Performance evaluation of hydrological model in simulating streamflow and water balance analysis: spatiotemporal calibration and validation in the upper Awash sub-basin in Ethiopia

Model calibration is critical for hydrologic modeling of large heterogeneous watershed environments. There is little guidance available for model calibration protocols for distributed models aimed at capturing the spatial variability of hydrologic processes in Ethiopia. Therefore, the main aim of this study was to evaluate the performance of the soil and water assessment tool (SWAT) hydrologic model using multi-site gauged data for simulating streamflow and analyzing water balance within upper Awash sub-basin in Ethiopia. On a monthly basis, the sequential uncertainty fitting version-2 (SUFI-2) algorithm embedded in the SWAT-calibration and uncertainty program (SWAT-CUP) was used for sensitivity analysis, calibration, and validation. The coefficient of determination (R2), Nash Sutcliffe efficiency (NSE), and percent bias (PBIAS) were used to statistically evaluate the SWAT model's performance in simulating streamflow. The monthly observed and simulated streamflow statistics revealed that values of R2, NSE, and PBIAS varied from 0.80 to 0.74 and 0.74 to 0.66, 0.74 to 0.66 and 0.71 to 0.62, -3.20 to 14.90 and 18.60 to 8.00 during spatial calibration and validation periods, respectively. In the entire sub-basin, the mean annual rainfall was approximately 1365.03 mm; of this amount, 11.61% flowed as surface runoff (SURFQ), 7.43% as lateral flow (LATQ), about 35.47% flowed as baseflow (GWQ), and 45.41% vanished as evapotranspiration. The sub-basin's average net annual water yield (WY), which includes the SURFQ, LATQ, and GWQ, contributes about 54.63% of the average annual rainfall. The multi-site calibration and validation-based performance evaluation results indicated that the SWAT model would simulate catchment hydrology very well at all gauged stations in the upper Awash sub-basin. According to the findings of the study, to achieve the required model performance efficiency and detect spatial variability within sub-basins, the performance of hydrological models should be evaluated using multi-site streamflow data, which is immensely useful for planning and designing proper water management strategies in the Awash River basin.