Atmospheric Environment Vol. 32, No. 21, pp. 3795 — 3799, 1998
( 1998 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
1352—2310/98 $19.00#0.00

Division of Environmental Science, School of Biological Sciences, Sutton Bonington Campus,
University of Nottingham, Loughborough LE12 5RD, U.K.
(First received 29 September 1997 and in final form 20 January 1998. Published August 1998)
Abstract—This note presents vertical concentration profiles of various size fractions of suspended particulate matter, measured over the first three metres from ground in a street canyon, using a novel sampling
system. The daily average percentage difference of airborne particulate matter concentration between
receptor heights 0.81 and 2.88 m, with the lower height taken as reference, was more than 35% for PM
and more than 12% for the inhalable fraction, for 5 out of the 7 days of measurement. These preliminary
measurements, carried out during June-August 1997, consisting of 37 h of data, corroborate the idea put
forward by Colls and Micallef (Atmospheric Environment 31, 4253—4254), that different height groups of the
population are exposed to different concentrations. Air-quality standards, based on human exposure
estimates, should take this variation into account. These measurements also have implications for the siting
of urban air quality monitoring systems. ( 1998 Elsevier Science Ltd. All rights reserved
Key word index: Vertical concentration profiles, suspended particulate monitoring, human exposure, air
quality standards, street canyon.


Airborne particulate matter has been shown to have
adverse effects on human health and the environment
(QUARG, 1996). Government agencies in different
countries have set limits and guide values for the
concentration of various size fractions of suspended
particulate matter (SPM) with the intention of protecting public health and the environment. At the
same time, monitoring of airborne particulate matter
concentration has continued and in some countries
even intensified especially in urban locations where
concentrations tend to be highest. The main driving
force behind this is the correlation that exists between
concentration of the PM and PM
fractions of
airborne particulate matter, and morbidity and mortality among human subjects (Vedal, 1997).
Air-quality monitoring data is used in estimating
human exposure to air pollution, and hence it is
crucial for the measured concentrations to be representative of what humans breath in. Most of the time
this is not the case (Ott, 1980; Vostal, 1994). Colls and

*Author to whom correspondence should be addressed.

Micallef (1997) generated vertical concentration profiles of SPM and PM on both the leeward and
windward sides of a street canyon over the first 3 m
from ground level, using the SLAQ model (Micallef
et al., 1997). Two scenarios were considered; the first
was for heavy traffic with relatively low turbulence,
and the second with low traffic and high turbulence.
The first scenario exhibited higher vertical concentration gradient and concentration as compared to the
second scenario. These modelled results indicated that
different height groups of the population might be
exposed to different concentrations and that air-quality monitoring systems, which are normally sited at
about 2.5 m or higher above ground, might not be
measuring concentrations that are representative of
the concentrations to which human subjects are exposed. The measurements presented here corroborate
these ideas and confirm that further investigation into
the problem is warranted.


Vertical concentration profiles of the inhalable,
thoracic, alveolic, PM and PM airborne particle

1. respectively. street orientation and aspect ratio. for the inhalable and PM fractions. but equally spaced. In this work. 29 exhibited higher concentration for the inhalable fraction at 0. 0.g. RESULTS AND DISCUSSION Fig. which include vehicle-generated (thermal and mechanical) turbulence. with one end close to a large pedestrianised area in the town centre. On completion of monitoring at the uppermost level. Photograph showing the general features of the kinetic sequential sampling (KSS) system and the location in high street. Nearly. 1. which has been described in detail elsewhere (Micallef et al.. the average flow is 600 vehicles per hour on each of the two lanes carrying traffic in opposite directions.88 m.81 and 2. the measured hourly 10 average PM concentrations are well below the limit 10 value of 50 kg m~3 as a 24 h running average. set by the United Kingdom’s Expert Panel on Air-Quality Standards (1995). Loughborough. Out of a total of 37 h. all of the profiles 10 shown exhibit a decrease in concentration with height. measured from ground. the lift returns directly to the lower level before starting to ascend. The extent to which these factors are capable of altering the pollutant concentration at a given height depends on the geometry of the street. United Kingdom.3796 Short Communication Measurements were carried out using KSS system on high street. The street is confined on both sides by buildings with an approximate height of 17 m. 1.23.0. once again.104/5.35. Data were collected at different times of the day. e. 2. Measurements were carried out close to this end of the street (see Fig. environmental (convective and mechanical) turbulence and variation in traffic flow. The analysis of these data is discussed below. It 10 is clear that higher concentrations of both fractions were measured at the lower receptor height. the response very much depends on the diameter of the aerosol particles. and 28 for PM . In Fig. A total of 37 hourly average concentration data were collected for each of the six receptor levels. The variation in the hourly average traffic count is typical of a trunk road cutting across an urban area. This sequence ensures that the time for monitoring at each level is constant and that the time interval between successive measurements at the same level is also identical. for seven days spread over the time period June—August 1997. United Kingdom where measurements were carried out. Germany) strapped to a ‘‘lift’’ which ascends and descends vertically at predetermined time intervals. the heights considered are 0. 1. Leicestershire. The street is bound by two signalised intersections.30 and 2. fractions were measured using the kinetic sequential sampling (KSS) system shown in Fig. 2 this variation is represented by an ‘‘error bar’’ of length equal to twice the standard deviation of concentration measured at each of the receptor heights.. and has a width-to-height ratio of 1. but one should bear in mind that as with all optical particle monitors employing the light scattering principle. Figures 3 and 4 show time series of hourly average concentration measured at 0.88 m heights. The whole operation is controlled by a purpose-built electronic circuit that was interfaced with the optical monitor in order to relate accumulated data to the appropriate receptor heights.81. This feature was common to the majority of the profiles measured in the street canyon environment. In the KSS system. Variation in concentration at a given height depends on several factors. This ensured that different meteorological and traffic conditions were considered. and on different weekdays. Loughborough. The lift stops at six separate. level by level. 1). Figure 2 shows a sample from the hourly averaged vertical concentration profiles for the inhalable and PM particle fractions.77. At both receptor heights. efficiency tends to be low for relatively large particles (which are . in press) and hence only a brief description is included here. Grimm Labortechnik Ltd. Ainring.81 m height. In most cases variation in concentration is not negligible but this does not undermine the hourly average concentration gradient and concentration. levels for monitoring to take place. airborne particle concentration is measured by a portable optical particle monitor (Model 1.

The length of each of the ‘‘error bars’’ is equal to twice the standard deviation of the measured concentrations at the particular receptor level for the given particle size fraction. The smallest particle diameter recorded by the optical particle monitor is 0. 1996). 1996). personal communication)). Particles from these sources have different aerosol distributions but certainly the majority of particles from vehicle exhaust tend to be smaller than 1 km diameter (Greenwood. there are five principal sources of airborne particulate matter: vehicle exhaust. resuspended road dust. pre-collected by a cyclone sampling head with a cutoff at 15 km in the case of the monitor used in this work) and also for particles that are small compared to the wavelength of the monochromatic light source used (in this case. secondary aerosol particles formed by gas-to-particle conversion and the background aerosol.35 km (Grimm Labortechnik Ltd. concentrations may be under measured. tend to be under measured. The under measurement of the optical particle monitor has implications for the absolute concentrations but less so for the concentration gradient. United Kingdom on 10 July and 7 August 1997. Sample of hourly average vertical concentration profiles for the inhalable and PM particle size 10 fractions of airborne particulate matter measured on High Street. Hence. 2. airborne particulate matter concentration measured in the street environment. Consequently. In the street canyon environment. Figure 5 shows the daily average (for those hours during which measurements were carried out) per- .Short Communication 3797 Fig. by an optical particle monitor. Loughborough. Grimm. 780 nm from a solid state laser (H. for environments where the mode of the airborne particle distribution lies well below this value. particulate matter originating from tyre wear and brakes.

Time series of the inhalable fraction of airborne particulate matter measured at heights 0. For 5 out of 7 days.88 m. concentration at 2. the difference was more than 35% for PM 10 and more than 12% for the inhalable fraction. and this exception occurred for the inhalable fraction. Loughborough.88 m from ground in high street.81 m receptor height. This exceptional case may also be explained by the variation in the calculated percentage difference. United Kingdom. part of which is background aerosol transported to the confines of the street and hence not associated with any particular height. .81 and 2.88 m 10 from ground in high street.81 and 2.3798 Short Communication Fig. with the lower height taken as reference.88 m height was on average slightly higher than that at the 0. Only on 5 August 1997. 3. centage difference of airborne particulate matter concentration between receptor heights 0. whose standard deviation (the length of the ‘‘error bar’’ in Fig. Fig.81 and 2. United Kingdom. 5) is at least five times the size of the average value. Time series of the PM fraction of airborne particulate matter measured at heights 0. This discrepancy in the percentage difference between the two fractions may be explained by the fact that the main source of PM in the canyon is vehicle exhaust 10 that is released near ground for most vehicles. 4. Loughborough.

Swansea. REFERENCES Colls. J.88 m. B. American Journal of Epidemiology 134. Ainring. and Micallef. Third edition. Quality of Urban Air Review Group (1996) Airborne Particulate Matter in the ºnited Kingdom. wishes to thank the Commonwealth Scholarship Commission in London for sponsoring his doctoral research. S. J. 101—106.g. Vostal. M. and Colls. Stanford University. (in press) Kinetic Sequential Sampling (KSS) system: an automated sampling system for measuring vertical concentration profiles of airborne particles. diesels. 1991).M. Micallef.. and Colls. PM and PM . Department of Statistics. pp. (1997) Towards better human exposure estimates for setting of air quality standards. M. ii#176 pp. Sunyer. Ott. (1991) Effects of urban air pollution on emergency room admissions for chronic obstructive pulmonary disease. J. vii #30 p. Germany. 1. J. Proceedings of the ‘Vehicle Emissions Meeting’ organised by The Aerosol Society at the University of Wales. and babies in pushchairs. J. 4253—4254.. J. A. . R. Murillo. held at the University of Warwick. e. Atmospheric Environment 31. C. W. 49—54.106. although small.81 and 2.104.105. 551—581. United Kingdom. C.. Coventry. The length of each of the ‘‘error bars’’ is equal to one standard deviation of the parameter for the given particle size fraction. A. which have shown that adverse health effects increase by 1% for each 1 kg m~3 increase in black smoke (Sunyer et al. (1997) Critical Review—Ambient particles and health: lines that divide. California. with the lower receptor height taken as reference. Figure 5 implies two things. Proceedings of the 10 2.Short Communication 3799 Fig. Environmental Health Perspectives 102 (Suppl. 277—286. Greenwood. Anto. 1.5 11th Annual Conference of The Aerosol Society entitled ‘Aerosols: Their Generation. A. gasoline and CNG fuelled engines. (1997) SLAQ: a street level air quality model for vehicle-generated suspended particulate matter. (1994) Physiologically based assessment of human exposure to urban air pollutants and its significance for public health risk evaluation. A. United Kingdom. 5. First. SIMS Technical Report Number 32. 18—20 March. 4). Singh. S. 43 p. Future setting of air-quality standards should take into account consideration relating to siting of air-pollution monitoring systems for realistic estimates of human exposure associated with different groups of the population. J. Department of the Environment. R. samples taken by urban air-pollution monitoring systems sited in the confines of streets may not be representative of the concentration to which human subjects are exposed. (1980) Models of Human Exposure to Air Pollution. Expert Panel on Air Quality Standards (1995) Particles. Daily average percentage difference of concentration for the inhalable and PM fractions of 10 airborne particulate matter between heights 0. It is a problem that presents science with practical difficulties in ensuring health protection to these individuals but surely merits consideration at the earliest.. Acknowledgements—The authors would like to thank Christopher Norton Deuchar for his help during the monitoring campaign and the anonymous referees for their comments. Behaviour and Applications’.. United Kingdom. (1996) Automotive particulate size distributions including 2-strokes. 16 April. Micallef. Vedal. N. J. should be seen in the context of epidemiological studies of urban air pollution. The measured concentration differences and gradients. Journal of the Air and ¼aste Management Association 47. Department of the Environment. Journal of the Air and ¼aste Management Association. J. Grimm Labortechnik Ltd (1996) Manual for the Dust Monitor 1. J. J. HMSO. different height groups of the population may be exposed to different airborne particle concentrations and distributions. The latter aspect has implications for the health of certain groups in society. London. and Saez. Deuchar. Third Report. Secondly. children. as most systems are located at heights greater than 2 m for ‘‘practical’’ purposes.