Estimation Of Runoff And Erosion Rates Using Agwa - Kineros2 Model: Application To A Mediterranean Watershed

Abstract

The Mediterranean area is classified as a region highly threatened by water erosion, which depends largely on climatic aggressiveness, erodible soils, sparse vegetation, and irrational human activity. Tleta watershed belongs to a young mountain ridge (Rif), readily erodible and highly vulnerable to water erosion. To assess the extent of this phenomenon, AGWA tool coupled with KINEROS2 (K2) model was used to predict the specific degradation and the hydrological functioning of the studied watershed. The data required consisted of topography, land use, soil type, and precipitation. The model simulations will make it possible to estimate the sediment load deposited at the Ibn Batouta dam, which results from upstream erosion. The evaluation criteria for the predictive skill of K2 model when comparing observed and simulated data are the coefficient of determination (R2) and the Nash–Sutcliffe model efficiency coefficient (NSE). These two coefficients are respectively (R² = 0.99, NSE = 0.71) and (R² = 0.99, NSE = 0.72) for runoff and sediments yield for calibration. For the validation process, these two coefficients are (R² = 0.99 and NSE = 0.98) for runoff and (R² = 0.99 and NSE = 0.97) for sediments yield. This allows us to conclude that the predictive accuracy of the model is promising.