Web based GIS for multicriteria vulnerability classification and prediction in urban area: the case of Yeumbeul (Dakar)

This study tries from rainfall variability, digital elevation model, and land cover change analysis to track the interaction between natural and human causes of flooding occurring in Senegal. We proposed a multilayer system for the study of the evolution of vulnerability in Yeumbeul a suburb located in Dakar. These layers include a spatial description and an urban model of the study area. The general architecture of the system allows users to access the application through the internet and to perform the operations on the study data. We also proposed solutions to perform a multicriteria classification of the vulnerability. Vulnerability prediction is performed using a nonlinear regression parametric model of rainfall. The results obtained are very promising and ultimately can help decision maker to better deal with floods in Dakar.