In the heart of Limpopo province, South Africa, the Grootegeluk coal mine is not just a hub of energy production but also a site of significant environmental transformation. A recent study, led by Nothando Boikanyo Dhlongolo from the School of Geography, Archaeology and Environmental Studies at the University of the Witwatersrand, delves into the long-term land cover changes resulting from mining activities, offering crucial insights for environmental management and sustainable practices in the energy sector.
Published in the journal *Environmental Systems Research* (translated to English as “Environmental Systems Research”), the study employs advanced machine learning algorithms to model and predict land use and land cover (LULC) changes over the years. Using Landsat satellite imagery from 1990 to 2025, Dhlongolo and her team utilized eXtreme Gradient Boost (XGB) and Random Forest (RF) algorithms to classify and analyze transitions among major land cover classes—built-up, vegetation, and water.
The findings are compelling. “We found that the XGB algorithm outperformed RF, achieving higher spatial precision and superior accuracy,” Dhlongolo explains. “This is crucial for understanding the spatial and temporal impacts of mining on the surrounding geo-environment.” The study’s change detection analysis revealed consistent vegetation-to-built-up transitions, confirming extensive environmental disturbance due to mining operations and urban expansion.
Looking ahead, the study employs the CA–Markov framework to project future LULC patterns for 2035. The projections suggest continued vegetation loss, increased fragmentation, and limited natural recovery. “Our predictions indicate that without effective management strategies, the environmental impacts of mining will persist and even worsen,” Dhlongolo warns.
The implications for the energy sector are significant. Understanding and mitigating the environmental impacts of mining is essential for sustainable energy production. The study highlights the importance of integrating remote sensing and machine learning techniques for effective environmental monitoring and impact assessment. “By continuously assessing LULC changes, we can support sustainable land use planning and environmental restoration in mining-affected regions,” Dhlongolo notes.
The research not only sheds light on the current environmental state of the Grootegeluk coal mine but also provides a roadmap for future developments in the field. As the energy sector continues to evolve, the integration of advanced technologies and sustainable practices will be crucial for balancing energy production with environmental stewardship. This study serves as a beacon, guiding the way towards a more sustainable and responsible future for the energy industry.