SECOND-GENERATION GLOBAL RISK MAP OF GROUNDWATER ARSENIC CONTAMINATION (Invited Presentation)

Naturally occurring arsenic in groundwater affects millions of people around the world. Odorless and tasteless, arsenic can present significant hazards to human health in the concentrations frequently found in nature. Due to generally not being measured in groundwater quality analyses and the fact that its health effects can be similar to those stemming from other causes, its presence can go undetected for a long time. In order to assess the extent of this problem, we have used machine learning to create an up-to-date global prediction map of groundwater arsenic concentrations exceeding the WHO drinking water guideline of 10 µg/L.

Over 200,000 measurements of arsenic concentration in groundwater were compiled from a wide variety of sources while excluding measurements known to have originated from a depth greater than 100 m. These were aggregated into 58,555 data points by taking the geometric mean of concentrations falling within 1-km square pixels, which corresponds to the resolution of the predictor variables used. A collection of 52 spatially continuous predictor variables with global coverage representing various climatic, geologic, soil and other parameters related to the dissolution and accumulation of arsenic in groundwater was assembled. Recursive feature elimination was employed to identify a subset of 11 variables, which were then used in a random forest model grown with 10,001 trees.

The resulting model predicts areas with high arsenic concentrations in groundwater on all continents. Known areas of groundwater arsenic contamination are identified as are new areas of potential geogenic arsenic contamination, including large sections of Central Asia, the Sahel region and Okavango Delta in Africa, and parts of the Arctic. Combining the global arsenic prediction model with household groundwater-usage statistics, we estimate that 94-220 million people are potentially exposed to high arsenic concentrations in groundwater. As groundwater is increasingly utilized to support a growing population and buffer against increasing water scarcity due to a changing climate, this model will help raise awareness, identify suitable areas for safe wells and guide where testing for arsenic should be prioritized.