近年来，京津冀地区的不均匀地表沉降日趋严重，对公路、铁路等基础设施安全造成严重威胁，已引起国内外广泛关注。合成孔径雷达时序差分干涉TS-DInSAR（Time Series Differential Interferometric Synthetic Aperture Radar）作为一种高效的广域形变测量手段，在地面沉降调查监测中已被广泛应用，但如何针对高现势性的影像数据高精准提取大范围区域地表形变是当前深化的热点。本文利用欧洲太空局发布的Sentinel-1A新型数据源，针对TOPS（Terrain Observation by Progressive Scans）模式影像间存在多普勒中心不一致问题，借助外部高精度POD（Precise Orbit Determination）轨道和DSM（Digital Surface Model）数据进行频移滤波迭代配准，并针对Sentinel-1A数据特征集成优化了基于点目标时序分析的大区域地表形变监测方法，以2015年—2016年期间的29景影像为实验数据开展了京津地区沉降监测研究，提取了北京东、廊坊及天津西等地区的沉降结果，并结合区域内典型区域人口密度、产业分布、地表覆盖和线路剖面等信息深入分析了沉降的时空分布特征和成因。结果表明，Sentinel-1A时序干涉结果在大范围地表沉降调查监测上具有可靠的应用精度。

In recent years, a differential subsidence of the Beijing-Tianjin-Hebei Plain has become an increasingly significant issue. The safety of the country's infrastructures, such as highways and railways, is seriously threatened, thereby arousing considerable attention at home and abroad. The time-series differential synthetic aperture radar interferometry (TS-DInSAR) has a wide-area deformation measurement relative to traditional measurement methods, such as GPS and leveling, and has been applied extensively. Therefore, the current research focuses on accurately extracting a large-scale surface deformation by using high-potential image data.
An integrated method is optimized in this research to apply Synthetic Aperture Radar interferometry (SAR) images of the Terrain Observation by Progressive Scan (TOPS) mode and accurately extract a large-scale surface deformation based on a novel Sentinel-1A source released by the ESA and time-series analysis of interferometric point targets (TSA-IPT). The present research uses an external high-precision digital surface model and POD and applies a spectral shift filtering iteration to suppress the phase jump caused by spectral decorrelation between bursts, thereby allowing the TOPS mode Sentinel-1A data to satisfy the requirements of registration accuracy of TS-DInSAR. Meanwhile, setting the regression analysis parameters and designing the processing flow are discussed using the TSA-IPT to determine the deformation of a large area.
Based on the method presented in this research, 29 Sentinel-1A images obtained from June 2015 to August 2016 are selected for processing. The subsidence result in East Beijing, Langfang, and West Tianjin is obtained. The maximum subsidence rate of the study area at Wang Qingtuo town in Tianjin is 224 mm/a, and its maximum subsidence is 265 mm. In the Hei Zhuanghu Town in Beijing, the maximum subsidence rate is 159 mm/a, and its maximum subsidence is 265 mm. In the Dong Fengwu Village in Langfang, the maximum subsidence rate is 161 mm/a, and its maximum subsidence is 191 mm. In the city district of Langfang in Yanjiao, the maximum subsidence rate is 108 mm/a, and its maximum subsidence is 120 mm. In the Yanjiao City in Hebei Province, the maximum subsidence rate is 104 mm/a, and its maximum subsidence is 121 mm.
The spatiotemporal distribution and cause of subsidence are thoroughly analyzed and combined with population density, industrial distribution, and surface coverage. A significant subsidence in the study area can be found on the basis of the monitoring results, which tend to concatenate. The main factors of subsidence are population density, development of industrial distribution parks, and cultivation of thirsty crops. The growth of thirsty crops is the main subsidence factor of a city suburb. Numerous differential settlement regions are found along a high-speed railway, which is in the boundary region of the city district and suburb. Results of this research show that using the TSA-IPT is effective for Sentinel-1A images. Therefore, Sentinel-1A images can be applied to monitor land subsidence over a large area in the future.