Regionalization of precipitation and the spatiotemporal distribution of extreme precipitation in southwestern China

Daily precipitation data from 1951 to 2010 at 33 meteorological stations in five provinces/cities, including Sichuan Province, Yunnan Province, the Guangxi Zhuang Autonomous Region, Guizhou Province and Chongqing City, are used to partition the study area based on precipitation and analyze the spatiotemporal distribution of extreme precipitation. The rotated empirical orthogonal function (REOF) analysis method is used to divide the study area into five parts according to precipitation. The precipitation exhibited greater fluctuations in the last two decades. This finding indicates that extreme precipitation events are increasing. The frequent occurrence of extreme precipitation has made drought and flood disasters more serious. The Mann–Kendall test (M–K test) and moving t test methods are used to analyze the jump and monotonic trends of extreme precipitation indices. It is determined that extreme precipitation indices, including Rx1d, Rx5d, R95p, R99p, CWD and R10 mm, exhibited a weak upward trend during the past 60 years, suggesting that the precipitation amount in the study area decreased slightly, but the maximum daily precipitation amount (Rx1d) and the extremely wet day precipitation (R99p) increased. This finding indicates that precipitation is more concentrated and the extreme precipitation is more serious. The jump for most of the extreme precipitation indices occurred in the 1990s. In terms of spatial scale, extreme precipitation indices, except CDD and CWD, exhibited an increasing trend from the northwest to the southeast. The regions with especially high or low values are easy to identify. Drought risk in northwest Sichuan and the junction of Sichuan and Yunnan is higher. Guangxi Zhuang Autonomous Region and southern part of Yunnan Province have a higher flood risk. The trends of nine extreme precipitation indices also demonstrated the spatial differences. There are more stations exhibiting upward trends than stations showing downward trends for six extreme precipitation indices. The risk of drought/flood may increase in Yunnan and Guangxi, and the storm flood risk in Chongqing exhibited an increasing trend.