4404
L.Y. Chan, W.S. Kwok / Atmospheric Environment 34 (2000) 4403}4412
their fuel for low cost. Hence, PM and PM
ꢀꢁ ꢂꢃꢄ
have
been measured in this study, although only two PM
measurement could be conducted at the same time.
ꢂꢃꢄ
become the major source of street-level air pollution.
Studies had been conducted on the dispersion of
vehicular emission in street canyons in the past two
decades. Variation in concentration at a given height
depends on several factors, which include vehicle-gener-
ated turbulence, variation in tra$c #ow, meteorological
parameter and geometry of the street such as the street
orientation and aspect ratio. Streets with uniform build-
ing heights and con"gurations were commonly selected
as observation sites in early studies. In recent years, the
e!ect of street con"guration and building shape on air
dispersion and initial dispersion in the vehicular wake in
the canyon have gained much attention. Qin and Kot
(1993) measured three di!erent typical street canyons in
Guangzhou City. This study compared the air dispersion
in di!erent types of road in urban canyons and its result
indicates that the con"guration of a street is an impor-
tant factor that in#uences air dispersion. Street measure-
ments are in#uenced by local conditions and care must
be taken in the interpretation of the result, in the estima-
tion of urban air pollution levels, and in the comparison
of air quality in di!erent cities (Ruwim et al., 1996). In
Hong Kong, due to the lack of space, buildings are
usually in the form of high-rise towers and they are in
close proximity. The residential, commercial and indus-
trial areas are close and sometimes mixed together.
Therefore, the road con"guration and street environment
are di!erent from district to district. In this study, the
vertical pro"les of suspended particulates were measured
in selected buildings within di!erent street con"gurations
in an urban area of Hong Kong. In addition to TSP and
Minivol portable air samplers (MiniVol) with a #ow
rate of 5 l minUꢀ were used in this study. They were chosen
in this survey study because they are portable, battery-
operated and easy to mount on roadside light poles or
outside a building. The #ow rates were checked before
and at the end of the study to ensure a constant #ow rate
through the size separator. Included in the daily quality-
control reviews were checking for battery power, #ow
rate, volume of air sampled, elapsed time, leaks, and
unusual "lter conditions including damage, odor, discol-
oration, or loose particles on the "lter. Any abnormal
conditions were corrected and noted on the "eld forms.
An open top balance (Mettler AE163) with accuracy of
0.01mg was used to weigh the "lter paper which was
conditioned in an electronic desiccator before and after
sample collection for 24 h. The balance was placed on an
anti-vibration table on top of a concrete bench. The "lter
was stored #at in a petri dish and covered after condi-
tioning or weighing.
The experiment was conducted from early November
1998 to mid-January 1999. These months were chosen
because they are the months with the highest suspended
particulate concentration due to dry weather. The pre-
vailing wind direction was NE and the average relative
humidity was about 70% during this period. The samp-
ling duration was 24 h on three weekdays for each samp-
ling site. The four mentioned streets represent di!erent
con"gurations with respect to di!erent dispersion condi-
tions. The location of sampling sites is shown in Fig. 1
and the main features are listed in Table 1. They are
within the urban area and there is heavy road tra$c next
to them or nearby. According to their street con"gura-
tion they are divided into two groups: street canyon and
open street. Two street canyons with di!erent height-to-
width ratios and two types of open streets were selected.
For the open street, the street is very board, with build-
ings only on one side of the street or far from each other
(more than 10 m).
PM , we also measured the vertical dispersion of "ne
ꢀꢁ
particles PM which is less studied. The vertical coe$c-
ꢂꢃꢄ
ient of variation of PM will be helpful for the estimation
of the representativeness of the HKEPD monitoring
stations' measurement which is claimed to represent
the exposure of the entire population living within the
district.
2. Field study
3. Site description
The vertical distribution of TSP, PM and PM in
ꢀꢁ
ꢂꢃꢄ
di!erent street environments was measured in the "eld
study. Minivol Portable Air Samplers (AIRMETRICS)
were used for the suspended particulate sampling. Due to
3.1. Street canyon
One of the important factors that in#uence the hori-
zontal and vertical air dispersions in the street canyon is
its con"guration. The concentration distribution pattern
varies with the canyon type and scale. To increase our
understanding of the vertical dispersion of suspended
particulates, two street canyons with di!erent street con-
"guration and street environment were studied. One is
located in the central commercial area, in Causeway Bay
(CWB) of Hong Kong Island while the other is located
in an old residential district, Sham Shui Po (SSP) of
the limitation of equipment (only two PM
minivol samplers), the vertical distribution study was
inlets for
ꢂꢃꢄ
mainly focused on TSP and PM in this study. How-
ꢀꢁ
ever, the PM has attracted attention recently due to its
ꢂꢃꢄ
potentially hampering health e!ect and it is anticipated
to be subjected to control in the near future in Hong
Kong, similar to the USA. Moreover, there are only
limited studies on the vertical dispersion of PM
ꢂꢃꢄ
had also
and
their correlation with other PM. Thus PM
ꢂꢃꢄ
Briner
