Autor: |
Fruin S; Keck School of Medicine, Division of Environmental Health, University of Southern California, Los Angeles, CA 90089, USA., Urman R; Keck School of Medicine, Division of Environmental Health, University of Southern California, Los Angeles, CA 90089, USA., Lurmann F; Sonoma Technology Inc., #C, Petaluma, CA 94954, USA., McConnell R; Keck School of Medicine, Division of Environmental Health, University of Southern California, Los Angeles, CA 90089, USA., Gauderman J; Keck School of Medicine, Division of Environmental Health, University of Southern California, Los Angeles, CA 90089, USA., Rappaport E; Keck School of Medicine, Division of Environmental Health, University of Southern California, Los Angeles, CA 90089, USA., Franklin M; Keck School of Medicine, Division of Environmental Health, University of Southern California, Los Angeles, CA 90089, USA., Gilliland FD; Keck School of Medicine, Division of Environmental Health, University of Southern California, Los Angeles, CA 90089, USA., Shafer M; Wisconsin State Laboratory of Hygiene, Environmental Health Division, University of Wisconsin-Madison, Madison, WI 53706, USA., Gorski P; Wisconsin State Laboratory of Hygiene, Environmental Health Division, University of Wisconsin-Madison, Madison, WI 53706, USA., Avol E; Keck School of Medicine, Division of Environmental Health, University of Southern California, Los Angeles, CA 90089, USA. |
Abstrakt: |
To characterize exposures to particulate matter (PM) and its components, we performed a large sampling study of small-scale spatial variation in size-resolved particle mass and composition. PM was collected in size ranges of < 0.2, 0.2-to-2.5, and 2.5-to-10 μm on a scale of 100s to 1000s of meters to capture local sources. Within each of eight Southern California communities, up to 29 locations were sampled for rotating, month-long integrated periods at two different times of the year, six months apart, from Nov 2008 through Dec 2009. Additional sampling was conducted at each community's regional monitoring station to provide temporal coverage over the sampling campaign duration. Residential sampling locations were selected based on a novel design stratified by high- and low-predicted traffic emissions and locations over- and under-predicted from previous dispersion model and sampling comparisons. Primary vehicle emissions constituents, such as elemental carbon (EC), showed much stronger patterns of association with traffic than pollutants with significant secondary formation, such as PM 2.5 or water soluble organic carbon. Associations were also stronger during cooler times of the year (Oct through Mar). Primary pollutants also showed greater within-community spatial variation compared to pollutants with secondary formation contributions. For example, the average cool-season community mean and standard deviation (SD) for EC were 1.1 and 0.17 μg/m 3 , respectively, giving a coefficient of variation (CV) of 18%. For PM 2.5 , average mean and SD were 14 and 1.3 μg/m 3 , respectively, with a CV of 9%. We conclude that within-community spatial differences are important for accurate exposure assessment of traffic-related pollutants. |