Two novel surface model based inversion algorithms using multi-frequency polSAR data

The aim of this paper is to present two novel surface model based inversion algorithms using multifrequency and polSAR data. The first part of this work introduces a polarimetric scattering model using the integral equation model (IEM) with a transition model for the reflection coefficient. Polarimetric descriptors: the entropy (H), the anisotropy (A), and the mean alpha angle obtained from the Cloude/Pottier polarimetric decomposition theorem are employed for surface characterization. A new polarimetric parameter: the eigenvalue relative difference (ERD) is developed to overcome the anisotropy limitations in case of rough surfaces. Built on relevant polarimetric descriptors: H, alpha1 and ERD, two novel surface model based inversion algorithms using multifrequency polSAR data are presented. The first algorithm is built up in a "low frequency" (P to S band) - "high frequency" (C to K band) scheme. This algorithm retrieves the soil parameters by using multifrequency least-square fit