Drought is a complex natural disaster affected by multiple climate factors and underlying surface. In recent years, drought monitoring indices of remote sensing have been widely applied to monitor drought in a certain region or global. However, some remote sensing drought monitoring indices do not consider the drought-causing factors enough to reflect the comprehensive drought situation of a region fully. In this paper, a new remote sensing drought monitoring index, called Remote Sensing Drought Evaluation Index (RSDEI), was constructed by combining Vegetation Condition Index (VCI), Temperature Condition Index (TCI), Precipitation Condition Index (PCI), and Soil Moisture Condition Index (SMCI) using the spatial principal component analysis (SPCA) method. The reasonableness of RSDEI was test and verified using Net Primary Productivity (NPP), Standardized Precipitation Evapotranspiration Index (SPEI), and unit area crop yield. The RSDEI was also applied to the drought condition monitoring of the northwest arid and semi-arid region from 2001 to 2019.The result demonstrated that the results showed that the RSDEI had a high correlation coefficient with SPEI-12 (R=0.85, p<0.01). It is concluded that the correlation coefficient between RSDEI and NPP is 0.74 at 95% confidence level, which indicates that RSDEI and NPP have a strong correlation. Then, the correlation between RSDEI and crop yield per unit area is 0.89. The results of RSDEI showed that the drought in northwest China started in May and lasted in September from 2001 to 2019. The lowest value of RSDEI appeared in May, which inflected the significant difference of drought level in different month in northwest China. The result of CV (coefficient of variation) showed that the drought variation in the study area had a stable low fluctuation condition as a whole, in the northwest and northeast of study area, which indicated that the changes of drought were different in the past 19 years. The Hurst exponent analysis showed that the area with the positive evolution of Hurst index (0.5<H<1) is 1,845,046.669 km²,which accounts for 75.9% of the total area, while the area with reverse evolution characteristics (H<0.5) accounts for 24.1% of the total area. The result obtained above reflected that the drought changes in most regions are better than that in the past 19 years. The trend gradually changes from drought to humid.

干旱是受多种气候因素和下垫面影响的复杂自然灾害。近年来，遥感干旱监测指标已被广泛应用于监测某一区域或全球的干旱。然而，一些遥感干旱监测指标没有充分考虑干旱成因，不能充分反映一个地区的综合干旱情况。本文结合植被状况指数（VCI）、温度状况指数（TCI）、降水状况指数（PCI）和土壤水分，构建了一种新的遥感干旱监测指数，称为遥感干旱评价指数（RSDEI）。条件指数 (SMCI) 使用空间主成分分析 (SPCA) 方法。RSDEI的合理性通过净初级生产力（NPP）测试验证，标准化降水蒸散指数 (SPEI) 和单位面积作物产量。RSDEI还应用于2001-2019年西北干旱半干旱区干旱状况监测。结果表明，RSDEI与SPEI-12具有较高的相关系数（R = 0.85, p<0.01)。得出的结论是，在 95% 的置信水平下，RSDEI 与 NPP 的相关系数为 0.74，表明 RSDEI 与 NPP 具有很强的相关性。那么，RSDEI 与单位面积作物产量之间的相关性为 0.89。RSDEI结果表明，2001年至2019年，西北地区干旱从5月开始，9月持续，RSDEI最低值出现在5月，反映了西北地区不同月份干旱程度的显着差异。CV（变异系数）结果表明，研究区的干旱变化总体上处于稳定的低波动状态，在研究区的西北和东北部，表明过去19年干旱的变化是不同的。年。²，占总面积的75.9%，具有逆向演化特征（H <0.5）的区域占总面积的24.1%。上述结果反映了大部分地区的干旱变化好于过去19年。趋势逐渐由干旱转为湿润。