Scholarlink

<p>Time-series Interferometric Synthetic Aperture Radar (TS-InSAR) enables precise, wide-area ground deformation monitoring but suffers from decorrelation and heavy computation with large archives of satellite imagery. To address these challenges, this study applies temporal dimension image compression to 199 Sentinel-1&#xa0;A scenes (March 25, 2017–May 11, 2024) covering the Jinchuan mining area, China. Specifically, through the construction of a covariance matrix, PL (Phase-Linking) for phase compensation, and dimensionality reduction and reconstruction processes, the time-series image datasets are compressed into 22 virtual images. These virtual images are then processed within the Persistent Scatterer Interferometry (PS-InSAR) framework, referred to as mini stack technology. Results show that (1) the time-series image compression mini stack technology significantly enhances computational efficiency compared to traditional time-series InSAR (TS-InSAR) methods, relieving the decorrelation issue caused by long time spans in conventional interferograms; (2) The average coherence coefficient obtained from the virtual image stack improved by 32.8%, and the sum of phase differences (SPD) decreased by 19.2% compared to the original image after full-group interferometric processing. Furthermore, the monitoring points (MPs) density extracted by mini stack technology in the deformation zone of the mining area increased by over 32 times more than the PS-InSAR method. Additionally, spatial deformation patterns derived from both approaches are consistent, with a Pearson correlation coefficient of 0.89, a mean deformation rate difference of 0.01&#xa0;mm/yr, and a standard deviation of 0.68&#xa0;mm/yr. These findings confirm that mini stack technology reliably captures ground deformation while enhancing processing efficiency and data quality, making it suitable for broader applications in long-term deformation monitoring.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Times series InSAR deformation monitoring of Jinchuan mining area based on mini stack technology

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月16日

查看全文
  • Jie Guo,
  • Gonghai Zhang,
  • Yewei Song,
  • Yuxing Bai,
  • Jie Wang,
  • Wenhao Wu,
  • Zixuan Ge

摘要

Time-series Interferometric Synthetic Aperture Radar (TS-InSAR) enables precise, wide-area ground deformation monitoring but suffers from decorrelation and heavy computation with large archives of satellite imagery. To address these challenges, this study applies temporal dimension image compression to 199 Sentinel-1 A scenes (March 25, 2017–May 11, 2024) covering the Jinchuan mining area, China. Specifically, through the construction of a covariance matrix, PL (Phase-Linking) for phase compensation, and dimensionality reduction and reconstruction processes, the time-series image datasets are compressed into 22 virtual images. These virtual images are then processed within the Persistent Scatterer Interferometry (PS-InSAR) framework, referred to as mini stack technology. Results show that (1) the time-series image compression mini stack technology significantly enhances computational efficiency compared to traditional time-series InSAR (TS-InSAR) methods, relieving the decorrelation issue caused by long time spans in conventional interferograms; (2) The average coherence coefficient obtained from the virtual image stack improved by 32.8%, and the sum of phase differences (SPD) decreased by 19.2% compared to the original image after full-group interferometric processing. Furthermore, the monitoring points (MPs) density extracted by mini stack technology in the deformation zone of the mining area increased by over 32 times more than the PS-InSAR method. Additionally, spatial deformation patterns derived from both approaches are consistent, with a Pearson correlation coefficient of 0.89, a mean deformation rate difference of 0.01 mm/yr, and a standard deviation of 0.68 mm/yr. These findings confirm that mini stack technology reliably captures ground deformation while enhancing processing efficiency and data quality, making it suitable for broader applications in long-term deformation monitoring.