PARTICULATE MATTER < 10 #m (PM10) AND TOTAL
SUSPENDED PARTICULATES (TSP) IN URBAN, RURAL AND
ALPINE AIR IN SWITZERLAND
CH. MONN, i" O. BRAENDLI,:~ G. SCHAEPPI,t CH. SCHINDLER,§
U. ACKERMANN-LIEBRICH,§ PH. LEUENBERGER¶ and SAPALDIA TEAM
i" Federal Institute of Technology, Institute for Hygiene and Applied Physiology, Clausiusstr. 25, CH-8092
Ziirich, Switzerland; :~ Ziircher H6henklink Wald, CH-8636 Faltigberg, Switzerland; § Institute for Social
and Preventive Medicine, Steinengraben 49, CH-4051 Basle, Switzerland; ¶ Centre Hospitalier
Universitaire Vaudois, Rue de Bugnon, CH-1011 Lausanne, Switzerland
Abstract—
Ambient concentrations of particulate matter of less than 10 μm aerodynamic diameter were measured in Switzerland for a 1 year period in 1993 at a dozen urban, rural and alpine sites. PM10 concentrations ranged between 10 μgm -a (alpine) and 33 μgm -3 (urban). Highest concentrations were found at Lugano, in the south of the Alps, and in urban sites of the Swiss Plateau (350-670 m.s.l.). Rural levels were one-third lower than the urban levels and relatively homogeneously distributed. Alpine sites showed considerably lower levels. The ratios between PM1o and TSP were found to be between 0.6 and 0.75, and the highest ratios were found in the most polluted urban sites. Seasonal variation of particulate pollution showed peak levels during winter and autumn, mainly during cold temperature-inversions.
Statistical analyses have shown a good correlation between PM10 and NO z and SO 2, and TSP in urban areas.
1. INTRODUCTION
Recent epidemiological studies have shown that suspended particulate matter considerably influences respiratory health. Associations between suspended particulate matter and lung function parameters, respiratory symptoms and mortality have been found (Braun et al., 1992; Pope and Dockery, 1992; Dockery et al., 1993). Sino," 1991, a study in Switzerland (SAPALDIA: Swiss Study on Air Pollution and Lung Diseases in Adults) has investigated the relationships between prevalence of chronic lung diseases and longterm exposure to ambient air pollution.
In epidemiological studies, respiratory healthrelated effects were found for TSP and distinct finer fractions (e.g. PMlo, PM2.s). Respiratory health effects are biologically expected to be associated with particles smaller than 10#m passing the nose and entering lung alveoli. In addition to the particle size, the number of inhaled particles can be of great importance: a change of the median particle diameter from 1 to 0.1 #m increases the number of particles by more than a factor of a thousand for a constant mass fraction. This can cause problems in the macrophage clearing mechanism. Macrophage clearing is more efficient for a smaller number of larger particles than for very high numbers of fine particles (Kreyling,1994).
Total suspended particle mass fraction does not seem to be the best parameter for the explanation of respiratory diseases; it can, however, be a good indicator if the particle size distribution does not vary greatly between regions.
In the U.S.A., an air quality standard for PMto (annual mean: less than 50 #g m-3) was introduced in the late eighties because of its known adverse effect on the human respiratory system. It is assumed that PMto represents anthropogenic pollution, whilst the origin of larger particles is mainly natural. As well as the size distribution, chemical composition (e.g. acidity, sulphates, nitrates, PAHs) of particles can induce health-related effects (Spengler et al., 1990). The greatest proportion of these chemicals is found within the PM10 fraction.
Within SAPALDIA, air pollutants and meteorological factors were measured routinely at continuously operated monitoring sites. Some sites have a long tradition of monitoring TSP and sulphur dioxide dating back to the mid-seventies. In 1987, after introducing new air quality standards for Switzerland, more than two dozen continuous monitoring sites, mainly for gaseous pollutants, were installed. The air quality standard for TSP was set at 70/zg m-3 (annual mean) which was not exceeded in the last few years. Whilst for gaseous pollutants the measuring techniques were standardised, the measuring methods for total suspended particulates varied greatly between different sites. Therefore, interpretations on regional differences in TSP levels have to be made carefully.
In 1993, a PM10 measuring programme was initiated in order to quantify pollution levels at sites reflecting differences in climate and gaseous pollution.
As we used identical sampling devices in all places,uncertainties in particulate pollution derived from the use of different techniques were removed. As fine particles undergo long-range transport, a similarity in PM10 levels at rural sites was expected.
The major objectives of the PM10 measuring programme during 1993 were as follows:
• to quantify PM10 levels in urban, rural and alpine air and their seasonal variation;
• to assess the fraction of PM10 within TSP and its variation over sites and time;
• to analyse correlations between other air pollutants and meteorological factors.
2. EXPERIMENTAL
Study sites: Figure 1 gives the map of the PM10 measuring sites. Eight of the regions are SAPALDIA study sites (underlined). In addition, PM10 data from three sites of an additional study (SCARPOL: Swiss Study on Childhood Allergy and Respiratory Symptoms with Respect to Pollution, Climate and Pollen) are presented (Langnan, Ziirich, Berne).
Monitors for PMlo were installed at fixed site monitoring stations in SAPALDIA. Measuring sites can be grouped into three major categories: urban and suburban regions (Lugano, Geneva, Basle, Berne, Ziirich, Aarau), rural regions (Binningen, Wald, Payerne, Langnau) and alpine regions (Davos, Montana).
Detailed description of monitoring sites--Aarau: town centre (383 m.s.l, rooftop, 25 m above ground), Basle (St. Johann): town centre, local road (3 m above ground, 260 m.s.l.); Binningen (dose to Basle: rural/suburban, hill, 3 m.a.g.,
300m.s.l.); Davos: outside town on flat valley plateau (3 m.a.g., 1580 m.s.l.); Berne: University of Berne, Terrace (2 m.a.g., 533 m.s.l.); Geneva: town centre, residential/office area, local road (3 m, 380 m.s.l.); Lugano: town centre, local road (3 m.a.g., 280 m.s.1.); Montana: remote on slope near forest (3 m.a.g., 1350 m.s.1.); Wald: outside village, hill slope, no road (3 m.a.g., 630 m.s.l.); Ziirich: close to town centre, local road (5 m.a.g., 460 m.s.l.); Payerne: outside village, no road (2.5 m.a.g., 460 m.s.l.); Langnau: rural, no road (2 m.a.g., 700 m.s.l.) PMt o measurements: PM10 was determined using sharpcut, low-volume cascade impactors (MEM: micro environmental monitors) which were developed at the Harvard School of Public Health. This method has been used in other U.S. studies (Marple et al., 1987; Lioy et al., 1988). Air is drawn through a two-stage impaction inlet, where particles greater than 10 #m are collected. Smaller particles pass the impaction stages and are collected on Teflon filters (Type Gelman Sciences PTFE R2PJ041, pore size 2 #m, diameter 41 mm). The air flow was 4 ~' min- 1. The sampling period was one week. Teflon filters were stored in an environmental chamber with constant air humidity (50% relative humidity) and at temperatures between 20 and 23°C. Filters were weighed before exposure (Mettler: limit of detection 0.01 mg, precision _+ 5%). After exposure, the filters were dried in a dry chamber (silica gel) for a minimum of 24 h and then stored in the environmental chamber for a minimum of 24 h before final weighing. Measurements took place between the 4th of January 1993 and the 4th of January 1994. At Basle, the monitoring period was between Febuary 1993 and February 1994.
PM10 concentrations were calculated for a standard atmosphere on the Swiss Plateau for 950 mbar and 283 K.
TSP measurements and gaseous measurement: Different techniques were used for measuring TSP. We will present TSP data only for SAPALDIA sites, where known technology, such as "High-Vol sampling" (Digitel, for 24 hourly means on glass fibre filters), beta attenuation (FAG, continuous monitor) and a tapered element oscillating microbalance (TEOM) were used. High volume samplers were run at a volume of 40 m a h- 1, the other monitors with a flow of 1-3 m 3 h- 1. The treatment of glass fibre filters was similar to the treatment of PM10 Teflon filters. From these data, daily means were calculated.
In order to compare data with PM10 levels, weekly averages were computed. When more than three days of any week was missing, the week was excluded.
(Instruments used in the SAPALDIA study during 1993-- "Digitel": Binnlngen, Geneva, Ziirich, Payerne, Lugano; High-Vol. self construction: Aarau. "FAG": Davos, "TEOM': Wald. "Elecos": Montana.) Analyses from technical field comparison studies between these monitors showed that data from Digitel and FAG monitor corresponded within a range of _+ 15%. Data from TEOM were always lower than for the other monitors. This monitor is actually designed for PM10 measurements. It can be run with an inlet for TSP but field experiments showed readings much lower.
The instrument readings were similar to monitors determining PM10 mass fraction. Readings were about 30% lower
than for the other TSP monitors. As these observations were consistent, we reassessed the Wald TSP data by multiplying
TEOM readings with a factor of 1.4.
Gaseous pollutant~; were measured using standard techniques: SO2 (fluorescence: "Monitor Labs", "Horiba"), NOx (chemiluminescence: "Monitor Labs", "Horiba"), O 3 (UV absorption: "Monitor Labs").
Meteorological parameters for air temperature, relative humidity, global radiation and wind speed were obtained from the Swiss Network of Meteorology (Swiss Institute for Meteorology) which uses standard technology. At Wald, air temperature and relative humidity were measured at the local air pollution monitor.
Statistical analysis: Statist