Abstract
A 3D computational fluid dynamics study using Reynolds-averaged Navier–Stokes modeling was conducted and validated with field data from the Joint Urban 2003 dispersion study in Oklahoma City, Oklahoma. The modeled flow field indicated that the many short buildings in this domain had a relatively small effect on the flow field, whereas the few tall buildings considerably influenced the transport and diffusion of tracer gas through the domain. Modeled values were compared with observations along a vertical profile located about 500 m downwind of the source. The isothermal base case using the standard k–ε closure model was within 50% of the concentration measurements, and a convective case with ground and building surfaces 10°C hotter than ambient temperatures improved the modeled profile to within 30% of observations. Varying wind direction and source location had a marked effect on modeled concentrations at the vertical profile site. Ground-level concentrations were 6 times the observed values when the approach flow wind direction was changed by +15° and were nearly zero when the wind direction was changed by −15°. Similar results were obtained when the source was moved 50 m to the east and to the west, respectively. All cases underestimated wind speed and turbulent kinetic energy near the surface, although adding heat significantly improved the magnitude of the modeled turbulent kinetic energy. Model results based upon a Reynolds stress closure scheme were also compared with the vertical concentration profiles. Neither the isothermal case nor the thermal buoyancy case resulted in an improvement over the standard k–ε model.
* Current affiliation: Pacific Northwest National Laboratory, Richland, Washington
Corresponding author address: Julia E. Flaherty, Pacific Northwest National Laboratory, P.O. Box 999, MSIN K9-30, Richland, WA 99352. Email: julia.flaherty@pnl.gov
This article included in the Urban 2003 Experiment (JU2003) special collection.
