Fine Particulate Matter and Black Carbon Land Use Regression Modeling and Source Apportionment across Downtown Pittsburgh, PA, USA

摘要

Intra-urban variability in sources of fine particulate matter (PM2.5) can result in differing exposures to public health, as certain areas have periods of heavy traffic congestion, frequent bus routes, and density of trucks. Land use regression (LUR) models and source apportionment can be used to help disentangle the complexity of pollution across a city. A saturation monitoring campaign was designed to capture spatial variability in PM2.5 and black carbon (BC) within the downtown Pittsburgh core, with an emphasis on diesel-related sources. Geographic information system-based (GIS) methods were used to characterize spatial distributions in local pollution sources, such as traffic density, truck density, and bus route frequency. PM2.5 and constituents (BC and metals) were sampled during winter and summer 2013 from 7AM to 7PM Monday through Friday to capture work-week exposures. Sampling locations (n=40) were systematically allocated to capture multiple pollution source profiles across downtown Pittsburgh. Four reference sites were used to capture long-range transport and meteorological drivers of local concentrations. Mean PM2.5 concentrations were 13.22 (SD=2.33 μg/m3) for winter and 13.28 (SD=1.99 μg/m3) for summer, and mean BC was 3.57 (SD=1.39 abs) for winter and 4.11 (SD=1.93 abs) for summer. For source apportionment, factors such as motor vehicle emissions, brake/tire wear, fuel/oil, and diesel were identified in the small domain. We are currently developing season- and pollutant-specific LUR models, and fine-scale concentration estimates. This spatial saturation monitoring campaign explored fine-scale variability in PM2.5, BC, and trace metal constituents across downtown Pittsburgh during winter and summer 2013. Intra-urban sampling within a small city allowed us to further understand and refine exposure assignment of source contributions to public health.