On variable density surface water-groundwater interaction: A theoretical analysis of mixed convection in a stably-stratified fresh surface water - saline groundwater discharge zone

Understanding the discharge behaviour of saline groundwater into fresh surface water can be critical for the effective management of water resources. While variable density flow has been studied intensely in a number of settings, the role it plays on the discharge behaviour of saline groundwater into freshwater streams is often neglected when calculating salt loads into a stream. The aim of this study was to determine what role variable-density flow behaviour plays in surface water/groundwater interaction in a stably-stratified fresh surface water/saline groundwater interface. The mixed convection ratio M, a measure of the ratio of density driven flow to advective driven flow, was defined for a matrix of one-dimensional numerical simulations that employed both varying hydraulic and density gradients. Vertical salt breakthrough into the surface water only occurred in the advection dominated cases (M < 1) and the salt flux into the surface water increased with increasing groundwater concentration until M reached a value of 1. Beyond this, when the flow was driven by the density difference between the two fluids (M > 1) vertical discharge of salt into surface water did not occur and the saltwater/freshwater interface migrated downwards with increasing density differences between the two fluids. This study therefore shows that there is a critical concentration difference that maximises salt loads to a surface water body and that a density-invariant approach to estimate the salt flux into the surface water (as the product of flow velocity determined through a potentiometric analysis and groundwater concentration) may be inadequate, especially where large density differences exist between the fresher surface water body and the underlying saline groundwater. The study is a purely theoretical approach and conclusions were drawn from simplified 1D simulations. Hence, further laboratory and modelling work is needed to confirm and test the plausibility of these findings for more realistic 2D and 3D cases.