Evaluation of flood hazards areas with fuzzy approach, Case study: Downstream of Neka catchment, Mazandaran province

Floods are among the hazards that affect millions of people around the world every year. The northern region of Iran has also faced numerous floods in recent decades and has caused a lot of damage to urban and rural areas. The northern regions of Iran have also experienced numerous floods in recent decades and have caused extensive damage to urban and rural areas. This study aimed to prepare a map of flood-prone areas by fuzzy method and to validate various maps downstream of the Neka River catchment in eastern Mazandaran. After the flood of 2019, the location of the areas that were affected by the flood was recorded in points. Data were randomly divided into two parts: training and validation 70% and 30%. Based on 70% of the observational data, fuzzy relationships of different information layers were determined, including elevation, slope, distance from the river, precipitation, lithology, and land use. After fuzzification of layers, using 9 different fuzzy operators, the layers overlapped and flood hazard maps were obtained. The validity of the resulting maps was performed using 30% of flooded points and the same number of non-flooded points by ROC method and Cohen's Kappa index. The results showed that operators of fuzzy Gamma 0.9, 0.8, and 0.7, had the highest accuracy of respectively. According to the map of the fuzzy Gamma 0.9, 15.7% of the area is in the very high and high class and 30% of rural and urban areas of the catchment are in these danger zones.