CFD Simulation of Urban Environment: Thermal Effects of Geometrical Characteristics and Surface Materials
Numerical simulations of a street-scale urban environment are utilized to investigate air flow and heat transfer that affect temperature distribution, urban energy use and Urban Heat Island formation. The following processes are considered: solar load, buoyancy effects, weather conditions, canyon geometry and surface radiative properties. Unsteady simulations are performed based on Reynolds-averaged Navier-Stokes equations (RANS) and the finite volume solver FLUENT 14.5 is used. For a clear summer day in southern California, thermal effects of urban geometry, surface radiative properties and wind direction are numerically investigated to examine surface and canopy air temperature and heat fluxes. Ground surface material (albedo) was found to have the most influence on urban facade temperature and energy balance. Replacing asphalt with concrete as ground material increased the surface temperature up to 4.5 K. For large canyon aspect ratios (canyon height to width = 3/2) peak wall temperature decreased by 4 K; while ground temperatures increased up to 4 K larger at night compared to AR = 2/3. An energy balance analysis suggests this to be the result of balancing radiation by convective fluxes and large shortwave radiation flux during the day. Rotating the wind direction to be 45° off canyon axis altered wall and roof temperature up to 4 K and 2.5 K respectively while ground temperature was not influenced due to the high density of the studied case.
The flow passes over a matrix of 3x3 equally spaced cubes placed on a flat wall boundary on top of several soil layers. For a fully-developed flow field, periodic boundary conditions are used in stream/span-wise directions and a zero shear velocity condition is applied at the domain top. In the stream-wise direction, the mass flow rate is specified according to a measured velocity profile in Rotach 1995  scaled by an average wind speed in the weather forcing data. Turbulence in the flow field is modeled using k-epsilon Realizable model  with Standard Wall Function for near wall treatment .
The surface energy balance consisting of longwave (L) and shortwave (S) radiation, conduction (Qc) and convection (Qh) heat flux is as follow:
(1-a)S – L=Qh + Qc
where a is surface albedo.
The coupling of convection and radiation is done by simultaneously solving the energy equation and Discrete Ordinate non-gray radiation model . The top of the domain is set to radiate downwards at a sky temperature of 255 K to simulate longwave radiation from clear sky conditions. Solar ray tracing algorithm is used to account for direct solar radiation and shading. A constant temperature of 295 K is applied at the base of the deepest soil layer and as internal temperature in roofs and walls to solve the 1-dimensional unsteady conduction. Natural convection and buoyancy effects are considered and Boussinesq approximation is used. Simulations are initialized at 0000 LST and solar and weather boundary conditions are updated every 5 minutes in the simulation.
Figure presents the diurnal temperature cycle of walls, roof, ground and air temperature that is prescribed at the inlet. Simulation is performed with asphalt as ground surface material and aspect ratio of 0.3 and results are compared with the data from Yaghoobian et al 2010  for validation. The largest temperature of the urban surfaces on this day are 335 K for the roof (1200 LST), 328 K for ground (1430 LST), 311 K for building wall (1400 LST). The large thermal inertia of soil layers caused the time of maximum ground surface temperature to shift from the maximum of solar irradiation. Despite the difference in initial surface temperatures, good agreement in both amplitude and timing of the maximum ground temperature with  is observed.
The effects of ground surface albedo on urban temperature and energy balance is analyzed by comparing the results for asphalt and concrete as ground surface material. During the daytime the average ground temperature of concrete is significantly lower than that of asphalt. The maximum ground temperature decreases 4.2 K relative to asphalt, which is 1.5 % decrease for a 0.17 increase in albedo.
To study the effect of urban built-up density, a sensitivity analysis on building aspect ratio (AR) is conducted for 4 cases of AR (=H/W) from 1/3 to 3/2. It is observed that the difference between ground temperature at AR of 3/2 and 1/3 is approximately 3 K at early in the morning and 4 K at night. Average wall temperature increased with AR and maximum temperature shifted toward solar noon in AR=3/2, due to the change in shortwave radiation absorbed by the surface.
To further address the effect of geometry characteristics of urban areas, different wind angles are simulated. The effect of wind direction on roof and wall surfaces is found to be more pronounced than on ground surfaces due to the high density of the studied case. The roof temperature for the building under 45 degree wind angle decreased approximately 2.6 K at solar noon which is due to the increase in velocity and sensible heat flux at roof level. However, the velocity magnitude adjacent to wall surfaces decreased and caused the averaged wall temperature to be larger at 45° wind direction.
 E. H. Chui, G. D. Raithby: ‘Computation of Radiant Heat Transfer on a Non-Orthogonal Mesh Using the Finite-Volume Method', Numerical Heat Transfer, 23:269-288, 1993.
 M.W. Rotach: 'Profiles of Turbulence Statistics in and Above an Urban Street Canyon', Atmospheric Environment, 29: 1473-1486, 1995.
 T.-H. Shih, W. W. Liou, A. Shabbir, Z. Yang, and J. Zhu: ‘A New k-epsilon Eddy-Viscosity Model for High Reynolds Number Turbulent Flows', Computers Fluids, 24(3):227-238, 1995.
 B. E. Launder and D. B. Spalding: ‘The Numerical Computation of Turbulent Flows', Computational Methods in Applied Mechanics and Engineering, 3:269 -289, 1974.
 N. Yaghoobian, J. Kleissl, E.S. Krayenhoff: ‘Modeling the thermal effects of artificial turf on the urban environment', Journal of Applied Meteorology and Climatology, 49:332-345, 2010.