Mosaicking Large-Scale InSAR Deformation Velocity Over Sparse GNSS Scenarios

SAR interferometry is commonly employed to map large-scale surface deformation. In applications spanning thousands of kilometers spatially, mosaicking multiple deformation velocity fields into a seamless entity is a critical process. This article analyzes the limitations of the existing mosaicking model, in particular the ill-conditioned problem posed by polynomial models over sparse Global Navigation Satellite System (GNSS) constraints. To resolve this issue and further extend InSAR measurement to regional and global scales, we propose a more robust solution within the framework of block adjustment, which effectively imposes constraints on the entire system even in the presence of limited GNSS data. Comprehensive testing on Sentinel-1 datasets over the North China Plain validates the efficiency of our approach in deriving a spatially continuous deformation field. Compared to traditional methods, our approach delivers a ~50% improvement in inner frame alignment. These advancements enable us to capture more accurate and continuous deformation features over wide areas, providing valuable insights for advancing large-scale deformation mapping within geodetic studies.