Integrated groundwater-surface water modeling at the neighborhood scale in urbanized hydrologic systems

Modification of the hydrologic cycle by urban development is influenced by fine-scale spatial characteristics of cut-and-fill topography, road networks, and subsurface utilities. To address impacts on both groundwater and surface water in an integrated manner, we are using ParFlow, a parallel distributed watershed model, to conduct high-resolution simulations. We are applying ParFlow across six watershed subbasins with drainage areas of 0.3-0.6 km² using a horizontal grid resolution of 10 m and vertical resolution of 1 m. Sites have been selected to represent a range of development intensity, age, and stormwater management practices, and each is instrumented for stage and discharge. A LIDAR-derived DEM defines model topography, and an orthoimagery and LIDAR-derived land cover classification from U. Vermont is used to develop model surface hydrologic properties. In some cases, portions of the watershed divide modified by large infrastructure elements, such as freeways, roads, and stormwater features, pose difficulties to overland flow routing within the model and to watershed delineation. In these cases, additional information, including the location of stormwater infrastructure, has been used to modify the DEM and represent where surface flow paths follow the storm drain network instead of topography. Results of these methods have improved estimation of domain extent and flow paths in overland flow tests of these basins. Boundary and initial conditions have been selected for each basin using legacy well data and a conceptual model of the Piedmont physiographic province hydrogeology. Steady-state simulations have been conducted in some cases to help refine model boundary conditions. Model spin-up has been conducted using surface forcing (P and ET) for the years 2008-2009 from NLDAS2 dataset. Ongoing analysis is focused on modeling the impact of development patterns and type of stormwater management. Challenges related to applying a coupled model in an urban setting have been addressed through developing and applying GIS processing techniques and mining additional data sources traditionally not utilized in watershed modeling. This approach has been successful in improving model representation of urban features within the existing conceptual framework.