Scholarlink

<p>Conventional landslide susceptibility assessment (LSA) methods fail to incorporate the variability of ground deformation characteristics, which limits their applicability in dynamically active areas. To address this limitation, this study proposes an optimized LSA framework that integrates Multi-Temporal Interferometric Synthetic Aperture Radar (MT-InSAR) with Boosting ensemble framework for clean energy transmission corridors (CETC) in China's Loess Plateau. First, an initial susceptibility model was constructed using a Boosting ensemble framework, incorporating comprehensive landslide inventory data and 12 influencing factors. Furthermore, the MT-InSAR technology and the K-Means clustering algorithm were employed to derive long-term surface deformation patterns for the period 2020–2024. Finally, the initial susceptibility assessment was optimized by combining deformation zoning through a slope-unit-based approach, generating the final landslide susceptibility map. The results demonstrate that the Categorical Boosting (CatBoost) model outperforms (AUC = 0.914) other methods within the Boosting ensemble framework. MT-InSAR analysis reveals a maximum deformation rate of 77&#xa0;mm/year and a cumulative displacement of 373&#xa0;mm in the study area. Time-series deformation clustering further indicates that Type 2 deformation patterns dominate the region. The enhancement matrix, which incorporated time-series deformation clustering, revised the initial assessment by reclassifying slope units from "very high" susceptibility, resulting in a net decrease from 1,496 to 394 units (a reduction of 1,102). This study optimizes the traditional landslide susceptibility model by incorporating varied surface deformation trends to address the risk overestimation of static models and support more precise hazard mitigation along the CETC.</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

Optimizing landslide susceptibility zoning for clean energy transmission corridors in loess regions using MT-InSAR and boosting ensemble framework

  • 期刊论文

  • 发表日期：

    2026年2月28日

查看全文
  • Bijing Jin,
  • Lianglong Zhang,
  • Xiaobo Liu,
  • Lei Gui,
  • Xinmin Li,
  • Binbin Zhao,
  • Yi Liu,
  • Aizhen Yang,
  • Jinhang Hu,
  • Kunlong Yin

摘要

Conventional landslide susceptibility assessment (LSA) methods fail to incorporate the variability of ground deformation characteristics, which limits their applicability in dynamically active areas. To address this limitation, this study proposes an optimized LSA framework that integrates Multi-Temporal Interferometric Synthetic Aperture Radar (MT-InSAR) with Boosting ensemble framework for clean energy transmission corridors (CETC) in China's Loess Plateau. First, an initial susceptibility model was constructed using a Boosting ensemble framework, incorporating comprehensive landslide inventory data and 12 influencing factors. Furthermore, the MT-InSAR technology and the K-Means clustering algorithm were employed to derive long-term surface deformation patterns for the period 2020–2024. Finally, the initial susceptibility assessment was optimized by combining deformation zoning through a slope-unit-based approach, generating the final landslide susceptibility map. The results demonstrate that the Categorical Boosting (CatBoost) model outperforms (AUC = 0.914) other methods within the Boosting ensemble framework. MT-InSAR analysis reveals a maximum deformation rate of 77 mm/year and a cumulative displacement of 373 mm in the study area. Time-series deformation clustering further indicates that Type 2 deformation patterns dominate the region. The enhancement matrix, which incorporated time-series deformation clustering, revised the initial assessment by reclassifying slope units from "very high" susceptibility, resulting in a net decrease from 1,496 to 394 units (a reduction of 1,102). This study optimizes the traditional landslide susceptibility model by incorporating varied surface deformation trends to address the risk overestimation of static models and support more precise hazard mitigation along the CETC.