Passive sampling to capture spatial variability in PM 10–2.5

Abstract

This work applied inexpensive passive sampling to measure airborne coarse particles with aerodynamic diameters between 2.5 and 10 μm (PM 10–2.5) over three 7-day periods at 33 sites in a medium-sized Midwest City (Iowa City, IA). The number of sites and their locations were selected using an optimal sampling design that captured 95% of the total variance in PM 10–2.5 as measured with real-time sampling equipment on a mobile sampling platform. Weekly averages of PM 10–2.5 were 15.9 μg m −3 (coefficient of variation between sites, CV=23%), 17.9 μg m −3 (CV=24%), and 6.1 μg m −3 (CV=30%). ANOVA showed that these means were statistically different ( p<0.001), and that the spatial variability plus random error accounted for 29% of the total variability. The maximum coefficient of divergence between sites—a relative measure of uniformity—ranged from 0.21 to 0.36. These values indicate that PM 10–2.5 was heterogeneous even on the fine spatial resolution studied in this work (average distance between sites was 4.4 km). The spatial patterns of PM 10–2.5 measured with the passive samplers closely matched with those of mobile mapping and corresponded with known coarse particle sources in the area. This work demonstrates that passive sampling coupled with effective sampling design may enhance our ability to assess exposure to PM 10–2.5 at a local scale. Compared to exposure estimates made with data from centrally located, filter-based samplers, these highly spatially resolved estimates should reduce exposure misclassification errors in epidemiological studies.