Robust estimation of the total unit hydrograph

Abstract

This paper investigates three techniques for estimating the parameter values for the unit hydrograph: full model optimisation (the CWI and CMD versions of IHACRES were used), simple rainfall scaling approaches for estimating effective rainfall and an inverse filtering approach. The approaches are tested using synthetic streamow data generated by the catchment moisture deficit (CMD) version of the IHACRES model, based on observed rainfall data for Cotter catchment (near Canberra, Australia) and the Singkarak catchment (Sumatra, Indonesia). The impact of rainfall and streamow errors are considered by reducing the raingauge density and introducting errors in the rating curve. The inverse filtering apprach performed well for the synthetic streamow derived from the rainfall data for the Cotter catchment. For the Singkarak catchment data, the differences between the approaches was small, due possibly to either the higher frequency of rainfall in that area, or to the lower correlation between the rainfall gauge locations.