Impact of climate change on SARS-CoV-2 epidemic in China

SP, where GR indicated the highest correlation with UV and T (RUV = -0.53, RT = -0.51), fol- lowed by P and WS (RP = -0.33, RWS = - 0.23). In addition, T indicated a strong positive correlation with P and a strong negative correla- tion with SP, both of which were significant (P-value < 0.05). T showed a weak positive corre- lation with WS, which was significant (P-value < 0.05). No correlation was indicated with TP. P indicated a significant negative correlation with SP (P-value < 0.05), whereas the remaining factors were positively correlated with SP. TP indicated a strong positive correlation with WS, whereas its correlations with the other factors were weak. Except for P and UV, SP showed weak negative correlations with other factors. The spatial distribution and cluster characteristics of the GR and each weather factor are shown in Fig 7. In terms of the GR and weather sensitivity, Guangdong and Guangxi indicated the characteristics of LL and LH clusters, respectively, indicating that the regions around these provinces showed high climate sensitivity. The cluster distribution characteristics of the GR and weather factors can be categorized into two regions, eastern and western China, with the eastern region showing HH and LH cluster characteristics, and the western region showing LL and HL clusters. Neither northern nor central China showed spatial cluster characteristics, which indicates that weather factor responses to the epidemic were evident in western and eastern China. Weather sensitivity analysis. Six weather factors were selected, and the individual analy-

ses presented certain limitations. To investigate the overall effect of weather factors on the SARS-CoV-2 epidemic, the six weather factors were represented in the form of a composite score for each province using principal component analysis, which was then classified into high, medium, and low levels using the K-means clustering algorithm. These levels represent the weather sensitivity, i.e., the level by which each region was affected by the weather. The spatial distribution of weather sensitivity is shown in Fig 8A, and the composite score of each province is shown in Fig 8B. Heilongjiang, Jilin, Hubei, and Guizhou were high-sensitivity provinces; 11 provinces were of medium sensitivity; and the remaining 18 provinces were low- sensitivity regions. LL clusters were indicated in Shandong, whereas HL clusters were indi- cated in Jiangsu and Shanxi. No cluster characteristics were discovered in other provinces, and the overall cluster characteristics were not significant (P-value > 0.05). Among the high-sensitivity regions, Heilongjiang and Hubei were classified as Group A, and Jilin and Guizhou were classified as Group B. The T in Groups A and B was low and high, respectively; the TP in Groups A and B was low and medium, respectively; the WS of all four regions were high; the SP of Heilongjiang and Jilin was high, whereas those of Hubei and Gui- zhou were medium and low, respectively, which is attributable to altitude. The remaining three regions, except Heilongjiang, did not indicate high GRs; in particular, Groups A and B

indicated high and low levels, respectively. Although Heilongjiang is adjacent to Jilin (in northern China), and Hubei and Guizhou are in close proximity to each other (in southern China), their spatial and temporal characteristics were different. Additionally, the spatial dis- tribution of the epidemic was not directly related to the south and north of China. Except for the eastern coastal region of China and the medium- and low-sensitivity areas, the GR and confirmed GR were at medium to low levels in most provinces, indicating a positive relation- ship between weather sensitivity and GR. The eastern coastal region is economically devel- oped, with a dense and mobile population dominated by human activities, where weather is only a secondary factor. Therefore, physical isolation was the most effective measure for con- trolling the outbreak and spread of the epidemic in this region. PLOS ONE Impact of climate change on SARS-CoV-2 epidemic in China PLOS ONE | https://doi.org/10.1371/journal.pone.0285179 July 27, 2023 12 / 21