The impact of wildfire smoke on compositions of fine particulate matter by ecoregion in the Western US.

Epidemiological studies of wildfire PM2.5 constituents are hindered by the limited information on the population exposure to ambient PM2.5 constituents during high-pollution episodes from wildfires ("smoke waves"). The chemical composition of wildfire-related PM2.5 can be affected by different ecosystems. Current literature assessing the differences in PM2.5 pollution from wildfire smoke by ecosystems often analyzes air samples collected from the smoke near the center of an individual fire, but the results might not represent the exposure of the general public living away from the fire center but affected by the smoke of the fire. We assessed the population-based exposure to wildfire-related PM2.5 species by integrating monitor measurements on 29 PM2.5 species and previous findings on smoke waves during 2004-2009 in 51 Western US counties across six ecoregions. We found that across all ecoregions, smoke waves were associated with an increase in the fraction of organic carbon of total PM2.5 by 20 percentage points (95% confidence interval (CI): 17, 23), an increase in the fraction of elemental carbon by 0.99 percentage points (95% CI: 0.43, 1.6), and decreases in fractions of sulfate and crustal species. While the PM2.5 mixtures were dominated by the same source (wildfires), compositions in North American Deserts and the Great Plains during smoke waves were distinct. Besides expanding the knowledge of wildfire PM2.5, our study has implications beyond wildfires and could aid future population-based epidemiological research on PM2.5 mixtures by source and region.