Author(s): Nemmert Johannes
Linked Author(s):
Keywords: Design flood; Time series; Rainfall-runoff processes
Abstract: In hydraulic engineering the determination of design floods with a high return period is required and hydraulic engineering structures such as spillways are designed for corresponding flood events. In Austria a flood with a return period of 5000 years is used for dam design. Considering the length of reliable data measurements, longest time series extend to about 100 years but often are much shorter and therefore the extrapolation with statistical procedures is inexact. Depending on the distribution functions chosen the results may vary significantly. In Austria a new design guideline has recently released (GUTKNECHT et al. 2003) which was used in a contractors work for the Austrian Railway Authorities to determine the 5000 year extreme flood of a group of dams in the Stubachtal area (Salzburg). In this guideline the rainfall-runoff model is charged with a precipitation corresponding to the return period of the design flood. However, in reality resulting runoff is dependent on various parameters such as distribution of the precipitation, temperature and therefore snow melt, saturation of the soil, available retention volume etc. This paper introduces an approach to determine the meteorological and hydrological factors causing a discharge of a certain return period. In a first step time series of precipitation for the entire Stubachtal area based on the Bartlett-Lewis-Model are generated for 1000 years. This time series have the same statistical properties as the measured ones. The project area Stubachtal is charged with the generated precipitation and the discharge is statistically analysed. In a second step the project area is charged with a combination of time series of temperature and time series of precipitation. Such it is also possible to account for snowfall in the winter and snow melt in the summer time. The discharge of this scenario is as well statistically evaluated. Discharges resulting form this procedure are then compared with those obtained using the Austrian guideline.
Year: 2007