Mapping flood susceptibility areas and assessing influential factors : case of the Chellif basin

Abstract

Floods are considered one of the most destructive catastrophic phenomena.Flood susceptibility is defined as the tendency to suffer damage caused by this phenomenon. However, accurately predicting flash floods remains challenging due to the complexity of the phenomenon. In this study, we adopted an approach based on geographic information systems (GIS), remote sensing techniques (RS), and machine learning classification models such as LGBM, AdaBoost, and the new machine learning technique called Stacking, to create a flood susceptibility map in the Chellif watershed. Fifteen hydrological and topographic factors were used as inputs for the flood susceptibility models. The results showed that Stacking was the most optimal model, with an AUC value of 0.99, followed by LGBM with 0.98 and AdaBoost with 0.96. The findings of this study are used for planning and implementing flood mitigation strategies in the region.