Scholarlink

<p>The Western Ghats of southwestern India experience frequent landslides during the monsoon season. The increasing extreme rainfall events associated with climate change and anthropogenic factors are the attributed causes of landslides in the region, which include major slides that occurred in 2018, 2019 and 2024. The present study employs an integrated approach and compares the Analytical Hierarchy Process (heuristic) and Logistic Regression Analysis (statistical machine learning method) to delineate landslide susceptibility zones in the Wayanad District of Kerala situated in the Western Ghats. The study aims to understand the role of various factors in conditioning and triggering the landslides in the region. Different sets of conditioning factors that fit the models are chosen and evaluated by area under the Receiver Operator Characteristic (ROC) curve method. Both the models yielded comparable results, based on which the study area is divided into five different classes, namely very high (9–13%), high (20–26%), moderate (31–33%), low (13–17%), very low (17–22%) in terms of landslide susceptibility. The generated susceptibility maps belong to&#xa0;the ‘very good’ category, with an accuracy assessment of 85% for AHP and 82% for&#xa0;the Logistic regression model. Both models can serve as tools to plan and adopt suitable strategies to reduce the vulnerability of landslides in the region.</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 Analysis of a Part of Western Ghats in South-Western India Using Geospatial Techniques: A Comparison of AHP and Logistic Regression Methods

  • 期刊论文

  • 发表日期：

    2025年6月3日

查看全文
  • Rashmi Chandan,
  • Kizhur Sandeep,
  • Chandan Kumar Boraiaha

摘要

The Western Ghats of southwestern India experience frequent landslides during the monsoon season. The increasing extreme rainfall events associated with climate change and anthropogenic factors are the attributed causes of landslides in the region, which include major slides that occurred in 2018, 2019 and 2024. The present study employs an integrated approach and compares the Analytical Hierarchy Process (heuristic) and Logistic Regression Analysis (statistical machine learning method) to delineate landslide susceptibility zones in the Wayanad District of Kerala situated in the Western Ghats. The study aims to understand the role of various factors in conditioning and triggering the landslides in the region. Different sets of conditioning factors that fit the models are chosen and evaluated by area under the Receiver Operator Characteristic (ROC) curve method. Both the models yielded comparable results, based on which the study area is divided into five different classes, namely very high (9–13%), high (20–26%), moderate (31–33%), low (13–17%), very low (17–22%) in terms of landslide susceptibility. The generated susceptibility maps belong to the ‘very good’ category, with an accuracy assessment of 85% for AHP and 82% for the Logistic regression model. Both models can serve as tools to plan and adopt suitable strategies to reduce the vulnerability of landslides in the region.