Investigating land subsidence and its causes along Beijing high-speed railway using multi-platform InSAR and a maximum entropy model

• Published in: International journal of applied earth observation and geoinformation 2021-04, Vol.96, p.102284, Article 102284
• Main Authors: Chen, Beibei, Gong, Huili, Chen, Yun, Lei, Kunchao, Zhou, Chaofan, Si, Yuan, Li, Xiaojuan, Pan, Yun, Gao, Mingliang
• Format: Article
• Language: English
• Subjects: High Speed Railway; Interferometric Synthetic Aperture Radar (InSAR); Land subsidence; Maximum entropy model; Time series fusion method
• ISSN: 1569-8432; 1872-826X
• DOI: 10.1016/j.jag.2020.102284

•A new time-series fusion method to blend SAR data from multi-platforms.•Adaption of the new fusion method in detecting subsidence impacting the BHSR.•Identification of the most severe subsidence sections along the BHSR.•Relationship between subsidence and its causes using a maximum entropy model. Beijing-Tianjin High Speed Railway is the first high-speed railway in China. The Beijing section of it runs through areas affected by subsidence which threaten its safe operation. This study develops a new time series fusion method based on the minimum gradient difference of a fitting curve to produce time series subsidence along this section. Through blending Envisat ASAR and TerrSAR-X time series, the InSAR-derived subsidence and its spatial–temporal development was analyzed along the railway. The relationship between subsidence and its causes was then explored using a maximum entropy model. The study reveals that: (1) The subsidence dynamics identified using the new fusion method agrees with the ground deformation measurements; (2) The sections of most severe subsidence occur between kilometer point KP 11 and KP 21; and (3) The main hydrogeological factors affecting subsidence are the compressible deposit thickness and the groundwater level in the second confined aquifer. The new fusion method proposed improves the accuracy and reliability of subsidence time series. It extends the time span of subsidence monitoring. The approach is mainly applicable to areas with significant vertical deformation, and is particularly suitable for integrating multi-platform data with overlapping in time or with a short time gap.