西南地区旱涝特征及其趋势预测

应用Z指数和标准化降水指数SPI,结合REOF方法和Hurst 指数对西南地区旱涝等级及其时空分布规律、演变趋势进行分析和预测,结果表明:SPI 和Z指数的计算结果和现实情况较一致,从多年平均值来看,云南的中甸及四川的石渠、德格等地干旱指数处于低值区,属重旱区。贵州和广西及云南南部部分地区属易涝区。逐年的SPI 和Z指数结果大体一致,2000 年以后,干旱指数波动较大。通过REOF分析,Z指数的分布大致分为川西区、川东(包括渝)区、云贵(包括广西北部)区、广南四个区域。云贵区Z指数总体呈下降趋势,涝情呈减弱趋势;川西区在1964 年至1973 年间干旱严重,1995 年以后变化相对稳定,但干旱也时有发生;川东区旱涝变化不显著。经Hurst 指数分析,西南五省市未来的旱涝情况大致为干旱程度呈逐步加大的态势,洪涝程度呈较弱的逐步减小态势。

Drought/Flood Characteristics and Trend Prediction in Southwest China

The Z index and Standard Precipitation Index (SPI) as well as the REOF method and Hurst exponent are used to analyze the features of droughts and floods. The spatial and temporal distribution of droughts and floods is investigated, and the long-term trend is detected. The result shows that the sensitivity of Z index is stronger, with large amplitude of changes, while the stability of SPI is better. As a whole, SPI and Z index are similar with the actual situation. According to the analysis, the result of Z index is more close to the actual in Southwest China. According to the multi- year average, the values of Z index at Zhongdian station in Yunnan, Shiqu station and Dege station in Sichuan are at the low level, which belongs to drought area. Guizhou, Guangxi and the south of Yunnan are flooding areas. SPI and Z index are broadly consistence each year, and the drought index showed a great fluctuation after 2000. The first three rotating load vectors and their corresponding principal components are given by REOF method. Through the analysis on REOF results, the southwestern five provinces/cities are divided into four areas: western Sichuan area, eastern Sichuan (including Chongqing) area, Yunnan- Guizhou (including northern Guangxi) area, and southern Guangxi area. Z index in Yunnan-Guizhou area shows downward trend, it means that waterlogging shows a decreasing trend in this area. Severe drought occurred in western Sichuan during the period of 1964-1973, and the Z index became relatively stable since 1995, but drought occurred frequently. The variation of droughts and floods in eastern Sichuan is not significant. Z index and SPI will be reduced in the future of the five provinces in Southwest China, and Hurst exponent is greater than 0.5, that is to say, the degree of drought in southwest region may continue for a period of time. As a whole, Hurst exponent analysis shows an upward trend for droughts, and shows a weak downward trend for waterlogging.