Soil moisture plays an important role in arid ecological system, and soil moisture content is an important parameter in determining soil drought. We analyzed the relationship between the spectral characteristic parameters of emissivity and the soil moisture content that were measured in field and laboratory, and extracted the best sensitive band of inversed soil moisture content by the parameters from spectral characteristic analysis. it was found that the wave length were in 9.480 2~9.7901 μm and the soil moisture content was the most sensitive to its emissivity. Based on the scope of the most optimal sensitive waveband, the thermal infrared wavebands of MODIS, ASTER and HJ-1B Satellites were simulated, a regression analysis was made for waveband, waveband ratio and soil moisture content, and a linear model of waveband ratio was ultimately constructed, which can effectively deduce soil moisture content. The model was verified so as to find the linear relation of soil moisture content with the emissivity of thermal infrared bands 12 to 14 (simulated ASTER satellite sensor). The inversion is so precise that an inversion model applicable to the relation of infrared emissivity and soil moisture content in the studied area was established.