Urban Air Quality during Smoke Events in the Western U.S.

Biomass burning events emit particulate matter (PM), carbon monoxide, and a wide range of volatile organic compounds, including many ozone (O₃) precursors. When smoke from these events is transported to urban areas, the impact on O₃ is highly variable. Under projections of increasing wildfire frequency and severity in the western U.S., many cities, which are already prone to poor air quality, are poised to experience greater O₃ impacts from smoke events. While urban smoke impacts on O₃ have been observed, they are not well-understood and vary by season, region, and city.

In this study, we investigate O₃ impacts from smoke from July-September, 2013-2017, in 18 cities in the western U.S. We use satellite-based Hazard Mapping System (HMS) fire and smoke product to identify overhead smoke, classifying days as nonsmoke, adjacent to smoke (immediately preceding or following a smoke event), first smoke (first day of smoke event), and consecutive smoke (any smoke days following the first day of the event). We identify four key findings:

1. Elevated O₃, PM_2.5, and NO₂ occur under smoke influence, with the greatest mean enhancement of all three pollutants occurring on consecutive smoke days.
2. The rate of morning O₃ production (dO₃/dt) is higher under smoke influence, indicating the importance of transported O₃ precursors, rather than O₃ itself.
   
3. Absolute enhancements in PM_2.5 and NO₂ under smoke influence are not well-correlated with O₃ enhancements across sites or smoke classifications. This may result, in part, from differences in chemical regime (e.g., NO_x sensitivity) among cities and/or non-linearity in the O₃ relationship with PM_2.5 (see next point).
4. We show that O₃ enhancements in smoke generally increase with 24-h PM_2.5 up to ~30 ug/m³ (smoke). Above this level, O₃ enhancements appear to decline.

We note that the HMS product is a useful tool for smoke identification, but elevated O₃ and PM_2.5 on days before and after smoke events suggests that some smoke events are not detected (e.g., low smoke levels, cloud cover), but still impact air quality.