Comparison of Subgrid Turbulence Closure Schemes in WRF-Chem for a Coastal Urban Area

The community WRF-Chem model has been widely used in the coastal urban regions such as the Greater Houston area. In a typical WRF model run, vertical mixing is parameterized within the boundary layer physics. It is assumed that there is a clear scale separation between the resolved and subgrid-scale eddies. This assumption may not be valid when horizontal grid spacing approaches 1 km or finer, and a fully three- dimensional subgrid turbulence closure should replace the parameterized mixing. While WRF has always had the option of applying horizontal and vertical diffusion explicitly in physical space, the surface fluxes provided by the surface layer and land-surface schemes were previously not coupled with the explicit diffusion. We have completed the coupling of heat and momentum fluxes with the subgrid turbulence closure schemes. We present results from a case study in which 200m grid spacing is used, and three different options for vertical mixing are tested: the explicit diffusion with eddy viscosities determined using a three-dimensional Smagorinsky turbulence closure, explicit diffusion with a prognostic turbulent kinetic energy closure, and a more conventional model set- up in which the Mellor-Yamada-Janjic boundary-layer parameterization is applied. The impact of these various closures on the transport and reaction of chemical constituents, as well as on the meteorological fields, is analyzed. It is found in this study that both the 1-D mixing parameterization and 3-D parameterization produce very similar vertical structure of the daytime ABL, but the simulated transport and dispersion of tracers and chemicals in the ABL are significantly.