Abstract:
Robust hydrologic models are needed to manage water resources for healthy aquatic ecosystems and reliable water supplies, but there is a lack of comprehensive model comparison studies that quantify differences in stream flow predictions among model applications developed to answer management questions. This study was initiated with the objective to compare and select the best conceptual rainfall-runoff model that can be used in the design, planning, and management of water resources in the watershed and estimation of monthly and annual water balance at the Shaya watershed in Bale mountainous area. Daily rainfall, as well as daily potential evaporation and discharge, for the watershed were collected and prepared as input for the model. In this study GLUE based Monte Carlo was used to assess the uncertainty of the model parameter. The conceptual rainfall runoff models chosen for this study were SMAR and HBV-light. Sensitivity analysis, model calibration and validation were made to evaluate the models. The calibrated SMAR and HBV model performed well for simulation of daily stream flow. The models statistically performance measures, were very good with coefficient of determination (R2) of 0.95 and 0.81, as well as the Nash-Sutcliffe efficiency (Reff) of 0.91 and 0.75 for both SMAR and HBV respectively. Performance of the model during validation period was with Reff 0.78 and 0.59 for SMAR and HBV respectively. The Mean monthly simulated discharge with the calibrated model were found to be 63.4 mm and 58.6 mm while the mean annual simulated discharge during calibration period found to be 790.29 mm and 721.6 mm for SMAR and HBV respectively. Overall, performance of the model demonstrated good in capturing the patterns and trend of the observed flow series. When we the performance of the two models SMAR gives better result than HBV model for selected watershed. For most parameters good results could be obtained over large ranges whereas a few parameters were well-defined over small ranges. Generally, an indication of the uncertainty in model simulations arising from the uncertainty in the parameterization was given by viewing the monthly simulation periods of discharge.
