Landslides, mass movements and slope instability pose significant geo-environmental issues in upland ecosystems, particularly in regions prone to seismic activity like the Kashmir Himalaya. However, the collective impact of diverse topography, intricate lithological composition, torrential rainfall, intense seismic shaking and human activities like as extensive deforestation and road construction on frequent landslide vulnerabilities is not thoroughly addressed in the Kashmir Himalaya. Therefore, this study has proposed a new perspective to map landslide hazard, vulnerability and risk factors in Kashmir Himalaya. A landslide inventory has been prepared that randomly splits into training (80% for model preparation) and testing (20% for validation) datasets. Thereafter, landslide susceptibility zonation (LSZ) has been carried out through Random Forest (RF) technique integrating different causative factors such as, surface geology, lineament density, soil, geomorphology, rainfall, slope, aspect, drainage density, Normalize Differences Vegetation Index, Landuse/landcover (LULC), epicenter proximity and road density defining the proneness of the terrain undergoing slope failure in terms of ‘None,’ ‘Low,’ ‘Moderate,’ ‘High,’ ‘Very High’ and ‘Severe.’ The relative landslide density index (R-index) and receiver operating characteristics (ROC) demonstrated an 82% accuracy level, indicating a significant association between LSZ and the testing landslide inventory data. The findings reveal that some densely populated and urbanized areas, namely Katra, Udhampur, Banihal, Ramban, Poonch and the Jammu-Srinagar Highway, are situated within zones characterized by ‘High to Severe’ susceptibility to landslides, underscoring the imperative for comprehensive risk assessment. Consequently, the classification of landslide risk zones has been performed by integrating landslide susceptibility distribution with vulnerability thematic layers such as population density and the number of households, within a hierarchical framework, which delineates approximately 22% of the study region falling in ‘Very High to Severe’ risk zones. This study provides a detailed analysis of landslide hazard and risk of the Kashmir Himalaya region, an area prone to frequent landslides. It underscores the necessity for the development and implementation of a strategic plan aimed at effectively managing hazards and executing rehabilitation protocols.

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

Integrated Landslide Susceptibility Zonation and Socio-Economic Vulnerability Assessment in the Kashmir Himalaya Using Machine Learning and GIS: Toward Precise Hazard Management and Planning

  • S. K. Nath,
  • A. Sengupta

摘要

Landslides, mass movements and slope instability pose significant geo-environmental issues in upland ecosystems, particularly in regions prone to seismic activity like the Kashmir Himalaya. However, the collective impact of diverse topography, intricate lithological composition, torrential rainfall, intense seismic shaking and human activities like as extensive deforestation and road construction on frequent landslide vulnerabilities is not thoroughly addressed in the Kashmir Himalaya. Therefore, this study has proposed a new perspective to map landslide hazard, vulnerability and risk factors in Kashmir Himalaya. A landslide inventory has been prepared that randomly splits into training (80% for model preparation) and testing (20% for validation) datasets. Thereafter, landslide susceptibility zonation (LSZ) has been carried out through Random Forest (RF) technique integrating different causative factors such as, surface geology, lineament density, soil, geomorphology, rainfall, slope, aspect, drainage density, Normalize Differences Vegetation Index, Landuse/landcover (LULC), epicenter proximity and road density defining the proneness of the terrain undergoing slope failure in terms of ‘None,’ ‘Low,’ ‘Moderate,’ ‘High,’ ‘Very High’ and ‘Severe.’ The relative landslide density index (R-index) and receiver operating characteristics (ROC) demonstrated an 82% accuracy level, indicating a significant association between LSZ and the testing landslide inventory data. The findings reveal that some densely populated and urbanized areas, namely Katra, Udhampur, Banihal, Ramban, Poonch and the Jammu-Srinagar Highway, are situated within zones characterized by ‘High to Severe’ susceptibility to landslides, underscoring the imperative for comprehensive risk assessment. Consequently, the classification of landslide risk zones has been performed by integrating landslide susceptibility distribution with vulnerability thematic layers such as population density and the number of households, within a hierarchical framework, which delineates approximately 22% of the study region falling in ‘Very High to Severe’ risk zones. This study provides a detailed analysis of landslide hazard and risk of the Kashmir Himalaya region, an area prone to frequent landslides. It underscores the necessity for the development and implementation of a strategic plan aimed at effectively managing hazards and executing rehabilitation protocols.