中国的降水量符合正态分布吗?

利用1961—2017年长江流域700个气象站点逐月降水资料计算长江流域9个子流域面雨量，采用基于Box-Cox正态分布转换后的百分位法对长江流域不同时间长度的极端降水气候事件阈值进行界定。结果表明，在数据序列长度发生变化的情况下，面雨量序列经Box-Cox正态转换后，计算得到的极端降水气候事件阈值的变化相较于常规百分位法明显减小，具有更为稳健的特性，从而使得相应极端降水气候事件个例的挑选更为稳定。根据该方法得到的阈值，对2018年汛期(6—8月)长江各子流域极端降水气候事件进行判定，岷沱江流域发生了极端多雨气候事件，而长江干流重庆-宜昌段、汉江及中游干流区间发生了极端少雨气候事件。

Increasing trend of extreme rain events over India in a warming environment

According to the geographical boundaries, we divided the Yangtze river basin into nine sub-basins, and calculated areal rainfall in each sub-basin based on monthly precipitation data at 700 meteorological stations in the basin from 1961 to 2017. We find threshold values of extreme precipitation climate events in each sub-basin by percentile method after applying Box-Cox transformation. The results show that compared with year number ranking method, the threshold value of extreme precipitation climate events in the Yangtze sub-basins defined by these normalized areal rainfall series have obvious advantage. It presents a more stable characteristic when the length of the areal rainfall series changes. The threshold value changes are reduced, and the threshold changes for vast majority of the sub-basins are within 10 mm. Thus the extreme years are stable. Using the threshold values by this method, we examined whether extreme precipitation climate events have occurred during flood season in 2018. As a result, extreme rainy climate event happens in the Mintuo river basin and extreme less rain climate events happen both in the stem stream from Chongqing to Yichang and in the basin that from Hanjiang river to the middle reaches of the Yangtze river.