Using scattering decomposition to improve the estimation of the topography-induced polarization orientation angle

ABSTRACT

Polarization orientation angle (POA) shifts produced by topography variations lead to rotating polarization around the radar beam direction. In addition to slope topography, depolarization induced by volume scattering and orientation of dihedral target impact POA which is used for deorientation and, in some cases, topography measurement. Unlike deorientation, to provide reliable topography measurements, it is essential to accurately estimate shifts in POA caused by topography variations, not volume scattering or oriented dihedral. The existing POA estimation methods approximately provide topography POA using an averaged POA. In this letter, a modified POA estimation method was proposed to achieve a high consistency with topography. The proposed method initially reconstructed some sub-scatterers using eigenvalue/eigenvector Cloude–Pottier decomposition. In the second step, the type of scattering mechanisms was determined using Cloude decomposition. Then, it compensated for the POA of the dihedral and the volume-reconstructed sub-scatterers using the POA of each of them obtained with the help of Huynen’s desying operation. Finally, topography POA was obtained from the new coherency matrix. The experimental results were evaluated by the digital elevation model (DEM)-derived POA. Results demonstrated that the proposed method L- and P-band derived POA are ${1.48}^{°}$ and ${1.55}^{°}$ better than the averaged POA in the existing method.