Comparative study of dual-polarized Sentinel-1 data in urban subsidence monitoring

MENG Ran; JIANG Yanan; LIAO Lu; XU Qiang; LI Weile; LUO Huiyuan

doi:10.16186/j.cnki.1673-9787.2022040048

Abstract：Objectives To address the problem of the ground subsidence only considering the VV polariza-tion Sentinel-1 data and ignoring the VH polarization data,Methods in this paper,PS-InSAR and SBAS-InSAR techniques were used to process the 84-view dual-polarization Sentinel-1 data in Yan'an New Area from August 2018 to May 2021,obtain the ground settlement information,compare the results of the settle-ment monitoring of the two polarizations,explore the significance of the VH polarization in different experi-mental methods,and put forward a new dual-polarized time-series InSAR fusion monitoring method.Results The results show that most of the new area is relatively stable with deformation rate of-8～8 mm/a;in addi-tion,three larger subsidence areas are detected,located in Qiaoergou Basin,Gaojiagou Basin,and the north-eastern filling area of the new area.The monitoring results of Sentinel-1 with different polarization data have high correlation and consistent deformation trends.The overall VV polarization data is better than the VH po-larization,but in some areas,the VH polarization data can get more detailed surface deformation.VH polar-ization data in some areas can get more detailed surface deformation information.In the VV-VH polarization PS-InSAR deformation monitoring,the VH polarization data can effectively supplement the insufficiency of the VV polarization data monitoring results,making the results more intensive,and therefore reflecting a more detailed inversion of surface deformation compared with the results of a single polarization monitoring.The VV-VH polarization SBAS-InSAR deformation rate results have higher coherence and stability com-pared to a single polarization data.Conclusions The time-series InSAR fusion monitoring technique based on dual-polarized Sentinel-1 data can obtain better urban surface deformation monitoring results than single-polarized data.