EGU26-13973, updated on 14 Mar 2026
https://doi.org/10.5194/egusphere-egu26-13973
EGU General Assembly 2026
© Author(s) 2026. This work is distributed under
the Creative Commons Attribution 4.0 License.
Poster | Thursday, 07 May, 16:15–18:00 (CEST), Display time Thursday, 07 May, 14:00–18:00
 
Hall X4, X4.51
EuroMineNet: Continuous Multitemporal Monitoring of Mining Dynamics in the European Union
Weikang Yu1,2, Vincent Nwazelibe1, Xiaokang Zhang3, Xiaoxiang Zhu2, Richard Gloaguen1, and Pedram Ghamisi1
Weikang Yu et al.
  • 1Helmholtz-Zentrum Dresden-Rossendorf, Helmholtz Institute Freiberg for Resource Technology, (w.yu@hzdr.de)
  • 2Technical University of Munich
  • 3Wuhan University

Mining activities are essential for the global energy transition, but they remain major drivers of land surface transformation and environmental degradation. Reliable, scalable monitoring of mining-induced land-use change is therefore critical for sustainable resource governance. In our earlier work, MineNetCD (2024) established the first global benchmark for mining change detection, enabling the identification of abrupt mining footprint changes from high-resolution bi-temporal imagery across 100 geographically diverse sites. While this provided a robust foundation for static change detection, sustainable mining oversight requires tracking the continuous and often gradual evolution of mining activities over time.

To address this limitation, we introduce EuroMineNet (2025), the first comprehensive multi-temporal mining benchmark designed for dynamic monitoring across the European Union. Leveraging a decade of Sentinel-2 multispectral imagery (2015–2024), EuroMineNet provides annual observations for 133 mining sites, enabling systematic analysis of both short-term operational dynamics and long-term land-use transformations.

The dataset supports two complementary, sustainability-oriented tasks: (1) Multi-temporal mining footprint mapping, producing temporally consistent annual delineations; and (2) Cross-temporal change detection, capturing gradual expansion, reclamation, and episodic disturbances.

To assess temporal consistency under evolving conditions, we propose a novel Change-Aware Temporal IoU (CA-TIoU) metric. Benchmarking 20 state-of-the-art deep learning models reveals that while current GeoAI methods perform well for long-term changes, they struggle with short-term dynamics crucial for early warning and mitigation. By advancing from global static detection to regional continuous monitoring, this work directly supports the European Green Deal and contributes to the development of transparent and explainable GeoAI tools for environmental resilience.

How to cite: Yu, W., Nwazelibe, V., Zhang, X., Zhu, X., Gloaguen, R., and Ghamisi, P.: EuroMineNet: Continuous Multitemporal Monitoring of Mining Dynamics in the European Union, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-13973, https://doi.org/10.5194/egusphere-egu26-13973, 2026.