Machine learning-enhanced monitoring of global copper mining areas.

Copper is one of the most critical minerals for the global transition to low-carbon energy. However, as copper mining activities expand worldwide, they often result in significant environmental impacts, yet the monitoring approaches and up-to-date databases remain limited. In this study, we present a high-resolution, site-specific database of global copper mining activities, developed using a machine learning approach that leverages Earth observation images and various dispersed data sources. Our database encompasses approximately 1,313 copper mines, covering an area of 7,267 km, and includes detailed monitoring of operational land use categories such as open pits, waste rock dumps, and tailings storage facilities as of 2022. Additionally, we analyse land use intensity at each mine site based on inferences of copper production levels to facilitate comprehensive comparisons and improved management strategies. This database can help to reveal the adverse impacts of copper mining behind the energy transition. The dataset is available for download from https://doi.org/10.6084/m9.figshare.28680863.v1 .