Impact of Seven Factors including Temperature, UV Radiation, and Ozone on the COVID-19 Epidemic in the United States

ABSTRACT

This paper is to investigate the impact of meteorological factors and ozone on the COVID-19 epidemic in the United States, and to analyze the annual change patterns of the major influencing factors. Random forest modeling was performed for the correlation between the meteorological factors including relative humidity, 2-meter temperature, ultraviolet (UV) radiation, surface pressure, total precipitation, and total cloud cover, as well as ozone variations, and the daily new cases of COVID-19 in 497 counties in the US, so as to obtain the ranking of the impact of the above factors on the COVID-19 epidemic in the US. Statistical processing and K-means cluster analysis were performed on the modeling results. The modeling results showed that 73.6% of the counties were related to the modeling factors, while 26.4% were not significantly related to the modeling factors. Counties with R2 less than 0.3 were mainly distributed in Connecticut, Louisiana, Michigan, North Carolina, and Texas;In the models with R2 greater than 0.3, the most important influencing factors were 2-meter temperature, UV radiation, and ozone variations. K-means cluster analysis showed that when the number of clusters K=3, the contour coefficient was the largest, and the clustering results at this point presented more obvious characteristics of geographic spatial distribution. Natural factors including relative humidity, 2-meter temperature, UV radiation, surface pressure, total precipitation, total cloud cover, and ozone could not fully explain the changes in the COVID-19 epidemic in the US. Temperature, UV radiation, and ozone were the most significant natural factors that affect the epidemic in the US. Higher 2-meter temperature and stronger UV radiation would help to curb the COVID-19 spread in the US. Lower temperatures and less ultraviolet radiation might be the cause of the epidemic surges in winter in the country. © 2021 ACM.