An artificial neural network-hydrodynamic coupled modeling approach to assess the impacts of floods under changing climate in the East Rapti Watershed, Nepal

Abstract

Recurring floods have devastating consequences on the East Rapti Watershed (ERW), but effective mitigation/adaptation measures are lacking. This article aims at establishing a rainfall-runoff (RR) relationship; estimating depth and extent of inundation under climate change scenarios; assessing impacts on the socio-economy; and identifying and evaluating adaptation strategies in the ERW. Artificial Neural Network (ANN) was used to generate peak flows which were then entered into a hydraulic model to simulate inundation. Results were validated with field survey. The calibrated and validated RR and hydraulic models were fed with projected future climate (2021–2050) derived from multiple regional-climate-models to assess the changes in inundation. Results showed the peak discharge likely exceeds 10,500 m3/s at the ERW outlet in the extreme future flood scenario with corresponding inundation of 80 km2 and up to a depth of 11 m sweeping away over 1000 houses and 19 km2 of agricultural land in the critical areas. Constructing a 17 km long embankment in the critical areas along the right bank of the East Rapti River could reduce the flood spread by 35%, safeguarding 78% of the houses and saving 51% agricultural land compared with the scenarios without the embankment.