西南雨季降水变化的区域特征

Regional Characteristics of Rainy Season Precipitation Variability in the Southwest

摘要: 为了更好地对西南雨季降水的时空变化特征进行研究，利用中国西南地区1960~2022年81个气象站点共63年的逐日降水量气象观测资料，通过旋转经验正交函数分解(REOF)、Mann-Kendall突变检验和Morlet小波分析等研究方法，揭示西南地区降水的时空分布特征、长期变化趋势以及周期变化特征。分析结果表明：1) 西南地区1960~2022年来主要有3个降水大值区，分别为云南南部、贵州西南部以及四川中东部地区，降水量分布自东、南向西、北逐渐递减。2) 在EOF分析的基础上进行REOF分解，将西南地区分为5个区：1区为云南区，2区为川西高原区，3区为川中区，4区为贵州区，5区为川东区。3) 5个区域的雨季降水除2区呈现缓慢上升趋势外，其余区域均呈现程度不同的下降趋势，5个区域的雨季降水突变主要集中在21世纪初期，且突变后呈现不同的变化趋势。变化周期主要以5~10a、准5a周期为主。

Abstract: In order to better study the spatio-temporal variation characteristics of rainy season precipitation in Southwest China, the day-by-day precipitation meteorological observations of 81 meteorological stations in Southwest China for a total of 63 years from 1960 to 2022 are utilized to reveal the spatio-temporal precipitation in Southwest China through the Rotated Empirical Orthogonal Function Decomposition (REOF), the Mann-Kendall Mutation Test, and the Morlet wavelet analysis and other research methods to reveal the spatial and temporal distribution characteristics of precipitation, long-term trends and cycle change characteristics in Southwest China. The analysis results show that: 1) there are three major precipitation areas in Southwest China from 1960 to 2022, namely, southern Yunnan, southwestern Guizhou, and east-central Sichuan, and the distribution of precipitation gradually decreases from east and south to west and north. 2) REOF decomposition is carried out on the basis of EOF analysis, and the southwest region is divided into 5 zones: zone 1 is Yunnan, zone 2 is the western Sichuan plateau zone, zone 3 is the central Sichuan zone, zone 4 is the Guizhou zone, and zone 5 is the eastern Sichuan zone. 3) The rainy season precipitation in the five regions shows a decreasing trend in varying degrees, except for region 2, which shows a slowly increasing trend. The sudden changes in rainy season precipitation in the five regions are mainly concentrated at the beginning of the 21st century and show different trends after the sudden changes. The change cycles are mainly 5~10a and quasi 5a cycles.