Monitoring Land Subsidence by Using Multi-Temporal Differential SAR Interferometry: a Use Case in Jiaxing, China

摘要

Decorrelation caused by temporal change and SAR imaging geometry prevents conventional SAR interferometry from monitoring land surface subsidence precisely. In addition, conventional methods don't deal with phase delay which caused by atmosphere. It negatively influences the accuracy of Differential Synthetic Aperture Radar Interferometry (D-InSAR) for small-scale deformation monitoring. In this paper, a multi-temporal D-InSAR method was used to extract the surface deformation of Jiaxing, China, automatically. At first, terrain phase has been subtracted from multi-temporal interferograms by using Digital Elevation Model (DEM) data. Then, the amplitude dispersion and phase stability analysis were combined to select stable high coherent pixels, which connected by Delaunay triangulation to form a spatial network, as reliable unwrap object. After spatial and temporal correlation analysis, the terrain error is corrected and the atmospheric delay phase and residual noise phase are removed. At last, the surface deformation map stack was created to generate land subsidence velocity map. The result has been proved to be reliable by comparing the ground monitoring data.