نبذة مختصرة : International audience ; This communication deals with a GIS-based urban flooding model operating at high spatial resolution (~10m) over a large urban conglomerate, for real-time forecasting purpose. Although floods involve multi-directional flows, 2D-models cannot be applied over a whole urban area, because these models require long computation times and detailed topographic information which may not be available. That is the reason why we put forward an alternative, simplified and easy-to–use model offering reasonable performance, and fast computation times in agreement with real-time forecasting requirements. The model operates on the basis of a 10m spatial resolution DEM, which is first modified to account for artificial drainage directions induced by the built-up areas and the channel network. Runoff conditions (e.g. land use) and routing conditions (e.g. friction, slopes and cross-sections) can be distributed in space. The flows are first computed at the outlet of elementary urban catchments of 1km2 area by the SCS-LR hydrological model, then the output of the catchments are injected as inputs into the drainage network and routed using the kinematic wave model. The outputs of the model are the flood hydrographs at the outlet of each urban elementary catchment, and the discharges at any time and any location of the drainage network, which allow estimating the overloading of the network and its impact in terms of flooded areas. The model was tested in African cities like Ouagadougou (Burkina-Faso) and Bamako (Mali). In addition to real-time forecasting, it allows impact assessments for future urban development or climate change scenarios, and also supplies the boundary conditions for 2D models to be used in restricted urban areas. The model is available within the open source ATHYS platform. ; Cette communication présente les principales caractéristiques d’un modèle d’inondations urbaines opérant à l’échelle de l’agglomération complète à une très fine résolution spatiale (~10m), dans une optique de prévision ...
Processing Request
No Comments.