Application and Evaluation of seasonal flood statistics in Germany - a contribution to the discussion

In many cases, the design of hydraulic infrastructure is based on statistically determined flood quantiles. One cause of inhomogeneities in the data may be different conditions under which flood events occur. Summer and winter flood events usually have different geneses. Here, the method of seasonal flood statistics can be used, in which the flood probabilities of the maximum discharge values are determined separately for the winter and summer period and then combined into an annual total probability. In the present study, the seasonal flood statistics is applied and evaluated throughout Germany on the basis of 218 gauging stations. The results shows that at about 25 % of all gauging stations there is no significant deviation between the flood quantile HQ100 from the annual statistics and the seasonal flood statistics. At a minority of gauging stations (about 6 %) a reduction of HQ100 results. At the majority the application of the seasonal flood statistics leads to an increase of HQ100. In principle, the application of the seasonal flood statistics can be recommended, but the results should be critically evaluated with regard to their plausibility. A special check should be made if there is a significant difference in the shape parameters of the distribution functions between winter and summer.