Hydrologic control of temporal variability in groundwater arsenic on the Ganges floodplain of Nepal

• We analyze the temporal variability of elevated groundwater arsenic in Nepal.
• Unique hydrostratigraphy with thin sand aquifers (2 m) in thick (>50 m) surficial clays.
• Causes strong temporal variability in arsenic at all depths.
• Elevated arsenic correlates with geochemical indicators of clay interaction.
• Indicates a clay source for arsenic, mobilized by transient hydrologic processes.

Elevated arsenic in groundwater affects some 100 million people in South Asia, yet mitigation efforts are hindered by persistent uncertainty about the proximal source of arsenic and mechanisms for its mobilization. At the core of this uncertainty are the relative roles of surficial organic clays vs. deeper aquifer matrix iron oxyhydroxides. Temporal variations in groundwater chemistry can serve to distinguish the contributions of these two sources, and such variation is especially pronounced in headwater areas of the Ganges floodplain immediately adjacent to the Himalayan foothills (e.g. the Terai of Nepal). Tubewells down to 50 m in the Terai commonly exhibit cyclical, temporally-correlated variation in dissolved arsenic, iron and other species. In Nawalparasi, the most arsenic-affected district, these wells tap thin (2 m) gray sand aquifers embedded in a thick (>50 m) sequence of organic clays. Monsoon recharge refreshes these aquifers, temporarily minimizing arsenic concentrations. Post-monsoon, average groundwater compositions exhibit increasing trends in water–rock interaction (higher TDS, with cation exchange to form increasingly Na–HCO3 waters), arsenic and iron. This cycle can be repeated during dry-season precipitation events as well, revealing direct correlation between trends in degree of clay interaction (sodium fraction of major cations) and arsenic concentrations. During the year, reversals in vertical head gradient yield reversals in arsenic temporal trend, and downward gradients in the dry season correlate with increases in arsenic. Collectively these observations strongly support a model of reductive mobilization of arsenic from adjacent clays into aquifers, tempered by repeated flushing during periods of appreciable rainfall.