Regionalization of Conceptual Rainfall-Runoff Model Parameters for Predicting Stream Flows of Ungauged Catchments in the Upper Blue Nile Basin

Abstract

Water resources development and research significantly suffered from lack of stream low data. Regionalization of model parameters was found very
useful in filling such data gaps. We therefore regionalized the parameters of the HBV model so that the model could be used in ungauged catchments of
the Upper Blue Nile (UBN) basin. Although we collected stream flow data for 76 stations from the Ministry of Water, Irrigation and Electricity, our data
quality assessment indicated that only 20 stations were suitable for calibrating the HBV model. We calibrated the model using hydro-climate data of 6 years and validated the calibrated model for independent data of 4 years. The calibrated model reproduced the overall pattern and base flow of the catchments. However, it noticeably missed several peak flows. The values of the calibrated parameters varied with the characteristics of the catchments. We therefore developed a multiple regression relationship between the parameters and catchment characteristics for the basin. The relationship was statistically significant and therefore could be used to apply the HBV model for un-gauged catchments in UBN for water balance studies, climate change impact assessment, and recharge estimation. However, additional work specifically improved data sets were needed to improve the regionalization results for peak flows.