Environmental Engineering Reference
In-Depth Information
1.7 Four Diagnostic Urban Wind Flow and Dispersion
Models Tested with Joint Urban 2003 Field Data
S. Hanna
1
, J. White
2
, J. Trolier
3
, R. Vernot
3
, M. Brown
4
, H. Kaplan
5
,
Y. Alexander
5
, J. Moussafir
6
, Y. Wang
7
, J. Hannan
8
, R. Fry
8
, C. Kiley
8
,
and E. Hendrick
9
1
Hanna Consultants, Kennebunkport, ME, USA
2
US Army Test and Evaluation Command, Dugway Proving Ground, UT, USA
3
Science Applications International Corporation, Allentown, PA, USA
4
Los Alamos National Laboratory, Los Alamos, NM, USA
5
IIBR, Tel Aviv, Israel
6
ARIA Tech., Paris, France
7
US Army Research Laboratory, Adelphi, MD, USA
8
US Defense Threat Reduction Agency, Alexandria, VA, USA
9
Epsilon Assoc., Maynard, MA, USA
Abstract
There are several diagnostic wind flow models available for urban areas.
Mass-consistent principles are used to solve for the mean wind flow on an
approximate 1 km domain, within which detailed 3-D building geometries are
defined. To initialize the models, simple vortex structures are parameterized behind
and around buildings and in street canyons. An upwind wind profile must be pre-
scribed. The wind flow results are used as inputs to a Lagrangian particle dispersion
model, where turbulent velocities and time scales are parameterized based on the
mean wind solution and the 3-D building geometry. The developers of four of
these models are collaborating in a study where their models are run for two tracer
releases (one daytime and one nighttime) during the Joint Urban 2003 (JU2003)
field experiment. The four models are: Microswift/Spray (MSS) by SAIC, QUIC
by LANL, the urban Lagrangian model by the Israel Institute for Biological Research,
and 3DWind by the Army Research Laboratory. Nearly identical domains and grid
systems are used, and model outputs are produced at agreed-upon horizontal and
vertical cross-section locations and monitoring sites. All models use the same
input wind profile. The simulated patterns of wind fields and tracer contours are in
good qualitative agreement. For wind speed near the surface, the mean model
biases are less than about 20% and RMS errors are about 1-2 m/s. The tracer
concentrations also show low bias but much scatter.
Keywords
Air quality model evaluation, mass consistent wind models, urban dis-
persion, emergency response, JU2003 field experiment
