基于主成分分析的中国南方干旱脆弱性评价

干旱脆弱性是干旱灾害形成的根本原因。遵循全面性、系统性和可操作性原则，选取水资源脆弱性、经济脆弱性、社会脆弱性、农业脆弱性和防旱抗旱能力脆弱性5个准则层，共32个指标，建立了中国南方农业干旱脆弱性评价指标体系，运用主成分分析的理论方法确定评价指标权重，建立中国南方地区的干旱脆弱性评价模型，得到不同省市的干旱脆弱性指数、分级阈值和区划，以期为南方地区防旱减灾工作提供基础数据和理论支持。得出以下结论，（1）通过主成分分析法得到四个主成分，第一主成分方差贡献率为54.90%，主要反映农业脆弱性和社会脆弱性；第二主成分方差贡献率为23.64%，主要反映水资源脆弱性和经济脆弱性；第三主成分和第四主成分所占比重较小，主要反映防旱抗旱能力脆弱性。（2）以4个主成分的方差贡献率为系数建立南方干旱脆弱性评价模型，得到中国南方干旱脆弱性综合评价得分及其排名，其中水资源脆弱性由高到低依次为云南、广西、贵州、四川、重庆和广东；经济脆弱性由高到低依次为广西、云南、贵州、四川、重庆和广东；社会脆弱性由高到低依次为贵州、云南、广西、重庆、四川和广东；农业脆弱性从高到低依次为贵州、云南、广西、重庆、四川和广东；防旱抗旱能力脆弱性由高到低分别为云南、贵州、四川、广西、重庆和广东；干旱脆弱性综合评价由高到低依次为云南、贵州、广西、四川、重庆和广东。（3）对干旱脆弱性指数进行正态分布性检验，发现该指数基本服从正态分布。根据正态分布原理，得到干旱脆弱性分级阈值，将干旱脆弱性指数小于0.3842定义为低风险，大于1.0758定义为高风险，介于两者之间的定义为中等风险，获得干旱脆弱性分级区划图。广东省位于干旱的低脆弱区，四川和?

Evaluation of Drought Vulnerability in Southern China Based on Principal Component Analysis

Drought vulnerability is the root of the formation of drought disaster. According to the principles of comprehensiveness, systematization and operability, and selecting five criterion layers, which included the vulnerability of water resources, economy, society, agriculture and the ability of drought prevention and resistance, the index evaluation system of agricultural drought vulnerability was established. By using principal component analysis to determine the weight of estimation indices, a drought vulnerability assessment model was constructed. Making use of the model, we got the drought vulnerability indices, threshold value and vulnerability zone in different provinces in southern China. These studies can provide the basic data and theoretical support for disaster prevention and reduction in South China. The results showed that： （1） through principal component analysis, we got four principal components. The first principal component mainly reflected the agricultural vulnerability and social vulnerability, and the variance contribution was 54.90%. The second principal component mainly reflected the water resources vulnerability and economical vulnerability, and the variance contribution was 23.64%. The proportions of the third and forth principal component were little, and mainly reflected the vulnerability of the ability of drought prevention and resistance. （2） The drought vulnerability evaluation model was constructed according to the variance contribution of four principal components. Then we got the comprehensive scores and ranking. The vulnerability of water resources in the sequence of descending were Yunnan, Guangxi, Guizhou, Sichuan, Chongqing and Guangdong. The vulnerability of economy from high to low were Guangxi, Yunnan, Guizhou, Sichuan, Chongqing and Guangdong. The vulnerability of society according to the order from high to low were Guizhou, Yunnan, Guangxi, Chongqing, Sichuan and Guangdong. The vulnerability of agriculture according to the order from high to low were Guizhou, Yunna