Building-resolving simulations in mesoscale models using the immersed boundary method

Computational fluid dynamics codes using body-fitted coordinates are commonly used for simulations of atmospheric boundary layer flow in urban terrain. While these models handle complex urban terrain well, they have several limitations. First, it can be extremely time consuming to produce a high quality body-fitted grid. In addition, these models are often forced at lateral boundaries with idealized boundary conditions that do not reflect regional scale weather effects, and these models lack representations of atmospheric physics processes. On the other hand, regional mesoscale models are capable of simulating regional weather and atmospheric processes, but cannot explicitly resolve urban terrain due to the use of terrain-following coordinates which are limited to moderate surface slopes.

To improve prediction of atmospheric flows in urban terrain, we have implemented an immersed boundary method (IBM) into the mesoscale Weather and Research Forecasting (WRF) model. IBM is a gridding technique which allows complex terrain to be modeled without body-fitted coordinates or a coordinate transformation by allowing the topography to intersect the background grid and applying boundary conditions along the immersed surface. IBM has been shown to effectively represent complex geometries in computational fluid dynamics codes, but additional issues arise in numerical weather prediction codes, such as modeling surface fluxes, handling under-resolved terrain features, dealing with pressure-based coordinates, and seamless nesting between terrain-following simulations and those using IBM. These issues have been successfully addressed within our IBM implementation, and flow and dispersion around urban terrain can be explicitly resolved in the WRF model. Furthermore, urban scale simulations can be seamlessly nested into larger mesoscale simulations.

To demonstrate the ability of the IBM-WRF code to simulate atmospheric flow around complex structures, we will present results from building resolving simulations in urban environments in a nested configuration. Specifically, flow and scalar dispersion in the downtown core of Oklahoma City is modeled, and the results are compared to the Joint Urban 2003 field campaign.