Land Use and Land Cover Change and its Projected Transition Analysis Using ArcGIS Machine Learning Algorithm of Shewa Robit Watershed, Middle Awash River Basin, Ethiopia

Abstract

In Ethiopia, the conversion of natural vegetation cover into agricultural land is the main factor driving changes in land use and land cover (LULC). Using Landsat imagery for both historical and present-day LULC mapping and a machine learning-driven CA-Markov model was used to predict future transitions, this study examines changes in LULC in the Shewa Robit watershed. In addition, ArcGIS 10.5 was used for supervised image classification, and IDRISI Selva version 17 was used for the land change modeling and test transition probability. Significant land conversion trends, primarily a 46.46 km² increase in cultivated land at the expense of grazing and forest areas, were found in the LULC investigation undertaken between 2003 and 2023. These increases were brought about by rapid urbanization and agricultural expansion. According to the LULC projected, between 2023 and 2098, there will be a 91.4% increase in settlements and a 68.3% increase in cultivated land. This change in LULC highlights the threats to biodiversity, water resources, and soil quality that come with the fast growth of agriculture and urbanization. This study emphasizes the need for sustainable land management practices to prevent further environmental harm, preserve ecosystem services, and ensure the watershed's resilience to the projected changes in land use.