Period of performance:
Over one-hundred million people in Bangladesh, India, Nepal, Pakistan, Cambodia, Vietnam, and Myanmar are chronically exposed to dangerous concentrations of arsenic because they rely on groundwater that is contaminated by naturally occurring arsenic. The most effective response to this crisis has been to encourage people to switch to safe wells.
In Bangladesh, where the problem is most severe, more than 50% have switched to either neighboring safe wells or newly installed deep wells that extract from below the contaminated level. The long-term risk of this strategy is clear: It relies on the continued production of safe water from regions of aquifers that are dangerously contaminated nearby. Although the strategy appears to be working for most wells, there is already evidence that contamination is increasing in some areas where groundwater flow is accelerated by municipal or irrigation pumping. At this point, we have little capacity to predict long-term groundwater safety even as aquifers are extensively pumped.
Harvey and his team seek to develop better predictive models of arsenic contamination by taking advantage of the large number of disjointed and uncoordinated data sets (including their own) that have been collected and published over the last two decades. They will identify features, spatial patterns, and temporal trends that are common across data sets of arsenic concentrations. Based on this analysis they will determine the types of models that either can explain, or are inconsistent with these factors.
The results of this work will be used to (1) devise future data collection methodologies targeted for testing important hypotheses, and (2) make recommendations for siting safe wells in areas with contaminated groundwater and for managing groundwater resources.