Identification Of Unit Hydrographs From Multi-Event Analysis

Abstract

Traditionally the catchment unit hydrograph has been determined by analysing several rainfall-runoff events separately to obtain unit hydrographs which are smoothed, to remove oscillations, and averaged to determine a unique catchment unit hydrograph. The smoothing is necessary because the set of equations describing the linear system are frequently ill-conditioned. Analysing several events simultaneously reduces the ill-conditioning. Therefore, an alternative method is presented in which mUltiple events are analysed simultaneously to determine the unit hydrograph which minimizes the difference between the predicted and observed quick-response runoff for all the events; the solution to the equations was found using a Householder transformation technique. This technique has been shown to be a good method for solving ill-conditioned sets of equations. The method was applied on two catchments of the River Tyne, NE England. For up to seven events the efficiencies of prediction, using these unit hydrographs, were compared with the efficiencies using unit hydrographs determined by the more traditional method described in the Flood Studies Report (FSR); the efficiencies by using the multi-event unit hydrograph were slightly higher than for the FSR method.