Scholarlink

<p>The green energy transition has intensified global demand for critical minerals, driving the expansion of mining activities with significant environmental consequences. In response, we present a globally consistent dataset of land use and land cover classification within mining areas, providing detailed information for over 80,000 recognised mining extents across 150 countries, spanning 95,644 km² and offering global-scale insights into the mining footprint. Developed through the integration of Sentinel-2 imagery and TanDEM-X elevation change data with a Random Forest classifier, this dual-source integration supports the differentiation of functionally different but spectrally similar land use types, such as open pits and waste dumps. This distinction is critical because different land uses pose varying environmental risks. By accurately identifying specific land use types, rather than treating all disturbed or adjacent areas as equally impacted, the dataset avoids overestimating mining-affected land. Ultimately, it provides a more accurate depiction of land use within mining areas and significantly improves the reliability of environmental impact assessments in the mining sector.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Classifying land use within 80,000 mining sites on a global scale

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月3日

查看全文
  • Yu-Tong Cheng,
  • Nguyen Tien Hoang,
  • Lou Maupu,
  • Keiichiro Kanemoto

摘要

The green energy transition has intensified global demand for critical minerals, driving the expansion of mining activities with significant environmental consequences. In response, we present a globally consistent dataset of land use and land cover classification within mining areas, providing detailed information for over 80,000 recognised mining extents across 150 countries, spanning 95,644 km² and offering global-scale insights into the mining footprint. Developed through the integration of Sentinel-2 imagery and TanDEM-X elevation change data with a Random Forest classifier, this dual-source integration supports the differentiation of functionally different but spectrally similar land use types, such as open pits and waste dumps. This distinction is critical because different land uses pose varying environmental risks. By accurately identifying specific land use types, rather than treating all disturbed or adjacent areas as equally impacted, the dataset avoids overestimating mining-affected land. Ultimately, it provides a more accurate depiction of land use within mining areas and significantly improves the reliability of environmental impact assessments in the mining sector.