Predicting Acid Mine Drainage Indicators Using Drone Data andMachine Learning Techniques

Acid Mine Drainage (AMD) poses significant environmental challenges, especially in mining-disturbed areas where sulfide-rich rocks oxidize, releasing acidic water with high concentrations of metals and sulfates. This issue underscores the urgent need for innovative and sustainable approaches to monitor and mitigate its effects on water quality and ecosystems.

To address these challenges, we integrated drone-derived multispectral data with machine learning (ML) techniques to predict key AMD indicators, including iron concentration, pH, and sulfate content. This approach enables efficient, high-resolution environmental monitoring, offering a scalable alternative to traditional resource-intensive methods. Our study, conducted in the Outokumpu mining area of Finland, demonstrates the potential of combining advanced technologies with strategic environmental management.

Given the limited availability of field-measured water quality samples (10 samples from three AMD-affected lakes and one non-AMD lake), we employed a novel data augmentation strategy. This included a window-based spatial data expansion method and the Synthetic Minority Oversampling Technique (SMOTE), significantly enhancing dataset variability and model robustness. These innovations align with the EU’s vision of leveraging cutting-edge technology for environmental resilience and sustainability.

Our findings highlight how integrating drone technology, ML, and data augmentation fosters a sustainable and efficient monitoring framework for AMD-affected regions. This approach aligns with the broader goals of the European raw material value chain, contributing to environmentally responsible resource management and innovation. By promoting cross-sector collaboration and showcasing the applicability of advanced monitoring techniques, our work supports the EU’s strategic objectives for a circular economy and sustainable development.

Acknowledgments: This work is part of the Secure and Sustainable Supply of Raw Material for EU
Industry (S34I) project, n.101091616, funded by European Health and Digital Executive Agency
(HADEA).