9.12
CFD MODELING OF FINE SCALE FLOW AND TRANSPORT
IN THE HOUSTON METROPOLITAN AREA, TEXAS
*Sang-Mi Lee, Harindra.J.S. Fernando,
Environmental Fluid Dynamics Program
Arizona State University, Tempe, AZ 85287-9809	EPA/600/A-04/087
Daewon W. Byun,
Institute for Multidimensional Air Quality Studies
University of Houston, Houston, TX
and
Jason Ching,
AMD/NERL, US EPA, RTP, NC 27711
(On assignment from the Air Resources Laboratory of NOAA to the NERLofthe US EPA)
1. INTRODUCTION
Fine scale modeling of flows and air quality
in Houston, Texas has been performed. A
computational fluid dynamics (CFD) model, a gridded
model in Eulerian frame, is applied to investigate the
influence of urban morphology on the sub-grid scale
transport and dispersion of pollutants in grid models
with grid sizes of the order of 1 km.
Meteorological flow fields of urban scales are
in the realm of meso and micro scales. According to
Orlanski (1975), meso a, p and y scales are of the
scale of 2000-200 km, 200-20 km, and 20-2 km,
respectively, and micro a, (3 and y scales 2 km - 200
m, 200-20 m, and less than 20 m. Mesoscale
phenomena including local thermally driven land-sea
and mountain-valley flows have been successfully
simulated by mesoscale meteorological models, such
as MM5 (Dudhia et al. 2003), RAMS (Pielke ef al.
1992), ARPS (Xue ef al. 1995), OMEGA (Bacon et at.
2000), and COAMPS (Hodur 1997). However, several
constraints limit the direct applicability of meso-scale
models for urban flows. While a mesoscale model
needs to be "scaled-down" and parameterized to
simulate urban wind flow, computational fluid
dynamics (CFD) models can be "scaled-up" and solve
for flow fields explicitly for the same purpose. CFD is
being widely utilized in engineering flow analysis and
building and structural design applications, and its
utility for urban wind flow predictions has been
increasingly recognized in recent years (Baik and Kim
1999). CFD has been used not only for urban flow
simulations, but also for estimating air pollutant
concentrations and human exposure (Cheatham et al.
2000; Emery et al. 2000; Chan et al. 2000, Huber ef
al. 2000). In the urban flow modeling community,
however, CFD is often viewed with caution as an
advanced, numerically expensive computational tool
with questionable utility in simulating meteorological
processes that are largely statistical rather than
deterministic in character (Lee et al. 2000). In
addition, specification of accurate boundary
conditions required for CFD is not tenable in urban
modeling. Recently, meso-scale and CFD models
have been jointly used to perform simulations of
urban wind flow in a nested configuration (Smith et al.
2000; Cox et al. 2000; Brown ef al. 2000). One of the
promising ways of simulating urban flows is to nest a
meso-scale model that generates a mean state of
meteorological variables with a CFD model dealing
with perturbations to the meso-scale flow by variability
within urban morphology.
Considering the inherent limitations and
algorithms of mesoscale models and CFD, we
simulated flow fields within urban morphology in a
nested application using the following methodology. A
mesoscale model provides an undisturbed
background flow field devoid of urban structures,
which reflects synoptic forcing as well as local
differential thermal forcing due to topography and
land-use type. Then, a CFD code is used to explicitly
resolve the flow fields around an urban building
canopy with initial and boundary values provided by
the mesoscale model and urban building morphology.
We employed one of the most widely applied
mesoscale meteorological model MM5 and a CFD
code (Lee and Park 1994, Kim and Baik 1999, Baik et
al. 2003) for examining flow and dispersion in a
commercialized area of Houston .
2. THE MODEL DESCRIPTION AND THE
CONFIGURATION OF THE SIMULATION
* Corresponding author address: Dr. Sang-Mi Lee
Arizona State University, Environmental Fluid
Dynamics Program/Civil & Environmental Eng,
P.O.Box 879809, Tempe, AZ 85287-9809
e-mail: smlee@asu.edu
web: http://www.eas.asu.edu/~pefdhome
The CFD code used in the study consists of
primitive governing equations, namely, the Reynolds-
averaged equations of momentum (with Boussinesq
approximation) as well as conservation equations for
heat, mass and passive scalar with closure realized
by 'eddy diffusivity' modeling of Reynolds stresses

-------
and turbulent heat and mass fluxes (Baik and Kim
1999, Baik et al. 2003). Eddy diffusivities, in turn, are
calculated using prognostic equations for Turbulent
Kinetic Energy (TKE) and the dissipation rate. The
governing set of equations is solved numerically on a
uniform grid system using a finite volume method. A
semi-implicit method is used for the pressure linked
equation (SIMPLE) algorithm (Patankar 1980).
A part of a commercialized area of Houston with
high buildings and deep street canyons were selected
for a CFD simulation (Fig. 1). The computational
domain has 352, 302, and 82 grid elements in x, y, z
direction, respectively. Grid spacing was 2 m in both
horizontal and vertical directions. A vertical profile of
winds from MM5 simulations was assigned as an
initial value for CFD modeling. The assigned initial
winds were westerly and slightly increased with
respect to height. The background meteorological
fields were assumed to be stationary during the 1-
hour period of urban simulation. The performance of
MM5 on various conditions has been validated by
numerous studies and thus here we focus mainly on
the CFD results.
3. PRELIMINARY RESULTS AND FUTURE
WORKS
Figure 2 represents the output of this urban
flow model generated under the given wind profile
produced by MM5. As evident from Figure 2, the
model predicts complex wind patterns such as flow
deflection, vortex wake zones, accelerations and
decelerations around the irregularly arranged building
clusters. Concentration fields are asymmetric rather
than following the Gaussian distribution due to the
building structures and are further modified by the
enhanced dilution at the lateral edges of buildings
parallel to the incident winds, possibly due to the
increase turbulence effect.
However, the trapping of pollutants within
street canyons located around at the position of (550,
300) from the origin in Fig. 3 that is perpendicular to
the incident wind was not very distinct in the
simulations. Enhanced mixing by turbulence at the
building top (which is suspected of overestimated)
and weak vortices induced by weak downward motion
at the leeside of the obstacles partly can be attributed
this phenomenon. A standard k-g model employed in
this study was found to overestimate the turbulent
kinetic energy around the frontal corner of bluff
obstacles and underestimate the lateral components
of normal stress in the recirculation region (e.g. see
Murakami et al. 1990).
In the future, this study will examine the
influence of lateral boundary conditions by utilizing the
output from an urbanized version of MM5 at 1 km grid
resolution which is to be implemented an advanced
urban module called DA-SM2U (Dupont et al., 2004).
This module requires an advanced set of urban
canopy parameters (UCPs) gridded at 1 km
resolution, which have been developed using the
same set of detailed, high resolution (order 1 m) fully
consistent data obtained for the CFD study. Note that
the current results presented in Figs 2 and 3 were
simulated by using the standard version of MM5
which does not have a sophisticated urban module.
During the examining of the sensitivity, the CFD
results will be contrasted against the use of other
meteorological inputs such as the obtained using (a)
the standard version of MM5, and (b) airport
observations.
Acknowledgment. This research has been funded in
part by the United States Environmental Protection
Agency through Grants R-82906801 and R-83037701
to the University of Houston. However, it has not
been subjected to the Agency's required peer and
policy review and therefore does not necessarily
reflect the views of the Agency and no official
endorsement should be inferred.
Disclaimer: This paper has been reviewed in
accordance with the United States Environmental
Protection Agency's peer and administrative review
policies and approved for presentation and
publication.
REFERENCES:
Bacon, D.P., Ahmad, N.N., Boybeyi, Z., Dunn, T.J.,
Hall, M.S., Lee, P.C.S., Sarma, R.A., Turner,
M.D., Waight, K.T., Young, S.H., and Zack,
J.W., 2000: A dynamically adapting weather
and dispersion model: The Operational
Multiscale Environment Model with Grid
Adaptivity (OMEGA), Mon. Wea. Rev., 128,
2044-2076
Baik, J.-J., J.-J. Kim, and H. J. S. Fernando, 2003: A
CFD model for simulating urban flow and
dispersion. J. Appl. Meteor., 42, 1636-1648
Baik, J.J., and Kim, J.J., 1999: A numerical study of
flow and pollutant dispersion characteristics
in urban street canyons, J. Appl. Meteor.,
38, 1576-1589.
Brown, M., Leach, M., Reisner, J., Stevens, D., Smith,
S., Chin, S., Chan, S. and Lee, B., 2000:
Numerical modeling from mesoscale to
urban scale to building scale, Proc. 3rd
Symposium on the Urban Environment, 14-
18 August, Davis, CA., 64-65.
Chan, S.T., Stevens, D.E. and Lee, R.L., 2000: A
model for flow and dispersion around
buildings and its validation using laboratory
measurements, Proc. 3rd Symposium on the
Urban Environment, 14-18 August, Davis,
CA., 56-57.
Cheatham, S.A., Cybyk, B.Z. and Boris, J.P., 2000:
Simulation of fluid dynamics around a
cubical building, Proc. 3rd Symposium on the
Urban Environment, 14-18 August, Davis,
CA., 46-47.

-------
Cox, C.F., Cybyk, B.Z., Boris, J.P., Fung, Y.T. and
Chang, S.W., 2000: Coupled microscale-
mesoscale modeling of contaminant
transport in urban environments, Proc. 3rd
Symposium on the Urban Environment, 14-
18 August, Davis, CA., 66-
Dudhia, J., D. Gill, K. Manning, W. Wang, and C.
Bruyere, 2003:PSU/NCAR Mesoscale
modeling system tutorial class notes and
user's guide: MM5 modeling system version
3, NCAR.
Dupont, S., T. L. Otte, and J. K. S. Ching, 2004:
Simulation of meteorological fields within and
above urban and rural canopies with a
mesoscale model (MM5), submitted to
Boundary Layer Meteorology.
Emery, M.H., Chtchelkanova, A.Y., Cybyk, B.Z.,
Boris, J.P., Shinn, J.H. and Bouveia, F.J.,
2000: Detailed modeling of air flow around a
complex building, Proc. 3rd Symposium on
the Urban Environment, 14-18 August,
Davis, CA., 48-49.
Hodur, R. M., 1997: The Naval Research Laboratory's
Coupled Ocean/Atmosphere Mesoscale
Prediction System (COAMPStm), Mon. Wea.
Rev., 125, 1414-1430.
Huber, A., Bolstad, M., Freeman, M., Rida, S., Bish,
E. and Kuehlert, K., 2000: Addressing
human exposures to air pollutants around
buildings in urban areas with computational
fluid dynamics (CFD) models, 3rd
Symposium on the Urban Environment, 14-
18 August, Davis, CA., 62-63.
Kim, J. J., and J. J. Baik, 1999, A numerical study of
thermal effects on flow and pollutant
dispersionin urban street canyons, Journal of
Applied Meteorology, 38, 1249-1261
Lee, I. Y., and H. M. Park, 1994, Parameterization of
the pollutant transport and dispersion in
urban street canyons, Atmos. Environ., 29,
2343-2349.
Lee, R.L., Calhoun, R.J., Chan, S.T., Leone, J., Jr.,
Shinn, J. and Stevens, D.E., 2000: Urban
dispersion CFD modeling, Fact or Fiction?,
3rd Symposium on the Urban Environment,
14-18 August, Davis, CA., 54-55.
Murakami, S., Mochida, A. and Hayashi, Y., 1990:
Examining the k-s model by means of a wind
tunnel test and large-eddy simulation of the
turbulence structure around the cube. J.
Wind Eng. ind. Aerodyn., 35, 87-100.
Orlanski, I, 1975: A rational subdivision of scales for
atmospheric processes, Bull. Am. Meteor.
Soc., 56, 527-530.
Patankar, S.V., 1980: Numerical Heat Transfer and
Fluid Flow. McGraw-Hill, 197 pp.
Pielke, R.A., Cotton, W.R., Walko, R.L, Tremback,
C.J.,	Lyons, W.A., Grasso, L.D., Nicholls,
M.E., Moran, M.D., Wesley, D.A., Lee, T.J.
and Copeland, J.H., 1992: A comprehensive
meteorological modeling system - RAMS,
Meteor. Atmos. Phys., 49, 69-91.
Smith, W.S., Reisner, J.M., Decroix, D.S., Brown,
M.J., Lee, R.H., Chan, S.T., and Stevens,
D.E.,	2000: A CFD model intercomparison
and validation using high resolution wind
tunnel data, 11th Conference on the
Applications of Air Pollution Meteorology with
A&WMA, 9-14 January, Long Beach, CA.,
41-46.
Xue, M., Droegemeier, K.K., Wong, V., Shapiro, A.,
and Brewster, K., 1995: ARPS Version 4.0
user's guide, pp380.

-------
600
500-
400
bJ)
Z;
300-
200
Easting (m)
0m<=z<7m 7m<=z<16m
19m <= z < 2\mU 21m <= z < 81m
700
16m <= z < 19m
Fig. 1. The building geometry in the computational domain.
Flow Pattern at 1 m agl
T T

\l )l v. 4>
1/ 2 / 2 i
I \C \C 2 2
3 300
700
Easting (m)
Fig. 2. Simulated flow field at 1 m above ground level. Filled polygons represents buildings.

-------
Concentration at 1 m agl
Easting (m)
< l.e-6	l.e-3 <= c < 3 | 3<=c<10
¦ 10<= c < 100 ¦ >= 100	unit (ppm)
Fig. 3. Concentration field of a tracer species 2300 sec after a release
at the positions of (100, 134) and (100, 420) from the origin.

-------