Scholarlink

<p>The assessment of landslide susceptibility is one of the most important steps in disaster management and the implementation of land use planning and development measures. We propose an approach to modelling landslide susceptibility based on statistical methods, namely frequency ratio (FR), information value (IV), weighted factor (Wf), and weight of evidence (WoE), using GIS software for the Wilaya de Mila (northeastern Algeria). As the Wilaya of Mila is highly affected by landslides, the main objective and expected impact of this study is to provide local authorities and decision-makers in this Wilaya with a decision-making tool. Prior to modelling, a landslide inventory map was prepared based on satellite imagery from various sources, documentary analysis, and field surveys. In addition, several thematic maps (lithologies, slopes, exposures, precipitation, distances to roads, etc.) were created. The recorded landslides (429 bodies) were divided into two samples: 80% for model training and 20% for model validation using ROC–AUC curves. The main results of the susceptibility maps showed a better accuracy of the AUC curve for the WoE model compared to FR, VI, and Wf, with success rate values of 0.774, 0.767, 0.754, and 0.749, respectively, and predictive values of 0.808, 0.785, 0.782, and 0.770, respectively, implying better decision making based on the WoE results.</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

Statistical-based methods for landslides susceptibility mapping in the Wilaya of Mila (northeast Algeria)

  • 期刊论文

  • 发表日期：

    2025年1月9日

查看全文
  • Serkhane Ahmed,
  • Oukid Fatma,
  • Saidi Yacine,
  • Mahdadi Fatna,
  • Bouhadad Youcef,
  • Guettouche Mohammed Said

摘要

The assessment of landslide susceptibility is one of the most important steps in disaster management and the implementation of land use planning and development measures. We propose an approach to modelling landslide susceptibility based on statistical methods, namely frequency ratio (FR), information value (IV), weighted factor (Wf), and weight of evidence (WoE), using GIS software for the Wilaya de Mila (northeastern Algeria). As the Wilaya of Mila is highly affected by landslides, the main objective and expected impact of this study is to provide local authorities and decision-makers in this Wilaya with a decision-making tool. Prior to modelling, a landslide inventory map was prepared based on satellite imagery from various sources, documentary analysis, and field surveys. In addition, several thematic maps (lithologies, slopes, exposures, precipitation, distances to roads, etc.) were created. The recorded landslides (429 bodies) were divided into two samples: 80% for model training and 20% for model validation using ROC–AUC curves. The main results of the susceptibility maps showed a better accuracy of the AUC curve for the WoE model compared to FR, VI, and Wf, with success rate values of 0.774, 0.767, 0.754, and 0.749, respectively, and predictive values of 0.808, 0.785, 0.782, and 0.770, respectively, implying better decision making based on the WoE results.