Urban-influenced coastal environments are an air quality management challenge because of complex wind patterns, shallow stable marine boundary layers, and the interaction of these meteorological features with ozone precursor (reactive nitrogen and volatile organic compound, VOC) emissions. Many of the counties in the eastern U.S. where ozone concentrations exceed the 2015 ozone National Ambient Air Quality Standards (NAAQS) of 70 ppb are along coastlines (USEPA 2015). Studies such as the Lake Michigan Air Quality Study 1991 (Dye et al. 1995), Lake Michigan Air Directors Consortium (LADCO) Aircraft Project (LAP 1994–2003) (Foley et al. 2011), Border Air Quality and Meteorology Study (BAQS-Met) (Makar et al. 2010), Ozone Water-Land Environmental Transition Study (OWLETS) (Sullivan et al. 2019), Long Island Sound Tropospheric Ozone Study (LISTOS) (Miller 2018; Zhang et al. 2020), and the Baltimore–Washington and Houston legs of DISCOVER-AQ (Mazzuca et al. 2016) have sought to improve understanding of coastal air quality and meteorological interactions. The Lake Michigan Ozone Study 2017 (LMOS 2017), a recent collaborative, multiagency field study targeting ozone chemistry, meteorology, and related air quality observations along the Wisconsin–Illinois Lake Michigan shoreline, built on these previous studies, and adapted new observing platforms and modeling capabilities. In this paper, we give an overview of the rationale for LMOS 2017 along with its study design and some principal results.
Around Lake Michigan, both rural and urban monitoring locations have persistently recorded high ozone concentrations that continue to exceed the ozone NAAQS as of 31 January 2021 (www.epa.gov/green-book). One- and eight-hour maximum concentrations have decreased substantially over the past three decades in conjunction with large decreases in emissions of ozone precursors. In EPA Region 5 (consisting of Illinois, Indiana, Michigan, Minnesota, Ohio, and Wisconsin) anthropogenic nitrogen oxides (NOx = NO + NO2) and VOC emissions dropped 65% and 52%, respectively (Adelman 2020) between 1997 and 2017. However, the ozone NAAQS was lowered to 75 ppb in 2008 and to 70 ppb in 2015 in response to health-focused science that showed further reductions of ozone were necessary for human health protection (USEPA 2006, 2013). As a result, current and future additional controls on NOx and VOC emissions will likely be required in the region to achieve attainment status.
Ozone concentrations for the 3-yr period from 2014 to 2016 (Fig. 1) show some of the locations exceeding the NAAQS (70 ppb), which motivated the field campaign. As Fig. 1 shows, the highest ozone concentrations in the Lake Michigan region have been found along the lakefront, consistent with elevated ozone concentrations over Lake Michigan (Dye et al. 1995; Foley et al. 2011; Cleary et al. 2015). Lake Michigan’s air quality episodes often occur when weak synoptic southerly winds, common during fair weather periods, interact with lake breezes (Lyons and Olsson 1973; Lennartson and Schwartz 2002). For the western coast of Lake Michigan, ozone exceedances are most frequent from late May to early July (Good 2017), and during this time period, the lake water temperatures (8°–17°C) are substantially colder than the surrounding land (Laird et al. 2001). Lake breeze flows commence in late morning, and are sometimes accompanied by offshore nighttime flows (driven by land radiative cooling) that lead to warmer temperatures over the lake relative to land. Under these conditions, emissions can be trapped when they are drawn out over the lake during night or early morning. This is followed by photochemical ozone production confined in a shallow lake inversion layer. Finally, processed plumes that have undergone substantial oxidation can return onshore with elevated ozone levels during midday and afternoon.
Map of ozone design values (2014–16) and Lake Michigan Ozone Study 2017 (LMOS 2017) ground site locations. Ozone design values (fourth-highest daily monitored 8-h average ozone concentration at that monitor, averaged over a 3-yr period; ppb) for 2014–16 from the U.S. Air Quality System (AQS). Purple circles indicate the primary ground sites hosting measurements for LMOS 2017. Color thresholds match ozone air quality standards, with 70 ppb as the 2015 standard, and 75 ppb as the 2008 standard.
Citation: Bulletin of the American Meteorological Society 102, 12; 10.1175/BAMS-D-20-0061.1
Dye et al. (1995) clearly showed the role of the shallow stable conduction layer over the lake during ozone episodes. The limited vertical mixing in this layer coupled with stability and slow deposition—due to lack of surface roughness—creates an excellent reactor for ozone production and accumulation. In classic Lake Michigan ozone pollution events, synoptic winds from the south and southwest push pollution plumes from the Chicago and Milwaukee metropolitan areas out over the lake, and late morning or afternoon lake breezes transport them back to the coastline. Emission sources outside the major population centers must also be considered against this conduction layer enhancement.
To best support ongoing air quality management decisions aimed at lowering ozone in the remaining nonattainment areas, LMOS 2017 was conceived to gather new observations of ozone, ozone precursors, and meteorology. Such measurements were required to evaluate air quality and meteorology models, which are run at increasingly high spatial resolutions with a very wide variety of choices for physics, chemistry, data inputs, and data assimilation options. These diverse characteristics make configuration selection challenging for the best forecasting and regulatory modeling of ozone episodes. Air quality models are also dependent on emission inventories that have large potential errors due to rapid changes in electrical generation and mobile sectors coupled with uncertainties in emission and speciation for VOC.
The measurement strategy for LMOS 2017 sought to take advantage of recent observational advances including increasingly fast and sensitive in situ chemical measurements, active and passive ground based remote sensing, and new airborne remote sensing capabilities (see sidebar “Preparing for New Air Quality Monitoring Capabilities”). Furthermore, the LMOS team aimed to use the study as an initial assessment of how improved satellite-based measurements of meteorology and air quality might be applied to ozone pollution management in airsheds such as Lake Michigan, which have heterogeneous land use and emissions coupled with complex meteorological transport. LMOS 2017 was also an example of the increasingly common grassroots field study structure, where early commitments for resources (e.g., NOAA satellite validation, NASA flight hours, EPA mobile laboratory, state and local personnel, equipment, and site access) seeded further investments (NSF, industry) around a coherent set of science objectives.
The scientific questions of the LMOS 2017 field campaign were the following:
What is the relative contribution of interstate and intrastate NOx and VOC emission sources on ozone production rates along Lake Michigan?
To what extent do lake breeze circulations affect ozone production?
How far inland does ozone-rich air penetrate during ozone events?
What is the spatiotemporal distribution of ozone and its precursors over Lake Michigan?
How can remote sensing products (e.g., measurements of nitrogen dioxide, NO2, and formaldehyde) be used to constrain ozone predictions?
How well do regional models capture ozone production chemistry as assessed through evaluation of critical observation-based indicators (e.g., formaldehyde:NOx ratio)?
In the remainder of the article, we first present the study design, measurement locations, and observation platforms. We then present a high-level overview of the air quality and meteorological conditions observed during LMOS 2017, and discuss multimodel underprediction of peak ozone in both forecast and postanalysis configurations. We conclude with a summary of study findings to date and invitation for the community to access the LMOS 2017 data at the NASA campaign repository. LMOS 2017 postanalysis work has made progress on all of the science questions, but additional work remains. We invite the scientific community to utilize the unique set of measurements collected during LMOS 2017 to continue to address these science questions in collaboration with university, federal, and state partners.
Study design
LMOS 2017 focused on ozone precursor emissions, ozone chemistry, and associated meteorology over the southwestern portion of Lake Michigan and along the western shore of Lake Michigan from Chicago, Illinois, to Sheboygan, Wisconsin. As shown in Fig. 2, the measurement strategy incorporated a combination of ground sites, aircraft and ship sampling, and mobile vehicles. The strategy relied on 1) ground-based sites close to the lake shore to capture onshore flows of oxidatively processed ozone-rich air, and 2) airborne platforms for large-scale mapping and vertical profiling. Supported by preliminary modeling using the Weather Research and Forecasting Model with Chemistry (WRF-Chem), and the locations of the highest ozone design values, two ground-based enhanced monitoring (EM) sites were used (design values are an EPA air quality evaluation based on the most recent 3 years of observations; lower values represent cleaner conditions; see Fig. 1 caption for more detail). A southerly ground-based EM site was established at Zion, Illinois (67 km north of Chicago), to intercept relatively fresh urban plumes with elevated precursor concentrations. A more northerly EM site was established at Sheboygan, Wisconsin (211 km north of Chicago), to intercept plumes at a greater distance from sources after more extensive oxidative aging. Contrasting ground sites of these types are critical for testing photochemical grid models (PGM) such as WRF-Chem, the Community Multiscale Air Quality (CMAQ) model, and the Comprehensive Air Quality Model with Extensions (CAMx).
Overview map showing spatial coverage of LMOS 2017 observations. Green and purple polygons indicate ship and aircraft operations areas, respectively. Enhanced monitoring sites at Sheboygan, WI (north), and Zion, IL (south), are mapped in more detail with insets.
Citation: Bulletin of the American Meteorological Society 102, 12; 10.1175/BAMS-D-20-0061.1
Preparing for new air quality monitoring capabilities
The global distributions of air quality (AQ) pollutants have now been observed by satellite instruments (e.g., the Ozone Monitoring Instrument; Levelt et al. 2018) for over two decades, collecting data for monitoring climatologic, economic, and regulatory impacts on air pollution globally. However, these instruments on satellites in low Earth orbits have provided observations at only one time of day, typically the early afternoon. AQ scientists and managers have long advocated for AQ observations from geostationary satellites, necessary for providing information about the “chemical weather” many times per day (IGACO 2004; Fishman et al. 2008, 2012). Such observations will help advance capabilities for monitoring and predicting air quality, analogous to how geostationary meteorological observations have become so fundamental within the weather community. This advancement will move forward when NASA launches the Tropospheric Emissions: Monitoring of Pollution (TEMPO) mission in 2022 (tempo.si.edu) providing the first geostationary AQ trace gas observations over greater North America, complementing the aerosol information available from the Advanced Baseline Imager (ABI) on the current generation of GOES satellites. TEMPO observations include key tropospheric pollutants such as nitrogen dioxide (NO2), formaldehyde, and ozone (O3) at unprecedented temporal (hourly) and spatial (∼2.1 km × 4.4 km) resolutions (Zoogman et al. 2017), giving a revolutionary perspective for addressing O3 air quality challenges. TEMPO is also a component of the developing global integrated AQ observing system, along with two other geostationary missions—South Korea’s Geostationary Environment Monitoring Spectrometer (GEMS) launched in February 2020, observing Southeast Asia, and the European Copernicus Programme’s Sentinel-4, to be launched in 2023 and observing Europe and Northern Africa.
To prepare for the never-before-captured spatial and temporal perspective of geostationary AQ observations, NASA supported the development of airborne instruments similar to TEMPO, GCAS (Kowalewski and Janz 2014; Leitch et al. 2014), and GeoTASO (Leitch et al. 2014), as well as the ground-based instrument, Pandora (https://pandora.gsfc.nasa.gov/). Working with partners such as the U.S. EPA over the past decade, NASA has deployed these instruments together over multiple regions of the country during air quality studies including LMOS. While satellite data excel at providing information in the gaps of surface measurement frameworks, they must be combined with additional information (including surface in situ concentrations and mixing layer heights) to enhance the relevance of the information to AQ managers. Pandora instruments are increasingly being integrated as long-term measurements within U.S. regulatory monitoring sites to provide this critical bridge between satellite observations and standard AQ measurements. The data collected during studies like LMOS are used to help AQ management prepare for using TEMPO data as soon as it becomes available. Observations of the spatial distribution of ozone precursors such as NO2 and formaldehyde repeatedly throughout the day in these studies are already altering conceptual models of the impacts of emissions, chemistry, and meteorology on AQ (e.g., Judd et al. 2018). Integrating all these observations with existing ground-based measurements and chemical transport models will enable AQ managers to better address ozone issues that continue to plague certain regions despite improving emissions.
Along with the EM sites, two aircraft platforms were fundamental to the LMOS 2017 observing strategy. The NASA Langley Research Center (LaRC) UC-12 flew 21 research flights during LMOS 2017. On board the UC-12 were the Geostationary Trace Gas and Aerosol Sensor Optimization (GeoTASO; Leitch et al. 2014) instrument from which vertical columns of NO2 and formaldehyde were retrieved, and the Airborne Hyper Angular Rainbow Polarimeter (AirHARP; McBride et al. 2020), which measured physical and optical properties of aerosols and clouds. A light aircraft (Scientific Aviation, SA, Mooney Ovation 2, N334FL) flew 22 research flights, measuring in situ NO2, ozone, carbon dioxide, methane, and meteorological parameters (temperature, pressure, horizontal wind speed and direction, and water vapor mixing ratio).
Table 1 lists the major measurement platforms and their locations or areas of operation for LMOS 2017. Brief descriptions of each platform or site’s capabilities are given below. Table ES1 in the online supplemental material contains a full list of measurement platforms and instruments, and measurements are publicly available at the NASA LaRC data repository (www-air.larc.nasa.gov/cgi-bin/ArcView/lmos). Many of the instruments and observations are discussed in previous LMOS team publications, such as the LADCO synthesis report (Abdi-Oskouei et al.2019), as well other LMOS 2017 publications (Judd et al. 2019; Vermeuel et al. 2019; Abdi-Oskouei et al. 2020; Doak et al. 2021; Hughes et al. 2021).
Major measurement platforms and locations for LMOS 2017
Ground-based enhanced monitoring sites.
The EM sites were chosen due to their proximity to the coast and high ozone design values that would benefit from more accurate emissions, chemistry, and transport modeling for design of attainment strategies.
The Zion EM site was located 1 km from the shoreline at Illinois Beach State Park. This site hosts an Illinois EPA State and Local Air Monitoring Station (SLAMS) that provides data to the U.S. Air Quality System (AQS). The Illinois Beach State Park station is site 17-097-1007 (42.468°N, 87.810°W). The Zion site also benefits from proximity to the Chiwaukee, Wisconsin, State and Local Air Monitoring Stations (SLAMS) station, which had one of the highest 2014–16 design values in the study domain (77 ppb). At Zion, Illinois, a dataset of continuous wind, temperature, and water vapor vertical profile was recorded using sodar and microwave radiometer (MR) instruments. These instruments captured thermal and wind characteristics of lake breeze penetration with high vertical and temporal resolution. Continuous or semicontinuous measurements of ozone, NOx, nitric acid, hydrogen peroxide, and many VOC and oxygenated volatile organic compounds (OVOC) were conducted. These measurements, in conjunction with detailed Lagrangian photochemical model calculations, have been used to assess the relative importance of NOx and VOC emissions in determining ozone production at the Zion EM site (Vermeuel et al. 2019). Finally, comprehensive chemical and physical aerosol characterization was conducted to enable source apportionment and study of oxidative chemistry (Hughes et al. 2021).
The Sheboygan EM site (43.745°N, 87.709°W) was located at the Spaceport science education facility near the harbor in Sheboygan; the site was 250 m from the shoreline and 8.75 km north of the WDNR Sheboygan Kohler Andrae (KA) monitoring station. The Sheboygan KA monitor had the highest 2014–16 ozone design value in the study domain (79 ppb, see Fig. 1). In this document, the Sheboygan site refers to the temporary campaign measurement site near the Sheboygan harbor; Sheboygan KA refers to the long-term monitoring site at Kohler Andrae.
The UW–Madison Space Science and Engineering Center (SSEC) Portable Atmospheric Research Center (SPARC; Wagner et al. 2019) was deployed at the Sheboygan site. It housed the Atmospheric Emitted Radiance Interferometer (AERI; Knuteson et al. 2004), Doppler lidar, and high spectral resolution (HSRL) aerosol lidar instruments to provide continuous profiles of temperature, water vapor, winds, and aerosol backscatter. This combination of instruments enabled detailed observation of many aspects of the lake breeze. In situ chemical measurements at the Sheboygan site included ozone, NOx, formaldehyde, and NOx plus its reaction products (NOy) at 1-min and finer time resolution. An experimental network of four low-cost ozone monitors (POM, 2B Technologies) was deployed over a 6-km2 area of Sheboygan, WI, to measure differences in concentrations with respect to distance from the lake.
Air pollution sources, both at ground-level and from elevated stacks, exist throughout the study domain. For example, both EM sites had nearby electric generation units (EGUs). The Zion site was located 11 km southeast of the Pleasant Prairie Power Plant while the Sheboygan site was located 3.35 km north of the Edgewater Generating Station. One of the study goals was to use the suite of meteorological and chemical measurements to characterize the impact of localized emissions during ozone episodes and other time periods. High resolution air quality modeling studies guided by the LMOS measurements, are currently underway to explore the impact of these localized emissions on ozone production at the Sheboygan and Zion EM sites.
Other ground-based remote sensing.
In addition to the meteorological sensors at the EM sites described above, the U.S. Environmental Protection Agency deployed remote sensing instruments for mixed layer height, cloud layer height, column NO2, and column ozone (Vaisala CL51 ceilometers and UV/visible Pandora spectrometers). The ceilometers were installed at Grafton, Milwaukee, and Zion. The Pandora spectrometers were installed at the Sheboygan site, as well as WDNR monitoring sites in Grafton, Milwaukee, and Illinois EPA monitoring sites in Zion and Schiller Park (Chicago). Data on mixed layer height, ozone column amounts, and NO2 column amounts are being included in ongoing analyses including comparison to models, comparison to aircraft in situ profiles, and comparison to other remotely sensed atmospheric composition products.
Aircraft measurements.
Primary flight objectives were to 1) conduct regional-scale coastal surveys along the western shore of Lake Michigan during ozone exceedance events; 2) conduct local flights over Chicago, Illinois, and Milwaukee, Wisconsin, to characterize the NO2 emissions; and 3) conduct local flights in the vicinity of Sheboygan, Wisconsin, and Zion, Illinois, to characterize coastal gradients in air pollutants and meteorological fields.
The NASA Langley Research Center UC-12 aircraft carrying GeoTASO focused on remote measurements of NO2 and formaldehyde using raster pattern flight plans to obtain maps of NO2 column abundances. In situ trace gas and meteorological measurements were made using instruments on board a Mooney Aircraft operated by Scientific Aviation. Flying at 28,000 ft, the UC-12 was able to perform remote sensing over the entire study domain, including over airport flight restriction zones, which can reach up to 10,000 ft. Scientific Aviation, on the other hand, flew low-level transects to sample the marine boundary layer along with in situ spirals to measure vertical profiles at selected waypoints. On most of the deployment days, the two aircraft flew over the same area at the same time to provide the best overlap between the two types of measurements. By providing information about the three-dimensional structure of O3 and NO2 the aircraft measurements have played a key role in evaluating the fidelity of air quality forecast model predictions during LMOS.
Other mobile platforms.
Three additional surface-based mobile platforms operated during LMOS 2017: 1) the Geospatial Monitoring of Air Pollution (GMAP) mobile vehicle sampled ozone in transects parallel and perpendicular to the shore; 2) mobile meteorological and ozone sampling was performed by the UW–Eau Claire team; and 3) an instrumented research ship operated out of Sheboygan, Wisconsin, with daily measurements of a number of relevant chemical and meteorological parameters. The UW–Eau Claire (UWEC) and EPA Region-5 GMAP vehicle-based samplers operated between the south and north EM sites. These measurements play a key role in determining the distribution of coastal ozone enhancements between the regulatory monitors, the inland penetration of the ozone enhancements, and the timing of the ozone increases along the shoreline.
Forecasting.
Daily forecasting for flight planning purposes utilized WRF-Chem v.3.6.1 at 4-km horizontal resolution (University of Iowa), as well as the operational air quality forecasts from the National Weather Service (NWS) National Air Quality Forecasting Capability (https://airquality.weather.gov/ and www.weather.gov/sti/stimodeling_airquality_predictions). The operational forecast is based on the 12-km North American Model (NAM) meteorology and CMAQ (photochemical grid model, v.5.0.2).
Meteorological and air quality context
The LMOS 2017 sampling period, 22 May–22 June, captured a typical increase in spring temperatures with periodic rainfall, regular lake breezes and three distinct high pollution episodes. Figure 3 summarizes key meteorological and air pollution variables for LMOS 2017. The LMOS 2017 team identified three episodes with high ozone concentrations for more detailed study, shown in Fig. 3b. Classification of days is discussed in detail in Doak et al. (2021), and generally required one or more routine ozone monitoring sites along the western shore of Lake Michigan to register a maximum daily 8-h average (MDA8) ozone concentration of 70 ppb or higher. The first period, extending from 2 to 4 June, had highest ozone concentrations on 2 June at both Sheboygan KA and Zion for the entire study period. The second and third episodes occurred 9–12 June and 14–16 June, respectively. During the ozone episodes, formaldehyde (Fig. 3c), fine aerosol mass (Fig. 3d), and organic aerosols (Fig. 3d) also increase.
Observational overview of LMOS 2017. Shown are daily time series of (a) winds, rainfall, and temperature; (b) ozone MDA8 concentrations; (c) peak hourly concentrations of selected gases; and (d) daytime PM2.5 concentrations and chemical composition. Gray vertical bands indicate ozone episodes. Temperatures represent maximum hourly air temperature at Lake Mills, WI, and daily lake water temperature at Wilmette Buoy, IL; wind directions represent the most frequent wind direction during 0800–1600 CST (bold for lake breeze and underlined for deep inland penetration lake breeze). Rainfall is the daily total averaged over 27 sites in WI and IL. Sheboygan ozone data are from the KA station. All variables for (c) and (d) are from Zion, IL, except formaldehyde is from the Sheboygan site; gas concentrations are peak hourly values for each day, and (d) is for daytime only (0700–1900 CST).
Citation: Bulletin of the American Meteorological Society 102, 12; 10.1175/BAMS-D-20-0061.1
The frequency of high MDA8 ozone concentrations at Zion and Sheboygan KA was similar to climatological averages for the study period. Climatology frequencies for events exceeding 85 ppb are 1 or less during the field campaign weeks of the year; the most severe ozone episode during LMO2017 did not reach that threshold.
Characterization of synoptic meteorology influences on air quality during LMOS 2017 was conducted by inspection of weekly anomalies in temperature, pressure, and precipitation fields of the North American Regional Reanalysis (Mesinger et al. 2006), and by consulting the NOAA National Centers for Environmental Information synoptic discussions for May and June 2017 (NOAA/NCEI 2017a,b). In general, May 2017, including the first 10 days of the campaign at the end of the month, were wetter and colder than climatological average conditions for Wisconsin and northeastern Illinois. These cold and wet conditions were related to a large-amplitude trough over the Northern Plains. Figure 3 shows campaign measurements consistent with these synoptic conditions—rainy with cool temperatures, low ozone concentrations, depressed biogenic primary and secondary hydrocarbon concentrations, and low aerosol concentrations.
As the trough over the Northern Plains moved to the northeast during the second week of the campaign (28 May–3 June), the weather became warmer, drier, and generally windier—coincident with ozone episode A. During the third week of the campaign (4–11 June), a weak ridge dominated the central plains leading to slightly above-average temperatures over southern and central Wisconsin.
During the fourth week of the campaign (12–19 June), a large-amplitude ridge over the central and eastern United States led to above average temperatures over Iowa, Wisconsin, Michigan, and northern Illinois. Ozone episodes B and C occurred during this period of warm weather and generally southerly winds (end of third week and middle of fourth week). The relative amount of biogenic compounds (isoprene, methyl vinyl ketone, methacrolein) increase relative to anthropogenic compounds (benzene, toluene, ethylbenzene, and xylene), marking a transition from low biogenic conditions (prior to 4 June) to higher biogenic conditions (after 12 June). This transition and its effect on aerosol particles (Figs. 3c,d), is discussed in more detail in Hughes et al. (2021).
In addition to being conducive for ozone formation and transport, the warm and humid conditions in northeastern Wisconsin resulted in increased afternoon cloudiness and occasional thunderstorms (see rainfall in Fig. 3a where there was some rainfall in the study area every day starting 12 June). This complicated flight planning on behalf of the airborne passive remote sensing platform and led to heterogeneous spatial and temporal patterns in air pollutants relative to clear sky episodes. The final week of the campaign had below-normal temperatures over Minnesota and northwestern Wisconsin and near-normal temperatures in Illinois, producing strong surface temperature gradients and somewhat stronger westerly winds over eastern Wisconsin and east-central Illinois. As shown in Fig. 3b, ozone concentrations were fairly low during this period.
Ozone pollution roses at Sheboygan KA are shown in Fig. 4 as derived from observations (left) and from the NOAA National Weather Service Operational NAM-CMAQ forecast. The highest ozone concentrations (>60 ppb) were most commonly observed when winds are from the south or southwest (lake breeze). At Sheboygan KA, the shoreline is angled, running about 35° clockwise of a pure north-to-south line; therefore, at Sheboygan KA, winds from the south and SSW approach from over the lake. Principal errors of the NWS NAM-CMAQ forecasts include both the failure to forecast the frequency of lake breeze circulations at Sheboygan KA, and a large underestimate in peak ozone concentrations.
Ozone (ppb) pollution roses based on (left) 1-min WDNR observations and (right) hourly model output from NAM-CMAQ at the Sheboygan KA from 22 May through 22 Jun 2017. Wind direction frequencies are indicated by the radial distance of the wind rose. The percentage of ozone values falling within 20-ppb bins are indicated by colors for each wind direction. The overall percentages of observed and modeled ozone during LMOS falling in each bin are indicated below the color bars.
Citation: Bulletin of the American Meteorological Society 102, 12; 10.1175/BAMS-D-20-0061.1
Featured measurements
Quantifying weekday/weekend emission variability.
GeoTASO raster pattern flights, in combination with continuous Pandora measurements, enabled emissions characterization by observing the spatial and temporal variations in NO2 columns during LMOS 2017. One stark example of this variability is provided by consecutive morning measurements from Sunday, 18 June, and Monday, 19 June, that highlight the weekend-to-weekday transition. Figure 5 shows spatial maps of NO2 column abundances during the morning (0730–0930 CST; 1330–1530 UTC) on Sunday and Monday over Chicago from GeoTASO. Wind speeds at Chicago Midway ranged from 5 to 9 m s−1 on 18 June and from 3 to 6 m s−1 on 19 June and were predominately from the west to WSW on both days. These GeoTASO measurements show strong signatures of weekday enhancements with peak tropospheric NO2 columns of (5–6) × 1015 molecules cm−2 on Sunday, 18 June, and over 20 × 1015 molecules cm−2 on Monday, 19 June. Diurnal time series of NO2 total column abundances from Pandora measurements located at the Schiller Park monitoring station show that the enhancement observed by GeoTASO on the morning of Monday, 19 June, is associated with the morning rush hour, which peaks between 0600 and 0900 CST. The absence of this peak the day before, along with much lower GeoTASO NO2 signal, suggests that these enhancements are consistent with weekday-associated emissions that are absent or much lower on the weekend. Combined, these GeoTASO and Pandora measurements provide valuable constraints on NO2 emission inventories used for air quality forecasting and assessment modeling.
(top) GeoTASO NO2 tropospheric vertical column densities (VCD; 1015 molecules cm−2) from 0730 to 1530 CST on (left) Sunday, 18 Jun 2017 and (right) Monday, 19 Jun 2017. The location of the Schiller Park AQS monitoring site is labeled. (bottom) NO2 total VCD (molecules cm−2) time series based on Pandora observations at the Schiller Park AQS station for (left) 18 Jun 2017 (blue) and (right) 19 Jun 2017 (red). The Pandora time series are shown as hourly averages (bold symbols) and at the instrument’s native temporal resolution (faint symbols), ∼88 s.
Citation: Bulletin of the American Meteorological Society 102, 12; 10.1175/BAMS-D-20-0061.1
Spatial characterization of an ozone exceedance.
In situ airborne measurements add additional horizontal and vertical characterization of the distribution of NO2 and of the relationship between NO2 and ozone enhancements over Lake Michigan. These measurements also provide data for evaluation of photochemical grid model performance and the GeoTASO NO2 column retrievals. Figure 6 highlights data from 2 June, which was one of the major coastal ozone exceedance days sampled during the 2017 LMOS campaign (also see Fig. 3). On this day GeoTASO reveals elevated onshore and offshore tropospheric NO2 columns over Milwaukee and north of Zion that extend out over Lake Michigan, as well as a narrow plume of enhanced tropospheric NO2 column that extends to the northwest of Sheboygan that is associated with emissions from the Edgewater Power Plant. The spatial pattern between the GeoTASO tropospheric NO2 columns and the integrated SA NO2 profiles are similar that afternoon. Slight difference of up to 2 × 1015 molecules cm−2 is within the uncertainty of the GeoTASO retrieval, and these samples are not entirely temporally collocated as is annotated in the Fig. 6a.
Aircraft mapping and profiling for 2 Jun. (a) NO2 tropospheric vertical columns (molecules cm−2) collected by GeoTASO (GT) with the Scientific Aviation (SA) flight track overlaid (orange). White circles indicate the five locations where surface-to-3-km spirals were executed to collect vertical profiles. Coincident SA and GT columns are labeled (molecules cm−2) along with the difference in time in minutes (SA minus GT). North-to-south transect (southbound legs) observations of (b) ozone (ppb) and (c) NO2 [log10 (ppb)] from SA are overlaid on WRF-Chem modeled O3 and NO2. South-to-north transect (northbound legs) observations of (d) O3 (ppb) and (e) NO2 [log10 (ppb)] from SA are overlaid on simulated O3 and NO2 mixing ratios from WRF-Chem.
Citation: Bulletin of the American Meteorological Society 102, 12; 10.1175/BAMS-D-20-0061.1
The order of the SA spirals are indicated by the red arrows in Fig. 6a. Offshore spirals near Sheboygan and Zion occurred during the north-to-south transect. The onshore spirals over Zion, offshore spirals near Milwaukee, and onshore spirals near Sheboygan occurred during the south-to-north transects. Low-level legs were conducted offshore of Milwaukee during both the north-to-south and south-to-north transects. SA observations show that ozone mixing ratios exceed 100–110 ppb below 100-m altitude offshore of Milwaukee; near Zion, ozone mixing ratios are near 100 ppb between 200- and 400-m altitude.
The largest ozone enhancements offshore of Milwaukee are not captured by the 4-km WRF-Chem simulation. The 4-km WRF-Chem simulations predict ozone mixing ratios of up to 70 ppb on the earlier north-to-south transects and up to 80 ppb on the later south-to-north transect, reflecting ozone production during this time period between the first spiral at 1307 and 1743 CST. SA observations show NO2 > 10 ppb offshore of Milwaukee during low-level legs of both transects that are also not captured by the 4-km WRF-Chem simulation. Short-duration decreases in NO2 at intermediate altitudes (near 200 m) are also observed by SA and not reproduced in WRF-Chem. The significant underestimate of both ozone and NO2 relative to the SA measurements within the marine boundary layer offshore of Milwaukee, and the general ozone underestimate by the 4-km WRF-Chem simulation, highlight the difficulties in predicting shoreline ozone exceedances along the western shore of Lake Michigan.
High temporal observations of the lake breeze.
Lake breezes are an important component of air pollution meteorology around Lake Michigan, and one of the study objectives was to characterize them in detail, including vertical profiling. The criteria for identifying lake breezes are a shift in surface wind direction from offshore to onshore that was nearly coincident with an abrupt cooling of surface temperatures and a rapid decrease in thermodynamic mixing height. In addition, cases with precipitation within 3 h of the wind shift are excluded. Using these criteria, five lake breezes are clearly identified at both of the EM sites (Wagner et al. 2020, manuscript submitted to J. Atmos. Sci.). As shown by the bold and underlined wind directions in Fig. 3, these occur on 2, 8, 11, 12, and 16 June. Lake breezes also occur at Zion on 15 June and at the Sheboygan site on 17 June. The average lake breeze arrival time is 0932 CST at the Sheboygan site and 1030 CST at Zion.
The cases examined are generally characterized by weak synoptic forcing and light winds. Three time series from the 2 June event are shown in Fig. 7. Winds at Zion are observed by sodar between 30 and 200 m AGL; winds shift direction abruptly at all observed levels (Fig. 7a) as surface temperatures decrease (Fig. 7c) with arrival of the lake breeze. Wind observations at the Sheboygan site from Doppler lidar begin at about 140 m AGL. Though the lidar does not capture the onshore flow at lower levels, Fig. 7b shows the upward growth of the lake breeze about two hours after arrival at the surface. For the set of cases examined, the depth of the onshore flow grows as the day progressed. Lake breeze arrival is also accompanied by increases in surface relative humidity, decreases in the water vapor mixing ratio, and increases in PM2.5 concentrations resulting from increases in both Aitken and accumulation mode particles.
Winds (kt; 1 kt ≈ 0.51 m s−1) observed by (a) sodar at Zion, IL, and (b) Doppler lidar at the Sheboygan site, and (c) surface air temperature (°C) during the 2 Jun 2017 lake breeze event. Note the different vertical scale for the two wind cross sections. In wind cross sections, wind barbs closest to the surface are from collocated 10-m wind; the 10-m wind barbs are displaced upward to enhance readability.
Citation: Bulletin of the American Meteorological Society 102, 12; 10.1175/BAMS-D-20-0061.1
Abdi-Oskouei et al. (2020) report mixed success in simulating lake breezes during LMOS 2017. In 4-km WRF-Chem modeling reinitialized daily with HRRR meteorological fields, good statistical performance was achieved compared to traditional benchmarks for wind speed, wind direction, temperature, and water vapor mixing ratio. However, model performance for lake breeze was inconsistent, with some days reproduced in fine details and other lake breezes missed. For Sheboygan, wind direction and speed shifts in WRF-Chem followed observed Doppler lidar patterns on all of the lake breeze days, but with errors in timing on some days and insufficient inland penetration on 11 June. For Zion, the model followed wind direction and speed shifts observed by sodar on all of the lake breeze observation days except 11 and 12 June. The failure of the WRF-Chem model to develop lake breeze on 12 June is also discussed in the next subsection. Errors in timing, usually the model being too late in lake breeze onset by 1–2 h, were common. More detailed analysis of 2 and 11 June, including quantitative metrics, are reported in Abdi-Oskouei et al. (2020). Simulation of lake breeze was found to be necessary but not sufficient for correct simulation of ozone concentrations at coastal sites on most ozone episode days.
Coastal gradient significance.
The GMAP vehicle contained a regulatory-grade commercial ozone monitor and performed preplanned driving routes to map ozone gradients parallel and perpendicular to the coast. The 12 June drive (episode period B), shown in Fig. 8, depicts several features of interest illustrating how the ozone events unfold, the importance of the lake breeze, and the difficulty of forecast modeling at high spatiotemporal resolution. The drive began in late morning and continued from south to north until late afternoon. Ozone concentrations sampled by the mobile platform generally increased during the day, and ranged from 40 to 89 ppb. Figure 8 shows the path of the drive, color coded by ozone concentration. The strong east–west gradient, most evident at Kenosha, is caused by a shallow lake breeze. The elevated ozone is confined to a narrow band near the lake shore. The minimum ozone (40.3 ppb) was found at a traffic intersection in Zion (1121 CST). The maximum ozone (89.4 ppb) was found between Kenosha and Racine (1358 CST), less than 200 m from the shoreline. In general, the distance of inland penetration of the lake breeze increased over this period.
Mapping of intense ozone spatial gradients using GMAP. The GMAP vehicle trajectory on 12 Jun 2017 was from south to north, pausing for collocated measurements at Zion (1114 CST), Chiwaukee (1214 CST), Kenosha (1231–1339 CST), and Racine (1423 CST). Inset (a) shows the area of detail relative to Lake Michigan, while inset (b) shows the WRF-Chem simulation relative to the GMAP drive.
Citation: Bulletin of the American Meteorological Society 102, 12; 10.1175/BAMS-D-20-0061.1
The gradient during the Kenosha east–west transect is particularly striking, with ozone levels of just 57.4 ± 1.6 ppb (±1 standard deviation) at distances more than 4.1 km from the shore. These can be contrasted with 81.4 and 87.4 ppb, respectively, at the beginning and end of the transect (closest to the shore). The gradient during the later (eastbound) portion of the transect is about 9 ppb km−1 over a 3.15-km distance. In other words, ozone varies by over 27 ppb within one model grid cell. The primary reason for the gradient is the limited inland penetration of the ozone-rich lake breeze. At the end of the drive, at Racine (1323 CST), the inland penetration distance has increased, leading to a decrease in the strength of the ozone gradient.
The WRF-Chem model output, shown in Fig. 8b, exhibits a good match with the inland portions of the drive where ozone is 40–60 ppb during the earlier (southern) east–west transect. Specifically, in the portion of the drive corresponding to the southernmost three grid cells, the observed and modeled ozone are 54.3 and 59.1 ppb, respectively. The simulation produces elevated ozone (up to 87.4 ppb) over the lake, but transport patterns are such that it does not reach the GMAP drive locations. Analysis in Abdi-Oskouei et al. (2020) explores the role of lake breeze and synoptic wind in positioning the ozone plume of 12 June. Figure 8b highlights the challenge of model reproduction of fine spatiotemporal features in coastal ozone. To reproduce the timing and magnitude of the ozone time series at coastal monitors, ozone production over the lake must be correctly simulated; furthermore, details of the lake breeze must be accurate—–timing, horizontal extent, and vertical structure.
LMOS 2017 findings
LMOS 2017, an observational field campaign focused on elevated ozone over Lake Michigan and the associated meteorology, ozone precursors, and oxidant chemistry, was successful in capturing relevant events. Three high ozone episodes were captured: 2–4 June, 9–12 June, and 14–16 June. The rich set of in situ and remotely sensed measurements collected during LMOS highlight the complex chemistry and meteorology that leads to ozone production along the western shore of Lake Michigan. These datasets resulted in valuable findings on their own but also serve as a guidepost to assess credibility of models and their representation of these complex photochemical and dynamical processes.
LMOS 2017 aircraft observed narrow plumes of high ozone and nitrogen dioxide (NO2) concentrations within the marine boundary layer over Lake Michigan, at altitudes ranging from about 50 to 370 m above lake level depending on date, time of day, and location. These measurements point to the need for air quality models to have high vertical resolution within the shallow marine boundary layer to adequately capture the vertical structure of mixing and photochemistry as the pollution plumes are transported over the lake.
GeoTASO remote sensing of NO2, supported by ground-based Pandora remote sensing (Judd et al. 2019) and in situ aircraft profiling, provided unprecedented mapping of NO2 over the study region, which provides new insight into the impact of satellite footprint sizes on NO2 column retrievals. This information is critical for use of these remotely sensed airborne, ground based, and satellite NO2 columns for top-down estimates of emissions inventories.
At the Sheboygan and Zion sites, valuable meteorological datasets were collected, including remote-sensed wind and temperature profiles that provide a detailed observational constraint on the evolution of the lake breeze (Figs. 3a and 7). At the Sheboygan site, vertical profiles of water vapor and aerosol were also recorded. These measurements can be used to improve the meteorological predictions of lake breeze circulations which are necessary to predict coastal ozone enhancements in this region.
Elevated ozone periods were predominantly, but not completely, associated with lake breeze airflow. Six lake breeze events have been identified at each of the EM sites and analyzed in detail for temporal features and connections between surface observations and the vertical profiles. An example of the sensitivity of concentrations to horizontal penetration depth of the lake breeze is shown in Fig. 8 as observed from GMAP.
LMOS 2017 shows that ozone air pollution can occur during both anthropogenically dominant periods (prior to 4 June) and during periods when both anthropogenic and biogenic emissions are significant (after 9 June). The transition between these two periods is apparent in time series of concentrations of NOx, formaldehyde, aerosol tracer species, anthropogenic VOCs, isoprene, and isoprene oxidation products.
Contemporary PGM models capture many features of air pollution and meteorology well during LMOS 2017 (e.g., the MDA8 temporal variability in Fig. 3b, the directional dependence of high ozone in Fig. 4, diel patterns of many species, and average spatial patterns of ozone MDA8). However, two classes of ozone simulation difficulties persist: airshed-wide bias on exceedance days (e.g., Figs. 3b, 6b,d, and 8), and consistent reproduction of fine spatiotemporal features (narrow plumes, vertically shallow layers, lake breeze inland penetration distance and timing, and NO2 and other ozone precursors in lake breezes). Using both in situ airborne and surface ozone measurements, we find that both the NAM-CMAQ 12-km modeling and the higher resolution (4-km) WRF-Chem modeling underestimate peak ozone concentrations and overestimate NO2 concentrations during ozone episodes. Model observation differences are reduced but persist in postanalysis modeling with improved meteorological fields. Statistics for WRF-Chem O3 MDA8 on days with observed MDA8 above 65 ppb (n = 6) are biases of −4.6 and −10.2 ppb, respectively, at Zion and Sheboygan KA (graphed in Fig. 3b). On low ozone days, a positive bias was found in WRF-Chem modeling at these two sites; ozone bias stratified by ozone MDA8 can be found in the supplemental material. For the NAM-CMAQ model (Fig. 4), ozone above 60 ppb is much more common in observations than in the model (12.5 and 6.2 times more common, respectively, at Sheboygan KA and Chiwaukee Prairie). NO2 is overpredicted in NAM-CMAQ relative to aircraft observations below 500 m by a factor of 2–4 (0.2–1 ppb interquartile range in observation, and 1–3.5 ppb interquartile range in model). Performance of modeled NO2 compared to aircraft in situ profiles, reported in Abdi-Oskouei et al. (2019), is similar to that of NAM-CMAQ. Additionally, both modeling systems underpredict NO2 at altitudes above about 2 km.
Further model sensitivity studies exploring emissions sensitivities, model resolution, model physics, and model chemical mechanisms need to be conducted to further quantify the reasons for these discrepancies so that they can continue to be reduced. For example, ongoing WRF physics sensitivity studies by the LMOS team suggest that the ability to capture the inland penetration of the lake breeze circulation is dependent on accurate estimates of the Lake Michigan water temperatures, soil moisture, model resolution, and the physics options chosen to represent boundary layer mixing and land surface exchange processes. Both Abdi-Oskouei et al. (2020) and McNider et al. (2018) found important influence of land surface models and data assimilation for land and atmospheric conditions. Abdi-Oskouei et al. (2020) found improvement in meteorological variables with the Noah land surface model (Chen and Dudhia 2001) and with initialization to the 3-km High-Resolution Rapid Refresh (HRRR) (Benjamin et al. 2016) while McNider et al. (2018) found best meteorological performance when assimilating insolation, satellite-derived vegetative greenness, and other land surface characteristics.
Ozone sensitivity to NOx and VOC was assessed through indicator chemical ratios (formaldehyde and NO2 at Sheboygan; H2O2 and HNO3 at Zion) supplemented by box modeling. The results suggest a complex system with some NOx limited periods, and some VOC limited periods. Discussion of the chemistry, emissions sensitivity, and HOx radical fates on 2 June can be found in Vermeuel et al. (2019) while the aerosol chemistry of the event is discussed in Hughes et al. (2021) and indicator ratios at the Sheboygan site are discussed in Abdi-Oskouei et al. (2019). Findings are most developed for 2 June within episode A. For 2 June, Vermeuel et al. (2019) conducted detailed Lagrangian box modeling. On that day, airflow went from Chicago (early morning), out over the lake, and finally to Zion (1600 CST). Ozone production peaked at solar noon at a rate of 10 ppb h−1. Conditions were initially strongly VOC limited. The degree of VOC limitation decreased during the evolution of the plume but remained VOC sensitive when this airmass moved onshore and was detected by instruments at the Zion EM site.
Each ozone event observed during LMOS 2017 was different. We believe these differences are significant for air quality model evaluation, future field campaign design, and air quality management. For example, the H2O2/HNO3 ratio during the 2 June ozone event at Zion (0.35) event was much lower than the study average of 3.3 (Vermeuel et al. 2019). Other observational and model-based lines of evidence support the variation in ozone event chemistry and meteorology. For example, 2 June occurred within the anthropogenically dominated portion of the campaign; modeled ozone concentrations at Zion on 2 June were highly sensitive to anthropogenic VOC emissions in Chicago as would be expected in a NOx rich plume. Finally, the 2 June ozone event had unique PM2.5 composition at Zion—with a strong primary combustion source influence (Hughes et al. 2021).
In addition to variability in NOx–VOC sensitivity, the ozone episodes vary in other aspects, including the importance of biogenics and the role of background and long-distance transport of ozone and ozone precursors. For example, the 11 June event differed from that on 2 June in terms of both aerosol chemistry and predominant air mass origin. On 11 June, high levels of organosulfates derived from isoprene oxidation were detected (Hughes et al. 2021). The source region was likely forested regions of Missouri and Arkansas. This air mass, transported over a relatively long distance, interacted with local emissions to contribute to the 9–12 June episode.
One finding from LMOS 2017 is that the conceptual model of a NOx rich and VOC sensitive urban core zone of high ozone production is useful. As plumes photochemically age, sensitivity of ozone production shifts—usually to a balanced regime where there is sensitivity to both NOx and VOC. The strength of the initial VOC sensitivity in the urban plume, and the location where the transition occurs, vary from episode to episode. This general pattern and associated variability from episode to episode should be kept in mind for future work on ozone air quality in the Lake Michigan airshed.
Another lesson learned from the analysis of LMOS 2017 data are that future campaigns need enhanced VOC measurements in source regions. A denser spatial network of high temporal resolution VOC sampling locations is needed. Measurements of both meteorology and chemistry of the urban plumes over the lake would be of great benefit to further development and evaluation chemical transport models for ozone applications at urban-influenced coastal environments such as Lake Michigan.
Data availability statement.
Measurements used herein are available at the NASA data repository (www-air.larc.nasa.gov/cgi-bin/ArcView/lmos). Model fields used herein are available upon request of the corresponding author.
Acknowledgments.
The LMOS 2017 Science Team acknowledges the NASA Airborne Science Program and the NASA GEOstationary Coastal and Air Pollution Events (GEO-CAPE) Mission Pre-formulation Science Working Group for supporting the airborne remote sensing instruments. This work was funded in part by the National Science Foundation under collaborative Grants AGS-1712909 (COS, EAS), AGS-1713001 (THB), and AGS-1712828 (DBM). DBM also acknowledges support from NSF under Grant AGS-1428257, and EAS acknowledges support from AGS-1405014. The UW Eau Claire team acknowledges the Student Blugold Commitment Differential Tuition program, and NSF Award 1400815. We acknowledge EPA and the NOAA GOES-R program office for supporting the measurements at Sheboygan, and the Electric Power Research Institute (EPRI) for supporting the Scientific Aviation airborne measurements. The University of Iowa modeling team acknowledges NASA support under NNX16AN36G.
We would like to acknowledge the strong support of EPA’s Office of Research and Developments A-E-Energy Research Program in the design and execution of LMOS. We would specifically like to thank Robert Kaleel, former direct of LADCO, for initial contributions to LMOS 2017. We also thank the team at Scientific Aviation, Donald R. Blake (UC Irvine, for VOC analyses), Michal Derliki, Andrew Habel, and Keith Kronmiller with the EPA support contractor Jacobs, and Nishanthi Wijekoon of Wisconsin DNR for mapping support and data.
Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation nor should they be construed as an official National Oceanic and Atmospheric Administration or U.S. Government position, policy, or decision. The views expressed in this paper are those of the authors and do not necessarily represent the views or policies of the U.S. Environmental Protection Agency. EPA does not endorse any products or commercial services mentioned in this publication.
References
Abdi-Oskouei, M. , and Coauthors , 2019: Lake Michigan Ozone Study (2017) Preliminary Finding Report. Lake Michigan Air Directors Consortium, 104 pp., www.ladco.org/wp-content/uploads/Research/LMOS2017/LMOS_LADCO_report_revision_apr2019_v8.pdf.
Abdi-Oskouei, M. , and Coauthors , 2020: Sensitivity of meteorological skill to selection of WRF-Chem physical parameterizations and impact on ozone prediction during the Lake Michigan Ozone Study (LMOS). J. Geophys. Res. Atmos., 125, e2019JD031971, https://doi.org/10.1029/2019JD031971.
Adelman, Z. , 2020: LADCO public issues. www.ladco.org/public-issues/.
Benjamin, S. G. , and Coauthors , 2016: A North American hourly assimilation and model forecast cycle: The Rapid Refresh. Mon. Wea. Rev., 144, 1669–1694, https://doi.org/10.1175/MWR-D-15-0242.1.
Chen, F. , and J. Dudhia , 2001: Coupling an advanced land surface-hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity. Mon. Wea. Rev., 129, 569– 585, https://doi.org/10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2.
Cleary, P. A. , and Coauthors , 2015: Ozone distributions over southern Lake Michigan: Comparisons between ferry-based observations, shoreline-based DOAS observations and model forecasts. Atmos. Chem. Phys., 15, 5109– 5122, https://doi.org/10.5194/acp-15-5109-2015.
Doak, A. G. , and Coauthors , 2021: Characterization of ground-based atmospheric pollution and meteorology sampling stations during the Lake Michigan Ozone Study 2017. J. Air Waste Manage., 71, 866– 889, https://doi.org/10.1080/10962247.2021.1900000.
Dye, T. S. , P. T. Roberts , and M. E. Korc , 1995: Observations of transport processes for ozone and ozone precursors during the 1991 Lake Michigan Ozone Study. J. Appl. Meteor., 34, 1877– 1889, https://doi.org/10.1175/1520-0450(1995)034<1877:OOTPFO>2.0.CO;2.
Fishman, J. , and Coauthors, 2008: Remote sensing of tropospheric pollution from space. Bull. Amer. Meteor. Soc., 89, 805– 822, https://doi.org/10.1175/2008BAMS2526.1.
Fishman, J. , and Coauthors, 2012: The United States’ next generation of atmospheric composition and coastal ecosystem measurements NASA’s Geostationary Coastal and Air Pollution Events (GEO-CAPE) Mission. Bull. Amer. Meteor. Soc., 93, 1547– 1566, https://doi.org/10.1175/BAMS-D-11-00201.1.
Foley, T. , E. A. Betterton , P. E. Robert Jacko , and J. Hillery , 2011: Lake Michigan air quality: The 1994–2003 LADCO Aircraft Project (LAP). Atmos. Environ., 45, 3192– 3202, https://doi.org/10.1016/j.atmosenv.2011.02.033.
Good, G. , 2017: Supplemental information for 2015 Ozone National Ambient Air Quality Standard (NAAQS) area designations. 102 pp., https://dnr.wisconsin.gov/sites/default/files/topic/AirQuality/OzoneTSD20170420.pdf.
Hughes, D. D. , and Coauthors, 2021: PM2.5 chemistry, organosulfates, and secondary organic aerosol during the 2017 Lake Michigan Ozone Study. Atmos. Environ., 244, 117939, https://doi.org/10.1016/j.atmosenv.2020.117939.
IGACO, 2004: The changing atmosphere: An integrated global atmospheric chemistry observation: Report of the Integrated Global Atmospheric Chemistry Observation Theme Team. GAW Rep. 159, WMO/TD-1235, 1282 pp., https://library.wmo.int/doc_num.php?explnum_id=9279.
Judd, L. M. , and Coauthors, 2018: The dawn of geostationary air quality monitoring: Case studies from Seoul and Los Angeles. Front. Environ. Sci., 6, 85, https://doi.org/10.3389/fenvs.2018.00085.
Judd, L. M. , and Coauthors, 2019: Evaluating the impact of spatial resolution on tropospheric NO2 column comparisons within urban areas using high-resolution airborne data. Atmos. Meas. Tech., 12, 6091– 6111, https://doi.org/10.5194/amt-12-6091-2019.
Knuteson, R. O. , and Coauthors, 2004: Atmospheric emitted radiance interferometer. Part I: Instrument design. J. Atmos. Oceanic Technol., 21, 1763– 1776, https://doi.org/10.1175/JTECH-1662.1.
Kowalewski, M. G. , and S. J. Janz , 2014: Remote sensing capabilities of the GEO-CAPE airborne simulator. Proc. SPIE, 9218, 92181I, https://doi.org/10.1117/12.2062058.
Laird, N. F. , D. A. R. Kristovich , X. Z. Liang , R. W. Arritt , and K. Labas , 2001: Lake Michigan lake breezes: Climatology, local forcing, and synoptic environment. J. Appl. Meteor., 40, 409– 424, https://doi.org/10.1175/1520-0450(2001)040<0409:LMLBCL>2.0.CO;2.
Leitch, J. W. , and Coauthors, 2014: The GeoTASO airborne spectrometer project. Proc. SPIE, 9218, 92181H, https://doi.org/10.1117/12.2063763.
Lennartson, G. J. , and M. D. Schwartz , 2002: The lake breeze-ground-level ozone connection in eastern Wisconsin: A climatological perspective. Int. J. Climatol., 22, 1347– 1364, https://doi.org/10.1002/joc.802.
Levelt, P. F. , and Coauthors, 2018: The Ozone Monitoring Instrument: Overview of 14 years in space. Atmos. Chem. Phys., 18, 5699– 5745, https://doi.org/10.5194/acp-18-5699-2018.
Lyons, W. A. , and L. E. Olsson , 1973: Detailed mesometeorological studies of air pollution dispersion in the Chicago lake breeze. Mon. Wea. Rev., 101, 387– 403, https://doi.org/10.1175/1520-0493(1973)101<0387:DMSOAP>2.3.CO;2.
Makar, P. A. , and Coauthors, 2010: Dynamic adjustment of climatological ozone boundary conditions for air-quality forecasts. Atmos. Chem. Phys., 10, 8997– 9015, https://doi.org/10.5194/acp-10-8997-2010.
Mazzuca, G. M. , and Coauthors, 2016: Ozone production and its sensitivity to NOx and VOCs: Results from the DISCOVER-AQ field experiment, Houston 2013. Atmos. Chem. Phys., 16, 14 463– 14 474, https://doi.org/10.5194/acp-16-14463-2016.
McBride, B. A. , J. V. Martins , H. M. J. Barbosa , W. Birmingham , and L. A. Remer , 2020: Spatial distribution of cloud droplet size properties from Airborne Hyper-Angular Rainbow Polarimeter (AirHARP) measurements. Atmos. Meas. Tech., 13, 1777– 1796, https://doi.org/10.5194/amt-13-1777-2020.
McNider, R. T. , and Coauthors, 2018: Examination of the physical atmosphere in the Great Lakes Region and its potential impact on air quality—Overwater stability and satellite assimilation. J. Appl. Meteor. Climatol., 57, 2789– 2816, https://doi.org/10.1175/JAMC-D-17-0355.1.
Mesinger, F. , and Coauthors, 2006: North American Regional Reanalysis. Bull. Amer. Meteor. Soc., 87, 343– 360, https://doi.org/10.1175/BAMS-87-3-343.
Miller, P. , 2018: Overview of the Long Island Sound Tropospheric Ozone Study (LISTOS). 2018Fall Meeting, San Francisco, CA, Amer. Geophys. Union, Abstract A34B-01.
NOAA/NCEI, 2017a: State of the Climate: Synoptic discussion for May 2017. Accessed 25 May 2021, www.ncdc.noaa.gov/sotc/synoptic/201705.
NOAA/NCEI, 2017b: State of the Climate: Synoptic discussion for June 2017. Accessed 25 May 2021, www.ncdc.noaa.gov/sotc/synoptic/201706.
Sullivan, J. T. , and Coauthors, 2019: The ozone water-land environmental transition study: An innovative strategy for understanding Chesapeake Bay pollution events. Bull. Amer. Meteor. Soc., 100, 291– 306, https://doi.org/10.1175/BAMS-D-18-0025.1.
USEPA, 2006: Air quality criteria for ozone and related photochemical oxidants. Final Rep., EPA/600/R-05/004aF-cF, U.S. EPA, 2118 pp., https://cfpub.epa.gov/ncea/risk/recordisplay.cfm?deid=149923.
USEPA, 2013: Integrated Science Assessment (ISA) for ozone and related photochemical oxidants. Final Rep., EPA/600/R-10/076F, 1468 pp., https://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=348522.
USEPA, 2015: Regulatory impact analysis of the final revisions to the national ambient air quality standards for ground-level ozone. U.S. EPA, 480 pp., www.epa.gov/sites/default/files/2016-02/documents/20151001ria.pdf.
Vermeuel, M. P. , and Coauthors, 2019: Sensitivity of ozone production to NOx and VOC along the Lake Michigan coastline. J. Geophys. Res. Atmos., 124, 10 989– 11 006, https://doi.org/10.1029/2019JD030842.
Wagner, T. J. , P. M. Klein , and D. D. Turner , 2019: A new generation of ground-based mobile platforms for active and passive profiling of the boundary layer. Bull. Amer. Meteor. Soc., 100, 137– 153, https://doi.org/10.1175/BAMS-D-17-0165.1.
Zhang, J. , M. Ninneman , E. Joseph , M. J. Schwab , B. Shrestha , and J. J. Schwab , 2020: Mobile laboratory measurements of high surface ozone levels and spatial heterogeneity during LISTOS 2018: Evidence for sea breeze influence. J. Geophys. Res. Atmos., 125, e2019JD031961, https://doi.org/10.1029/2019JD031961.
Zoogman, P. , and Coauthors , 2017: Tropospheric Emissions: Monitoring of Pollution (TEMPO). J. Quant. Spectrosc. Radiat. Transfer, 186, 17– 39, https://doi.org/10.1016/j.jqsrt.2016.05.008.
