Flood susceptibility modeling of the Karnali river basin of Nepal using different machine learning approaches

The Karnali River Basin (KRB) comprises the longest river in Nepal, located south of Himalayas. Despite its high susceptibility to floods, basin lacks detailed studies. Proper floodplain management is essential reduce impacts due rising flood frequency, magnitude, and severity aggravated by climate change. This research applies three machine learning techniques, Support Vector Machine (SVM), Random Forest (RF), Artificial Neural Networks (ANN), event data from KRB. Ten conditioning factors; Aspect, curvature, distance a (DTR), normalized difference vegetation index (NDVI), elevation, slope, rainfall, soil, stream power (SPI), topographical wetness (TWI) were selected based on multicollinearity test. parameter performance was evaluated using Cohen Kappa Score, with NDVI having greatest influence, followed DTR, TWI. Based Area Under Curve Receiver Operating Characteristics (AUROC), SVM outperformed RF, ANN, ANN for FSM. area very susceptible areas ranges 0.8 2.5% area, most them low slopes elevations. results this study suggest use FSM help proper management.