Flooding is a complex phenomenon and hence prediction of flood occurrence is difficult. For predicting the probability of a flood and for mitigating and managing future floods, mapping flood susceptibility is a first essential step. Engineering solutions associated with the conventional approach often create a sense of false security to communities in hazard prone-areas. In addition, traditional flood coping mechanisms can be disrupted or destroyed. It would be wrong, however, to argue that traditional coping mechanisms have been in ecological balance over the years. This conceptualization of the research problem demands the need for a unique methodological framework for integrating local knowledge with geo-spatial information to study flood vulnerability using machine learning algorithms, which they can predict the future floods events in the Haraz watershed with aid satellite images and Radar data. Main objective of this study is the comparison, assessing of flood detection and susceptibility mapping using Radar data including Multi-temporal Sentinel-1 and Sentinel-2 images and Advanced Land Observing Satellite (ALOS)- Phased Array type L-band Synthetic Aperture Radar (PALSAR) Digital Elevation Model (DEM) with new GIS-based machine learning such as: Adaptive neuro-fuzzy inference system (ANFIS), Imperialistic competitive algorithm (ICA), Firefly algorithm (FA), Classification and regression trees (CART), Flexible Discriminant Analysis (FDA), Generalized Additive Model (GAM), Boosted Regression Trees (BRT), Generalized Linear Model (GLM), Multivariate adaptive regression splines (MARS), Maximum entropy (Maxent), Cultural algorithm (CA), Bees algorithm (BA), Invasive Weed Optimization algorithm (IWO) and their ensembles.
