TEMPORAL VARIABILITY IN ARSENIC CONCENTRATIONS IN SHALLOW GROUNDWATER: THE ROLE OF LOCAL HYDROLOGY

Our recent work in Bangladesh suggests that local recharge of shallow aquifers plays a role in setting the spatial variations of As in shallow aquifers (Stute et al. in review). This may be why monitoring of a suite of tube wells <10 m deep over a 3-year period indicates seasonal variations in groundwater chemistry, including dissolved As, that appear to be linked to the monsoon cycle (Cheng et.al, 2005). The objective of the present study was to take a closer look at the subsurface processes that control As concentrations in shallow aquifers by installing a series of very shallow (3-9 m deep) monitoring wells near (within 2m) three private wells that had been monitored previously. The composition of groundwater from the first of these private wells (No. 4115) showed large seasonal variations in dissolved As, Fe, Mn, S, and major cation concentrations. In the second private well (No. 816), the variations were less pronounced but there was a gradual increase in dissolved As from 50 to 70 ug/L over 3 years. The third site of shallow monitoring wells is a control site located near private well No. 808, where no detectable seasonal or gradual variations in groundwater As were detected over the 3 years.

The physico-chemical properties of groundwater at sites X (No. 4115), Y (No. 816), and Z (No. 808) including conductivity, Eh, major cation and anion concentrations, and dissolved As, Fe, Mn, and P concentrations have been measured since January 2005. At site X, average As concentrations in the shallowest and the deepest well were very different but varied relatively little over time ((2±1 ug/L at 4.5 m and 230±60 ug/L at 9 m, respectively). The chemical data from site X indicate that larger variations in the composition of groundwater at intermediate shallow well can be explained by conservative mixing of groundwater sampled by the shallower and the deeper well. In contrast, As concentrations at sites Y and Z were fairly uniform with depth (39±17 ug/L and 30±15 ug/L, respectively) and showed little indication of seasonal variability. We are investigating differences in the hydrology of the three sites that could explain these different spatio-temporal patterns.