Morphometry-based flood hazard zonation of Ajay River basin using coupled TOPSIS–VIKOR models, India

Abstract

The main aim of the study is to delineate flood-hazard-prone zones in the Ajay River basin using advanced multi-criterion decision-making (MCDM) models. TOPSIS and VIKOR are the two models which were employed, due to their complex decision-making ability and efficient integration of multiple influencing factors. The Ajay River is a severely flood-prone river; flooding of the river is concentrated within the monsoon season, triggered by successive episodes of the high-intensity monsoonal precipitation that inundates the lower floodplain. Post-independence, the river has experienced several flood events, which continue to constitute a major issue in the lower floodplain of the river. Fourteen flood conditioning factors, i.e. drainage density, drainage texture, drainage frequency, Normalized Difference Water Index (NDWI), confluence density, soil type, geology, elevation, relief, dissection index, ruggedness, slope, length of overland flow and infiltration number, were used to generate the final flood hazard maps. Parameter weightages were calculated using the Analytic Hierarchy Process (AHP). The two models showed very similar results; while the TOPSIS model classified 8% and 24% of the basin area as highly and very highly hazard-prone, respectively, the VIKOR model assigned 10% and 23% of the basin area to the same respective classes. These areas are spread throughout the floodplains of the lower Ajay River basin. The validation flood points were obtained from the annual flood report of West Bengal (2023). The Receiver Operating Characteristics (ROC) curve method was used, and both models revealed high accuracy, with area under the ROC curve (AUC) values of 0.827 and 0.837. These results confirm the models' suitability for similar environmental conditions, making them valuable tools for strategic flood-hazard management planning in the study area.