buildings that are loosely aligned in rows or continuous canyons ((ii), (iii), and (iv) of Table 3).
Where the downwind buildings are nearly aligned
with the wind (Fig. 3e, cos f w/d), they distort
the wake’s vorticity so that the wakes tend to be
aligned with the building [29]. The recirculating
flow in the wake of the upwind row tends to extend
across the “street,” and there is usually a significant
mean velocity V s along the street. The mean flow
has a component along the street and, when the
external flow is not aligned with the “canyon,” a
swirling motion in the perpendicular direction that
tends to extend up to the top of the buildings. The
peak mean and fluctuating velocities near the
ground are comparable with those at the level of
the top of the buildings. The key quantitative
parameter for dispersion modelling is the ratio of
the mean wind along the street V s to the mean wind
U H above the buildings at the height H (in the
approach flow). Computation and field observations [23] have suggested a large range of this
ratio from 0.3 to 1.0. The analysis also suggests
that the value of V s can be quite sensitive to the
value of V s at the upwind end of the canyon. Thus a
major crossroad can have a large effect on the
canyon flows downwind.
In conclusion it has been shown that most of
the usual types of building/street shape and configuration found in urban areas can be categorized
in geometrical terms that also correspond to the
characteristic air flow patterns found in these situations for typical wind conditions. In particular
there is a major distinction between, on the one
hand, flows determined by random interactions of
wakes of well-separated buildings and, on the
other hand, flows where the interaction of wakes
is of less relevance, for example, highly organized
“canyon” like flows that occur when the buildings
form street canyons or, if they are separated, when
they are positioned along streets and flows in
enclosed areas (e.g., courtyards) or between very
densely packed buildings.
Urban Data
Experiments and dedicated simulations have
shown features of urban dispersion at the neighborhood and street buildings scale. This includes
how the mean transport direction differs from the
mean flow above the buildings, the rapid vertical
and horizontal spreading of plumes in wakes and
canyons, the more complete mixing by the
interacting wakes (with lower relative levels of
fluctuation compared with rural terrain), the transition of dispersion below the buildings level to
above them and the variation in concentration
downwind of sources and within the buildings.
In the context of urban air quality, the full use
of models based on an up-to-date understanding
of the flow as described in the previous sections is
somehow limited by the scarcity of available data.
Urban flow and dispersion models require input
information and source distribution data about the
mean flow field, turbulence, and source distribution. Until very recently velocity field data in
urban areas were quite sparse and typically
derived from the experiments in the USA. Some
progress has been made in recent years with some
field experiments performed in some European
cities even though one has to accept that datasets
from field experiments are yet not fully complete
given the inherent difficulty and the cost of urban
flow and dispersion experiments. This aspect has
motivated the increase of laboratory experiments
of flow and dispersion over simplified groups of
buildings (building arrays) as well as over scaled
reproduction of real city quarters. In addition to
field studies, laboratory experiments allow easier
validation of physical or numerical models
because of carefully controlled flow conditions
and the possibility of making numerous measurements. The laboratory experiments can be either
wind-tunnel experiments [30–32] or waterchannel experiments [33]. Some field studies
have focussed on real cities, releasing a tracer
gas in selected neighborhoods, for example:
Copenhagen [34]; Salt Lake City, “Urban 2000”
[35]; Oklahoma City, “Urban Joint 2003” [36];
Basel (BUBBLE experiment) [37]; London (the
Dispersion of Air Pollution and its Penetration
into the Local Environment known as the “Dapple” experiments) [38, 39].
The advantage of such experiments is to simulate flow and pollution dispersion in real conditions, but an important drawback is that the
complexity of the urban morphometry can
obscure the “true” impact of buildings from
Urban Air Quality: Meteorological Processes
173
Where the downwind buildings are nearly aligned
with the wind (Fig. 3e, cos f w/d), they distort
the wake’s vorticity so that the wakes tend to be
aligned with the building [29]. The recirculating
flow in the wake of the upwind row tends to extend
across the “street,” and there is usually a significant
mean velocity V s along the street. The mean flow
has a component along the street and, when the
external flow is not aligned with the “canyon,” a
swirling motion in the perpendicular direction that
tends to extend up to the top of the buildings. The
peak mean and fluctuating velocities near the
ground are comparable with those at the level of
the top of the buildings. The key quantitative
parameter for dispersion modelling is the ratio of
the mean wind along the street V s to the mean wind
U H above the buildings at the height H (in the
approach flow). Computation and field observations [23] have suggested a large range of this
ratio from 0.3 to 1.0. The analysis also suggests
that the value of V s can be quite sensitive to the
value of V s at the upwind end of the canyon. Thus a
major crossroad can have a large effect on the
canyon flows downwind.
In conclusion it has been shown that most of
the usual types of building/street shape and configuration found in urban areas can be categorized
in geometrical terms that also correspond to the
characteristic air flow patterns found in these situations for typical wind conditions. In particular
there is a major distinction between, on the one
hand, flows determined by random interactions of
wakes of well-separated buildings and, on the
other hand, flows where the interaction of wakes
is of less relevance, for example, highly organized
“canyon” like flows that occur when the buildings
form street canyons or, if they are separated, when
they are positioned along streets and flows in
enclosed areas (e.g., courtyards) or between very
densely packed buildings.
Urban Data
Experiments and dedicated simulations have
shown features of urban dispersion at the neighborhood and street buildings scale. This includes
how the mean transport direction differs from the
mean flow above the buildings, the rapid vertical
and horizontal spreading of plumes in wakes and
canyons, the more complete mixing by the
interacting wakes (with lower relative levels of
fluctuation compared with rural terrain), the transition of dispersion below the buildings level to
above them and the variation in concentration
downwind of sources and within the buildings.
In the context of urban air quality, the full use
of models based on an up-to-date understanding
of the flow as described in the previous sections is
somehow limited by the scarcity of available data.
Urban flow and dispersion models require input
information and source distribution data about the
mean flow field, turbulence, and source distribution. Until very recently velocity field data in
urban areas were quite sparse and typically
derived from the experiments in the USA. Some
progress has been made in recent years with some
field experiments performed in some European
cities even though one has to accept that datasets
from field experiments are yet not fully complete
given the inherent difficulty and the cost of urban
flow and dispersion experiments. This aspect has
motivated the increase of laboratory experiments
of flow and dispersion over simplified groups of
buildings (building arrays) as well as over scaled
reproduction of real city quarters. In addition to
field studies, laboratory experiments allow easier
validation of physical or numerical models
because of carefully controlled flow conditions
and the possibility of making numerous measurements. The laboratory experiments can be either
wind-tunnel experiments [30–32] or waterchannel experiments [33]. Some field studies
have focussed on real cities, releasing a tracer
gas in selected neighborhoods, for example:
Copenhagen [34]; Salt Lake City, “Urban 2000”
[35]; Oklahoma City, “Urban Joint 2003” [36];
Basel (BUBBLE experiment) [37]; London (the
Dispersion of Air Pollution and its Penetration
into the Local Environment known as the “Dapple” experiments) [38, 39].
The advantage of such experiments is to simulate flow and pollution dispersion in real conditions, but an important drawback is that the
complexity of the urban morphometry can
obscure the “true” impact of buildings from
Urban Air Quality: Meteorological Processes
173