Scholarlink

<p>Landslides pose a geotechnical risk to the Carajás Railway (EFC) in the state of Maranhão, Brazil, requiring the identification and quantification of conditioning factors for effective risk management. However, limited understanding of the relative influence of these factors hinders the implementation of suitable preventive measures. This study integrates the Weight of Evidence (WoE) method with Interferometric Synthetic Aperture Radar (InSAR) data to model landslide susceptibility along a 41&#xa0;km segment of the EFC. Eleven conditioning factors were analyzed, including slope, elevation, terrain curvature, Topographic Position Index (TPI), soil and lithology classes, proximity to geological faults and rivers, as well as land use and land cover. The WoE model was used to compute the Landslide Susceptibility Index (LSI), which was validated using Receiver Operating Characteristic (ROC) curves, yielding an area under the curve (AUC) of 84.6%. Results indicate that 10.57% of the area has very high susceptibility, encompassing 79.41% of recorded landslides. The integration of InSAR data revealed a strong correlation between highly susceptible zones and ground deformation patterns, with subsidence rates reaching − 47.3&#xa0;mm/year and uplift up to 30.9&#xa0;mm/year. The most influential conditioning factors were high terrain roughness (Wf: 3.73), proximity to geological faults (&lt; 1000&#xa0;m, Wf: 3.67), and steep slopes (&gt; 25°, Wf: 2.78). The quantitative assessment of these parameters was essential for predictive landslide susceptibility modeling, enabling optimized strategic planning and efficient railway management. By combining statistical modeling with InSAR monitoring, this study provides a robust framework to support proactive landslide susceptibility management in transportation infrastructures.</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

Landslide Susceptibility Modeling in a Railway in the Amazon Region: Integration of the Weight of Evidence (WOE) Method and InSAR Monitoring Data

  • 期刊论文

  • 发表日期：

    2025年9月19日

查看全文
  • Felipe Pacheco Silva,
  • Luiz Felipe Goulart Fiscina,
  • Felipe Santos de Almeida,
  • Marcos Timóteo Rodrigues de Sousa,
  • Winicius Brito Cordeiro,
  • Gabriela Vitelli,
  • Marcos Massao Futai

摘要

Landslides pose a geotechnical risk to the Carajás Railway (EFC) in the state of Maranhão, Brazil, requiring the identification and quantification of conditioning factors for effective risk management. However, limited understanding of the relative influence of these factors hinders the implementation of suitable preventive measures. This study integrates the Weight of Evidence (WoE) method with Interferometric Synthetic Aperture Radar (InSAR) data to model landslide susceptibility along a 41 km segment of the EFC. Eleven conditioning factors were analyzed, including slope, elevation, terrain curvature, Topographic Position Index (TPI), soil and lithology classes, proximity to geological faults and rivers, as well as land use and land cover. The WoE model was used to compute the Landslide Susceptibility Index (LSI), which was validated using Receiver Operating Characteristic (ROC) curves, yielding an area under the curve (AUC) of 84.6%. Results indicate that 10.57% of the area has very high susceptibility, encompassing 79.41% of recorded landslides. The integration of InSAR data revealed a strong correlation between highly susceptible zones and ground deformation patterns, with subsidence rates reaching − 47.3 mm/year and uplift up to 30.9 mm/year. The most influential conditioning factors were high terrain roughness (Wf: 3.73), proximity to geological faults (< 1000 m, Wf: 3.67), and steep slopes (> 25°, Wf: 2.78). The quantitative assessment of these parameters was essential for predictive landslide susceptibility modeling, enabling optimized strategic planning and efficient railway management. By combining statistical modeling with InSAR monitoring, this study provides a robust framework to support proactive landslide susceptibility management in transportation infrastructures.