Scholarlink

<p>Soil erosion, a predominant form of land degradation, exacerbated by human activities and the accelerating impacts of climate change, poses a significant environmental challenge across the diverse landscapes of Bangladesh. The growing threat calls for enhanced erosion prediction. This study presents a novel, climate-informed soil erosion prediction framework that integrates both current environmental conditions and future climate dynamics. The scientific contribution of this work is threefold: (i) it incorporates Coupled Model Intercomparison Project Phase 6 (CMIP6)-based future rainfall, temperature, and land-use projections under two Shared Socioeconomic Pathways SSP1–2.6 and SSP5–8.5 for the periods 2021–2040 and 2041–2060, marking one of the first forward-looking soil erosion assessments in Bangladesh; (ii) it systematically compares knowledge-driven (Fuzzy-AHP) and data-driven ensemble machine learning techniques including AdaBoost and LightGBM within a unified modeling framework to evaluate methodological strengths and limitations; (iii) it employs a comprehensive set of eighteen causative variables encompassing topographic, hydrological, soil, geological, and climate-projected factors, thereby establishing a holistic and robust erosion prediction framework. Applied to the erosion-prone region of Chattogram, the ensemble ML models achieved accuracies exceeding 80%, outperforming the Fuzzy-AHP approach and underscoring the advantages of data-driven methods. Spatial projections indicate an increase of approximately ~ 8–14% in areas classified as highly to very highly erosion-prone under future climate conditions. These findings highlight the critical importance of integrating climate projections into erosion modeling and offer a robust, evidence-based tool to support sustainable land use planning and soil conservation policy in Bangladesh.</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

Predicting Climate-Driven Soil Erosion in Bangladesh Using Fuzzy-AHP and Ensemble Machine Learning

  • 期刊论文

  • 发表日期：

    2026年3月24日

查看全文
  • Monira Jahan Tania,
  • Sanzida Murshed,
  • Musabbir Ahmed Arrafi,
  • Amy L. Griffin,
  • Mehedi Hasan Peas,
  • Mansour Almazroui,
  • Mahesh Bade,
  • Md. Ashraful Islam

摘要

Soil erosion, a predominant form of land degradation, exacerbated by human activities and the accelerating impacts of climate change, poses a significant environmental challenge across the diverse landscapes of Bangladesh. The growing threat calls for enhanced erosion prediction. This study presents a novel, climate-informed soil erosion prediction framework that integrates both current environmental conditions and future climate dynamics. The scientific contribution of this work is threefold: (i) it incorporates Coupled Model Intercomparison Project Phase 6 (CMIP6)-based future rainfall, temperature, and land-use projections under two Shared Socioeconomic Pathways SSP1–2.6 and SSP5–8.5 for the periods 2021–2040 and 2041–2060, marking one of the first forward-looking soil erosion assessments in Bangladesh; (ii) it systematically compares knowledge-driven (Fuzzy-AHP) and data-driven ensemble machine learning techniques including AdaBoost and LightGBM within a unified modeling framework to evaluate methodological strengths and limitations; (iii) it employs a comprehensive set of eighteen causative variables encompassing topographic, hydrological, soil, geological, and climate-projected factors, thereby establishing a holistic and robust erosion prediction framework. Applied to the erosion-prone region of Chattogram, the ensemble ML models achieved accuracies exceeding 80%, outperforming the Fuzzy-AHP approach and underscoring the advantages of data-driven methods. Spatial projections indicate an increase of approximately ~ 8–14% in areas classified as highly to very highly erosion-prone under future climate conditions. These findings highlight the critical importance of integrating climate projections into erosion modeling and offer a robust, evidence-based tool to support sustainable land use planning and soil conservation policy in Bangladesh.