Unprecedented urban development poses significant subsidence challenges, primarily attributed to underground construction activities. Conventional approaches for monitoring these underground operations can be cumbersome. Therefore, time-series InSAR techniques emerges as a promising approach to quantify these deformations with high spatio-temporal details. However, distinguishing subsidence from city-scale deformation patterns often presents difficulties. Additionally, most previous studies concentrated on vertical deformation neglecting horizontal motions, thus hinder the comprehensive understanding of deformation dynamics. Lastly, the monitoring of deformations during ongoing subway construction has rarely been conducted. Therefore, this study employed PS-InSAR techniques to integrate Sentinel-1A ascending and descending datasets over the period from 2018 to 2023, with the aim of comprehensively extracting 3D deformation patterns, incorporating both vertical and horizontal components. Automatic methodologies were employed to distinguish the deformation caused by groundwater fluctuations and subway construction. The robustness of the results was statistically validated using geodetic leveling datasets. The results of the case study revealed vertical and horizontal deformations, with the vertical component being more prominent. Within the vertical deformation zones, localized horizontal deformations were evident due to their abrupt directional shift, albeit with a lesser magnitude. Persistent uplift was identified through East-North-Up (ENU)-derived vertical deformation at the rates of 8.04 and 6.85 mm/yr, respectively, in the test area. Long-term groundwater rebounds were observed, corresponding to a 1 mm uplift for every 1 m rise. The maximum cumulative subsidence recorded was 1 mm at the G07 MRT station, well within safe allowable limits. The findings are grounded in methodological advancements that hold the potential for comprehending deformation patterns on a global scale. Consequently, a much comprehensive and efficient monitoring of the ground deformation of a subway construction project can be achieved by employing the proposed approach.

中文翻译：

---

使用时间序列 InSAR 进行地铁建设过程中大都市区的三维变形分析


前所未有的城市发展带来了重大的沉降挑战，这主要归因于地下施工活动。监测这些地下作业的传统方法可能很麻烦。因此，时间序列 InSAR 技术成为一种很有前途的方法，可以量化具有高度时空细节的这些变形。然而，区分沉降与城市规模的变形模式往往存在困难。此外，以往的研究大多集中在垂直变形上，而忽略了水平运动，从而阻碍了对变形动力学的全面理解。最后，在正在进行的地铁建设过程中，很少进行变形监测。因此，本研究采用 PS-InSAR 技术在 2018 年至 2023 年期间整合了 Sentinel-1A 升序和降序数据集，旨在全面提取三维变形模式，包括垂直和水平分量。采用自动方法来区分地下水波动和地铁建设引起的变形。使用大地测量水准数据集对结果的稳健性进行了统计验证。案例研究的结果显示垂直和水平变形，其中垂直分量更为突出。在垂直变形区内，由于它们的突然方向偏移，局部水平变形很明显，尽管幅度较小。在测试区域内，通过东-北-上 （ENU） 衍生的垂直变形分别以 8.04 和 6.85 mm/yr 的速率识别出持续隆起。观察到地下水的长期回弹，相当于每上升 1 m 就会上升 1 mm。 G07 地铁站记录的最大累积沉降量为 1 毫米，远在安全允许的范围内。这些发现以方法论的进步为基础，这些进步具有在全球范围内理解变形模式的潜力。因此，采用所提出的方法可以实现对地铁建设项目的地面变形进行非常全面和有效的监测。