Data on land cover areessential for many facets of life, including politics, economics, and science. Since timely and precise land cover information is essential to the correctness of all subsequent applications, it is highly sought after. The purpose of this work is to select the better LULC classifier and investigate change detection using a more accurate classifier. Support vector machine (SVM) and random forest (RF) algorithms were applied to categorize LULC satellite data in the watershed. Six LULC classifications comprised the research area: bare lands, forests, shrublands, waterbodies, settlements and agricultural lands. SVM and RF have overall classification accuracies of 87.46% and 91.19%, respectively, and RF was selected for change detection analysis. According to the results, there was a growth in agricultural land of 6.44% between 2002 and 2012 and 14.94% between 2012 and 2022. Between 2002 and 2012, the settlement area grew by 72.17%, and between 2012 and 2022, it expanded by 21.44%. The forest saw a 48.27% decrease from 2002 to 2012 and a 14.94% gain from 2012 to 2022. Shrub land decreased by 8.16% between 2002 and 2012 and by 26.30% between 2012 and 2022. Additionally, there was a change in bare land between 2002 and 2012, when it increased by 74.05%, and between 2012 and 2022, when it decreased by 41.42. Consequently, utilizing an RF algorithm to detect changes in LULC is a crucial method for classifying multispectral satellite data to comprehend the best way to exploit natural resources, implement conservation measures, and make decisions regarding sustainable development. The study results provide useful information for policymakers and planners in the implementation of sustainable land resource planning and management in the context of environmental change.