Abstract View
How Spatially Correlated Are Ultrafine Particle Number and Fine Particle Mass at Urban Scales?
PROVAT SAHA, Shayak Sengupta, Peter Adams, Allen Robinson, Albert A. Presto, Carnegie Mellon University
Abstract Number: 252
Working Group: Urban Aerosols
Abstract
The epidemiological evidence for ultrafine particles (UFP; particles with a diameter less than 100 nm) causing chronic health effects independent of fine particulate matter (PM2.5) mass is inconclusive. A prevailing view is that UFP and PM2.5 mass have different spatial patterns in urban areas, which should allow epidemiological studies to distinguish their independent, chronic health effects. Here, we investigate the intra-urban spatial correlation of PM2.5 mass and UFP exposures in Pittsburgh, Pennsylvania. Measurements and predictions of a land-use regression model indicate a moderate spatial correlation between particle number concentrations (PNC; a proxy for UFP) and PM2.5 mass (R2 of 0.38 and 0.41, respectively). High-resolution (1-km) chemical transport model simulations predict a much stronger spatial correlation (R2 ~ 0.8). The true spatial correlation likely lays within this range, the width of which is due to the combined effects of measurement noise and incomplete representation of emissions and other processes in the model. The finding of moderate to strong spatial correlation was initially surprising because secondary aerosol contributes the vast majority of PM2.5 mass in cities like Pittsburgh. However, intra-urban spatial patterns are strongly influenced by local emissions. The spatial correlation of PNC and PM2.5 mass concentrations indicates that, in Pittsburgh, concentrations of these pollutants are driven by emissions from the same sources, especially traffic and cooking. Furthermore, both pollutants largely behave as passive tracers at time scales of one day or less required for transport across most urban environments. Although previous research has shown little temporal correlation between PNC and PM2.5 mass (which may allow epidemiological studies to differentiate acute health effects), our finding of moderate to strong spatial correlation may complicate the use of epidemiological analysis to separate the chronic health effects of PNC from PM2.5 mass.