Polarimetric SAR interferometry as applied to fully polarimetric rain forest data

Abstract

The scope of this paper is to establish a polarimetric SAR interferometry approach to estimate interferometrically observable forest and ground parameter. Three optimized interferometric coherences can be found through a properly weighted superimposition of all the interferometric coherences available from a fully polarimetric single baseline experimental set-up in repeat-pass mode. These proper weights can be found by solving an eigenvalue problem for each resolution cell [Cloiude and Papathanassiou 1998]. The result serves as the right hand side of a least square optimization problem which makes use of a scattering model based on a randomly oriented volume over an impenetrable surface [Treuhaft and Siqueira 2000] to estimate a set of physical forest parameters transformed through this model. A conjugate direction set iteration scheme is used to minimize the cost function of the optimization problem and to solve for the pertinent candidate parameters, which are the forest volume thickness, the volume extinction coefficient, the interferometric phase related to the underlying topography, and the effective ground-to-volume amplitude ratios of the related interferometric coherences. Due to the non-uniqueness of this inversion problem an additional constrained is introduced via a Tikhonov regularization scheme. The results of the presented schemes are shown for the first time on P-band data takes acquired over tropical rain forest in Brazil.