According to the latest Sentinel-1 radar system parameters and the surface characteristic parameters of the study area, this paper established a database of surface microwave scattering characteristics under the sparse vegetation cover condition by AIEM model. On the basis of AIEM model numerical simulation analysis, soil moisture inversion model was constructed. The results showed that: 1) Under the different incident angles and polarization mode, the backscatter coefficient response of soil moisture (Mv) and the composite surface roughness (Zs) have obvious logarithmic correlation, respectively. Soil moisture response of VV polarization is sensitive, the optimal response range is for Mv 0~30%, Zs 0~0.06 cm. 2) GammaMAP filtering has good denoising effect in the sentinel-1 radar data preprocessing. The spatial distribution of soil moisture extracted from sentinel-1 images has good consistency with the actual situation in the same period field in the study area. In April, the soil moisture spatio-temporal distribution of Weigan-Kuqa river basin shows spring drought characteristics. 3) For 0~10 cm surface soil moisture, the correlation coefficient between the measured value and simulated value reaches 0.76, showing that this model is suitable for the extraction of soil moisture in the arid area of a certain regional scale.