A multivariate assessment of meteorological influences on inhalable particle source impacts

Abstract

This paper identifies the sources of fine and coarse inhalable particles at a site in metropolitan Boston and investigates their respective relationships to meteorological conditions. In this work, Principal Component Analysis (PCA) is applied to: fine particle mass elemental data; coarse particle mass elemental data, and; meteorological measurements (primarily collected at nearby Logan International Airport). In addition to local surface observations, air mass trajectory information concerning each sampling day is included in the meteorological data set, allowing the consideration of air mass transport as one factor in particle impacts. As part of these PCA analyses, four different objective component slection criteria are examined and compared. The Scree test of eigenvalues is found to result in the most interpretable components for the specific air pollution and meteorological data sets considered in this work. The rotated PCA analyses result in the identification of five fine mass particle source classes (soils motor vehicle, coal related, oil and salt aerosols); six coarse mass particle source classes (soil, motor vehicle, refuse incineration, residual oil, salt and coarse sulfate aerosols); and five meteorological components (air mass transport over the eastern seaboard to Boston, moisture influences, poor local dispersion, season and air mass transport over the Ohio Valley/Midwest to Boston). It is found that local combustion sources (e.g. residual oil burning and motor vehicles) recorded their highest impacts during poor local dispersion and mesoscale advection from downtown Boston. Coarse soil impacts are highest at this site during dry, summertime conditions (when fugitive dust generation is at a maximum). Coal combustion-related fine particle impacts, however,are at a maximum during air mass transport into the Boston vicinity (especially form the Ohio Valley/Midwest direction), regardless of local dispersion conditions. It became obvious that meteorology plays an important role in the extent to which these paricle source classes impact the monitoring site on a given day, and that it is, in large measure, the highly variable weather conditions in Boston which allow statistical approaches to discriminate the various source classes. Based on these analyses, it is concluded that both local and transported aerosols have an influence on inhalable aerosol exposures at this site, with the fulfate portion being most correlated with polluted air mass transport into the Boston area.