Event-based sediment yield modelling using MUSLE in north-central Nigeria

Constantine Mbajiorgu, Peter Ameh Adegede

Abstract


The misapplication of empirical hydrologic models with disregard to the spatial context of their development and application is rife. The Modified Universal Soil Loss Equation (MUSLE) model was calibrated and validated for Ofuloko Watershed (612ha) in North-Central Nigeria. Hydro-meteorological data for 20 rainfall events (occurring from 2nd June to 14th July 2017) were obtained from a station installed adjacent the watershed. The control points of the Ofuloko Watershed were obtained at 10m intervals from a survey conducted with a hand-held GPS receiver to delineate the watershed and generate the Digital Elevation Model (DEM) in Surfers 10 environment. Sediment yield was measured at the watershed outlet and soil samples obtained from the watershed was analysed in a laboratory. Curve Number method was used to estimate the runoff volume, peak runoff for the watershed was estimated using the Rational Method and the time of concentration was computed using the method developed by Kirpich. The slope length factor and the slope steepness factor were obtained using McCool equations. Correction factors were applied to the soil erodibility factor to obtain the actual values for tropical soils. Data from the first 10 storm events were used to calibrate the MUSLE while data from the remaining 10 storm events were used to test and validate the model. A comparison of the measured (observed) sediment yield and the simulated sediment yield using the Chi-square goodness-of-fit test showed that they are not significantly different at 5% significance level. The Nash-Sutcliffe Efficiency Coefficient value of 0.8805 which is acceptable for most hydrologic applications was also obtained after the comparison. Strict adherence to the application requirements of MUSLE as documented in this research has demonstrated its utility as a watershed management tool in the derived ecological zone of Nigeria.


Full Text:

PDF