Modelling Saltwater/Freshwater Interface

I am developing a 3-D model for the freshwater-saltwater interface in an unconfined aquifer on a small island. I am able to calibrate head levels for the island, however the model has the saltwater concentrations far too high. For example, at one well, with a head of 3m, our geoelectrical soundings have shown 300mg/l chloride concentration to be at -120m, where as the model calculates that it is at a depth of -12m.

I have adjusted the boundary conditions for the head along the coastline to compensate for the density difference, as suggested in the white papers. Is there another parameter that I need to adjust in the model that I am overlooking? Any advice on how to deal with this problem is definitely appreciated.

Thanks

Talking about other parameters, you have to make sure that the density ratio and reference concentrations are put correctly. I would try to not use upwinding techniques and keep the dispersivity in a reasonable range.