperature in shaping influenza seasonality Effects of temperature variation and humidity on the death of COVID-19 in Wuhan Weather variability and COVID-19 transmission: A review of recent research Effects of temperature and humidity on the spread of COVID-19: A systematic review Temperature and precipitation associate with Covid-19 new daily cases: A correlation study between weather and Covid-19 pandemic in Oslo Effect of temperature and humidity on the dynamics of daily new cases and deaths due to COVID-19 outbreak in Gulf countries in Middle East Region Evidence that higher temperatures are associated with a marginally lower incidence of COVID-19 cases COVID-19 time series forecast using transmission rate and meteorological parameters as features The temperature and regional climate effects on communitarian COVID-19 contagion in Mexico throughout phase 1. Science of the Total Environment Correlation between weather, population size and COVID-19 pandemic: A study of Brazilian capitals Effect of weather on COVID-19 transmission and mortality in Lagos Association of COVID-19 pandemic with meteorological parameters over Singapore A review of the impact of weather and climate variables to COVID-19: In the absence of public health measures high temperatures cannot probably mitigate outbreaks Effects of temperature, humidity, and diurnal temperature range on influenza incidence in a temperate region Doubling time and its interpretation for COVID 19 cases. Statistical Update pISSN Air transportation, population density and temperature predict the spread of COVID-19 in Brazil Development of an assessment method for investigating the impact of climate and urban parameters in confirmed cases of COVID-19: A new challenge in sustainable development Climatic factors influence the spread of COVID-19 in Russia Temperature significantly changes COVID-19 transmission in (sub) tropical cities of Brazil Influence of absolute humidity, temperature and population density on COVID-19 spread and decay durations: Multi-prefecture study in Japan Association between meteorological indicators and COVID-19 pandemic in Pakistan COVID-19: relationship between atmospheric temperature and daily new cases growth rate Short-term effects of specific humidity and temperature on COVID-19 morbidity in select US cities Impact of weather on COVID-19 pandemic in Turkey. Science of the Total Environment Temperature, humidity, and latitude analysis to estimate potential spread and seasonality of coronavirus disease 2019 (COVID-19) Impact of meteorological factors on COVID-19 pandemic: Evidence from top 20 countries with confirmed cases Impact of extreme hot climate on COVID-19 outbreak in India Distribution of the SARS-CoV-2 pandemic and its monthly forecast based on seasonal climate patterns Asymmetric nexus between temperature and COVID-19 in the top ten affected provinces of China: A current application of quantile-on-quantile approach The impact of temperature and absolute humidity on the coronavirus disease 2019 (COVID-19) outbreak-evidence from China A gradient boosting machine learning approach in modeling the impact of temperature and humidity on the transmission rate of COVID-19 in India Correlation between temperature and COVID-19 (suspected, confirmed and death) cases based on machine learning analysis Association between climatic variables and COVID-19 pandemic in National Capital Territory of Delhi Association between climate variables and global transmission of SARS-CoV-2 ) and Israeli districts A four-year seasonal survey of the relationship between outdoor climate and epidemiology of viral respiratory tract infections in a temperate climate Environmental predictors of seasonal influenza epidemics across temperate and tropical climates An initial investigation of the association between the SARS outbreak and weather: with the view of the environmental temperature and its variation Correlation of ambient temperature and COVID-19 incidence in Canada Correlation between weather and Covid-19 pandemic in Jakarta Investigation of the Importance of Climatic Factors in COVID-19 Worldwide Intensity Effect of temperature on the infectivity of COVID-19 A seasonal model to simulate influenza oscillation in Tokyo Stability of Middle East respiratory syndrome coronavirus (MERS-CoV) under different environmental conditions Eurosurveillance Humidity is a consistent climatic factor contributing to SARS-CoV-2 transmission Q&A: Climate change and COVID-19 Effects of temperature and humidity on the daily new cases and new deaths of COVID-19 in 166 countries Association between ambient temperature and COVID-19 infection in 122 cities from China Influence of Meteorological Factors on the COVID-19 Transmission with Season and Geographic Location No association of COVID-19 transmission with temperature or UV radiation in Chinese cities GIS-based geostatistical approaches study on spatial-temporal distribution of ozone and its sources in hot, arid climates Dust storms backward Trajectories' and source identification over Kuwait Assessment of the atmospheric mixing layer height and its effects on pollutant dispersion Stochastic analysis of concentration field in a wake region Ozone background levels and trends in the state of Kuwait A climatologic investigation of the SARS-CoV outbreak in Beijing Do humidity and temperature impact the spread of the novel coronavirus? Frontiers in Public Health Meteorological impact on the COVID-19 pandemic: A study across eight severely affected regions in South America The authors declare no conflict of interest. Mohamed F. Yassin: data collection, data curation; author of original paper, Methodology, Results, and Discussion.Hassan A. Aldashti: data collection; author of Review, Results, and Discussion. The supporting information contains five figures showing the relationship between the infection cases for SARS-CoV-2 and atmospheric variables. Data, associated metadata, and calculation tools are available from the corresponding author (mohamed_f_ yassin@hotmail.com). Mohamed F. Yassin https://orcid.org/0000-0001-5059-2954