中文摘要：

干旱具有成因复杂、影响范围广等特点,全球气候变化等因素使得干旱发生的频率日益增加。干旱指数是定量化干旱的工具,可用来分析干旱频率、程度及其时空变化特征。本文选取标准降水指数(SPI)与水文干湿指数这两个指标分别代表气象干旱指数与水文干旱指数,利用西北干旱区黑河上游祁连、野牛沟、托勒和肃南4个气象站1968-2009年的降水数据,采用子流域划分和空间插值等方法,分月、季节和年尺度计算出各站气象干旱指数及相应年份的水文干旱指数,并对两者进行了对比分析。结果表明:标准降水指数SPI与水文干湿指数都对干旱反应灵敏,均能较准确地发现和评价干旱,且二者在年尺度上以及空间上都具有较好的一致性;但在月尺度上,因水文循环过程,水文干旱滞后于气象干旱。

英文摘要：

Drought is a natural phenomenon and a natural hazard characterized by complex formation and an extended scope of influence.With the expanding human population,water demands are increasing along with global climate change,leading to an increase in drought frequency.Drought indices are useful tools to quantify drought and can be used to analyze the frequency,severity,and spatial and temporal pattems of drought.Here we chose the standard precipitation index (SPI) and hydrological index (representing the meteorological and hydrological drought index respectively) to calculate meteorological and hydrological drought at monthly,seasonal and annual scales in the Upper Reaches of the Heihe River in northwestem China from 1968-2009.Meteorological data from four meteorological stations (Qilian,Yeniugou,Tuole and Sunan) were interpolated to the entire upper reaches by Kriging and Inverse Distance methods.Digital Elevation Model (DEM) data were used to delineate subbasins within the study area by geographical information system (GIS) to compare spatial patterns of the two drought indices.The meteorological drought index was calculated,then we calculated the hydrological drought index according to historical records and drought reflected by the meteorological drought index.Results show that the SPI and hydrological index are both sensitive enough to detect the severity and frequency of drought; it is feasible to use these indices to assess drought.The SPI and hydrological index are consistent with each other at annual and spatial scales.However,at the monthly scale,hydrological drought lags behind meteorological drought due to the time it takes for precipitation to transfer to runoff and water vapor through the hydrological process at the watershed scale.