Preparation of flood susceptibility mapping using an ensemble of frequency ratio and adaptive neuro-fuzzy inference system models

Abstract

Floods are among the most common natural disasters that impose severe financial and human losses every year. Therefore, it is necessary to prepare susceptibility and vulnerability maps for comprehensive flood management to reduce their destructive effects. This study is trying to provide a flood susceptibility mapping in Jahrom (Fars Province) using a combination of frequency ratio (FR) and adaptive neuro-fuzzy inference system (ANFIS) and compare their accuracy. Totally, 51 flood locations areas were identified, 35 locations of which were randomly selected to model flood susceptibility and the remaining 16 locations were used to validate the models. Nine flood conditioning factors namely: slope degree, plan curvature, altitude, topographic wetness index (TWI), stream power index (SPI), distance from river, land use/land cover, rainfall, and lithology were selected, and the corresponding maps were prepared using ArcGIS. After preparing the flood susceptibility maps using these methods, the relative operating characteristic (ROC) curve was used to evaluate the results. The area under the curve (AUC) obtained from the ROC curve indicated the accuracy of 89% and 91.2% for the ensembles of FR and ANFIS-FR models, respectively. These results can be useful for managers, researchers, and designers in managing flood vulnerable areas and reducing their damages.