Scholarlink

<p>The Heifangtai loess terrace in northwest China is prone to frequent loess landslides, driven by complex interacting factors. Accurate stability assessment of these landslides is crucial for mitigating and even avoiding landslide risks, yet challenges remain when using purely physical-based or data-driven techniques alone. Hence, this study proposes a new approach integrating physical and data-driven methods to evaluate the stability of the Dangchuan landslide group on the Heifangtai loess terrace. A detailed three-dimensional geological model of the landslide was first constructed using multiple data sources, including high-precision Global Navigation Satellite System (GNSS) data and high-resolution Digital Elevation Models (DEM), to accurately represent its natural characteristics. Then a coupled Finite Difference and Discrete Element (FD-DE) method was developed for landslide stability evaluation. High-precision, long-term GNSS displacement monitoring data points (HF06/07, HF09, HF05) surrounding the landslide were utilized to constrain the FD-DE model. Three actual landslides from the Dangchuan landslide group were adopted for model validation. The results demonstrate that the simulated instability sequence—initiation first at HF06/HF07, followed by HF09, and finally HF05—corresponds with predicted Factor of Safety (FoS) contours. This sequence was consistently observed in actual ground movements, substantiating the model findings. Consequently, the integration of landslide displacement data with a physical-driven model can yield accurate loess landslide stability assessment.</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

Integrating physical and data-driven approaches for evaluating the stability of landslides on the Heifangtai loess terrace

  • 期刊论文

  • 发表日期：

    2026年2月11日

查看全文
  • Qing Ling,
  • Junguang Ren,
  • Qin Zhang,
  • Wei Qu,
  • Jing Zhang,
  • Guangwei Jiang,
  • Jiebo Qu,
  • Qi Guo,
  • Lingjie Kong,
  • Yuming Wei

摘要

The Heifangtai loess terrace in northwest China is prone to frequent loess landslides, driven by complex interacting factors. Accurate stability assessment of these landslides is crucial for mitigating and even avoiding landslide risks, yet challenges remain when using purely physical-based or data-driven techniques alone. Hence, this study proposes a new approach integrating physical and data-driven methods to evaluate the stability of the Dangchuan landslide group on the Heifangtai loess terrace. A detailed three-dimensional geological model of the landslide was first constructed using multiple data sources, including high-precision Global Navigation Satellite System (GNSS) data and high-resolution Digital Elevation Models (DEM), to accurately represent its natural characteristics. Then a coupled Finite Difference and Discrete Element (FD-DE) method was developed for landslide stability evaluation. High-precision, long-term GNSS displacement monitoring data points (HF06/07, HF09, HF05) surrounding the landslide were utilized to constrain the FD-DE model. Three actual landslides from the Dangchuan landslide group were adopted for model validation. The results demonstrate that the simulated instability sequence—initiation first at HF06/HF07, followed by HF09, and finally HF05—corresponds with predicted Factor of Safety (FoS) contours. This sequence was consistently observed in actual ground movements, substantiating the model findings. Consequently, the integration of landslide displacement data with a physical-driven model can yield accurate loess landslide stability assessment.