As a key factor in the drought monitoring, the soil moisture is always the important content in the dynamic drought monitoring with remote sensing. In this paper, the drought model of EPDI was constructed based on the spec-trum feature space of Nir-Red by using the data of EVI、infrared and red reflectance so as to conduct the soil moisture inversion. In order to test the accuracy of the inversed data by EPDI, it was compared the three drought index models with the field observation data in the LhasaRiver Basin. The results showed that the model of EPDI could better inverse the soil moisture with the relative errors of only 0 . 1040 and linear correlation coefficient of 0 . 9181 . The precisions of the inversed soil moisture for EPDI has been greatly enhanced by 38. 83% and 29. 35% respectively compared to those of PDI and MPDI. Therefore,the EPDI could provide a more effective approach for the dynamic monitoring of drought in large scale. In addition,the parameters for EPDI were more convenient for construction with less artificial interferences, so that the model provided the possibility for the broader application for the prevention and control of the drought.%土壤水分作为干旱监测的重要指标,一直是干旱遥感监测研究的重要内容。本研究利用MODIS数据的EVI、红波段反射率、近红外波段反射率数据构建了基于Nir-Red反特率光谱特征空间的EPDI模型进行土壤水分的反演。利用野外同步测量数据对PDI、MPDI、EPDI三种干旱指数模型获取的拉萨河流域土壤水分进行了验证和对比分析,研究结果表明EPDI能够更准确地反演土壤水分,其样本点的相对误差仅为0．1040,线性相关系数为0．9181,反演精度相比PDI、MPDI (0．1646、0．1472)分别提高了36．83%和29．35%,为利用遥感影像数据进行大尺度的干旱动态监测提供了新途径。并且相比MPDI,EPDI模型参数更容易获取,模型构建受人为因素影响小,从而为模型的大范围推广提供了可能,具有很好的应用意义。
展开▼
