Polarization-Optimized D-TomoSAR for Urban Building Monitoring With PAZ Data
Abstract
Differential tomographic synthetic aperture radar (D-TomoSAR) has emerged as a powerful technique for 3-D reconstruction and deformation monitoring of urban buildings. However, most existing spaceborne SAR tomography studies have only used single-polarization channels, leaving the advantages of multipolarization information inherent to dual-polarimetric data insufficiently exploited. To address this limitation, this letter proposes a polarimetric phase-optimized D-TomoSAR method for building reconstruction and deformation monitoring, utilizing dual-polarimetric spaceborne SAR data acquired by the PAZ satellite. The proposed approach employs the exhaustive search polarization optimization (ESPO) algorithm to enhance polarimetric signal quality and integrates a compressive sensing (CS) framework based on orthogonal matching pursuit (OMP) for the joint estimation of 3-D scatterer elevations and temporal deformation rates. Experimental results demonstrate that the proposed method significantly improves monitoring point density, point cloud reconstruction quality, and noise suppression compared with conventional single-polarization approaches, validating the applicability and practical potential of PAZ dual-polarimetric SAR (PolSAR) data for D-TomoSAR-based urban infrastructure monitoring.