The effect of multiple reservoirs on flood peak reduction: the case of two austrian catchments

In Austria there are 78 large reservoirs and, in most cases, they are located in nested catchments. The presence of a reservoir generally produces an attenuation effect on flood peaks downstream, however the effect of multiple reservoirs in a catchment remains unclear. The aim of this work is to analyze the combined effect of multiple dams on flood peaks along the river network at the regional scale. Its evolution is studied in space, considering how the effect changes along the river segments and as a function of the return period of the flood event. The method used consists in two steps: (i) the design hydrographs are evaluated at several locations along the river network for 30, 100 and 300 years return periods, based on interpolated flood quantiles in ungauged catchments; (ii) the peak reduction is estimated based on the information on dams in the catchment (i.e. number, position, storage capacity and drainage area) and using the concept of equivalent reservoir, protection ratio and filling discharge. The reservoirs effect is here evaluated as the percentage of reduction of the flood wave peak in river sections downstream of one or several dams. This approach is applied to the Salzach and the austrian part of the Drau catchments, where there are a fairly large number of dams placed both in series and in parallel with each other. The results obtained show that the relative peak reduction is greatest at the reservoirs and decreases rapidly downstream for each of the return periods analyzed and for the different combinations of dams. The method is validated in both time and space, using data from a number of gauging stations available in these two catchments. Overall the method is able to estimate accurately flood peak attenuation of reservoirs even though the values are generally lower than the observed one, the latter being however based on short samples compared to the return periods of design events. Overall the method can be considered a valuable instrument for estimating reservoir induced flood peak reduction in ungauged basins, over the river network at the regional scale.