Scholarlink

<p>High-altitude ecosystems face growing threats from natural hazards and human activities, intensifying socio-economic and environmental risks. The Nilgiris District, Tamil Nadu, is a hotspot where steep terrain, fragile ecosystems, climate variability, and anthropogenic pressures converge. This study integrates geospatial technologies and machine learning (XGBoost) to map multi-hazard risk zones and assess their implications for ecosystem stability and contaminant susceptibility. The GIS-based multi-moderated evaluation was applied using slope, elevation, land use/land cover (LULC), drainage density and proximity to roads and settlements. Improved the accuracy of the XGBoost classification by capturing complex spatial relationships. The multi-hazard risk zone map identified five classes with very low-risk zones with high-risk zones cantered near Coonoor and Kotagiri, which are associated for landslides and contaminated mobility, while in the lower-way areas cluster around AU. Combining crisis weakness with environmental fragility, this structure supports durable land-use management, ecosystem conservation and reducing pollution. The integration of geospatial analytics and machine learning provides a strong tool for disaster preparedness, risk reducing and elasticity in ecological sensitive hill districts.</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 machine learning and geospatial approaches for multi-hazard vulnerability mapping: implications for environmental health and contaminant risk in fragile ecosystems

  • 期刊论文

  • 发表日期：

    2025年10月28日

查看全文
  • Ayinshaer Kuannaxiaer,
  • Jiacheng Liu,
  • N. Kasthuri

摘要

High-altitude ecosystems face growing threats from natural hazards and human activities, intensifying socio-economic and environmental risks. The Nilgiris District, Tamil Nadu, is a hotspot where steep terrain, fragile ecosystems, climate variability, and anthropogenic pressures converge. This study integrates geospatial technologies and machine learning (XGBoost) to map multi-hazard risk zones and assess their implications for ecosystem stability and contaminant susceptibility. The GIS-based multi-moderated evaluation was applied using slope, elevation, land use/land cover (LULC), drainage density and proximity to roads and settlements. Improved the accuracy of the XGBoost classification by capturing complex spatial relationships. The multi-hazard risk zone map identified five classes with very low-risk zones with high-risk zones cantered near Coonoor and Kotagiri, which are associated for landslides and contaminated mobility, while in the lower-way areas cluster around AU. Combining crisis weakness with environmental fragility, this structure supports durable land-use management, ecosystem conservation and reducing pollution. The integration of geospatial analytics and machine learning provides a strong tool for disaster preparedness, risk reducing and elasticity in ecological sensitive hill districts.