Monitoring Ground Deformation In Urban Major Project Area With High-Resolution Persistent Scatterer SAR Interferometry

Large-scale constructions during the rapid development of city aggravate the local land subsidence to a certain degree, which affects the important urban infrastructures maintaining the basic function of the city, e.g., high buildings, rail transport, viaducts. In order to mitigate and reduce this effect, a scientific, efficient and comprehensive monitoring technology is needed. Differential synthetic aperture radar interferometry (D-InSAR) technology is a common technique for the extraction of ground subsidence information. Specifically, time series D-InSAR techniques, represented by the Permanent Scatterer Interferometry technique (PSI), have been proposed to resolve the problem of loss coherence of temporal and spatial and atmospheric retardation, which affect the conventional D-InSAR technique. With the launch of new high-resolution radar satellite with short revisit time and high spatial resolution, high resolution PSI technique offers a prompt and accurate method for monitoring of urban infrastructures deformation. However, the major problem for this technique is that the high phase gradient between the high-rise buildings and the background easily results in phase unwrapping errors lacking of urban Digital Surface Model (DSM), especially for the long baseline. In order to solve this problem, it is desirable to propose new methods to reduce phase discontinuous caused by phase unwrapping error in steep area. Furthermore, there are some technical obstacles in geocoding PS from the radar slant-range coordinate to the optical image, and using traditional two-dimensional expression method affects the extraction of deformation information of PS on single building. Therefore it has a practical significance to do the research on geocoding and three-dimensional visualization of PS.This thesis aims to provide new solutions to the above issues to improve the applicability and feasibility of high resolution PSI for ground deformation monitoring in urban major project Area. Major contributions of this thesis are as follows.(1) An iterative combination of short and long baseline PSI method is developed. In contrast to traditional PSI approach, the new approach can significantly reduce the impact of high phase gradient on the precision of building elevation and deformation monitoring, which indicates that the new approach can be used to monitor the ground deformation of dense urban built-up areas in large region.(2) A separation of the building from the background PSI method is proposed for the suppression of the high phase gradient between building and background. An evaluation based on comparison with the iterative combination of short and long baseline PSI method shows that the new technique can dectect the subtle deformation characteristics of buildings which can’t be detected by the former approach. But limited by the separation techniques, this approach is more suitable for single building deformation monitoring.(3) A PS geocoding technique is applied, which incorporates the Range-Doppler model and reference optical image. Experiments with high resolution SAR image prove the registration accuracy within a pixel. A PS elevation recovery algorithm is used to show three-dimensional information of the building in ArcGlobe GIS platform. The experimental results for the above method indicate that the advantage is more obvious compared with traditional two-dimensional expression method.(4) An iterative combination PSI method is conducted over central city of Shanghai. The results show a good agreement with the ground-based leveling measurements. A comparision between the subsidence measurements derived by the CSK PSI solution and those derived by the ENVISAT PSI solution is carried out, which indicate that the above results can both reveal the similar subsidence characteristics of the research area, but the PS density of CSK results is higher than ENVISAT results and CSK results can detect the maximum subsidence value of subsidence center more precisely than ENVISAT results. The reasons for the land subsidence of the research area are discussed and results show that load increase brought by the urban construction and construction activities aggravates the local land subsidence.(5) A separation of the building from the background PSI method is applied to monitor the deformation of3high-rise buildings along Shanghai Metro Line10. The four-dimentional results indicate that high-rise buildings are affected by subway excavation less than low-rise buildings.In summary, the main innovations of this paper are:(1) the use of iterative combination of short and long baselines PSI method to detect land subsidence in dense urban built-up area;(2) the use of separation of the building from the background PSI method for monitoring the single building deformation;(3) the application of PS geocoding technique incorporating Range-Doppler model and reference optical image to geocode PS from the radar slant-range coordinate to the optical image;(4) the PS elevation recovery algorithm to reaveal four-dimensional information of the building in ArcGlobe GIS platform.