We are now in a position to solve the puzzle of why dissolved arsenic concentrations are dangerously high in the groundwater of the Ganges Delta. Over the last five years several research groups have provided detailed characterizations of the static geochemical characteristics of groundwater and sediments in arsenic-contaminated aquifers. The challenge now is to determine how groundwater flow transports chemicals in and out of the subsurface, and hence controls subsurface biogeochemistry. We propose to develop novel hydrologic methods to characterize the complex spatial and temporal patterns of groundwater flow, and then to employ hydrogeologic models to study the evolution of groundwater geochemistry. By combining dynamic hydrological models with geochemical characterization we intend to answer key scientific questions that have thus far eluded us: Why has arsenic not been flushed from some aquifers? Combined estimates of groundwater residence times and arsenic retardation factors indicate that arsenic residence times are only decades to centuries, implying that arsenic should be flushed from the aquifers that are thousands of years old. Is dissolved arsenic supplied by a continuous source, or are high concentrations transient? Will arsenic concentrations change in the future? Our field injection-withdrawal experiments show that arsenic concentrations respond within days to biochemical perturbations. The adoption of dry-season rice cultivation has dramatically altered geochemical input and outputs from aquifers. How does this change affect subsurface geochemistry and dissolved arsenic concentrations? Why do arsenic concentrations differ between neighboring wells? Arsenic concentrations at nearby locations in grey-colored anoxic aquifers often differ greatly, despite similar sediment characteristics. Do these dramatic gradients result from the pattern of groundwater flow and recharge? What are the intellectual merits of the proposed activity? These questions can only be resolved by determining how groundwater dynamics control chemical input and output to aquifers over two timescales: (1) Seasonal cycle: The hydrology of Bangladesh annually cycles between Monsoon flooding and dry-season arid conditions when evapotranspiration greatly outstrips precipitation and irrigation water is pumped from aquifers to meet the transpiration demands of crops. This cycle drives water table oscillations that create seasonally varying oxic/anoxic conditions in soils and also drives water exchange between aquifers and surface water (rice paddies, ponds and rivers). (2) Anthropogenic changes over decades: The Ganges Delta has been dramatically altered over the last three decades by population growth and the advent of irrigated agriculture. Groundwater irrigation has changed the location, timing and chemical content of recharge. Anoxic irrigation water is ponded in rice fields over much of the land, thereby changing both the hydrologic budget and the biogeochemistry of recharge, potentially mobilizing arsenic from soil layers that may be rich in arsenic and iron (oxy)hydroxides. Furthermore, pumping changes flow-paths deep in aquifers, affecting both the rates and locations of recharge as well as groundwater exchange with surface water bodies that now receive much higher loads of untreated waste. We propose to: (1) Build on our successful field program in Bangladesh by extending our field characterization from vertical geochemical profiles at one location to three dimensional flow around this location; (2) Characterize recharge and discharge and map transient flow-paths through the aquifer by applying novel combination of natural isotope data and numerical inverse methods for groundwater flow; (3) Conduct a detailed study of geochemical fluxes through the bottom of a rice field, now a principle source of groundwater recharge at our site, and a very likely source of dissolved arsenic; (4) Construct predictive numerical models that couple groundwater flow and recharge with the biogeochemical transformations that control arsenic. What are the broader impacts of the proposed activity? This collaborative project is built on our successful and productive partnership over the last five years. We will continue to place a significant emphasis on the education and transfer of technology, with further exchange of students between BUET, MIT and Tufts. We also will continue to collaborate with other research groups including Stanford, EAWAG in Switzerland, UBC in Vancouver and UCLA. Our research findings should answer some key scientific questions and also help evaluate alternative arsenic mitigation strategies and better manage water resources in Bangladesh.
