MODELING RAINFALL-RUNOFF AND SEDIMENT YIELD IN THE MILLE RIVER WATERSHED, LOWER AWASH SUB-BASIN, ETHIOPIA

Abstract

Rainfall-runoff and sediment yield modeling of watersheds is very essential to solve waterassociated problems and alleviating reservoir sedimentation problems caused by the soil erosion of the watershed. The general objective of this study was to model the Rainfall-runoff and sediment yield and to evaluate the performance of the HEC-HMS model of the Mille River watershed. Hydro-meteorological data collected from the National Meteorological Institute and Ministry of Water and Energy were used for model calibration and validation, whereas spatial data collected from different sources. The main materials and software like ArcGIS, HEC-GeoHMS, Arc Hydro, HEC-HMS, Rainbow, Mendeley, MS-Excel, MS-word and XLSAT were used. The methodology of this work was the following: data collection, data preparation, and quality assurance, data analysis and model setup, running model, model performance evaluation, and model result interpretation. In this study the simple canopy, simple surface, soil moisture accounting (SMA), Clark Unit Hydrograph, recession, Muskingum-Cunge, Modified USLE and volume ratio methods were adopted for canopy, surface, precipitation loss modeling, transform modeling, base flow, flood routing, modeling, erosion modeling and sediment modeling respectively. The Gage weight meteorological method was selected to assign the areal rainfall computed by Thiessen polygon method to each sub-basin and the monthly average meteorological method was used for evapotranspiration modeling. The model was calibrated from 1993-2008 and validation was done using observed data from 2009-2016. Storage coefficient and time of concentration are the most sensitive parameters for simulation of runoff. The model performance was checked and the results of stream flow calibration and validation R² = 0.74, NSE = 0.65, RSR = 0.60, PBIAS =-7.10 and R² = 0.84, NSE = 0.69, RSR = 0.60, PBIAS = +14.18 respectively and for sediment yield calibration and validation R² = 0.72, NSE = 0.56, RSR= 0.67, Pbias = -15.50 and R² = 0.80, NSE = 0.54, and RSR = 0.68 respectively. Based on these model performance results, the computed value of both the stream flow and sediment yield shows good agreement with the observed values appropriately. In this study, the HECHMS model simulates the average annual stream flow and average annual sediment yield of 11.42 m3/s and 8.34 Mt/year or 18.36 ton/ha/year respectively, for the period of 1993 to 2016 at the outlet of the Mille River watershed. The results of this study will be used by water resource managers and decision makers to conserve soil and water and taking appropriate flood and sediment control measures in the watershed.