Towards an integrated model of the surface-subsurface water cycle
De Maet, T.; Hanert, E. (2012). Towards an integrated model of the surface-subsurface water cycle, in: Crine, M. et al. Ph.D. Student Day ENVITAM (08.02.2012). pp. 21 In: Crine, M.; Vanclooster, M. (2012). Ph.D. Student Day ENVITAM (08.02.2012). UCL (Louvain-la-Neuve): Gembloux. 85 pp., more |
| Abstract | Surface water and ground water have for a long time been approached as separate components by hydrologists, engineers, and decision makers. In consequence, the relevance of interactions between groundwater and surface water for the aquatic ecosystems has frequently been underestimated. Recent years have experienced a crucial paradigm shift, progressing from defining rivers and aquifers as discrete, separate entities towards an understanding of groundwater and surface water as integral components of a continuum with strong mutual influences between ronoff/river, vadoze zone/aquifer and their interface. We have developed a coupled model of the surface water and ground water. Our model explicitly couples the 2D surface runoff with the 3D saturated-unsaturated subsurface hydrodynamics. We use the false transient method on the saturated zone, where the hydrodynamics is described by an elliptic equation. The surface-subsurface coupling is ensured by setting physical continuity of mass and pressure by appropriated boundary conditions (BC). Th model solves the governing equations with the discontinuous Galerkin finite element method (DGFEM) and an explicit time integration scheme. The DGFEM method allows for unstructured meshes and physical discontinuities between different types of soils, and is locally conservative. The mixed formulation of the Richards equation uses both the pressure head h and the water content w. On the one hand, w is used for the unsaturated zone, where it is know to be more efficient. On the other hand h is used for the saturated zone, where w is constant. The shallow water equation is approximated by the diffusive wave approximation. We will present the overall structure of our model and highlight its advantages. Then we will show its performances for some test cases. |
|
