An estimated 23 M people live on floodplains affected by potentially dangerous concentrations of toxic waste derived from past and present metal mining activity. We analyze the global dimensions of this hazard, particularly Pb, Zn, Cu and As, using a geo-referenced global database detailing all known metal mining sites, and intact/failed tailings storage facilities. We then use process-based and empirically tested modelling, to produce a global assessment of metal mining contamination in river systems, and the number of human populations, and livestock exposed. Worldwide, metal mines impact 479,200 km of river channels and 164,000 km² of floodplains. The number of people exposed to contamination sourced from long-term discharge of mining waste into rivers is almost fifty times greater than the number directly impacted by tailings dam failures.

Data on active (defined as still in operation in database sources published/accessed before August 29^th 2022) and inactive (defined in database sources as closed) metal mines worldwide, including their location, mineral commodities, and operational status, were compiled into the Water and Planetary Health Analytics (WAPHA) global metal mines database using QGIS. Mine information was acquired from the United States Geological Survey Mineral Resources Data System (73,917 mines worldwide), the BritPits database of the British Geological Survey (8,459 mines in the United Kingdom), the S&P Global Market Intelligence database (2,584 mines worldwide), and our own compilation of c. 100,000 additional mines from academic and gray literature, including regional data published by government agencies and industry. We also compiled a georeferenced global database of metal mining tailings storage facilities and tailings dam failures based on the ICOLD/UNEP 2001 compilation (Bulletin 121), the World Information Service on Energy, the World Mine Tailings Failures and Global Tailings Portal databases, in conjunction with our own compilation of source literature published by government and non-government organizations.Together these spatial data represent, to our knowledge, the most comprehensive compilation of metal mine locations to date. 

Data for (i) active mine sites and (ii) inactive mine sites are stored are stored as Excel spreadsheets. NB the number of active/inactive mines shown in the spreadsheets is less than that reported in Table S1, because proprietary data sources have not been included (i.e. MRDS, BRITPITS and S&P). Each spreadsheet lists mine names (column A), mine status i.e. active or inactive (column B), the principal commodity mined (column C), and lat/long coordinates (columns D & E).

Data for (iii) TSFs and (iv) TDFs are stored as zipped Shapefiles. Data should be uncompressed and then imported into any GIS programme that can read Shapefiles.

Modelling was implemented procedurally in MATLAB v9.9.0 (R2020b) with the open source TopoToolbox MATLAB program for the analysis of digital elevation models (https://topotoolbox.wordpress.com). Modelling workflow is presented in SI Figure S8 with example code available in the WAPHA database (Macklin et al code.pdf). Citations to software sources are given in the manuscript and SI.