Ground-based synthetic aperture radar (GBSAR) interferometry for deformation monitoring

Abstract

Ground-based synthetic aperture radar (GBSAR), together with interferometry, represents a powerful tool for deformation monitoring. GBSAR has inherent flexibility, allowing data to be collected with adjustable temporal resolutions through either continuous or discontinuous mode. The goal of this research is to develop a framework to effectively utilise GBSAR for deformation monitoring in both modes, with the emphasis on accuracy, robustness, and real-time capability. To achieve this goal, advanced Interferometric SAR (InSAR) processing algorithms have been proposed to address existing issues in conventional interferometry for GBSAR deformation monitoring. The proposed interferometric algorithms include a new non-local method for the accurate estimation of coherence and interferometric phase, a new approach to selecting coherent pixels with the aim of maximising the density of selected pixels and optimizing the reliability of time series analysis, and a rigorous model for the correction of atmospheric and repositioning errors. On the basis of these algorithms, two complete interferometric processing chains have been developed: one for continuous and the other for discontinuous GBSAR deformation monitoring. The continuous chain is able to process infinite incoming images in real time and extract the evolution of surface movements through temporally coherent pixels. The discontinuous chain integrates additional automatic coregistration of images and correction of repositioning errors between different campaigns. Successful deformation monitoring applications have been completed, including three continuous (a dune, a bridge, and a coastal cliff) and one discontinuous (a hillside), which have demonstrated the feasibility and effectiveness of the presented algorithms and chains for high-accuracy GBSAR interferometric measurement. Significant deformation signals were detected from the three continuous applications and no deformation from the discontinuous. The achieved results are justified quantitatively via a defined precision indicator for the time series estimation and validated qualitatively via a priori knowledge of these observing sites.