A multi-frame deformation velocity splicing method for wide-area InSAR measurement based on uncontrolled block adjustment: A case study of long-term deformation monitoring in Guangdong, China

Abstract

Wide-area deformation measurements require multi-frame InSAR datasets for joint monitoring due to the limited swath width of SAR images. However, variations in the stability of reference points, error signal magnitude and distribution, and spatio-temporal filtering parameters result in different deformation rates between adjacent frames, leading to centimeter-level errors in the results. To obtain spatially consistent wide-area results, here we propose a multi-frame deformation velocity splicing method that combines InSAR datasets and block adjustment. We convert the line-of-sight deformation of each frame to vertical using an improved vertical deformation solution method. We construct an uncontrolled block adjustment model and use the results in the automatically identified overlapping areas as redundant observations to solve the deviation of each frame in a wide area. Finally, the multi-frame seamless results are sampled into a uniform spatial resolution cell network, to generate a wide-area deformation product with consistent reference. We validate the proposed method in Guangdong Province, China, covering an area of 1.96 × 105 km2. By correcting and fusing a total of 78 ALOS-1/PALSAR frames (2007–2010) and 11 Sentinel-1 frames (2015–2021) separately, we obtain the wide-area deformation products for Guangdong during the two periods. Compared with the original results, the inner precision of the ALOS-1/PALSAR and Sentinel-1 data is improved by 82.4% and 73.7%, respectively. Compared with GNSS, the accuracy of the corrected Sentinel-1 results is improved by more than 19%. The proposed method contributes to standardizing InSAR deformation production and facilitating the engineering application of wide-area InSAR measurements.