Key messages
- Arsenic and fluoride naturally occurring in groundwater potentially affect the drinking water and health of hundreds of millions of people worldwide.
- Spatial prediction maps of the occurrence of these elements prioritize the areas for groundwater quality testing and allow safe and unsafe groundwater sources to be determined more quickly.
As groundwater is generally protected from the biological contamination found in surface water bodies it is frequently used as a drinking-water source. But some chemical elements can occur naturally in concentrations that pose a health threat when consumed on an ongoing basis.
The most significant natural (or geogenic) groundwater contaminants in terms of health effects and geographical extent are arsenic and fluoride. Globally, up to about 220 million people1 are potentially exposed to hazardous concentrations of arsenic, and 180 million2 to fluoride. In both cases, any negative health effects typically develop only after long-term consumption.
Any amount of arsenic can be detrimental to health, with chronic exposure leading to skin, vascular and nervous-system ailments, or cancer.3 To help guard against the worst effects, but also reflecting the practicalities of measuring and regulating arsenic, the World Health Organization (WHO) has set a guideline concentration of 10 µg/L for drinking water.4
Fluoride, on the other hand, gets added to toothpaste and, in some countries, to drinking water, for its positive effects on dental health and the prevention of cavities. However, too much fluoride can cause dental mottling, which is a largely cosmetic problem, albeit sometimes with negative social effects. Higher concentrations may lead to skeletal fluorosis,5 which is a debilitating stiffening of the joints. To reflect this balance between beneficial and detrimental fluoride concentrations, the WHO maintains a guideline of 1.5 mg/L for fluoride in drinking water. In warmer regions, some countries, such as India,6 opt for a lower limit of 1.0 mg/L to account for higher drinking-water needs in hotter weather.
Where is the contamination?
Both arsenic and fluoride are odourless and tasteless, so groundwater sources must be sampled and analysed to determine their presence. However, this requires special procedures and equipment and so typically is not included in a standard suite of basic water quality analyses. As such, the global coverage of measurements of arsenic and fluoride in groundwater is spotty, with many data points in some places and few to none in others, as can be viewed on the Groundwater Assessment Platform (GAP).7 As a consequence, the quality of groundwater sources in many regions remains unknown, particularly if the wells are relatively new, or if any common ailments in the community cannot be linked to a specific source, such as the drinking water.
To help identify groundwater with high arsenic and fluoride concentrations, prediction maps have been created using machine learning and large global datasets of arsenic and fluoride concentrations along with a variety of environmental predictor variables, including geology, climate, topography and soil properties.
Overlaying these maps with population and taking into account the rate of use of untreated groundwater, global risk maps of arsenic1 (Figure 1) and fluoride2 (Figure 2) have been produced that identify problem areas where groundwater quality testing should be prioritized. In the case of arsenic, most of the affected population (94 per cent) resides in Asia (Figure 1), whereas the fluoride-affected population is split mainly between Asia (51 per cent) and Africa (46 per cent). These maps can be viewed interactively on GAP or downloaded for arsenic8 or fluoride.9
What to do about it
Once groundwater sources containing high concentrations of arsenic or fluoride have been identified, the most straightforward and simplest mitigation measure is to switch to a water source that is confirmed to be safe. This may be possible as safe and unsafe sources can exist in close proximity, or at different depths, due to complex subterranean groundwater travel paths and varying geochemical conditions.
Other options include blending water sources to achieve a contaminant concentration that is compliant with local regulations, or employing filtering processes, though this can be expensive and is often pursued only if necessary.
Figure 1: Estimated population at risk to natural arsenic pollution in groundwater. Population consuming groundwater in risk areas potentially containing aquifers with arsenic concentrations >10 µg/litre. The bar chart indicates the distribution of affected population by continent, whereby nearly all (94 per cent) are in Asia. (Adapted from Podgorski and Berg, 2020.1)
Figure 2: Estimated population at risk to natural fluoride pollution in groundwater. Population consuming groundwater in risk areas potentially containing aquifers with fluoride concentrations >1.5 mg/litre. The pie chart shows the distribution of affected population by continent, whereby most of the population is found in either Asia (51 per cent) or Africa (46 per cent). (Adapted from Podgorski and Berg, 2022.2)
------------------------------------
1 Podgorski, J. & Berg, M. Global threat of arsenic in groundwater. Science 368, 845-850, doi:10.1126/science.aba1510 (2020).
2 Podgorski, J. & Berg, M. Global analysis and prediction of fluoride in groundwater. Nature communications 13, 1-9 (2022).
3 Smith, A. H., Lingas, E. O. & Rahman, M. Contamination of drinking-water by arsenic in Bangladesh: a public health emergency. Bulletin of the World Health Organization 78, 1093-1103 (2000).
4 WHO. Guidelines for drinking-water quality. WHO chronicle 38, 104-108 (2011).
5 Fawell, J., Bailey, K., Chilton, J., Dahi, E. & Magara, Y. Fluoride in drinking-water. (IWA publishing, 2006).
6 Reddy, K. N. Revised guidelines of National Water Quality Sub-Mission. (Government of India, Ministry of Drinking Water and Sanitation, New Delhi, 2017).
7 Eawag, Swiss Federal Institute of Aquatic Science and Technology. Groundwater Assessment Platform (GAP), <https://www.gapmaps.org> (2022).
8 Podgorski, J. & Berg, M. Podgorski_and_Berg_2020. ERIC/open, doi:https://doi.org/10.25678/0001ZT (2020).
9 Podgorski, J. & Berg, M. Podgorski_and_Berg_2022. ERIC/open, doi:https://doi.org/10.25678/0006GQ (2022).