Improving the Signal-to-Noise Ratio for Polarimetric and Interferometric SAR

This thesis develops new methods to improve the accuracy and reliability of polarimetric scattering parameters and interferometric synthetic aperture radar (InSAR) measurements of surface deformation. Three methodological advances are presented to address orientation-induced ambiguity in polarimetric SAR (PolSAR), atmospheric phase artefacts in InSAR, and phase-unwrapping errors in sparse persistent scatterer InSAR (PS-InSAR) results.

For PolSAR, a novel phenomenological de-orientation method is developed using Huynen's target decomposition theory and polarisation nulls to identify and independently de-orient orientation-perturbed scattering components. The method reduces orientation-induced ambiguity and improves the physical consistency of model-based target decompositions, particularly in built-up areas.

For InSAR, new tropospheric and ionospheric correction methods are developed to reduce atmospheric phase artefacts. Applied to hundreds of Sentinel-1 interferograms, the proposed methods provide greater reductions in interferometric phase variability than established approaches, with improvements of approximately 10.5\% and 12.7\% relative to D-LOS and GACOS, respectively. The corrected measurements also reduce the average GNSS--InSAR RMSE by approximately 28\%, compared with approximately 22\% for existing corrections.

For PS-InSAR in the Antarctic Peninsula, a range-constrained approach is developed to mitigate phase-unwrapping errors between sparse clusters of persistent scatterers. Incorporating sub-pixel SAR range measurements improves the internal consistency of the derived deformation velocity field in an environment characterised by spatially isolated coherent targets.

Together, these contributions demonstrate that exploiting the physical characteristics, external datasets and complementary information content of SAR measurements can reduce key sources of uncertainty in SAR-derived measurements. The developed methods improve the reliability of polarimetric parameter retrieval and InSAR deformation measurements, extending the applicability of SAR for geophysical monitoring in challenging environments.