Scholarlink

Abstract <p>The Manimala River Basin (MRB) and its hinterland were one of the severely battered river basins in Kerala during the 2018 floods. This study aims to demarcate the flood-risk zones in the MRB through five different models: Analytic Hierarchy Process (AHP), Fuzzy-AHP (F-AHP), Frequency Ratio (FR), Statistical Index (SI), and Height Above Nearest Drainage (HAND). Furthermore, this modelling also aims to suggest specific recommendations. A total of 10 conditioning factors (CFs), namely, geomorphology, slope, soil texture, land use and land cover (LULC), stream density, stream power index (SPI), normalized difference water index (NDWI), topographic ruggedness index (TRI), soil adjusted vegetation index (SAVI), and enhanced built-up and bareness index (EBBI), have been employed for the hazard modelling. Ultimately, the data on vulnerability (deprivation) and composite exposure are merged with the hazard data to generate the flood risk maps. With an area under the curve (AUC) value above 0.900, the hazard maps created employing the AHP (AUC: 0.913), F-AHP (AUC: 0.914), FR (AUC: 0.935), and SI (AUC: 0.934) models are found to have excellent performance, whereas the HAND (AUC: 0.861) model has good performance. Of these five&#xa0;models, the data-driven (FR and SI) models show more efficacy than the knowledge-driven (AHP and F-AHP) models, and the hydrological (HAND) model. According to the best performed model (FR), 5.89% (63.11 km<sup>2</sup>) of the MRB is categorized as having a very high hazard, while 0.17% (1.79 km<sup>2</sup>) is classified as having a very high risk. Thus, through this modelling, one will be able to identify which type models are suitable for producing hazard and risk maps in limited data conditions and for developing guidelines, which will further help in reducing the loss and in administering land-use zoning and building codes.</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

Flood Risk Assessment in a Data-Scarce Tropical River Basin of the Global South: A Multi-Model Approach

  • 期刊论文

  • 发表日期：

    2026年5月4日

查看全文
  • Chandini P. C. Senan,
  • R. S. Ajin,
  • A. Rajaneesh,
  • Jitendra K. Nagar,
  • K. S. Sajinkumar

摘要

Abstract

The Manimala River Basin (MRB) and its hinterland were one of the severely battered river basins in Kerala during the 2018 floods. This study aims to demarcate the flood-risk zones in the MRB through five different models: Analytic Hierarchy Process (AHP), Fuzzy-AHP (F-AHP), Frequency Ratio (FR), Statistical Index (SI), and Height Above Nearest Drainage (HAND). Furthermore, this modelling also aims to suggest specific recommendations. A total of 10 conditioning factors (CFs), namely, geomorphology, slope, soil texture, land use and land cover (LULC), stream density, stream power index (SPI), normalized difference water index (NDWI), topographic ruggedness index (TRI), soil adjusted vegetation index (SAVI), and enhanced built-up and bareness index (EBBI), have been employed for the hazard modelling. Ultimately, the data on vulnerability (deprivation) and composite exposure are merged with the hazard data to generate the flood risk maps. With an area under the curve (AUC) value above 0.900, the hazard maps created employing the AHP (AUC: 0.913), F-AHP (AUC: 0.914), FR (AUC: 0.935), and SI (AUC: 0.934) models are found to have excellent performance, whereas the HAND (AUC: 0.861) model has good performance. Of these five models, the data-driven (FR and SI) models show more efficacy than the knowledge-driven (AHP and F-AHP) models, and the hydrological (HAND) model. According to the best performed model (FR), 5.89% (63.11 km2) of the MRB is categorized as having a very high hazard, while 0.17% (1.79 km2) is classified as having a very high risk. Thus, through this modelling, one will be able to identify which type models are suitable for producing hazard and risk maps in limited data conditions and for developing guidelines, which will further help in reducing the loss and in administering land-use zoning and building codes.