Assessing soil erosion and sediment yield in the Bou Rouina watershed (Northeast of Algeria): a RUSLE-SDR modeling approach

ABSTRACT Evaluating soil erosion is vital due to its significant environmental and societal impact. This research addresses the scarcity of measured data on soil erosion in North Africa by introducing a methodology using the Revised Universal Soil Loss Equation (RUSLE) and the sediment delivery ratio (SDR) to estimate soil erosion and sediment yield. Serving as an illustrative example and case study, this study contributes to the broader context of erosion research in subhumid regions. Key factors including rainfall erosivity (R-factor), soil erodibility (K-factor), slope length-steepness factor (LS-factor), and vegetation coverage factor (C-factor) were computed and mapped using GIS tools. Validation of the generated soil erosion map was performed using Receiver Operating Characteristic (ROC) analysis. Results showed mean values of rainfall erosivity at 148.02 MJ.mm.ha-1.h-1.yr-1 and soil erodibility at 0.034 t.ha.h.ha−1.Mj−1.mm−1. Steeper slopes exhibited higher LS-values exceeding 20, with a notable positive impact of vegetation coverage factor contributing approximately 16% to mitigating soil erosion. The mean annual soil loss was 6.81 t∙ha−1∙yr−1 with an SDR of about 45%, resulting in a sediment yield of 3.07 t∙ha−1∙yr−1. The Williams and Berndt model emerged as the most suitable for estimating SDR, with an AUC value of 0.736, indicating good accuracy. This study underscores the importance of using modeling frameworks to better understand erosion and sediment processes and guide sustainable conservation practices.