Historical and Future Land Use and Land Cover Dynamics in the Erer Sub-Basin Wabi-Shebelle River Basin, Ethiopia

Abstract

Rapid land use and land cover (LULC) change threatens ecosystem sustainability, yet its future patterns and dynamics remain insufficiently understood in developing countries like Ethiopia. The study aimed to evaluate the historical LULC changes and predict its future changes in the northeast of the Wabi-Shebelle River Basin (WSRB), Ethiopia. We used the Support Vector Machine (SVM) algorithms in ArcGIS Pro to classify LULC classes from multi-sensor Landsat imagery (MSS, ETM+, OLI). The Cellular Automata (CA)-Markov model and Land Change Modeler (LCM) in TerrSet with Idrisi software were used to predict future LULC changes. Between 1985 and 2022, water bodies, forests, and shrub lands decreased by 84.1%, 47.4%, and 22.6%, respectively, whereas built-up, agricultural, and barren lands increased by 2862.4%, 44.6%, and 75.5%, respectively. By 2062, under the business-as-usual scenario, water bodies, forests, and shrub lands are projected to decline by 80.1%, 40.1%, and 19.8%, respectively, while developed, agricultural, and barren lands are expected to increase by 566.3%, 18.5%, and 4.9%, respectively. The SVM and CA–Markov models effectively classified historical LULC and simulated future changes. Although LULC conversion is projected to continue, its rate is expected to slow as much of the highly suitable land, particularly in the upper region, has already been converted. Sustainable land management and targeted conservation measures should be prioritized to limit further LULC conversion and protect the remaining natural ecosystems, particularly in the upper region.