关键词: 土壤含水量, 植被指数, TVDI, 特征空间

Abstract: [Objective] The aim of the study was to study topsoil moisture inversion condition of desert steppe based on surface temperature and vegetation index characteristics spaces. [Method] Three periods of MODIS L1B data in 2014 and 2015 was used to construct the characteristics spaces of surface temperature Ts-vegetation index NDVI from which temperature vegetation drought index(TVDI) was calculated by the equation of dry and wet line. TVDI carries on regression analysis with synchronization field of the measured soil moisture data of different depth, establishing the model of TVDI estimates soil moisture and to verify this model. [Results] TVDI was correlated with each soil layer moisture of measured data, with the correlation coefficient was above 0.5. Based on TVDI model, relative errors of soil moisture inversion 0-10 cm, 0-20 cm and 0-30 cm depth in the area were small. The relative error increased with increasing soil depth. The average relative error between models and measured values in three periods were approximately 10%. [Conclusion] The model of TVDI can more accurately reflect the spatial variations and distribution of topsoil moisture of 0-10 cm depth.

Key words: soil moisture, vegetation index, TVDI, characteristic space