Keywords:Coronavirus disease (COVID-19), Spatial pattern, Moran's I, Spatial autocorrelation
Coronavirus disease (COVID-19) broke out in Wuhan city and expanded rapidly throughout China. The government has taken a series of tough unprecedented interventions to prevent the spread of the disease. Now, COVID-19 in China has passed the peak and entered the last stage of the pandemic. The aim of this study is to identify the spatial distribution pattern of new confirmed cases in China, in order to understand how interventions affected the COVID-19 spread. First, four-time periods were categorized according to the proportion of new confirmed cases in Hubei. Then, spatial autocorrelation analyses were employed to explore geographic patterns in these four periods. Finally, outliers were analyzed by calculating an Anselin local Moran's I statistic. The results show that the spatial patterns of the COVID-19 in the first three periods spread are the high cluster and random in the fourth period. As the proportion of new confirmed cases in Hubei has dropped significantly, the spatial pattern of the epidemic also has a major transformation. Many economic and culturally developed cities became the focus in addition to Wuhan city.