Abstract:
Understanding and predicting water quality require the concomitant knowledge of water origin and flow paths as stream and catchment hydrology are intimately linked. Bridging of various temporal and spatial scales is a key challenge in the numerical investigation of the terrestrial hydrologic cycle. A physically based numerical model for coupled surface and subsurface water flow with heat and mass transport has been developed in the software toolbox OpenGeoSys. The hydrological processes surface water flow, unsaturated, and saturated flow are described by diffusion type equations and solved with finite element and finite volume methods. New is the application of Lagrangian stochastic particles (random walk particle tracking) for the simulation of advective-diffusive/dispersive transport in coupled hydrosystems. Alternatively, Euler methods can be used. The coupling concept is a compartment approach. Typically, the hydrosphere is subdivided in surface, soil, and aquifer compartments, which interact via exchange fluxes at common interfaces. Each process is numerically solved with its own spatial and temporal discretization and an additional coupling loop is executed (partitioned coupling). A key to the object-oriented implementation of the compartment approach is a hierarchy of geometric, topologic (discretization meshes), and process libraries designed for multiphysics problems. The object-oriented environment of OpenGeoSys for high performance computing was used in the development of a regional hydraulic soil model. A central part of this work is the examination of the novel model with several application examples spanning hydrological and transport processes from laboratory to catchment scales: Two benchmark tests on Horton and Dunne overland flow, a modeling study on Cryptosporidium parvum oocysts, and three case studies - at the Lahn river basin in Germany, the Borden site in Canada, and the Beerze-Reusel drainage basin in the Netherlands.