Concurrent 24-hr samples of particulate matter of median aerodynamic diameter less than 10 m (PM 10 ) were collected over a 10-day period in August 2000 at four sites along a transect in west-central Scotland, UK (passing from the coast through the city of Glasgow) in line with the prevailing southwesterly wind. Each sample was analyzed for chloride (Cl Ϫ ), nitrate (NO 3 Ϫ ), sulfate (SO 4 2Ϫ ), ammonium (NH 4 ϩ ), calcium (Ca 2ϩ ), iron (Fe), and organic hydrocarbon material (OHM). The contribution from elemental carbon (EC) was estimated. Sampling days were categorized according to local wind direction, synoptic flow, and air mass back trajectories. Chemical mass balance (CMB) reconstruction of the following PM 10 components was derived for each wind direction group and at each transect location: ammonium sulfate ((NH 4 ) 2 SO 4 ), ammonium nitrate (NH 4 NO 3 ), sodium chloride (NaCl), gypsum (CaSO 4 ), OHM, EC, soil/surface dusts, and particle-bound water. The results showed that PM 10 at the coastal site was dominated by the marine background (NaCl) compared with the urban sites, which were dominated by local primary (EC and soil/resuspension) and IMPLICATIONS This study highlights the utility of a multisite transect aligned with the prevailing wind to identify changes in PM 10 chemical composition in air masses moving inland from a coastal region, through urban and periurban regions. Local, marine, and long-range PM 10 components were identified. The CMB calculation offers a pragmatic method for accounting for the major chemical sources contributing to PM 10 at a receptor. Interpreting the CMB of PM 10 is of policy relevance not only in helping identify sources but also because of the implications for health impacts. secondary sources (NH 4 NO 3 , (NH 4 ) 2 SO 4 , and OHM). There was evidence of Cl Ϫ depletion as NaCl aerosol passes over urban areas. There was also evidence of longrange transport of primary PM 10 (EC and OHM); for example, at the coastal site from transport from Ireland. The work demonstrates how the general approach of combining mass reconstruction along a transect with other information such as wind/air-mass direction generates insight into the sources contributing to PM 10 over a more extended spatial scale than at a single receptor.
TECHNICAL PAPER
INTRODUCTIONNumerous studies published over the last 20 yr have demonstrated a significant, linear association between the concentration of airborne particulate matter (PM) of median aerodynamic diameter less than 10 m (PM 10 ) and cardiopulmonary mortality and morbidity. 1-3 There appears to be no lower threshold in the dose response between PM 10 concentrations and health effects. 4 Evidence suggests that particle size, number, and chemical composition all ultimately determine the PM toxicological impact. 5 PM 10 has many sources, including anthropogenic, biogenic, geogenic, primary and secondary, local, and longrange. 6,7 To determine the source contribution to PM 10 present at any given receptor, it is necessary to gain as much k...