基于标准降水指数的多尺度叠加干旱监测指标及其等级划分

针对中国不同地区的区域气候及其变化特征，使用信息熵方法从降水观测资料中提取出区域降水的两个（均态和变化）本征尺度，提出一种基于区域气候系统层次性内在特征的气象干旱指数MSPI。使用各站1981-2010年的逐日MSPI序列，基于互信息算法，确定各站气象干旱指数MSPI的干旱等级划分规则，即按某一临界点集合对MSPI序列进行划分，得到的等级序列和MSPI序列之间的互信息最大时，此临界点集合是最优的，该临界点集合也就是干旱和湿润等级的合理划分。使用互信息方法对全国2 211站1981-2010年逐日MSPI序列进行等级划分，分为“特别干、严重干、中等干、轻微干、正常、轻微湿、中等湿、严重湿、特别湿”9个等级；进一步分析了全国1981-2010年各月不同等级干旱日数和站数百分比、各季节中等干、严重干、特别干事件日数比率的空间分布及不同等级干事件发展所缺降水量的距平百分率及缓解所需降水量。最后以2011年长江中下游春季严重气象干旱的发生、发展、持续、缓解过程及山西省1981-2010年逐月干旱事件为例，对MSPI指数及其等级划分规则的干旱监测能力进行检验，发现两者对不同程度的干旱都有较好的表征能力，并且对于干旱过程也有较好的识别能力。

Research on Multi-scale Superposed Drought Index and Its Classification Based on Standardized Precipitation Index

By taking account of the climatic characteristics in different regions, in this study the comentropy method was used to obtain the zonal precipitation at two intrinsic scales, i.e. the homogeneous state intrinsic scale and variational state intrinsic scale, and a new meteorological drought index named MSPI was developed. The mutual information method was used to divide the drought ranks based on set of critical points for MSPI series at all 2 211 stations in China and get the standards of different drought and flood ranks, because this set of critical points was optimal when the mutual information between the MSPI series and the rank order reached their maximum. The MSPI series was divided into 9 ranks, i. e. the “extremely severe drought, severe drought, moderate drought, slight drought, normal situation, slight rainy, moderate rainy, severe rainy and extremely severe rainy”. On which the percentages of monthly days and stations at different drought ranks were calculated, the spatial distribution of the ratio of days at moderate, severe and extremely severe drought ranks at all the stations in different seasons during the period from 1981 to 2010 was analyzed, and the anomaly percentages of precipitation for  alleviating drought were estimated. The occurrence, development, duration and alleviation of severe meteorological drought in the middle and lower reaches of the Changjiang River in 2011 were discussed, and a case study on moderate or more severe drought events in Shanxi Province during the period from 1981 to 2010 was carried out based on MSPI so as to examine the capability of MSPI in monitoring meteorological drought. The result shows that MSPI can be used to monitor meteorological drought at different ranks, and it can also be used to perfectly develop the meteorological drought process.