The influence of climate factors on COVID-19 transmission in Malaysia: An autoregressive integrated moving average (ARIMA) model

Background A unique concern pertaining to the spread of COVID-19 across countries is the asymmetry of risk and the irrational fear of a new pandemic and its possible serious consequences. This study aimed to perform a time series analysis on the association between climate factors and daily cases of COVID-19 in Malaysia up to 15 July 2020. The second objective was to predict daily new cases using a forecasting technique. To address within-country variations, the analysis was extended to the state level with Sarawak state as an example. Methodology/Principal Findings Datasets on the daily confirmed cases and climate variables in Malaysia and Sarawak state were obtained from publicly accessible official websites. A descriptive analysis was performed to characterize all the important variables over the study period. An autoregressive integrated moving average (ARIMA) model was introduced using daily cases as the dependent variable and climate parameters as the explanatory variables. For Malaysia, the findings suggest that, ceteris paribus, the number of COVID-19 cases decreased with increasing average temperature (p=0.003) or wind speed (p=0.029). However, none of the climate parameters showed a significant relationship with the number of COVID-19 cases in Sarawak state. Forecasts from the ARIMA models showed that new daily COVID-19 cases had already reached the outbreak level and a decreasing trend in both settings. Holding other parameters constant, a small number of new cases (approximately a single digit) is a probable second wave in Sarawak state, Conclusions/Significance The findings suggest that climate parameters and forecasts are helpful for reducing the uncertainty in the severity of future COVID-19 transmission. A highlight is that forecasts will be a useful tool for making decisions and taking the appropriate interventions to contain the spread of the virus in the community.