Speaker
Description
Due to intense precipitation, rapid snowmelt and rising sea and lake waters, river levels and groundwater, urban areas can be flooded. These flood risks are associated with several factors of a topographical, geological, hydrological, climatic and anthropic nature. These factors must be taken into account when managing floods and especially when delimiting vulnerable areas. This research aims to assess and compare Frequency Ratio, Weighting Factor, and Weight of Evidence Models for landslide susceptibility mapping using Geographic Information Systems and Remote Sensing data in Beni Mellal City, Morocco. A set of 5000 landslides were identified and mapped by evaluating observations from satellite images (Google Earth images) and fieldwork from 2018 to 2022. The landslide inventory data was arbitrarily divided into two groups for training (70%) and validation (30%). Thirteen landslide conditioning factors were selected for landslide susceptibility modeling, based on multicollinearity analyses and the information gain method. Validation of the results is based on statistical rules for the Spatial Effective Method, Statistical Measures, and Receiver Operating Characteristics Curve (ROC).
