Impact of Wildfires on O3 and PM in the Western U.S

As part of our NOAA-FIREX supported work, we have focused on understanding the mixing and influence of smoke on air quality in the Western U.S., particularly urban areas. This analysis uses data from the Mt. Bachelor Observatory (MBO), and the IMPROVE and EPA air quality networks, augmented with additional observations in Boise, ID, in the summer of 2017. During the 2017 campaign, smoke was observed at ground level in Boise on 28 out of 61 days.

Key conclusions from this work are:

1. Over the last 3 decades PM_­2.5 has increased at the 98^th percentile due to wildfires over a large portion of the western U.S. (McClure and Jaffe, 2018, PNAS).
2. The NOAA Hazard Mapping System-Fire and Smoke Product (HMS-FSP) is a very good tool to identify transported smoke plumes. However the HMS-FSP is not indicative of surface smoke. By examining the distribution of PM_2.5 on days with and without overhead smoke, we can identify days with a high probability of influence at the surface.
3. Analysis of a large number of smoke events from 18 monitoring locations in the Western U.S. shows that average O₃ production rates are enhanced by ~1-2 ppb/hour during the daytime, compared to non-smoke days. During days with smoke influence, maximum daily 8-hour averaged O₃ (MDA8) generally increases with the daily mean PM_2.5 with an average slope of ~0.5-1 ppb per ug/m3 up to approximately 60 ug/m3, but with a large degree of variability. Above this level, O₃ production appears to be suppressed.
4. O₃ from fires is very difficult to predict from typical Eulerian photochemical models. For this reason we have developed a statistical approach using Generalized Additive Modeling (GAM) that can be used to estimate the smoke influence on O₃. This approach shows that fire emissions can result in enhancements in the MDA8 of up to 30 ppb, but again with a large degree of variability. The GAMs also show that the strongest influence from fires is on photochemically active days, with much lower influence on other days.
5. In 2017, smoke plumes enhanced the average MDA8 O₃, NOy, PAN, and CO mixing ratios in Boise by 19, 1.7, 0.25, and 179 ppb or 49, 13, 42, and 70%, respectively, compared to the non-smoke days. Although VOCs were not measured during this campaign, we can estimate an average VOC/NOx molar enhancement ratio of ~100/1 due to smoke plumes. This suggests that wildfire VOCs are largely responsible for enhancing O₃ in urban areas. While PAN is a significant NOy component in free tropospheric smoke plumes, it plays a smaller role in urban areas, likely due to the higher temperatures and subsequent thermal decomposition.
6. Since 2015 the Mt. Bachelor Observatory has observed over 80 smoke plumes from small local prescribed burns to large regional wildfires, and even Siberian wildfires. The physical and optical properties of these smoke plumes vary significantly. We found aerosol size distribution to be primarily dependent on plume concentration, with thicker smoke leading to larger particles. The amount of Black Carbon (BC) and aerosol absorption (σabs) in the smoke plumes varied widely as well. Enhancement ratios of ΔBC/ΔCO and Δσabs/ΔCO ranged from 2.40 to 17.6 ng/m3/ppbv and 0.007 to 0.13 Mm-1/ppbv, respectively. We found mass absorption coefficients (MACs: Δσabs/ΔrBC) to be positively correlated with absorption angstrom exponent (AAE) values (an indicator of Brown Carbon) and negatively correlated with rBC mass fraction.