Search

The estimation of design floods in ungauged small and medium basins in Slovakia was often performed by simple regional flood formulae derived for various geographical regions. Envelope curves were also applied in order to get an acceptable regional safety factor for design purposes. Usually the 100-year flood was estimated from the basin area; correction factors were used to account for deviations from the average of local runoff forming conditions in the region. Floods with shorter return periods were computed by regional frequency factors. Comparisons of design floods derived from such formulae with at site values in 260 small and mid-sized basins [1,2] showed a case-sensitive, highly variable safety factor in these schemes. They are not generally applicable for river training and restoration since they tend to overestimate the design discharges in almost all basins of a particular region. For these tasks other regional approaches, which do not include the regional safety introduced by the envelope curve concept, but allow for site specific over-and under-estimation resulting from the regional average concept, seem to be more appropriate. Some of the recently introduced concepts of regional homogeneity as e.g. in [3,4,5,6,7] were suggested to be used together with regional flood formulae composed of catchment characteristics other than the basin area only. To eliminate the heterogeneity in the annual flood data series due to the diverse genetic origin of flood events, it was proposed to consider annual maximum peak discharges separately for two seasons - the summer season with rainfall-induced discharges and the winter season with floods originating from snowmelt and mixed events.