Hyper-local to Regional Exposure Contrast of Source-Resolved PM2.5 Components across the Contiguous United States: Implications for Health Assessment

Background: Improved understanding of what sources and processes drive exposure contrast of fine particulate matter (PM2.5) is essential for designing and interpreting epidemiological study outcomes. Objective: We investigate the contribution of various sources and processes to PM2.5 exposure contrasts at different spatial scales across the continental United States. Methods: We consider three cases: exposure contrast within a metro area, nationwide exposure contrast with high spatial resolution, and nationwide exposure contrast with low spatial resolution. These three cases correspond to common epidemiological study designs. Using high spatial resolution (census-block-level) national empirical model estimates of source- and chemically-specific PM2.5 concentration predictions, we quantified the contribution of various sources and processes to PM2.5 exposure contrasts in these three cases. Results: At the metro level (i.e., metropolitan statistical area; MSA), exposure contrasts of PM2.5 vary between -1.8 to 1.4 µg m-3 relative to the MSA-mean with about 50% of within-MSA exposure contrast of PM2.5 caused by cooking and mobile source primary PM2.5. For the national exposure contrast at low resolution (i.e., using MSA-average mean concentrations), exposure contrasts (relative to the national mean: -3.9 to 3.2 µg m-3) are larger than within an MSA with ~80% of the variation due to secondary PM2.5. National exposure contrast at high resolution (census block) has the largest absolute range (relative to the national mean: -4.7 to 3.7 µg m-3) due to both regional and intra-urban contributions; on average, 65% of the national exposure contrast is due to secondary PM2.5 with the remaining from the primary PM2.5 (cooking and mobile source 26%, other 9%). Discussion: While national epidemiological studies that use high-spatial-resolution exposure estimates maximize the exposure contrast of total PM2.5, other study designs may offer advantages to investigate the health impacts of specific components. City/metro scale studies better isolate the health impacts of primary PM2.5 from local sources, while national studies with low-spatial resolution can help to infer the health impacts of secondary PM2.5.