Effect of Variable-Density and Constant-Density Representations of Flow on Simulating Terrestrial Groundwater Discharge into a Coastal Lagoon

Terrestrial groundwater discharge (TGWD) can be an important pathway for pollutants into coastal water bodies. Thus, a reliable way to quantify it is essential for efficient coastal management practices. This study evaluated the feasibility of using constant-density models for estimating TGWD amounts into the Indian River Lagoon, which is a variable-density estuarine environment. Constant-density models were developed using MODFLOW, while variable-density models were developed using SEAWAT. The numerical models were calibrated to match the field measured head data under the lagoon. The amounts of TGWD into the IRL and hydraulic head distributions calculated by the two codes were compared over eight pairs of numerical experiments. Two of those numerical experiments used the calibrated model and field measured conditions, while the rest of them used modified versions of the calibrated models, including variable anisotropy ratio k, variable lagoon salinity $L_S$, and increased water table elevation by 5%. The results showed that the constant-density model is fairly accurate in estimating TGWD and head distributions at the calibrated k in the range of 1000–20,000 with an error not exceeding 9.4% under the actual measured field conditions. Even when $L_S$ was assumed to increase to ocean salinity value of 1.0, a case that rarely occurs in IRL, the calibrated constant-density model’s accuracy was not affected substantially. However, the constant-density model failed to represent the physics of the variable-density environment at k values lower than 1000, where the error exceeded 129%. Generally, the accuracy of the constant-density model was found to increase substantially at lower $L_S$ and higher water table elevations