REGIONAL ESTIMATION OF DESIGN FLOOD DISCHARGES FOR RIVER RESTORATION IN MOUNTAINOUS BASIS OF NORTHERN SLOVAKIA

摘要

The methods (Ⅰ- Ⅳ) are valid in the region, on the average. Typically, underestimation of the design values in approximately one half of the sites and overestimation in the other half occurs. Therefore, regionalisation approaches with the lowest variability over the regional reference values should be preferred in a practical application. However, no preference can be stated regarding the choice of the regionalisation method. They are generally applicable for river training and restoration since they do not tend to overestimate the design discharge in all basins of a particular region as the traditional approach. The traditional approach has to be preferred in tasks where safety of design is required, since under-design cannot be compensated by over-design. Several questions remain to be investigated such as the preference of a particular regionalisation approach, aspects under which the concept of regional homogeneity can be used for design purposes involving safety, and the applicability of the methods in other regions. Regional frequency analysis following principles of Hosking and Wallis could lead to the selection of more appropriate regions than subjective regionalisation methods relying on the concept of geographical regions. However, with the climatic and physiographic catchment parameters used in this paper it can lead to difficulties, when assigning ungauged catchments to regional types. Inclusion of other variables controlling the variability of flood formation, describing the intensity of overland flow formation, upper layer permeability, catchment storage, etc., should be further investigated in order to test the practical applicability of the concept of regional homogeneity based on physiographic characteristics.