Soil Moisture Mapping with Polarimetric SAR in Huanghe Delta of China

Abstract

The objective of this paper is to use the partial least squares regression as a modeling method to establish the relationship between soil moisture and polarimetric decomposition parameters which are obtained by T3 matrix and $\mathrm{H}/\mathrm{A}/\alpha$ polarimetric decomposition with full-polarization RADARSAT-2 SAR image. The results show that the parameter set of T3 matrix have the minimum accuracy comparing with the other three decomposition ways of the $\mathrm{H}/\mathrm{A}/\alpha$. The best results of the $\mathrm{H}/\mathrm{A}/\alpha$ is the Eigenvector parameter set. However, it is still lower than the inversion of parameter combinations with R2 of 0.84 for calibration and R2 of 0.77 for validation. It indicates that the polarimetric decomposition parameters can be used for mapping soil moisture in Huanghe Delta.