Polarimetry In Radar Backscattering from Soil and Vegetated Surfaces

Thesis (Ph.D.)--University of Washington, 2015

The unique contributions of this thesis include (1) 3D numerical simulation using Maxwell Method (NMM3D) for soil surface for L-, C-, X-, and Ku-bands. (2) Multiple scattering models with cyclical correction for vegetated surface with large optical thickness and albedo (3) Simulate polarimetry SAR speckle using fully numerical method (NMM3D) for soil surface. (4) Combine NMM3D with distorted Born approximation to give coherency matrix for both polarimetry SAR and interferometric SAR. The thesis starts with the large scale computation of rough surface for the interests increased in the earth observation using high frequency bands. The numerical methods shows the merits not only agrees the co- and cross-polarization backscatters but also acts as polarimetric SAR simulator to capture the radar speckle characteristics from soil. For the vegetation model, first we introduce the multiple scattering model which uses the scattering path to account trace the multiple scattering effects. Cyclical correction is applied to account the coherent addition at backscattering direction. Second, NMM3D is processed as coherency matrix and is combined with distorted Born approximation. The detail derivation is included for coherency matrix for both PolSAR and PolInSAR. Finally, cross-polarization and polarization ratio are validated with measurement data from SMAPVEX12 to address their importance for the focus of future study.