Flood Vulnerability Assessment Using Geoinformation Data and Multicriteria Evolution Approach in Dakar Urban Environment, Senegal

During this last decade, urban flooding has become recurrent in Senegal, particularly in the city of Dakar. The causes are multiple and varied, with devastating impacts on population well-being and the urban environment. For sustainable and harmonious urban development, flood risk mapping should be integrated into urban planning policies. This study aims to map areas vulnerable to flooding at regional, departmental and communal scale, using Remote Sensing, Geography Information System (GIS) and Multicriteria Evolution Approach (MCEA). Twelve criteria layers were defined to identify flood-prone areas, including Rainfall, Elevation, Slope, Groundwater level, Soil type, Impervious surfaces, Land use/cover, Population density, Housing type and structure, Humid zones, Drainage density and Standard of living. A Weighted Linear Combination (WLC) technique was applied and adapted in this study to map flood vulnerability. The criteria were scored and weighted according to their relative importance, as defined by local experts. ArcGIS 10.1 software through its extension Weighted sum overlay tool, was used to aggregate criteria layers into three main indicators: “Socio-economic”, “Physical environment” and “Land use/cover”. The results indicate that approximately 60% of the Dakar urban area is highly vulnerable to flooding according to the "Physical environment" indicator, compared to 12% and 10% for the "Socioeconomic" and "Land use/cover indicators, respectively. Overall, the combination all indicators shows that about 50% of Dakar urban environment is vulnerable to flooding. This vulnerability mainly concerns the departments of Pikine, Guediawaye and Keur Massar, where more than 80% of areas are classified as highly vulnerable. At the municipal and departmental scales, vulnerability indices are higher (greater than 0.7) in these areas compared to the department of Dakar (0.37) and Rufisque (0.14). Validation of final flood vulnerability map was conducted by comparing the most vulnerable areas identified in this study with those severely affected by floods observed using Earth Observation Satellite (SPOT) imagery in September 2005 and 2009. The validation results show an overall accuracy approximately 0.95 and 0.90 for the two respective dates. Understanding the level of vulnerability of the Dakar urban environment, as well as the contribution of each factor, can serve as a decision-support tool for rational and localized management of flood-prone areas in the context of climate change.