Encyclopedia of Environmental Science and Engineering, Volume I and II

1164 URBAN AIR POLLUTION MODELING

sources and imperfect mathematical formulations as well as

our imperfect sampling of air pollution levels, causes dis-

crepancies between the observed and calculated values. This

makes the verification procedure a very important step in the

development of an urban air pollution model. The remain-

der of this chapter is devoted to these four components, the

verification procedures, and recent research in urban air pol-

lution modeling.

Accounts may be found in the literature of a number of

investigations that do not have the four components of the

mathematical urban air pollution model mentioned above,

namely the source inventory, the mathematical algorithm,

the meteorological network, and the monitoring network.

Some of these have one or more of the components miss-

ing. An example of this kind is the theoretical investigation,

such as that of Lucas (1958), who developed a mathematical

technique for determining the pollution levels of sulfur diox-

ide produced by the thousands of domestic fires in a large

city. No measurements are presented to support this study.

Another is that of Slade (1967), which discusses a megalop-

olis model. Smith (1961) also presented a theoretical model,

which is essentially an urban box model. Another is that of

Bouman and Schmidt (1961) on the growth of pollutant con-

centrations in the cities during stable conditions. Three case

studies, each based on data from a different city, are pre-

sented to support these theoretical results. Studies relevant

to the urban air pollution problem are the pollution surveys

such as the London survey (Commins and Waller,^ 1967), the

Japanese survey (Canno et al., 1959), and that of the capital

region in Connecticut (Yocum^ et al., 1967). In these studies,

analyses are made of pollution measurements, and in some

cases meteorological as well as source inventory informa-

tion are available, but in most cases, the mathematical algo-

rithm for predicting pollution is absent. Another study of this

type is one on suspended particulate and iron concentrations

in Windsor, Canada, by Munn et al. (1969). Early work on

forecasting urban pollution is described in two papers: one

by Scott (1954) for Cleveland, Ohio, and the other by Kauper

et al. (1961) for Los Angeles, California. A comparison of

urban models has been made by Wanta (1967) in his refresh-

ing article that discusses the relation between meteorology

and air pollution.

THE SOURCE INVENTORY

In the development of an urban air pollution model two

types of sources are considered: (1) individual point sources,

and (2) distributed sources. The individual point sources are

often large power-generating station stacks or the stacks of

large buildings. Any chimney stack may serve as a point

source, but some investigators have placed lower limits on

the emission rate of a stack to be considered a point source

in the model. Fortak (1966), for example, considers a source

an individual point source if it emits 1 kg of SO 2 per hour,

while Koogler et al. (1967) use a 10-kg-per-hour criterion.

In addition, when ground concentrations are calculated from

the emission of an elevated point source, the effective stack

height must be determined, i.e., the actual stack height plus

the additional height due to plume rise.

Level of

uncertainty

Predicting the future

Modelling the science

Describe case

using available data

3

2

1

Evaluation of

model quality

Approximation to urban boundary layer

Representation of flow in urban canopy

Parameterization of roadside building geometry

representative?

Air quality

monitoring data

Meteorological

monitoring data

Modelled past

air quality

Past situation

Traffic

flow

data

precise?

accurate?

Atmospheric Dispersion

Model

Emissions per vehicle

Measured past air

quality

Future prediction

Modelled future air

quality to inform

AQMA declaration

Will climate change?

Will atmospheric oxidation capacity change?

How will traffic flow change?

How fast will new technology be adopted?

Emissions data

FIGURE 1 Schematic diagram showing flow of data into and out of the atmospheric dispersion model, and three categories

of uncertainty that can be introduced (From Colvile et al., 2002, with permission from Elsevier).

C021_001_r03.indd 1164C021_001_r03.indd 1164 11/18/2005 1:31:34 PM11/18/2005 1:31:34 PM