Groundwater Quality Deterioration Driven by Urban Pumping in Mexico and Bangladesh

Abstract

Aquifers are the primary water source for many cities. As urban groundwater demand increases the water quality in the aquifers underlying the city and surrounding areas are nevertheless vulnerable to geogenic pollutants. Here I investigate the impact of intensive urban pumping on the groundwater quality within and around two cities: San Miguel de Allende (SMA) located in semi-arid central Mexico, and Dhaka, located in a humid monsoonal climate in central Bangladesh. In San Miguel de Allende, municipal supply wells are constructed as perforated casing across nearly their entire length. When unpumped, these long wellbores act as highly permeable conduit through which non-negligible ambient flow passes from one aquifer to another and causes cross-aquifer contamination. Gaining information on the chemical stratification of an urban aquifer will help understand the future water quality for this region with rapidly falling water tables. To do this, in Chapter II, I measured and modeled the changing temperature and concentrations of chemical constituents in the pumped water. I propose a feasible and low-cost approach to estimate the ambient flow rate and inferred the layered (two end-members) chemical composition across the aquifer that the long-screened well is screened. In Dhaka, the intensive urban pumping caused water levels underling Dhaka dropped more than 60 m over the last half century. The cone of depression extends 30 km beyond the city limits. This threatens the household access to safe water in and around Dhaka. In Chapter III, I simulate the expansion of drawdown cone with a numerical model and demonstrate that the downward migration of groundwater from ground surface to intermediate aquifer through the break in the clay aquitard was accelerated by the Dhaka pumping. I go on to show that Dhaka pumping captured recharge from the Meghna River to the underlying intermediate aquifer. Finally, in Chapter IV, I estimated the seasonal and annual water exchange flux between the Meghna River and the adjacent shallow aquifer to test the hypothesis that mass fluxes of dissolved arsenic (As) are the source of observed highly concentrated solid-phase As deposits in riverbank sediments. I did this by combining the newly estimated groundwater discharge with local dissolved arsenic (As) concentrations measured in 1962 wells. The mass fluxes over a 200-year time frame since the avulsion of the Brahmaputra River away from the area, which halted the building of new aquifer sediments, approximately equaled (with a factor of 2) the amount observed in the sediments.