Spatial Variability of Fine Particle Mass, Components, and Source Contributions during the Regional Air Pollution Study in St. Louis
详细信息 查看全文
- 作者：Eugene Kim ; Philip K. Hopke ; Joseph P. Pinto ; and William E. Wilson
- 刊名：Environmental Science & Technology
- 出版年：2005
- 出版时间：June 1, 2005
- 年：2005
- 卷：39
- 期：11
- 页码：4172 - 4179
- 全文大小：343K
- 年卷期：v.39,no.11(June 1, 2005)
- ISSN：1520-5851
文摘
Community time-series epidemiology typically uses either24-hour integrated particulate matter (PM) concentrationsaveraged across several monitors in a city or dataobtained at a central monitoring site to relate PM concentrations to human health effects. If the day-to-day variationsin 24-hour integrated concentrations differ substantiallyacross an urban area (i.e., daily measurements at monitorsat different locations are not highly correlated), thenthere is a significant potential for exposure misclassificationin community time-series epidemiology. If the annualaverage concentration differs across an urban area, thenthere is a potential for exposure misclassification inepidemiologic studies that use annual averages (or multi-year averages) as an index of exposure across differentcities. The spatial variability in PM2.5 (particulate matter 
2.5 
m in aerodynamic diameter), its elemental components,and the contributions from each source category at 10monitoring sites in St. Louis, Missouri were characterizedusing the ambient PM2.5 compositional data set of theRegional Air Pollution Study (RAPS) based on the RegionalAir Monitoring System (RAMS) conducted between 1975and 1977. Positive matrix factorization (PMF) was applied toeach ambient PM2.5 compositional data set to estimatethe contributions from the source categories. The spatialdistributions of components and source contributions to PM2.5at the 10 sites were characterized using Pearsoncorrelation coefficients and coefficients of divergence.Sulfur and PM2.5 are highly correlated elements betweenall of the site pairs Although the secondary sulfate is the mosthighly correlated and shows the smallest spatial variability,there is a factor of 1.7 difference in secondary sulfatecontributions between the highest and lowest site on average.Motor vehicles represent the next most highly correlatedsource component. However, there is a factor of 3.6 differencein motor vehicle contributions between the highest andlowest sites. The contributions from point source categoriesare much more variable. For example, the contributionsfrom incinerators show a difference of a factor of 12.5 betweenthe sites with the lowest and highest contributions. Thisstudy demonstrates that the spatial distributions of elementalcomponents of PM2.5 and contributions from sourcecategories can be highly heterogeneous within a givenairshed and thus, there is the potential for exposuremisclassification when a limited number of ambient PMmonitors are used to represent population-average ambientexposures.
2.5 m in aerodynamic diameter), its elemental components,and the contributions from each source category at 10monitoring sites in St. Louis, Missouri were characterizedusing the ambient PM2.5 compositional data set of theRegional Air Pollution Study (RAPS) based on the RegionalAir Monitoring System (RAMS) conducted between 1975and 1977. Positive matrix factorization (PMF) was applied toeach ambient PM2.5 compositional data set to estimatethe contributions from the source categories. The spatialdistributions of components and source contributions to PM2.5at the 10 sites were characterized using Pearsoncorrelation coefficients and coefficients of divergence.Sulfur and PM2.5 are highly correlated elements betweenall of the site pairs Although the secondary sulfate is the mosthighly correlated and shows the smallest spatial variability,there is a factor of 1.7 difference in secondary sulfatecontributions between the highest and lowest site on average.Motor vehicles represent the next most highly correlatedsource component. However, there is a factor of 3.6 differencein motor vehicle contributions between the highest andlowest sites. The contributions from point source categoriesare much more variable. For example, the contributionsfrom incinerators show a difference of a factor of 12.5 betweenthe sites with the lowest and highest contributions. Thisstudy demonstrates that the spatial distributions of elementalcomponents of PM2.5 and contributions from sourcecategories can be highly heterogeneous within a givenairshed and thus, there is the potential for exposuremisclassification when a limited number of ambient PMmonitors are used to represent population-average ambientexposures.