Abstract:
The main objective of the study was to model the hydrology of Maybar watershed using the
SWAT model in order to understand hydrological fluxes of Watershed and to evaluate the
performance of SWAT-WB & SWAT-CN for simulating stream flow. The Arc SWAT
interface implemented in the ArcGIS software was used to delineate the sub-basin, combine
the data layers and edit the model database. The watershed was discretizing into 14 sub basins.
89 HRU was created on the watershed by a threshold combination of 20% land use, 10% soil
and a 20 % slope. EDC values for seven soil types were assigned based on the topographic
index of the watershed. The hydro- meteorological data collected from nearby stations and the
missed data filed by arithmetic mean and simple interpolations. Sensitivity analysis was
conducted for both SWAT-WB & SWAT-CN. The sensitivity result showed ALPHA_BF &
CANMEX are the most sensitive parameters for SWAT-WB &SWAT-CN respectively. The
models were calibrated and validated on Maybar watershed using SUFI-2 and GLUE for
SWAT-WB &SWAT-CN. SUFI-2 gives good result compared to GLUE. SWAT-WB was
performed slightly better during calibration than SWAT-CN (NSE of 0.91 &0.87) respectively.
In the validation period SWAT-WB was also performed better than SWAT-CN (NSE of
0.85&0.78) respectively. All sources of uncertainty captured by bracketing more than 33% for
SWAT-WB & 51% of SWAT-CN. Base- flow (64.2%-71.46%) is an important component of
the total discharge compared to surface runoff in the study area. Based on the statistical
indicators, the evaluation indicates that SWAT-WB &SWAT-CN had very good performance
for both month calibration and validation period in Maybar watershed. The statistical result
revealed that if properly calibrated SWAT-WB can be used more effectively than SWAT-CN
in Monsoonal climate like Ethiopia.
