APPLICATION OF THE MODIFIED UNIVERSAL SOIL LOSS EQUATION (MUSLE) IN THE PREDICTION OF SEDIMENT YIELD IN AGEWMARIAM EXPERIMENTAL WATERSHED; TEKEZE RIVER BASIN, NORTHERN ETHIOPIA

Abstract

The severe effect of soil and water resource degradation in Ethiopia is putting pressure to take soil and water erosion control measures. To support the planning, management, and appropriate use of the soil and water resources the Modified Universal Soil Loss Equation (MUSLE) was used to model eventual sediment loss in Agewmariam experimental watershed. The model uses different methods for the estimation of the runoff energy, soil erodibility, slope length & steepness, cover management, and support practice factors input parameter values; so that calibration, evaluation, and validation of parameter values become mandatory. The model input parameters were collected from a combined process of field observation, laboratory analysis, Google Earth Pro, and GIS processing while cover management and support practice are adopted from literature. Analysis of input parameters was done with the help of ArcGIS and MS-EXCEL. The input parameter maps were overlaid to assign a value and the combined effect of the parameters for the suspended sediment yield was calculated on a raster calculator. The runoff energy factor is the most sensitive parameter followed by the slope length and steepness factor. For the calibration period, the event-based mean observed and estimated suspended sediment yields were 0.2 and 0.23 ton/ha respectively with a little overestimation; while for the validation period, is 0.7 and 0.53 ton/ha with great underestimation. The model evaluation shows a 0.85 coefficient of determination, a 0.85 coefficient of efficiency, and an index of agreement value of 0.96 for the calibration period which can be considered a good model performance, whereas a 0.84 coefficient of determination, 0.65 coefficient of efficiency, and an index of agreement value of 0.83 for the validation period. The result revealed the use of the MUSLE model in its original form without calibration is not proper. The evaluation result shows the model with the appropriate calibration is a good estimator of sediment yield so that it can be used for spatial prioritization of soil and water conservation need within the watershed or extrapolated to the neighboring similar watersheds