<b>Machine learning-enhanced monitoring of global copper mining areas</b>,
Bibliographic record
Abstract
This dataset provides a comprehensive, site-specific global assessment of land use areas associated with copper mining activities as of 2022. Using machine learning methodologies applied to multispectral remote sensing data, we mapped and classified operational land-use features, including open-cut pits, waste rock dumps, and tailings storage facilities. The dataset covers a total of 1,313 copper mines across 80 countries, encompassing a combined spatial extent of approximately 7,267 km².Observations were made using Sentinel-2 satellite imagery, characterized by high spectral (13 bands, 443–2203 nm) and spatial (10 m) resolution. Parameters measured include spectral reflectance values, Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI), Bare Soil Index (BSI), Enhanced Vegetation Index (EVI), Index-Based Built-up Index (IBI), and topographic information derived from Digital Elevation Models (DEM). Operational data, including historical copper production, production capacity, mining intensity, start and end dates of operations, were integrated from Standard & Poor’s and the US Geological Survey databases.The temporal coverage of the dataset is the year 2022, ensuring temporal consistency and accuracy across global mining areas. Spatial coverage is global, with significant data density in regions such as Canada, Australia, China, the United States, Chile, Peru, and Mexico.The primary purpose of this dataset was to quantify and evaluate the environmental impacts of global copper mining, particularly land use intensity and potential ecological consequences of mining activities. This dataset enables detailed environmental assessments, aids in ecological risk management, supports supply chain sustainability studies, and assists policy-makers and stakeholders in improving resource management and minimizing ecological impacts.Data collection involved preprocessing Sentinel-2 satellite imagery via the Google Earth Engine (GEE) platform, applying cloud-free median composites, and training a Random Forest classification algorithm using manually collected sample points for accurate delineation of mining features. Model validation achieved an overall accuracy of 91.08%.The dataset is provided in vector format, containing polygons annotated with detailed operational attributes, specifically: 1. Name: Copper mine name; 2. Latitude: Latitudinal coordinates of the mine centroid (WGS84, decimal degrees); 3. Longitude: Longitudinal coordinates of the mine centroid (WGS84, decimal degrees); 4. P_C: Primary commodity extracted from the mine; 5. List_of_C: List of secondary commodities extracted; 6. A_S: Activity status of the mine (e.g., Active, Inactive); 7. State/Prov: State or province location of the mine; 8. Country: Country location of the mine; 9. Land_use: Area altered by mining activities (square meters); 10. Cum_Prod: Historical cumulative copper production (metric tons); 11. MI: Mining intensity (100 m²/ton)This structure facilitates easy integration and usability across multiple research and management disciplines. The Google Earth Engine (GEE) script used for remote sensing classification is also included to facilitate replication and further analysis.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.002 | 0.003 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.005 | 0.004 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".