中国气候变化检测归因研究进展

孙颖, 王东阡, 张学斌

Progress in climate change detection and attribution studies in China

SUN Ying, WANG Dong-Qian, ZHANG Xue-Bin

图6 人为辐射强迫对2021年9月中国北方创纪录降水偏多事件的影响[82] （a~e）月平均降水距平百分率（MPPA）的概率密度分布，重现期分布和风险比及其bootstrap检验结果，ALL（红色)、NAT（蓝色)、GHG（紫色)、AA（橙色）和控制试验（CTL，绿色）情景结果使用颜色区分；（f~j）为最大日降水距平百分率（Rx1day%）结果；（k~o）为最大连续5 d降水距平百分率（Rx5day%）结果

Fig. 6 Fitted distributions, return periods, risk ratios, and exceedance probabilities of domain-averaged MPPA (a-e) by the empirical probability formula and Rx1day% (f-j) and Rx5day% (k-o) by a GEV distribution in September 2021 over northern China based on ALL (red), NAT (blue), GHG (purple), AA (orange), and CTL (green) ensembles. Black lines indicate the observed threshold values of the September 2021 event, i.e., 140.5%, 83.87%, and 88.82% for MPPA, Rx1day%, and Rx5day%, respectively. Panels (e), (j), and (o) are best estimates and 90% confidence intervals of risk ratios (left, gray boxes) and exceedance probabilities (right, color bars). The error bars and boxes mark 5%-95% uncertainty ranges estimated via the bootstrapping method (N=1000)[82]