On Multibaseline Polarimetric Interferometric Techniques for Natural Media

Within the framework of the DRAGON project, a study of Polarimetric SAR Interferometry (PolInSAR) techniques for forests, and natural media in general, has been conducted. This paper focuses on two points. The first is a review of a general PolInSAR coherence model with respect to polarization diversity. The coherence constituents are identified and examined. The second part introduces two general multibaseline PolInSAR coherence optimization methods. The coherence model is utilized to discuss the advantages and applications of these newly developed multibaseline coherence optimization methods. A framework for a multibaseline orthogonal optimal scattering mechanisms decomposition is presented. As the experimental results indicate, multibaseline coherence optimization does improve the accuracy in the estimation of dominant scattering mechanisms and the associated interferometric phases. The methods are evaluated on real data acquired by DLR’s ESAR sensor at L–band.