Correlation Study of Global Weather Conditions and COVID-19 Transmission: A Clustering Approach

ABSTRACT

The COVID-19 pandemic is having a drastic impact around the world. Much research has showed that weather conditions (e.g., temperature, humidity) may correlate with trans-mission and spread of the virus. However, there exist past studies with inconsistent conclusions. In this research, we analyzed a correlation between weather conditions and the COVID-19 transmission using a globally collected database. We explored the data and found that the correlation is not linear;therefore, we employed a non-linear model called Random Forest to analyze the relationship between weather conditions and the number of COVID-19 daily reported cases on a global scale, extracting only strongly correlated variables for further analysis. Subsequently, we applied K-Means clustering to identify clusters of weather conditions that influence COVID-19 dissemination. The environmental conditions that found the highest transmission of COVID-19 were-0.605 degrees Celsius, 70.63 percent relative humidity, and 17.15 km/hr wind speed, according to the results of our research. The experimental results also indicate a negative correlation between temperature and the number of new cases, while a positive relationship exists between relative humidity and wind speed. © 2022 IEEE.