Distinctive Framework Incorporating Multi-Criteria Decision Analysis Technique for Flood Susceptibility Mapping in an Urban Watershed of Bagmati Basin, Nepal

Abstract

The Hanumante Watershed is a northeastern tributary of the Bagmati Basin in Kathmandu Valley which faces recurrent flooding challenges. Hydraulic modelling and statistical extrapolations of flood data are the conventional methods of flood assessment in Nepal. These conventional methods are more objective and lack comprehensively reflecting ground-level information. However, this study introduces a distinctive approach employing two-tier Multi-Criteria Decision Analysis (MCDA) for GIS-based flood mapping incorporating maximum possible ground-level data. Flood susceptibility was mapped using the Analytical Hierarchy Process (AHP) method, incorporating 15 parameters grouped under four criteria: Hydro-meteorological, Morphological, Soil-permeability, and Land-use/land-cover dynamics. The map was validated against the HEC-RAS model output, a flood benchmarking survey, and ground-level photographs of the 2023 flood, achieving an accuracy of AUC >90. Flood susceptibility map indicated that 10.48% of the area is highly prone to floods, with 99.29% of this area being within 200 meters of the river. This holistic framework of flood assessment offers potential application in similar urban/peri-urban watersheds.

Keywords: Urban Flood, MCDA, AHP, HEC-RAS, AUC, Flood Hazard