摘要： Soil moisture content (SMC) is the important information for the crop land irrigation management and drought warning. The study of SMC quantitative inversion based on hyperspectral remote sensing technology has become the hot spot. Because of using statistical modeling methods and without considering soil bidirectional reflection characteristics, the SMC inversion accuracy and model applicable scope were limited. The purpose of this study was to use the soil radiation transfer model (SOILSPECT) to invert SMC in order to not only improve the SMC hyperspectral inversion accuracy, but also make the model applying widely. Taking the black soil and chernozem in Gongzhuling city of Jilin province as the study object, spectrometry measurement for different SMC from 5% to 45% (5% interval) to get measured soil reflectance by using the ASD Fieldspec Pro spectrometer under the indoor condition. The influences of the light source zenith angle, observing zenith angle, azimuth angle and SMC on soil bidirectional reflectance analyzed. Then distribution SOILSPECT model parameters were obtained by using Particle Swarm Optimization (PSO) method under the different SMC gradients, and SMC inversion by using SOILSPECT model was conducted. The results showed that soil BRDF declined with SMC rising in the range of 400~1400 nm wavelength, when SMC was less than the field water holding capacity, and soil BRDF rose with SMC rising when SMC was larger than the field water holding capacity. In the range of 1400~2400nm wavelength and under different observing zenith angles, there was no rules for soil BRDF changing. SMC inversion accuracy based on SOILSPECT model was higher. R2 value was above 0.98 compared the estimated values with the measured values. The inversion accuracy for 15% and 30% of SMC were higher at every sensitive band, but that for other SMC was unsteadiness. SMC inversion accuracy based on SOILSPECT model increased with the decreasing of the observational zenith angle. The vertical observation can get the highest SMC inversion accuracy.