Impact of Stratospheric Intrusions, Regional Wildfires, and Long-Range Transport Events on Air Quality in the Western United States

Thesis (Ph.D.)--University of Washington, 2017-08Baseline ozone refers to observed concentrations of tropospheric ozone at sites that have a negligible influence from local emissions. In 2004, the Mount Bachelor Observatory (MBO) was established to examine baseline air masses as they arrive to North America from the west. In the western U.S., air quality is impacted by local, regional, and trans-Pacific sources, which can be natural or anthropogenic. Because of increasing baseline O3 and tighter air quality standards, it is therefore important to understand the nature and the impact of high-ozone events on urban air quality in the western U.S. This dissertation is focused on three types of high-ozone events that we typically observe at MBO and in other surface sites in the western U.S.: (1) upper troposphere/lower stratosphere (UT/LS) episodes, (2) biomass burning (BB) plumes, and (3) long-range transport (LRT) events. First, I looked at an anomalously high-O3 springtime episode in May 2012, where I observed an O3 increase of 2.0–8.5 ppbv in monthly average maximum daily 8–hour average O3 mixing ratio (MDA8 O3) at MBO and numerous other sites in the western U.S. compared to previous years. This shift in the O3 distribution had a strong effect on the number of exceedance days. I also observed a good correlation between daily MDA8 variations at MBO and at downwind sites. This is consistent with previous studies that show that under specific meteorological conditions, synoptic variation in O3 at MBO can be observed at other surface sites in the western U.S. At MBO, the elevated O3 concentrations in May 2012 are associated with low CO values and low water vapor values, consistent with transport from the upper troposphere/lower stratosphere (UT/LS). The Real-time Air Quality Modeling System analyses indicate that a large flux of O3 from the UT/LS in May 2012 contributed to the observed enhanced O3 across the western U.S. Results from this component suggest that a network of mountaintop observations, LiDAR and satellite observations of O3 could provide key data on daily and interannual variations in baseline O3. Second, I studied BB events that I observed at MBO during the summer of 2015. Regional BB has the potential to increase ozone in the western U.S. especially during summer. I explored the photochemical environment in BB plumes, which remains poorly understood. Because I am interested in understanding the effect of aerosols only (as opposed to the combined effect of aerosols and clouds), I carefully selected three cloud-free days in August and investigate the photochemistry in these plumes. At local mid-day (solar zenith angle, SZA = 35o), j(NO2) values were slightly higher (0.2-1.8%) in the smoky days compared to the smoke-free day, presumably due to enhanced scattering by the smoke aerosols. At higher SZA (70o), BB aerosols decrease j(NO2) by 14-21%. I also observe a greater decrease in the actinic flux at 310-350 nm, compared to 350-420 nm, presumably due to absorption in the UV by brown carbon. I compare my measurements with results from the TUV5.2 model and find a good agreement during cloud-free conditions. I perform sensitivity runs and find that j(NO2) is not sensitive to O3 column input. If I keep single scattering albedo and aerosol optical depth constant, an increase in the total ozone from 280 to 400 DU leads to only a 0.8% decrease in j(NO2). Finally, I used the extended Leighton relationship to estimate mid-day HO2 and RO2 concentrations and P(O3) in the fire plumes. I calculate HO2 and RO2 values from 49-185 pptv, and compute ozone production rates of ~2 ppbv/hour in these fire plumes. Finally, I looked at an LRT event on Spring 2015. I observed O3 and CO enhancements of up to 40 ppbv and 80 ppbv, respectively, at MBO. I also used measurements from the NOAA WP-3D Orion research aircraft during the Shale Oil and Natural Gas Nexus (SONGNEX) campaign in Spring 2015. One of the flights during the SONGNEX campaign intercepted the Siberian plume. Ground-based, satellite, and LiDAR data suggest that the Siberian plume was transported at high elevation, did not encounter a high-pressure system that would have led to air subsidence, and therefore did not cause any surface O3 enhancement. I compare our measurements at MBO with aircraft observations and conclude that the Siberian airmass split into two plumes in the eastern Pacific. One plume moved eastward and was sampled by MBO. The other moved over to Alaska and then down to the U.S. Midwest; this second plume was intercepted by the aircraft. sp/CO enhancement ratios for the Siberian 2015 plume were higher than similarly aged plumes in previous studies, owing to dust and the absence of particulate matter loss because of the relatively intact nature of the Siberian plume. O3/CO ratio observed at MBO was higher than the aircraft because of PAN decomposition. I look at the plume’s reactive nitrogen speciation using data from the aircraft and find that ~75% of the NOy is stored as PAN. For ozone production to take place, the plume has to warm up (i.e., descend) to re-form NOx. But given the high elevation of the plume and stable atmospheric conditions, this likely did not take place. The results of this dissertation have important policy implications. They suggest that at the current standard, high-ozone events such as BB plumes, UT/LS episodes, and LRT events would affect the attainment status of a site if they were not identified as exceptional events, which the EPA defines as an uncontrollable event that affected air quality. Understanding the nature and year-to-year variability of these events is therefore critical for an effective implementation of the US NAAQS. Long-term measurements aimed at observing exceptional events would be valuable. MBO is the only high-elevation site on the U.S. West Coast that routinely observes high-O3 events in the FT; however, it provides measurements at only a single point. High-frequency measurements of O3, water vapor and CO at a network of mountaintop sites would be valuable at observing exceptional events. Vertical profiles of O3 and water vapor from LiDAR and ozonesondes, and satellite retrievals would also be helpful. This network of mountaintop observations, LiDAR and satellite observations of ozone could also provide key data on daily and interannual variations in baseline O3. Forecasting exceptional events would be possible using high-resolution chemical transport models that have been evaluated and verified with free troposphere observations