Author(s): Majone Ugo; Tomirotti Massimo; Galimberti Giacomo
Linked Author(s):
Keywords: Peak flood flows; Regional estimation; Unpredictable events; Maximum probable flood discharge
Abstract: The present paper describes the results of a regional analysis of peak flood flows that has been carried out on the basis of the data recorded in more than 8500 gauging stations belonging to different geographical areas of Europe (Italy, Switzerland, Austria, Great Britain, Portugal), North America (USA, Canada), Oceania (Australia) and to other sites of Africa, Asia and South America. The regionalization procedure is based on the definition of a dimensionless variable Y (obtained by a suitable standardization of the peak discharge) whose probability distribution can be assumed constant for all the considered sites. The probabilistic model (MG model) obtained by interpolation of the empirical non exceedance frequencies of the maximum values of Y observed at each gauged site gives the probability distribution of peak discharges for return periods ranging from 30-50 years (nearly equal to the average length of the considered series) up to about 4000 years. With respect to other probability distributions that are currently used for the same purpose, the MG model provides a better interpolation of the empirical non exceedance frequencies of peak discharges, especially for long return periods. The value Y=7 – corresponding to about a 4000 year return period – is the upper bound for which the interpolation of the frequency distribution of Y provided by the MG model can be considered reliable; moreover, the value Y=9 is the absolute maximum of the empirical values of the standardized variable with reference to the 8500 gauging stations considered in the analysis. These values can be assumed both to define the “predictability bound” of flood events and to estimate the design peak discharges in cases of particularly high risk levels.
Year: 2007