LAUSR

Over decades, the planning of flood management was based on a safety-oriented approach. A design flood was estimated by probabilistic means to specify the limit up to which a flood should be controlled completely by technical measures. A case of failure was expected only in such cases where the design flood is overtopped. As design floods were specified by very small probabilities, the risk of a flood beyond the design flood was seen as negligible. Devastating flood events all over Europe raised the public awareness of remaining flood risks in the last two decades. Risk management became a political task in the EU. According to the European Flood Directive geographical areas, which could be flooded “with a low probability or under extreme event scenarios”, have to be specified. The combination of “low probability” and “extreme event scenarios” demonstrates the problem of modern flood management. The existing probabilistic tools are not sufficient to specify the risks of failures, which result from critical combinations of multiple characteristics of hydrological loads. Scenarios are one option to specify them, but their probabilities stay unknown. Multivariate statistics could offer a way to fill this gap, but some problems of their practical application are still unresolved.