Abstract Manganese (Mn) is an essential element for human nutrition with the bulk of intake typically coming from food. However, Mn concentrations in drinking water can be substantial, and high levels of overall Mn consumption may be toxic. Recognizing detrimental health effects particularly for infants, the World Health Organization lowered its guideline for Mn in drinking water from 400 µg l −1 to 80 µg l −1 in 2021. Here, to assess the implications of this change on affected regions and populations, machine learning was used with a large dataset of groundwater Mn concentrations and relevant environmental parameters to generate global prediction maps of manganese. Considering populations reliant on untreated groundwater, we estimate that about 180–220 million people are at risk to Mn contamination >80 µg l −1 in drinking water, a four to fivefold increase from about 45 million people at the previous guideline of 400 µg l −1 . The number of people potentially impacted by high Mn is comparable to that for geogenic arsenic (As) and fluoride, though Mn receives much less attention. Despite Mn and As accumulating under reducing geochemical conditions, the occurrence of high concentrations of Mn and As generally does not overlap. Since groundwater buffers climate impacts and supports half the world’s population, the presented novel risk maps help locate and secure safe sources.
Podgorski et al. (Tue,) studied this question.