Uncertainty estimate of flood hazard for an extreme flood

Abstract

Hydrodynamic models are usually calibrated from historical events observations and then used for estimating the flood hazard for more extreme floods. In such a case, the uncertainty related to the calibration process is likely to be emphasized by the inadequacy of the model to describe hydraulic processes met during the flooding of additional areas. Comparing the results of various 1D and 2D models of the Besançon flooding by Doubs river for a 10,000 years return period, an estimate of the resulting uncertainty on both water depth and flow velocity is carried out. A difference of more than one metre is found for peak water levels and can be related to the features of the hydraulic models. However, this latter difference is smaller than the uncertainty range linked to the estimate of the peak discharge even if a detailed hydrological study is performed such as during the Flowres ("Predicting the flow in the floodplains with evolving land occupations during extreme flood events") project. Some conclusions are linked to the specific case of Besançon (e.g. the effect of the loop of Doubs river) but other ones are more general (e.g. representation of the urban pattern).