Remote sensing-based monitoring of land use and cover dynamics in surface lignite mining regions: a supervised classification approach

Monitoring land use/cover changes (LUCCs) in mining areas is vital for evaluating the environmental impact of mining operations and assessing the effectiveness of land reclamation works. Mining activities directly influence the surrounding environment, shaping the social and economic development of the region in which they occur. The present research aims to investigate the spatiotemporal dynamics of the LUCCs within the Ptolemais lignite surface mines and the surrounding region in Northern Greece. In this context, satellite imagery from 1988 to 2023 was analyzed using supervised classification techniques at five-year intervals. Specifically, four major land use/cover classes were evaluated regarding their expansion employing the Support Vector Machine (SVM) classification model within a Geographic Information System (GIS) environment for the three classes: "high vegetation cover", "low vegetation cover", and "barren soil". Additionally, the "urban areas" class was manually incorporated into the classification results to enhance model accuracy. Subsequently, statistical relationships among land cover areas inside and outside the environmental boundaries of the mine were investigated through the Ordinary Least Square (OLS) regression model. The obtained results underscore the presence of both positive and negative correlations attributed to mining activities not only within the designated mining area but also to the surrounding landscape. This research demonstrated the effectiveness of remote sensing techniques for monitoring lignite surface mining landscapes.