Estimating Runoff And Sediment Yields And Effect Of Best Management Practices: A Case Study On Bilate Watershed

Abstract

Watershed runoff which is highly influenced by watershed characteristics (LULC, Hydrologic soil group, slope, and Climate condition) plays an important role in designing hydraulic structures, controlling soil erosion, and assessing the water yield potential of the watershed. Surface runoff and sediment loading are immense problems that have threatened water resources development in the Bilate watershed. The main objective of this study was estimating runoff and sediment yields and effect of best management practices in Bilate watershed. DEM, LULC, Soil and Weather data are inputs to the SWAT Model. The consistency and homogeneity of data were checked by using double mass curve and Rainbow respectively. Simulation was carried out for the period of 30 years (1986 to 2015) and divided the watershed in to 29 sub basins and 478 Hydrologic Response Units (HRUs). The flow calibration and validation were carried out for the period of (1989-2003) and (2004-2015) respectively. The calibration and validation of sediment yield was carried out for the period of (1990-1998) and (1999-2004) respectively. The SWAT-CUP with the Sequential Uncertainty Fitting (SUFI-2) algorithm was used for calibration and validation of both flow and sediment. The average annual runoff simulated in the watershed was 292mm/ha/y and that of sediment was 59.8t/ha/y. The maximum runoff took place in the month of August and the minimum runoff took place in the month of January The highest sediment yield was observed during the months of August and whereas the lowest sediment yield was observed during the month of December The Model performance evaluation was checked by using a coefficient of determination (R 2 ), Nash-Sutcliffe model efficiency (NSE), Root mean Square Error Standard Deviation Ratio (RSR) and Percent bias (PBIAS) for both flow and sediment. The result showed that R 2=0.82 for calibration and 0.77 for validation. NSE= 0.7 for calibration and 0.65 for validation. RSR= 0.6 for calibration and 0.54 for validation. PBIAS= -15.4% for calibration and -14.1% for validation for flow. The Model performance evaluation of sediment yield showed that R 2= 0.74 for calibration and 0.71 for validation, NSE= 0.71 for calibration and 0.68 for validation, RSR= 0.6 for calibration and 0.62 for validation and PBIAS= -34.8% for calibration and - 31.3% for validation. The average annual runoff simulated in the watershed was 292mm/ha/y and that of sediment was 53.48t/ha/y. Runoff varies from place to place in this watershed. The average yearly maximum runoff generated from Sub basin 14 was 426mm and the minimum average yearly runoff was generated from Sub basin 20 was 49mm.The result also showed that, runoff varies randomly in the catchment and the maximum runoff took place in the month of August and the minimum runoff took place in the month of January. Sediment yield varies both spatially and temporally. The Sub basin 23 has small amount of sediment yield which was 4t/ha/y and that of 9 has high sediment yield (83t/ha/y). It was witnessed that the highest sediment yield was observed during the months of August and whereas the lowest sediment yield was observed during the month of December. Three management scenarios were carried out in this study and the result showed that the average annual sediment yield at the entire watershed and Sub basin after application of grassed waterway, filter strips, and contouring was highly reduced. After comparing the scenarios results, contouring was more effective than the others to reduce sediment yield entering in to the catchment (59.8 to 6.4t/ha/y). The result showed that the calibration and validation results of both flow and sediment have good correlation with observed data.