Atmospheric pressure and population density as super-factors inuencing the transmission of coronavirus disease 2019 (COVID-19)

The novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the etiologic agent of coronavirus disease (COVID-19), recently emerged and led to a global pandemic with enormous consequent losses to global health and economies. To date, more than 30 million cases have been reported globally and have affected almost every with varying degrees. Meteorological and non-meteorological factors such as temperature, relative humidity, atmospheric pressure, population density, and latitude, are considered critical in virus transmission. To explore the correlation of environmental factors with the transmission of SARS-CoV-2 based on parameters including infection rate, effective reproduction number, and compound growth rate, we analyzed data of conrmed cases from 487 counties in the United States. We found a small impact of temperature and humidity on virus transmission, but observed a considerable positive inuence of atmospheric pressure and population density on virus transmission. Geographic areas and seasons (autumn and winter), with exposure to higher atmospheric pressure, are more likely at higher risk of an outbreak. Social distancing and other measures could be effective strategies to combat COVID-19 outbreaks in densely populated areas. Additional studies are needed to explore the mechanisms underlying the relationship between meteorological parameters and transmission of SARS-CoV-2. various and critical in the spread of different viruses, including inuenza virus , respiratory syncytial severe acute East mechanisms involved in


Introduction
The outbreak of coronavirus disease 2019 , caused by the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), was declared a global pandemic by the World Health Organization on 11 March 2020 1 . And this pandemic has led to a global pandemic with enormous consequent losses to global health and economies 2 . By September 29, 2020, there were 33,417,386 con rmed cases of COVID-19, including 1,002,864 deaths, worldwide. The United States is the most affected country, with 7,150,824 con rmed cases and 205,107 deaths (Johns Hopkins Coronavirus Resource Center, https://coronavirus.jhu.edu). COVID-19 has touched nearly every country on the planet; however, its impact has varied widely. For example, the disease causes high mortality in Iran but lower mortality has been reported just across the border in Iraq, Bahrain, and Kuwait 3  East respiratory syndrome coronavirus (MERS-CoV) 7 . It is generally accepted that such environmental factors affect the stability of these pathogens and facilitate their spread. However, little is known about the underlying mechanisms involved in SARS-CoV-2 transmission.
Many countries around the world, such as China, have successfully controlled the outbreak using strategies such as effectively implementing social distancing and lockdowns, thereby minimizing humanto-human transmission of SARS-CoV-2. It is expected that subsequent outbreaks may occur during the autumn and winter seasons owing to the changes in meteorological conditions. Therefore, understanding the effect of meteorological factors on the spread of SARS-CoV-2 could help us devise effective strategies to prevent further outbreaks and spread of COVID-19. Many recent studies have reported a correlation between the COVID-19 pandemic and numerous meteorological variables [8][9][10][11][12] ; however, the in uence of meteorological factors on the spread of COVID- 19 is not yet completely understood. In this study, we used data from 487 counties in the United States (US), collected from March 5, 2020 to May 1, 2020, to determine the association of meteorological and non-meteorological factors with the early transmission of SARS-CoV-2 in the US. Our study provides a comprehensive analysis of factors associated with the spread of COVID-19.

Results
We calculated R proxy , IR, and CG by analyzing the data from 487 counties in the US from March to May 2020. Data from 486, 482, and 303 counties (R proxy ) and 487, 487, and 307 counties (IR and CG) were used to calculate the three parameters (R proxy , IR, and CG) for 17, 24, and 31 days after the 50th con rmed case in each county, respectively. The AT in these counties during this period ranged from −3.43 °C to 26.56 °C, and the AAH varied from 2.56 to 24.18 g/m 3 . Latitude ranged from 21.46°N to 64.81°N, and PD ranged from 1.2 to 11,103 inhabitants/mi 2 ( Fig. 1 and S.1).
Potential relationships between early transmission of SARS-CoV-2 and environmental variables, including temperature, absolute humidity, atmospheric pressure, latitude, and PD were explored using univariate regression analysis. In detail, AAP and PD were found to be key factors associated with virus transmission. Particularly, the average R proxy for 17 days showed a relationship with AAP (R 2 =0.259, P<0.01) and PD (R 2 =0.200, P<0.01); however, relatively small in uences of AT, AAH, and latitude were observed (Table 1) (Fig. 2). Similarly, the CG for 17 days showed a relationship with AAP (R 2 =0.176, P<0.01) and PD (R 2 =0.325, P= P<0.01), but small in uences of AT, AAH and latitude were observed (Table  S.1). The IR of COVID-19 displayed slight in uences owing to all ve environmental parameters (Table  S.2). The overall results for 24 days also depicted a similar pattern but the in uences of AAP and PD on R proxy and CG were comparatively less than those for 17 days ( The sMLR models showed that SARS-CoV-2 transmission could be explained by different combinations of parameters (Table 2). Speci cally, latitude, AAH, AAP, and PD explained 37.2% of the total variance for the 17-day R prxoy whereas latitude, AT, AAP, and PD explained 41.0% of the total variance for the 17-day CG ( Table 2). In contrast, latitude, AT, AAH, AAP, and PD explained only 20.5% of the total variance for the 17-day IR (Table 2). However, in the results for 24 and 31 days, the pattern of variance explained by these parameters was not followed (Tables S.9 and S.10).

Discussion
The present study was based on a relatively large dataset of the early transmission of COVID-19 in 487 US counties from March to May 2020. Our results suggest that both temperature and absolute humidity have a relatively small in uence on the transmission of the virus that causes COVID-19, SARS-CoV-2.
These ndings are consistent with those of several recent studies conducted in China 13 {cmarchant:2020ki} and Spain 14 . These results are also in good agreement with the fast-growing incidence of COVID-19 during the summer season, despite higher temperatures.
A novel nding of this study is the relationship between AAP and SARS-CoV-2 transmission. This study showed that higher AAP led to the rapid spread of COVID-19 during the early stage of the pandemic. Our results are consistent with a previous study conducted in Hubei Province of China, which demonstrated that the incidence of COVID-19 was positively correlated with AAP 15 ; however, our results are in contrast with other studies conducted in Japan 16 and China 17 . Use of a larger dataset than in those previous studies and systematic approaches may lead to more robust conclusions. A positive correlation of atmospheric pressure with the spread of in uenza virus 18 and SARS-CoV 19 has been found in previous studies. The underlying mechanism behind the relationship of AAP with the transmission of COVID-19 is unclear. One explanation might be that increased AAP may lead to an increase in the number of viruses per unit area when the human body expels respiratory viruses 15 . A positive correlation was observed between COVID-19 transmission and PD, consistent with previous study 20 , which also strengthens the notion that lockdown policies can help to atten the epidemic curve 21 . The in uence of PD on the transmission of COVID-19 suggests that lockdown measures in cities with high PD, along with other measures including social distancing, isolation, and quarantine, are effective strategies to combat the transmission of COVID-19 in population-dense areas.
Our study is based on a comprehensive statistical analysis of a relatively large dataset. However, there are some limitations associated with this study. First, the observed R proxy across counties were estimated using available data, which may include incomplete records and this would add noise to the analysis results. Secondly, accurate estimates of SARS-CoV-2 infection are critical in our study; however, con rmed COVID-19 case counts in the US do not capture the total burden of the pandemic because testing has been primarily restricted to individuals with moderate to severe symptoms owing to limited test availability 22 . Third, our study included nearly 500 counties in the US; however, a much larger dataset including different countries and seasons worldwide is needed to better assess the relationships of meteorological and non-meteorological variables with virus transmission.
In conclusion, we found a minimal in uence of temperature and absolute humidity on SARS-CoV-2 transmission. We, therefore, disagree with the expectation that outbreaks of COVID-19 are likely during the coming autumn and winter seasons owing to a fall in temperature. However, we observed a considerable positive in uence of atmospheric pressure and population density on COVID-19 transmission in this study. Therefore, our results imply that increased atmospheric pressure during the coming autumn and winter may increase the risk of COVID-19 outbreaks in the northern hemisphere this year. The positive correlation of population density with COVID-19 transmission observed in this study shows that lockdown measures in population-dense cities are effective approaches to minimize disease transmission. The mechanisms underlying the relationship between meteorological parameters and the spread of COVID-19 need to be further explored. Additional studies are needed to better elucidate the role of meteorological and non-meteorological variables and to better forecast outbreak events during the ongoing COVID-19 pandemic.

Study region and data
We collected cumulative information of con rmed cases of COVID-19 in 487 US counties, reported by the Johns Hopkins Coronavirus Resource Center (https://github.com/CSSEGISandData/COVID-19) from March 5, 2020 to May 1, 2020. We used three parameters, including infection rate (IR), effective reproduction number (R proxy ), and compound growth rate (CG), to evaluate the spread of COVID-19.
IR was calculated using the following equation: where C(t) denotes the cumulative number of con rmed cases per day t, and N represents the total population of the county. IR during time interval T was calculated using the following equation: The effective reproduction number (R proxy ) was used, as described in a study by Luo et al. 13 . Brie y, a proxy for the reproductive number R in 5-day intervals was calculated using cumulative incidence data for each county. A proxy for R, R proxy , indicates the occurrence of cases from time (t) to time (t + d) onto cases reported from time (t + d) to time (t + 2d), where d is the calculated serial interval (i.e., the interval between successive cases in a series of disease transmissions). For multiple time points, t, values of R proxy (t, d) were obtained using the equation below: Taking d as 5, we estimate the R proxy of D days (where D 10); for example, to calculate the R proxy of 17 days, we used the following formula: Compound growth rate (CG) was calculated using the following equation 23 : Where C 1 represents the number of con rmed cases on the rst day after the 50 th case, the C 2 represents the number of con rmed cases on the last day of the investigation period, and D represents the duration (days) of the period.
Meteorological data, including temperature, relative humidity, and atmospheric pressure were collected from Reliable Prognosis (https://rp5.ru/Weather_in_the_world). We calculated the "absolute humidity" using temperature and relative humidity for each county with the following formula, which is an approximation of the Clausius-Clapeyron equation 13 : where AH refers to absolute humidity, T is the temperature in Celsius, RH is the relative humidity (%), and e is the base of the natural log.
We calculated average temperature (AT), average absolute humidity (AAH), and average atmospheric pressure (AAP) for 17, 24 and 31 days after the 50th con rmed case in each county. Population data at county level at the start of the year 2020 were obtained from the U.S. Census Bureau, and population density (PD) was calculated (inhabitants/mi 2 ) for each county. To perform statistical analysis, we converted the population data logarithmically.

Statistical and modeling analysis
Statistical analysis was performed using the R statistical platform, v. 3.6.1 (The R Project for Statistical Computing, Vienna, Austria). First, univariate linear regression analysis was used to identify relationships between the measured environmental variables and R proxy , IR, and CG of COVID-19. R-squared values (R 2 ) were calculated for the regression model to evaluate the percentage of variance in R proxy , IR, and CG of COVID-19 that could be explained by each environmental variable. Second, stepwise multiple linear regression (sMLR) models were developed. Model t was assessed using R 2 , and the Akaike information criterion was used to determine whether to add or remove variables during the stepwise procedure 24 . Before their use in sMLR, the data were standardized to their z-scores.

Declarations
Author contribution statement

Declaration of competing interests
The authors declare no con ict of interest. Table 2 Statistics table for