Evaluation of Various Polarimetric Parameters for Soil Moisture Inversion using Polarimetric SAR Data

Abstract:
Polarimetric Parameters from Polarimetric Synthetic Aperture Radar (PolSAR) data are very useful in land and ocean remote sensing applications such as target detection, classification and geo- and bio-physical parameters estimations. In this paper, the recent polarimetric surface scattering models such as X-Bragg model and Polarimetric Two Scale Model (PTSM) have been analyzed for soil moisture estimation over bare soil surface using fully polarimetric AIRSAR data. The various polarimetric parameters used in the two dimensional inversion space have been evaluated for soil moisture estimation using the above polarimetric surface scattering models. These parameters include both Eigenvalue based polarimetric parameters derived using Eigenvalue/Eigenvector decomposition and Non-Eigenvalue based parameters derived using some statistical analysis and simple ratios. The Eigenvalue based parameters are entropy, ‘Cloude-Pottier α’, Touzi alphas and dominant α parameters of Cloude-Pottier (alpha1). The Non-Eigenvalue based parameters are copol ratio, crosspol ratio, scattering diversity and surface scattering fraction. It is observed that X-Bragg model shows less RMS error as compared to PTSM. Scattering diversity and surface scattering fraction shows less and closer RMS error to H-α method.