Spatial assessment of soil erosion for prioritizing efforts on SDG15.3 in Ethiopia

Soil erosion has emerged as a significant global concern, posing a critical challenge particularly affecting natural resources and agriculture in emerging nations. Understanding the extent and spatial pattern of soil erosion is vital for effective planning and the implementation of targeted soil conservation strategies, especially under limited resource conditions. This research was carried out in the Gununo watershed, where soil erosion endangers agricultural productivity and environmental health. Primary and secondary datasets such as coordinate points, soil samples, digital soil map, meteorological data, digital elevation model (DEM), and Landsat images were collected. Using RUSLE model in the GIS environment, this study calculated a mean annual soil loss, identified high-risk areas, and prioritized subwatersheds (WHs) for intervention. The overall analysis was carried out by multiplying input factors together in a raster calculator to quantify soil loss of the entire watershed. The analysis revealed that annual erosion varied between 0 and 360 t ha −1 yr −1 , with an average of 22 t ha −1 yr −1 . Approximately 36% of the area was classified as experiencing moderate to very severe classes, contributing 72.2% of the annual soil loss. The finding indicated that cultivated and bare lands are the most vulnerable land use classes which comprise 73% of the annual loss. The northeast and central-west zones of the study area emerged as erosion hotspots. Based on average annual erosion rate, the subwatersheds WH-4, WH-5, WH-7, WH-8, WH-3, WH-9, WH-6, WH-2, WH-10, and WH-1, were assigned sequential priority levels from 1–10. Among these, the first six consecutive WHs, covering 57.5% of the total landmass, exceeded tolerable soil loss rates, highlighting their urgent need for intervention. This research highlights the significance of earth observation in advancing sustainable land management and contributing to the goals of the SDG 2030 agenda.