Among all kinds of environmental hazards, floods are one of the most destructive natural highs that cause a lot of damage. Flash flooding and flooding of residential areas is one of the most common natural disasters in Iran after the earthquake that endangers human life. Every year, terrible floods occur in the northern cities of Iran, such as the provinces of Mazandaran, Gilan and Golestan. Due to the occurrence of these floods, the places with the highest potential for floods (sensitive areas) should be identified by flood sensitivity maps before planning to reduce the human and financial losses caused by future floods. Due to the importance of flood hazards and its increasing trend in recent years, the preparation of flooding maps and flood susceptibility mapping has been given special attention by researchers and experts. There are different hydraulic (by HEC-RAS) and hydrological methods for preparing flooding maps, and in recent years many statistical and probabilistic models have been tested for flood susceptibility mapping. GIS software has also been used as a basic analysis tool for spatial management and data edition due to its ability to handle large amounts of spatial data and the combination of statistical and probabilistic models with RS and GIS is very important for researchers. Among the various watersheds in the north of the country, in this project, Haraz watershed has been selected as a study area, due to the location and proximity of key cities in the north of the country, including Amol, Mahmoud Abad, Babol, Babolsar, Ghaemshahr. , Sari, Pol-e Sefid, Shirgah, Neka, Behshahr, Gulogah and Bandar-e-Gaz, as well as hundreds of rural points and thousands of hectares of agricultural and garden lands and part of the Caspian Sea road (Rasht to Gorgan) Tehran and Ghaemshahr to Tehran are in this basin, and on the other hand, the occurrence of floods in recent years in this geographical area, which brings with it numerous social, economic and environmental damages and challenges, is necessary. Preparation of flood susceptibility mapping in Haraz watershed becomes more and more necessary. In this international project, distance sensing data including Sentinel radar data and field studies were initially used to identify and monitor flood points, and also various artificial intelligence models including ANFIS-CA were used to flood sensitivity mapping in the study area including ANFIS-BA, ANFIS-IWO, Firefly algorithm (FA), Imperialist Competitive Algorithm (ICA), Logistic Regression (LR), Evidential belief function (EBF), K-Nearest Neighbor (KNN), Deep Belief Network (DBN), Logistic Model Tree (LMT), Bayesian Logistic Regression (BLR), Alternating Decision Tree (ADT), Reduced Error Pruning Tree (REPTree), Bat Algorithm (BA), Artificial Neural Networks (ANN), Classification and regression trees (CART), Flexible Discriminant Analysis (FDA), Generalized Linear Model (GLM), Generalized Additive Model (GAM), Boosted Regression Trees (BRT), Multivariate adaptive regression splines (MARS), Maximum entropy (MaxEnt) were used. In this project, ten effective factors including slope, land curvature, distance from river, altitude, precipitation, river power index (SPI), topographic moisture index (TWI), lithology, land use and vegetation dispersion index (NDVI) are used. The impact weight of each factor was determined using data mining algorithms and the ROC curve was plotted and the subsurface area (AUC) was calculated to validate the flood susceptibility map. The results showed that in order to prepare a flood susceptibility map, the models used artificial intelligence have high accuracy and the high accuracy of these models indicates its reliability, especially in mountainous areas and lacks accurate statistics.
