CONCLUSION

A combined water and solute mass balance approach with uncertainty analysis was used to estimate time-varying ground-water fluxes at a constructed wetland (ENR). The combined water and solute mass balance was needed in order to estimate both ground-water discharge and recharge, rather than just the net exchange. However, reliable estimation of both ground-water recharge and discharge required comparisons with other independent approaches, because of the high uncertainty in estimating ground-water discharge. In the ENR, the main hydrologic interaction with ground water is recharge of surface water to the underlying aquifer system (approximately 31% of water supplied for treatment). Ground-water discharge was almost negligible in comparison (approximately 2.8%). In this constructed wetland, we found that time-varying pumping rates from a supply canal and highly fluctuating water levels within the wetland greatly affected the rate of recharge of surface water (and possibly contaminants) into the subsurface aquifer. Our finding suggests that hydrologic managers of wetlands constructed for water treatment might be able to use this knowledge to construct accurate water and contaminant mass balances and possibly to optimize performance of constructed wetlands by minimizing the migration of contaminants from treatment wetlands into adjacent hydrologic systems.