Plain Language Summary
Because man-made emissions are decreasing, concentrations of harmful ozone pollution have also decreased in many areas of the USA. Not all sources can be easily controlled however. For example, biomass burning emits lots of pollutants to the atmosphere, and descending air from the stratosphere can bring with it high levels of ozone. These sources can pollute first the air above us, and then when the air is transported to the surface it can pollute the air we breathe. In this study we used balloons to measure ozone pollution as it changes from the surface to high in the atmosphere. We then used computer models to identify the sources responsible for higher pollution levels. Our study was part of a major field campaign that took place in the Midwest U.S. in the summer of 2013.
1 Introduction
Even moderate concentrations of ground-level ozone (O3) can adversely affect human health and the environment across the United States (U.S. EPA, 2015). To lessen these impacts, the U.S. Environmental Protection Agency (U.S. EPA) has lowered the National Ambient Air Quality Standard (NAAQS) for O3 from 75 to 70 parts-per-billion-per-volume (ppbv) (U.S. EPA, 2015). Ozone is a key secondary air pollutant associated with a large number of health issues ranging from asthma to premature deaths (Fann et al., 2013, 2018; Reid et al., 2016; Rappold et al., 2017), leading to 12,300 to 52,000 annual premature deaths in North America alone (Silva et al., 2013). Due to more stringent regulations over the past twenty years, regional ozone levels have declined by 7 to 13% (Simon et al., 2015; Strode et al., 2015). New evidence has demonstrated that even at lower concentrations ozone can still be very toxic (U.S. EPA, 2013). Despite overall progress, ozone concentrations may still be increasing in some areas (Fishman et al., 2014; Cooper et al., 2015; Jaffe et al., 2018), particularly in fire prone regions (McClure & Jaffe 2018).
Regulating locally formed ozone is complicated by the fact that ozone has significant background levels in the troposphere typically ranging 30 to 50 ppbv (Jaffe et al., 2018). Background concentrations of ozone can be enhanced due to noncontrollable ozone sources (NCOS), defined by Jaffe et al. (2018) to include recent local pollution influence from Stratosphere-to-Troposphere Transport (STT), lingering biomass burning emissions, long-range transport from international sources, lightning, or photochemical production from precursor emissions (e.g. nitrogen oxides and volatile organic compounds). Therefore, in this study we define background ozone to include NCOS. While foreign sources of pollution and wildfires are theoretically controllable, these are beyond the control of any local jurisdiction, so for this discussion, we can justify including these as background ozone. Over the past decade, significant progress has been made in our efforts to understand aspects of the U.S. background ozone problem for example: episodic stratospheric sources (Lin et al., 2015), interannual variability (Strode et al., 2015), and wildfire contributions (Jaffe et al., 2013; Westerling, 2016). However, these efforts have lacked coordination, and are largely focused on the impacts to the western U.S. where NCOS are considered to be the greatest (Jaffe & Wigder, 2012; Jaffe et al., 2018). Despite a 15 to 33% reduction in ozone precursors from anthropogenic emissions across the Eastern and Midwestern U.S. (Strode et al., 2015), the Midwestern background ozone levels continue to increase ~0.23 ppbv per year (Fishman et al., 2014) on par with the trends in the intermountain western U.S.
The Midwestern U.S., like the intermountain west U.S., is frequently impacted by biomass burning (McCarty et al., 2007; Liu et al., 2016) and stratospheric intrusions (Lin et al., 2015; Langford et al., 2018). Although biomass burning in the Midwest consists mostly of smaller agricultural fields, their combined emitted pollution and higher frequency can result in emissions double that of wildfires (Larkin et al., 2014). Understanding stratospheric intrusions and biomass burning contributions for this region will be critical as they can confound NAAQS attainment (e.g., Jaffe, 2011; Lin et al., 2012; Hess and Zbinden, 2013), where summertime daily ozone maxima already range from 70 to 80 ppbv and occasionally reach mixing ratios higher than 140 ppbv (Fishman et al., 2014). The periodic nature of some NCOS makes it difficult to target these sources with dedicated field campaigns, but opportunistic measurements have been made during field studies with other objectives (Sullivan et al., 2015; Ott et al., 2016). For example, the influence of wildfires, long-range transport, and SST were the foci of the Las Vegas ozone Study (LVOS) where it was demonstrated that NCOS contributed ~30 ppbv to three ozone exceedances (Langford et al., 2015).
The SEAC4RS (Studies of Emissions and Atmospheric Composition, Clouds and Climate Coupling by Regional Surveys) field campaign in August-September 2013 presents a unique opportunity to study Midwest U.S. background ozone (Toon et al., 2016). During the flight mission several ozonesondes were launched nearly every day to measure vertical profiles of column ozone and meteorology over seven U.S. stations in a network known as SEACIONS (SouthEast American Consortium for Intensive Ozonesonde Network Study). The SEAC4RS campaign sparked a growing number of studies concerning ozone in the southeast U.S. (Travis et al., 2016; Wagner et al., 2015; Zhu et al., 2016). Although the research flights did not cover the Midwest U.S., this area is of interest because the time of the field campaign was an active period for NCOS with signatures for STT, large fires in the western U.S., and smaller prescribed fires in the central U.S. (~0.2 million hectares burned). During the 2013 ozone season (1 April through 31 October) fifteen total ozone exceedances occurred with six high-ozone days occurring during SEAC4RS. Additionally, a satellite-based detection method identified smoke plumes with confirmed injection heights above the boundary layer, initiated by smoke-infused thunderstorms known as pyrocumulonimbus or pyroCb (Fromm et al., 2010, 2019; Peterson et al., 2015, 2017a,b, 2018). The parametrization of plume rise from biomass burning can vary widely and cause undue biases and errors within model simulations of pollution transport (Paugam et al., 2016; Wilkins et al., 2018; Val martin et al., 2012, 2018). The detection algorithm used for the smoke plumes reduced the uncertainty of the injection heights and gave improved estimates based on remote sensing of each specific case.
Using a chemical transport model, it has been shown that fires can contribute ~14% to daily max ozone levels in the central U.S. (Wilkins et al., 2018). This has driven the need for developing accurate tropospheric ozone budgets for a typical summertime in the Midwest as suggested by Wilkins et al. (2018). Several approaches have been used to partition or identify the individual contribution of a naturally occurring ozone rich STT intrusions (Langford et al., 2018), a biomass burning plume (Westerling, 2016), or their combined impacts (Brioude et al., 2007) to elevated ozone levels. These approaches include the use of intensive balloon-borne ozonesonde field campaigns (Cooper et al., 2007; Thompson et al., 2019); lidar and aircraft remote sensing (Brioude et al., 2007; Langford et al., 2018); and Lagrangian trajectory and chemical transport models (Morris et al., 2006; Baker et al., 2016, 2018). However, the determination of the contribution from individual sources to local ozone enhancement with a single point measurement remains difficult (Thompson et al., 2019), and it can be even more challenging to quantify accurately the contribution from multiple sources (Lin et al., 2015). In particular, the partitioning of these sources is critical for NAAQS attainment investigations along with long-term trend determinations, as they are two of the most critical NCOS imported (Parrish et al., 2012, 2014). Further, 30% of the present-day atmospheric ozone burden is attributable to human activity and these emissions are still rising (Granier et al., 2011; Parrish et al., 2010). For example, Li et al. (2019) using ~1,000 surface ozone sites over a five year period (2013-2017) showed that East Asian concentrations are continuing at +1 to 3 ppbv per year. In this study, we use ozonesonde measurements from the SEAC4RS campaign and models to identify and characterize contributions from biomass burning and stratospheric intrusions to tropospheric ozone columns above the Midwest background ozone levels. Additionally, we evaluate the contribution from fires observed in the western and Midwest U.S. to ozone concentration in St. Louis.
2 Tropospheric Ozone and Meteorological Soundings
Vertical profiles of ozone, pressure, humidity, and temperature were measured using balloon-borne ozonesondes coupled with radiosondes. Each ozonesonde consists of a Teflon pump, an ozone sensing electrochemical concentration cell (ECC) (Komhyr et al., 1986, 1995), attached alongside a Vaisala RS-8015N radiosonde. The instruments were prepared in accordance with the Southern Hemisphere Additional Ozonesondes (SHADOZ) protocol (Thompson et al., 2011a,b, 2019). The ECC-type ozonesonde is currently the most widely used due to its well-characterized behavior and working capability under various sky conditions (Kuang et al., 2012). Comparisons with other O3 measuring instruments have demonstrated the ECC sonde precision to be ±5% near the ground and ±10% in the upper troposphere, although errors can reach 17% (Thompson et al., 2007; Stauffer et al., 2014). The radiosonde temperature and pressure sensors have accuracies below 20 km of ±0.3°C and ±0.5 hPa, respectively (Lal et al., 2014). The heights were calculated based on the observed pressure and are above sea level (asl). The humidity sensor has an accuracy of about ±2% near the ground which decreases to ±15-30% in the 5 to 15 km altitude range (Kley et al., 1997; Hurst et al., 2011).