中文摘要：

介绍了一种新的基于广义极值分布（Generalized Extreme Value，GEV）的干旱指数的构建过程及其原理。并利用1951—2010年中国地区160个气象台站的逐月降水资料，选择6个代表台站，检验1951_2010年60年间预测的干旱和实际干旱事件发生的吻合程度；利用2009--2010年资料检验GEV指数和实际发生的干旱事件在月尺度上的监测效果。结果表明，利用GEV方法定义的干旱指数能较好地监测干旱事件的发生和发展，与目前广泛使用的ci指数监测结果较为一致。同时这种指数也能用于某一地区干旱事件的预测和干旱等级的划分。对干旱业务监测和预测有重要的参考意义。但该指数对站点疏密有一定的依赖性，站点稀疏的区域存在着一些误差。

英文摘要：

A new drought index, which is based on generalized extreme value （GEV） distribution, was proposed. The GEV index＇ s theory and method was introduced. By using the monthly precipitation data of 160 sites during 1951-2010, the drought events calculated and observed are compared and validated. By predicting the drought events occurred during 2009-2010, the prediction performances of the GEV drought index is validated on monthly scale. The results show that the GEV method has well performances for drought monitoring and prediction, which also has good agreement with CI index, which is widely used for drought monitoring practices. Meanwhile, this drought index can be used for the predication of drought climatology and categorizing of the drought severity. The results also show that the index has good performances in observation sites density regions, but it has some errors in site-sparse regions.