Prediction of flood hazard map based on hybrid fuzzy geographic information system and its application for Ayamama watershed

Abstract

This research aims to bring a new methodology to early flood management and warning. It proposes modelling one of Istanbul watersheds, the Ayamama watershed in Istanbul, Turkey for possible flood hazard using a new fuzzy-geographical information system hybrid approach. It expands the domain of flood hazard early warning and management that usually uses conventional hydraulic and hydrology approaches to a newly developing area of artificial intelligence in flood early warning and management. The research opted for a more technical study using both GIS and MATLAB software to model the flood hazard levels in Ayamama watershed. The methodology takes into account three factors to model the flood hazard map. Elevation, Euclidian distance from Ayamama creek and local urbanisation degree are the chosen factors. Results on how to identify flood hazard were demonstrated by providing a map of flood hazard zones with their respective negative levels of impacts.