In recent years, larger floods and resulting damage has occurred in Germany repeatedly, for example in the years of 2002, 2006 and 2013. Flood hazard is caused by a combination of different meteorological, hydrologic and hydraulic processes. At present, there is no defined methodology to describe flood models, with which the overall flood risk - and not only specific water levels or for specific catchments - can be evaluated and quantified.
The aim of the cooperation project FLORIS (Flood Risk SV Sparkassen Insurance; IMK together with IWG and J.E.Daniell) is to quantify precipitation and flood scenarios by using a stochastic modelling approach. This will make it possible to determine maximum damage over realistic distribution functions of the precipitation or the runoff/ effluent flow. In the meteorological part, the spatial-temporal variability of past heavy precipitation events with different durations (1 to 20 days) will be characterized and compared with past (actually occurred) flood events. It is expected that these analyses will already produce several statistically-valid indications for flood-causing factors for different sizes and characteristics of different catchment areas.
For the extreme value statistic description of the heavy precipitation hazard, a stochastic calculation method will be developed. It will be based on a sufficiently large sample from past events to generate datasets of realistic heavy precipitation events. By using such a stochastic approach all heavy precipitation events are to be covered that could at least be possible theoretically, but may not necessarily actually have occurred in the past. Based on theses simulations, the project partners (hydrology and hydraulic) will be able to estimate the flood risk for a larger area, such as Baden-Württemberg, for a specific portfolio.
Abb.: Total precipitation from 28.7.2013 6 UTC to 29.7.2013 6UTC based on REGNIE data.