The effects of anisotropy and inhomogeneity on turbulence in urban areas

Understanding the turbulence characteristics of Atmospheric Surface Layer (ASL) flow in and through urban areas is essential for such problems as predicting transport and dispersion of contaminants in cities. Urban areas, however, have very different turbulence properties than their rural counterparts. Due to the complexity of the flow within the urban canopy, much of the past research has focused on the flow above the buildings in the urban roughness sublayer (URSL) and inertial sublayer (ISL) or has applied analyses intended for ISL flow on the flow within the urban canopy layer (UCL). In order to gain a better understanding of UCL flow the data from two field campaigns were analyzed. The first was the Mock Urban Setting Test (MUST) which was a simulated urban area consisting of 120 shipping containers at Dugway Proving Ground. The second was carried out in a real urban area (Oklahoma City, OK). A comparison of the spectra measured over the MUST array with spectra, measured in the undisturbed ISL over the salt playa in Utah's West Desert revealed that buildings broke down the large scale motions present in the ISL flow and redistributed the energy, momentum and heat fluxes over a small band of scales associated with building vortex shedding. Similar to previous research, both the MUST and the Oklahoma City experimental data show a decoupling between the Reynolds shear and normal stresses with increasing depth into the canopy from the proportional relationship that they share in the ISL. Spectral analyses from the Oklahoma City experiment reveal that in contrast to the suppositions of previous researchers, this decoupling was not due to a randomization of the turbulence within the canopy. Instead, the large flow scales associated with building vortex shedding were found to act against the smaller flow scales associated with gradient transport processes reducing the net momentum fluxes. The momentum and turbulent kinetic energy budget analyses of the Oklahoma City data revealed that the horizontal terms to the budgets were not negligible while the buoyancy and storage terms were found to be negligible within the UCL.