DInSAR Monitoring of Surface Subsidence by Fusing Sentinel-1A and -1B Data to Improve Time Resolution in a Mining Area

Abstract

Monitoring large gradient ground deformation due to temporal and spatial image decoherence has long been a challenge. We attempted an improvement using Sentinel-1A and Sentinel-1B C-band data fusion methods based on the variation law of subsidence velocity of ground leveling monitoring points. This approach improved the temporal resolution from 12 to 6 days. Using a mine in Datong, Shanxi Province, China as an example, 13 scenes of Sentinel-1A data and 12 scenes of Sentinel-1B data were fused and compared with ground-measured data. The results obtained were closer to the measured values than those obtained by single data set (Sentinel-1A or -1B only). Simultaneously, 61 scenes of Sentinel-1A data and 12 scenes of Sentinel-1B data were used to calculate the subsidence of the mining area over two years. The subsidence map was consistent with the actual leveling trend, which reflected the dynamic change of surface subsidence range in the mining area. This study provides an approach using DInSAR technology to monitor large gradient deformation in mining areas and provides an effective method to monitor of surface dynamic deformation.