Statistical Analysis of the Mechanisms of COVID-19 Spread Among Japan's Prefectures

In this study, we analyzed daily time series data of new COVID-19 positive cases in each prefecture of Japan to investigate the mechanisms behind the spread of the virus. First, we decomposed the time series data for each prefecture into trend, weekly variation, and short-term components. Using the trend components, we estimated the time lag of infection spread between prefectures and found that Okinawa and Tokyo were leading the spread compared to other prefectures. We also analyzed factors affecting the lag values. Furthermore, through cluster analysis, we grouped all prefectures into 13 categories and conducted a detailed analysis of the infection transition structure within each group. The results suggested that regions centered around Tokyo in the Kanto area served as the epicenter, influencing nationwide spread through areas centered around Osaka and Kyoto. Additionally, we examined the effects of holidays and seasonal variations in the short-term components using regression analysis. The findings confirmed that the effect of holidays was negative immediately after the holidays but became significantly positive one week later. Regarding seasonal effects, the greatest positive impact occurred in November, while a negative impact was observed in March.