other complex phenomena due to the specific
configuration of the selected city or neighborhoods. In the case of the Mock Urban Setting
Test (MUST, [40]) it was determined that the
focus would be on a simplified representation of
a city by modelling it as a regular array of shipping containers, while keeping meteorological
conditions real. This study has helped to clarify
and give scientific evidence of specific flow phenomena as outlined in the previous sections. One
important aspect for dispersion application has
been the quantification of the deflection of the
mean plume axis relative to the incoming wind
direction induced. This can be significant, being
up to 23° for a wind direction of 27° relative to the
axis of the array, as found in the MUST experiment. Extrapolations to the real conditions are not
always straightforward. The full field experiments
mentioned above confirmed that, depending on
building density and morphology, transport and
dispersion may lead to complications not
observed in simpler building array distributions.
Near street level, for example, the dispersing
plume may travel several blocks in a direction
opposing the prevailing wind and many blocks
laterally. This topological dispersion can lead to
secondary sources and significantly alter the rate
of lateral dispersion. A ground-level source can
rise several hundred meters in depth in less than a
block when caught in the updraft just downwind
of a tall building. Buildings also alter the timing of
the transport and dispersion, generally resulting in
much longer residence times as compared to open
terrain. Even though it is difficult to generalize the
time of residence of a pollutant cloud, the Birmingham experiments suggested a few hours for a
single source. However, the Dapple experiments
did not give a definite conclusion on this.
A further observation important for air quality
applications is that the experiments have shown
the presence of stable stratification during the
night when consequently street level flow is less
well coupled to the upper level reference. This
might be more frequent in US cities than in
European cities. As observed by Barlow et al.
[39], stable conditions occur in London only occasionally. The analysis by Allwine et al. [35] carried out using data from the Joint Urban campaign
in July 2003, Oklahoma City, showed that a stable
layer is often observed overnight, with a nocturnal
jet. Thus, there is no common pattern regarding
the frequency of stable conditions; such conditions are a strong function of thermal advection
due to regional scale flow processes and also
depend on the specific materials used in the different cities.
The above are only some examples and are far
from being a complete discussion of specific
results obtained from the urban experiments.
Computational Models for Meteorology
and Air Flow in Urban Areas
This section discusses those computational
models for both meteorology and air flow in
urban areas that are used as input to models for
calculating dispersion. The numerical and physical basis and operational attributes are summarized in Table 4 for the three ranges of length
scale introduced in section “Characteristic
Regions of the Flow and Drivers”, and for the
different types of model, ranging from off-line,
fully computational models (FCM) to the fastest
approximate models (FAM) that can be run on
PCs. The choice of the model depends on their
practical purpose (in relation to dispersion modelling), the required level of spatial/temporal detail
and scientific understanding involved, and on the
detail and accuracy of meteorological and topographical input data. For research and for off-line
validation studies, a variety of types of model tend
to be used: some studies require great detail and
accuracy about particular or commonly occurring
situations; others require simpler models for rapidly calculating a wide range of meteorological
and topographic boundary conditions. When
models are analytically based, their predictions
can be easily understood and may be expressed
in useful formulae.
The reason why it is necessary to have different
types of model for the different distances is
because every type of model can only represent
a finite range of length scales, limited by the
capacity and speed of the computational systems
and data input.
174
Urban Air Quality: Meteorological Processes
configuration of the selected city or neighborhoods. In the case of the Mock Urban Setting
Test (MUST, [40]) it was determined that the
focus would be on a simplified representation of
a city by modelling it as a regular array of shipping containers, while keeping meteorological
conditions real. This study has helped to clarify
and give scientific evidence of specific flow phenomena as outlined in the previous sections. One
important aspect for dispersion application has
been the quantification of the deflection of the
mean plume axis relative to the incoming wind
direction induced. This can be significant, being
up to 23° for a wind direction of 27° relative to the
axis of the array, as found in the MUST experiment. Extrapolations to the real conditions are not
always straightforward. The full field experiments
mentioned above confirmed that, depending on
building density and morphology, transport and
dispersion may lead to complications not
observed in simpler building array distributions.
Near street level, for example, the dispersing
plume may travel several blocks in a direction
opposing the prevailing wind and many blocks
laterally. This topological dispersion can lead to
secondary sources and significantly alter the rate
of lateral dispersion. A ground-level source can
rise several hundred meters in depth in less than a
block when caught in the updraft just downwind
of a tall building. Buildings also alter the timing of
the transport and dispersion, generally resulting in
much longer residence times as compared to open
terrain. Even though it is difficult to generalize the
time of residence of a pollutant cloud, the Birmingham experiments suggested a few hours for a
single source. However, the Dapple experiments
did not give a definite conclusion on this.
A further observation important for air quality
applications is that the experiments have shown
the presence of stable stratification during the
night when consequently street level flow is less
well coupled to the upper level reference. This
might be more frequent in US cities than in
European cities. As observed by Barlow et al.
[39], stable conditions occur in London only occasionally. The analysis by Allwine et al. [35] carried out using data from the Joint Urban campaign
in July 2003, Oklahoma City, showed that a stable
layer is often observed overnight, with a nocturnal
jet. Thus, there is no common pattern regarding
the frequency of stable conditions; such conditions are a strong function of thermal advection
due to regional scale flow processes and also
depend on the specific materials used in the different cities.
The above are only some examples and are far
from being a complete discussion of specific
results obtained from the urban experiments.
Computational Models for Meteorology
and Air Flow in Urban Areas
This section discusses those computational
models for both meteorology and air flow in
urban areas that are used as input to models for
calculating dispersion. The numerical and physical basis and operational attributes are summarized in Table 4 for the three ranges of length
scale introduced in section “Characteristic
Regions of the Flow and Drivers”, and for the
different types of model, ranging from off-line,
fully computational models (FCM) to the fastest
approximate models (FAM) that can be run on
PCs. The choice of the model depends on their
practical purpose (in relation to dispersion modelling), the required level of spatial/temporal detail
and scientific understanding involved, and on the
detail and accuracy of meteorological and topographical input data. For research and for off-line
validation studies, a variety of types of model tend
to be used: some studies require great detail and
accuracy about particular or commonly occurring
situations; others require simpler models for rapidly calculating a wide range of meteorological
and topographic boundary conditions. When
models are analytically based, their predictions
can be easily understood and may be expressed
in useful formulae.
The reason why it is necessary to have different
types of model for the different distances is
because every type of model can only represent
a finite range of length scales, limited by the
capacity and speed of the computational systems
and data input.
174
Urban Air Quality: Meteorological Processes