Estimating Soil Loss in the Gelda Watershed, Lake Tana Sub-basin, Ethiopia, Using an Integrated Approach (GIS, Remote Sensing, and the Revised Universal Soil Loss): Implications for the Lake Tana Ecosystem

Soil loss poses a serious threat to agricultural productivity and causes environmental degradation, particularly in Ethiopia. It depletes soil fertility, resulting in reduced crop productivity and accelerated sedimentation in Lake Tana, which directly threatens the lake’s ecosystem. While previous studies have estimated erosion at regional scales, localized assessments using high-resolution data remain scarce, limiting the precision of conservation planning. This study assessed soil erosion in the Gelda Watershed by integrating the Revised Universal Soil Loss Equation (RUSLE) with high-resolution Sentinel-2 (10 m) imagery and ALOS PALSAR DEM (12.5 m) to enhance spatial accuracy. Land use land cover (LULC) analysis, which achieved an overall accuracy of 84.33% and a kappa value of 0.812, revealed that cropland dominates the landscape (50%), while forest cover is restricted to 10%. The study estimated a mean annual soil loss of 176.4 tons per hectare per year (t·ha−1·yr−1), with extreme hotspots reaching 880 t·ha−1·yr−1 at steep slopes and cropland, significantly exceeding the tolerable limit of 18 t·ha−1·yr−1. Erosion risk mapping indicated that while 79.82% of the watershed faces low risk, 8.69% is subjected to moderate to extreme erosion. These high-risk zones are primarily concentrated in the steep, cultivated upper reaches of the watershed. By utilizing high-resolution sensors, this research provides a more precise identification of erosion hotspots than previous coarse-scale models, offering a targeted spatial framework for urgent soil and water conservation interventions.