Continuous-based Comparison of Two Hydrological Models for Modeling a Watershed, in the Case of Robi Watershed, Middle Awash River Basin, Ethiopia

Abstract

Continuous-based comparison of the hydrological model is applied in different parts of a watershed across the world. However, little has been done on hydrological modeling comparison in Ethiopian watersheds, and nothing was done to the Robi watershed. This study aimed to compare two hydrological (SWAT and HEC-HMS) models for modeling the Robi watershed using monthly continuous data. To achieve this objective climate data, spatial and hydrological data were collected. The model was calibrated and validated using observed streamflow data from 1995 to 2004 and 2005 to 2011, respectively. The performance of the model shows that, during the calibration of the SWAT model, R 2 , Nash-Sutcliff efficiency, PBIAS, p-factor, and r-factor indicate that .82, 0.81, 2%, .83, and 1.03, and in the validation, it indicates that .73, 0.71, 4%, .88, and 1.5, respectively. Similarly, in the HEC-HMS model, R 2 , Nash-Sutcliff efficiency, and PBIAS indicate .8, .79, and −3 for calibration and .7, 0.68, and −6, respectively in validation. The model was compared in the time domain and quantile domain. The magnitude of observed monthly stream flows for 10%, 50%, and 95% exceedance percentages is 10.7, 2, and 0.14 m 3 /s. Simulated flow shows exceedance probabilities of 10%, 50%, and 95% as 9.766, 1.8, and .15 for the SWAT model and 9.34, 1.7, and 0.6 m 3 /s for the HEC HMS model, respectively. Overall the SWAT model performed better than the HEC-HMS model in the Robi Watershed.