|
Screening models are
the simplest and there is no need to consider meteorology. They will calculate worst case concentrations based on empirical
results from field observations, wind tunnel experiments and more advanced
models. There are quick and cheap so enable an initial impression of
concentrations which are likely to occur under certain conditions and so
determine whether further more advanced work is necessary.
Intermediate models are
mainly desktop computer based. They require more information than screening
models, such as diurnal traffic flow data, and can often include varying meteorology.
They may also be able to deal with more than one point, area or volume source.
They lack the versatility of the more advanced models but are less expensive
and require less staff resources.
Advanced models are
computed based and can produce more accurate results if accurate meteorological
and emissions data are available. They can deal with many point, area and line
sources simultaneously. They can have a variety of output types and different
averaging times. Special effects such as photochemistry, complex terrain and
building effects can be taken into account.
The choice of model
will depend on the size and nature of the municipality. Screening models may be
used to look at specific sites, either point or road, in isolation. When the
results are added to a known background concentration it may be apparent that
an air quality objective is likely to be breached and so further examination
with a more sophisticated model is required.
A small municipality
which has a limited number of polluting sources which are large enough to cause
problems may be able to use intermediate models to determine the scale of the
problem. However, a typical city will have many point sources and a complex
road system. It may be located in a valley or have complex terrain and so may
require an advanced model.
In some cases if an
environment which has very complex geometry needs to be assessed then physical
modelling using wind tunnels or three-dimensional computational fluid dynamic
using high powered computers may be chosen. These, however, are very expensive
and require specialist knowledge.
The choice of model is
also determined by the type and quality of data which are available. See Topic What data is required by atmospheric dispersion models.
Simpler models should consider
worst case conditions and not be used for detailed analysis. They may not be
appropriate for predicting future scenarios, especially over a large area such
as a city. Whatever model is used, some validation or testing usually by
monitoring will be required. |