Flood Susceptibility Modeling in the Aland Chai Basin using New Ensemble Classification Approach (FURIA-GA-LogitBoost)

In the beginning of spring, floods are the most important geomorphic hazards in Iran, destructing propertiesas and human lives. Aland Chai basin, located in Khoy County (northwest Iran), is also known as one of the basins with high potential for flood hazard due to its special geographical situation. This study tried to model spatial variation in flood hazard susceptibility in this basin using the ensemble model, FURIA-GA-LogitBoost. For this purpose, 13 effective parameters of flooding including lithology, soil hydrological groups, NDVI, land use, slope, aspect, elevation, distance to the river, river density, precipitation, topographic wetness index, stream power index, and sediment transport index were used. WEKA software was used to implement the research model and the final flood hazard susceptibility map was prepared. The study found that downstream areas of the basin have a high flood hazard susceptibility. These areas contain the most important human settlements (Khoy city) and agricultural lands and flood as a geomorphic hazard can seriously damage them. Considering the ROC curve and area under the curve (AUC), it was found that the FURIA-GA-LogitBoost model performed well in the preparation of flood hazard susceptibility map with coefficients of 0.861 and 0.895, respectively, in training and validation data.