Inversion of soil moisture using bi-temporal ASAR images

Soil surface roughness and soil moisture play critical roles in influencing backscattering coefficient, in order to investigate the application of soil moisture estimation by ENVISAT-ASAR data in China, therefore based on the ASAR image, with resolution of 30 m, the soil surface roughness characteristics and the bare soil moisture content, near Nanpi ecological experimental station of Chinese Academy of Sciences, were inversed by using an empirical model. The results show that, surface roughness mostly ranges from 0.05～0.50 cm, surface volumetric moisture mainly distributes in the range of 10%～34%, and minority show high, owing to some accumulated water ditches, which were all consistent with ground investigation. Meanwhile, the inversed results of soil moisture were validated by ground truth measured data, and a very good agreement between simulated and measured data was observed, with the residual RMSE error of 3.7%. In this study, a soil moisture inversion methodology was introduced using backscattering model which deducts the effect of surface roughness, without a-prior information on surface roughness. The methodology just needs two HH polarized ASAR images, with a short time for inversion and different incidence angles. Roughness and moisture can be estimated by the difference of radar signal, which provides an available method for monitoring soil moisture quickly over local areas.