EVALUATION OF LEAKAGE TO THE MEMPHIS AQUIFER IN THE DAVIS WELL FIELD, SHELBY COUNTY, TENNESSEE, USA USING TRITIUM - HELIUM-3 AND HYDROCHEMICAL DATA AND AN INVERSE APPLICATION OF AGE-DISTRIBUTION MODELING

Recharge to the partially confined Memphis aquifer in Shelby County, Tennessee, occurs, in part, through “windows” in the upper Claiborne confining unit. Windows are discrete zones of hydraulic connectivity in an otherwise continuous confining unit where leakage from an overlying or underlying aquifer can recharge a confined aquifer. In the present study, we use 3H -3He, geochemical, and hydrostratigraphic data along with an inverse application of age-distribution modeling to evaluate the quantity and source of leakage to the Memphis aquifer in the Davis well field, Shelby County, Tennessee.

Existing geochemical and hydrostratigraphic data were combined with geochemical and 3H-3He data collected during this study to identify trends in water quality and relationships with physical parameters such as depth to well screen, distance to known windows and distance from the bluff line. Two end-member water sources were identified from the geochemical and environmental tracer data: a pre-modern, low-TDS, mixed-cation bicarbonate water and a modern, moderate-TDS, Ca-Mg—bicarbonate water. Geochemical modeling was employed to determine mixing ratios of end-member waters in the Davis wells observed in this study.

Geochemical mixing ratios and 3H-3He –age data from multiple years were used with an inverse exponential age-distribution model to determine recharge flux through and distance to a window. The model results were subjected to a sensitivity analysis and compared to geochemical and tritium maps as well as hydrostratigraphic data. The results suggest that leakage to the Memphis aquifer in the Davis well field originates from the Mississippi Alluvial aquifer to the west of the well field; however, leakage through faults or hydrostratigraphic windows cannot be ruled out for wells in the easternmost part of the well field.