| Popis: |
In coastal regions, the higher density of the saline groundwater influences the flow pattern of the lighter fresh/brackish groundwater as it makes its way to the sea. In order to get the best possible understanding of the flow patterns when modelling such coastal aquifers it is necessary to take the density difference into account, e.g. by setting up a SEAWAT model. However, SEAWAT can easily become unstable if the cell volumes across the model are not uniform, and it does best if several model layers represent each geological unit [Guo & Langevin, 2002]. These requirements make it difficult to assign values of hydraulic parameters to the model, as the geological units rarely are uniformly distributed across the modelled area. Besides this, when considering density differences and solute transport, the complexity of the calculations means that the runtime of such a model is much longer than for a density-independent model of the same system. In my Master’s thesis I set up a 5-layer density-independent MODFLOW model with non-uniform cell volumes, and calibrated it from measurements of hydraulic head. Subsequently, this model was “translated” into a 10-layer density-independent model with uniform cell volumes, and then further developed into a SEAWAT model, and it was shown that these three models produced very similar results, as far as hydraulic heads go. In this talk I will present the technique that was used for translating the model, and show that the models do indeed produce similar results. |
