From Expert Opinion to Data: Statistical Models and Performance-Based Classification for Flood Susceptibility Mapping

Floods are natural disasters that cause socioeconomic and environmental losses in both urban and rural areas. Within the framework of planning, precautionary measures against flood disasters can be undertaken through the application of analytical methods based on various modeling approaches. In this context, using the example of the Melen Basin in Turkey, flood-prone areas were identified and mapped using five methods: Frequency Ratio (FR), Shannon Entropy Index (SE), Evidential Belief Function (EBF), and the hybrid models EBF-SE and EBF-FR. Elevation, slope, aspect, land use, plan and profile curvature, drainage density, distance to river, curve number, long-term average precipitation, geological formation, soil depth, topographic wetness, sediment/sediment transport, and stream power index were evaluated in the spatial database. To measure the accuracy of these prediction models for determining flood risk, AUC (area under the curve) values were compared using the ROC (Receiver Operating Characteristic) curve. In terms of method validity, the SE method showed the highest efficiency (AUC = 0.979), followed by the FR (AUC = 0.974), EBF-SE (AUC = 0.972), EBF-FR (AUC = 0.968), and EBF (AUC = 0.966) methods. According to the results of the FR and SE methods, elevation, lithology, and slope are the most influential factors in flood formation.. In the evaluation of the success index of the models, the following values were determined according to their size: EBF-SE (96.0), SE (94.4), EBF (91.8), FR (81.9), and EBF-FR (79.4). In the classification of salinity sensitivity maps, Natural Breaks (Jenks) is the most successful method according to the success index.