Sediment Yield Modeling And Evaluation Of Best Management Practices In Sediment Yield Reduction In Angar Watershed, Abbay River Basin, Ethiopia.

Abstract

Soil erosion is the major cause of land degradation and the most problem in high lands of Ethiopia. Blue Nile River Basin is one of the Ethiopian high land which faces high soil erosion rate and sediment transport by the river. Angar watershed is one among the sixteen sub basins of Abbay River Basin, which is affected in high rate of soil erosion and sediment transport. Therefore, sediment yield modeling for its entire watershed or analysis of spatial distribution of sediment yield at its sub basin level is paramount to develop best management practices for most prone sub basins in soil erosion. So, the objective of this study was to determine sediment yield of Angar watershed and to carry out simulations for different scenarios of best management practices by using ArcSWAT2012 model. The data collected include, spatial data (Digital Elevation model, soil and land use land cover), meteorological data and hydrological data (stream flow and sediment yield) were used as input data for the model in this study. Hence, these data were checked and prepared as per requirement for ArcSWAT2012 using ArcGIS 10.1 to setup the model in this work. For the simulated hydrologic output of ArcSWAT2012 and measured hydrologic data, sensitivity analysis, calibration and validation was done using SWAT-CUP, SUFI2, to evaluate performance of the model on the watershed. For sensitivity analysis, 20 parameters for stream flow and 12 parameters for sediment yield were analyzed. The first 8 and 7 from most sensitivity ranked parameters were considered as most sensitive for stream flow and sediment yield respectively. The model calibration was for the period of 1995 to 2004 and the validation period was 2005 to 2011 for both sediment and stream flow modeling at near Nekemte gauging station. Model performance indicators, coefficient of determination (R2 ), Nash-Sutcliffe coefficient (NS) and percent of bias (PBIAS) shows well performance of the model under calibration and validation for both monthly stream flow and monthly sediment yield. The statistical value of these performance indicators for stream flow were, R2 =0.87, NS =0.77 and PBIAS =-2.1 for calibration and, R2 =0.89, NS =0.83 and PBIAS =1.2 for validation, and for sediment yield, R 2 =0.87, NS =0.86 and PBIAS =-7.3 for average monthly sediment yield calibration and, R2 =0.82, NS =0.81 and PBIAS =-6.2 for validation. After model calibration and validation has performed mean annual spatial sediment yield of 36.904 Ton/ha/yr was estimated. It is also indicated that about 15.87% of total area was under severely affected area in soil erosion. These potentially affected sub basins (36, 39, 40, 42,44,54,55 and 57) were obtained priority to implement best management practices. These best management practices were filter strips, parallel terraces with stone/soil bund and reforestation. The simulation results showed that applying filter strips, parallel terrace/stone bunds and reforestation scenarios reduced sediment yield by 43.61%, 44.1% and 20.82% respectively from the existing condition scenario at entire watershed level. Hence, among these three filter strips and parallel terrace/stone bunds reduce sediment yield in large amount and considered as the best to reduce sediment yield of the watershed.