ACPAtmospheric Chemistry and PhysicsACPAtmos. Chem. Phys.1680-7324Copernicus PublicationsGöttingen, Germany10.5194/acp-17-2943-2017US surface ozone trends and extremes from 1980 to 2014: quantifying the roles
of rising Asian emissions, domestic controls, wildfires, and climateLinMeiyunmeiyun.lin@noaa.govhttps://orcid.org/0000-0003-3852-3491HorowitzLarry W.PaytonRichardFioreArlene M.https://orcid.org/0000-0003-0221-2122TonnesenGailAtmospheric and Oceanic Sciences, Princeton University, Princeton, NJ
08540, USANOAA Geophysical Fluid Dynamics Laboratory, Princeton, NJ 08540, USAUS Environmental Protection Agency, Region 8, Air Program, Denver, CO 80202, USALamont-Doherty Earth-Observatory and Department of Earth and
Environmental Sciences, Columbia University, Palisades, NY 10964, USAMeiyun Lin (meiyun.lin@noaa.gov)1March2017174294329704December20167December20162February20176February2017This work is licensed under a Creative Commons Attribution 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/This article is available from https://acp.copernicus.org/articles/17/2943/2017/acp-17-2943-2017.htmlThe full text article is available as a PDF file from https://acp.copernicus.org/articles/17/2943/2017/acp-17-2943-2017.pdf
US surface O3 responds to varying global-to-regional precursor
emissions, climate, and extreme weather, with implications for designing
effective air quality control policies. We examine these conjoined processes
with observations and global chemistry-climate model (GFDL-AM3) hindcasts
over 1980–2014. The model captures the salient features of observed trends
in daily maximum 8 h average O3: (1) increases over East Asia (up to
2 ppb yr-1), (2) springtime increases at western US (WUS) rural sites
(0.2–0.5 ppb yr-1) with a baseline sampling approach,
and (3) summertime decreases, largest at the 95th percentile, and
wintertime increases in the 50th to 5th percentiles over the eastern
US (EUS). Asian NOx emissions have tripled since 1990, contributing as
much as 65 % to modeled springtime background O3 increases
(0.3–0.5 ppb yr-1) over the WUS, outpacing O3 decreases attained
via 50 % US NOx emission controls. Methane increases over this
period contribute only 15 % of the WUS background O3 increase.
Springtime O3 observed in Denver has increased at a rate similar to
remote rural sites. During summer, increasing Asian emissions approximately
offset the benefits of US emission reductions, leading to weak or
insignificant observed O3 trends at WUS rural sites. Mean springtime WUS
O3 is projected to increase by ∼ 10 ppb from 2010 to 2030 under
the RCP8.5 global change scenario. While historical wildfire emissions can
enhance summertime monthly mean O3 at individual sites by 2–8 ppb,
high temperatures and the associated buildup of O3 produced from
regional anthropogenic emissions contribute most to elevating observed
summertime O3 throughout the USA. GFDL-AM3 captures the observed
interannual variability of summertime EUS O3. However, O3
deposition sink to vegetation must be reduced by 35 % for the model to
accurately simulate observed high-O3 anomalies during the severe drought
of 1988. Regional NOx reductions alleviated the O3 buildup during
the recent heat waves of 2011 and 2012 relative to earlier heat waves (e.g.,
1988, 1999). The O3 decreases driven by NOx controls were more
pronounced in the southeastern US, where the seasonal onset of biogenic
isoprene emissions and NOx-sensitive O3 production occurs earlier
than in the northeast. Without emission controls, the 95th percentile
summertime O3 in the EUS would have increased by
0.2–0.4 ppb yr-1 over 1988–2014 due to more frequent hot extremes
and rising biogenic isoprene emissions.
Introduction
Within the United States, ground-level O3 has been recognized since the
1940s and 1950s as an air pollutant detrimental to public health. Decreases
in summertime O3 were observed in parts of California and throughout the
EUS (e.g., Cooper et al., 2012; Simon et al., 2015), following regional
NOx controls after the lowering of the US National Ambient Air Quality
Standard (NAAQS) for O3 in 1997 to 84 ppb. On the basis of health
evidence, the NAAQS level for O3 has been further lowered to 75 ppb in
2008 and to 70 ppb in 2015 (Federal Register, 2015). There are concerns that
rising Asian emissions and global methane (Jacob et al., 1999; Lin et al.,
2015b), more frequent large wildfires in summer (e.g., Jaffe, 2011; Yang et
al., 2015; Abatzoglou et al., 2016), and late spring deep stratospheric
O3 intrusions (Lin et al., 2012a, 2015a; Langford et al., 2014) may pose
challenges in attaining more stringent O3 standards in high-elevation
WUS regions. A warming climate would also offset some of the air quality
improvements gained from regional emission controls (e.g., Fiore et al.,
2015). Quantitative understanding of sources of O3 variability on daily
to multi-decadal timescales can provide valuable information to air quality
control managers as they develop O3 abatement strategies under the
NAAQS. Here we systemically investigate the response of US surface O3
means and extremes to changes in Asian and North American anthropogenic
emissions, global methane, regional heat waves, and wildfires over the course
of 35 years from 1980 to 2014, using observations and chemistry-climate model
(GFDL-AM3) hindcasts (Lin et al., 2014, 2015a, b).
Rapid economic growth has led to a tripling of O3 precursor emissions
from Asia in the past 25 years (e.g., Granier et al., 2011; Hilboll et al.,
2013). Observed 1 h O3 mixing
ratios can frequently reach 200–400 ppb during regional pollution episodes
in eastern China (Wang et al., 2006; Li et al., 2016), with a seasonal peak
in the late spring to early summer (Wang et al., 2008; Lin et al., 2009). A
synthesis of available observations from the mid-1990s to the 2000s indicates
increases of 1–2 ppb yr-1 in spring to summer O3 in China (Ding
et al., 2008; Ma et al., 2016; Sun et al.,
2016). Long-range transport of Asian pollution plumes towards
western North America has been identified by aircraft and satellite
measurements and in chemical transport models (e.g., Jaffe et al., 1999;
Fiore et al., 2009; Brown-Steiner and Hess, 2011; Lin et al., 2012b; Huang et
al., 2013; Verstraeten et al., 2015).
Systematic comparison of observed and modeled long-term O3 trends over
Asia is lacking in the published literature but is needed to establish
confidence in models used to assess the global impacts of rising Asian
emissions.
Model simulations indicate that import of Asian pollution enhances mean WUS
surface O3 in spring by ∼ 5 ppb (Zhang et al., 2008; Lin et al.,
2012b), and occasionally contributes 8–15 ppb during springtime pollution
episodes observed at rural sites (Lin et al., 2012b) as supported by in situ
aerosol composition analysis (VanCuren and Gustin, 2015). Stratospheric
intrusions can episodically increase daily 8 h average surface O3 by
20–40 ppb, contributing to the highest observed O3 events at
high-elevation WUS sites (Lin et al., 2012a, 2015a), in
addition to pollution transport from California (e.g., Langford et al.,
2010). In the densely populated EUS, both changes in regional anthropogenic
emissions and air pollution meteorology have the greatest impacts on summer
surface O3 during pollution episodes (e.g., Jacob and Winner 2009;
Rieder et al., 2015; Porter et al., 2015; Pusede et al., 2015). Discerning directly the effect of climate
change on air quality from long-term observation records of O3 would be
ideal, but concurrent trends in precursor emissions and large internal
variability in regional climate impede such an effort. It is difficult to
separate the impacts of changes in global-to-regional precursor emissions and
different meteorological factors on O3 at given locations without the
benefit of multiple sensitivity experiments afforded by models.
On the other hand, process-oriented assessments of the models are needed to
build confidence in their utility for assessing pollution control strategies,
estimating tropospheric O3 radiative forcing and projecting pollution
extremes under future climate scenarios (e.g., Monks et al., 2015). A number
of studies show that global models capture observed decreases in summertime
O3 over the EUS during 1990–2010, but have difficulty simulating
O3 increases measured at remote high-elevation sites that are believed
to represent hemispheric-scale conditions with little influence from fresh
local pollution (hereafter referred to as “baseline”) (e.g., Lamarque et
al., 2010; Koumoutsaris and Bey, 2012; Parrish et al., 2014; Brown-Steiner et
al., 2015; Strode et al., 2015). Recently, Lin et al. (2015b) examined the
representativeness of O3 trends derived from sparse measurements in the
free troposphere over the WUS, originally reported by Cooper et al. (2010)
and used in prior model evaluations. They found that discrepancies between
observed and simulated O3 trends reflect measurement sampling biases.
Here we seek additional insights into the causes of the model–observation
disagreement at the WUS rural sites with continuous, high-frequency
measurements. Notably, we reconcile observed and simulated O3 trends at
these sites with a baseline sampling approach in the model.
Our goal in this paper is 2-fold: first, to systematically evaluate how well
the GFDL-AM3 model represents trends and variability of surface O3
observed at rural sites across the US; second, to examine changes in US
surface O3 means and extremes in a suite of multi-decadal hindcast
simulations designed to isolate the response of O3 to increases in Asian
anthropogenic emissions, North American emission controls, rising global
methane, wildfires, and interannual variability in meteorology. We examine
trends across the entire probability distribution of O3 concentration,
which is crucial to assessing the ability of models to simulate the surface
O3 response under different temperature and chemical regimes depending
on seasons, geographical location, and regional transport patterns.
Specifically, we evaluate the trends separately for the 5th, 50th and 95th
percentiles of the O3 concentration distribution in spring (MAM),
summer (JJA), autumn (SON), and winter (DJF).
Section 2 briefly describes the observational records, model experiments, and
analysis approach. As a first step towards assessing our understanding of the
impacts of rising Asian emissions, we briefly review Asian O3 trends
from observations in recent publications and evaluate modeled trends
(Sect. 3). We then focus our analysis on the US, using both observations and
models to assess the response of US surface O3 to changes in background
O3, regional anthropogenic emissions and meteorology (Sect. 4). In
Sect. 5, we further separate the influence of background on WUS O3 into
components driven by rising Asian anthropogenic emissions, global methane,
and wildfires. We quantify the contribution of these factors to surface
O3 in both rural areas such as national parks (Sect. 5.1 to 5.3) and in
densely populated regions such as the Denver metropolitan area (Sect. 5.4).
After evaluating historical trends, we additionally draw upon two simulations
following the 21st century RCP4.5 versus RCP8.5 global change scenarios to
project WUS O3 through 2050 (Sect. 5.2). Section 6 examines how the EUS
summertime O3 probability distribution and pollution extremes respond to
large-scale heat waves, droughts, and regional NOx reductions over the
past decade, and how well our model simulates the observed features. Finally,
we summarize in Sect. 7 the key drivers of US surface O3 trends and
extremes and discuss the implications of this study.
Model and observationsChemistry-climate model experiments
Summary of forcings and emissions used in AM3 hindcasts and CM3
projections.
ExperimentTime periodsMeteorologyRadiative forcingsCH4 (chemistry)Anthropogenic emissionsFire emissionsBASE1979–2014Nudged to NCEPHistoricalHistoricalHistoricalHistoricalBackground1979–2014As BASEHistoricalHistoricalZeroed out in N. America;Historicalas in BASE elsewhereFIXEMIS1979–2014As BASEHistorical2000Constant1Constant1IAVFIRE1979-2014As BASEHistorical2000Constant1HistoricalIAVASIA1979–20122As BASEHistorical2000Varying in Asia as in BASE;Constant1as in FIXEMIS elsewhereIAVCH41979–20122As BASEHistoricalHistoricalConstant1Constant1CM3_RCP4.52005–2050Free runningRCP4.5RCP4.5RCP4.5RCP4.5CM3_RCP8.52005–2050Free runningRCP8.5RCP8.5RCP8.5RCP8.5
1 Averaged over the whole 1970–2010 period.
2 Note that the IAVASIA and IAVCH4 simulations only extend to
2012.
The GFDL-AM3 model includes interactive stratosphere–troposphere chemistry
and aerosols on a cubed sphere grid with a resolution of approximately
200 × 200 km2 (Donner et al., 2011). Table 1 summarizes the
meteorology, radiative forcing agents, and emissions used in each experiment.
The hindcast simulations (1979–2014) are nudged to the NCEP/NCAR reanalysis
zonal and meridional winds using a height-dependent nudging technique (Lin et
al., 2012b). Biogenic isoprene emissions and lightning NOx are tied to
model meteorology (Guenther et al., 2006; Rasmussen et al., 2012) and thus
can respond to changes in climate, whereas soil NOx and chemical dry
deposition velocities are set to a monthly climatology (Naik et al., 2013),
with a diurnal cycle applied for O3 dry deposition. To investigate the
possible influence of drought on O3 removal (e.g., Emberson et al.,
2013), we additionally conduct a sensitivity simulation for 1988 with reduced
O3 deposition velocity (see Sect. 6). Our BASE simulation and two
additional simulations with modified emissions (FIXEMIS and IAVFIRE) were
previously used to interpret the causes of increasing autumnal O3
measured at Mauna Loa Observatory in Hawaii since 1974 (Lin et al., 2014),
interannual variability of springtime O3 (Lin et al., 2015a) and the
representativeness of free tropospheric O3 measurements over the WUS
(Lin et al., 2015b).
With anthropogenic emissions and methane held constant (Table 1), the FIXEMIS
and IAVFIRE simulations isolate the influence from meteorology and wildfire
emissions, respectively. In IAVASIA, anthropogenic emissions from East Asia
(15–50∘ N, 95–160∘ E) and South Asia (5–35∘ N,
50–95∘ E) are allowed to vary from year to year as in BASE, while
anthropogenic emissions in the other regions of the world, global methane and
wildfire emissions are held constant as in FIXEMIS. In IAVCH4, global
methane is allowed to vary over time as in BASE, but with anthropogenic and
wildfire emissions held constant as in FIXEMIS. The IAVASIA and IAVCH4
simulations thus isolate the role of rising Asian anthropogenic emissions and
global methane, respectively, by contrasting with the FIXEMIS simulation.
Both BASE and IAVCH4 simulations apply observed time-varying methane
concentrations as a lower boundary condition for chemistry (Supplement
Fig. S1). Thus, underestimates in historical methane emissions reported
recently by Schwietzke et al. (2016) do not affect our
results. We quantify the total contributions to surface O3 from
meteorological variability, stratosphere-to-troposphere transport, pollution
from foreign continents and O3 produced by global methane, lightning
NOx, wildfires and biogenic emissions with the Background simulation, in
which North American anthropogenic emissions are zeroed out relative to BASE.
We additionally draw upon two simulations with the GFDL Coupled Model CM3
following the 21st century RCP global change scenarios to project changes in
WUS O3 through 2050. Details of these CM3 simulations were described in
John et al. (2012).
Anthropogenic and biomass burning emissions
Changes in NOx emissions. (a–b) Mean annual vertical
column densities of tropospheric (VCDtrop) NO2 normalized to the year
2000 for the eastern China and eastern US domains (black boxes on map) from
GOME (1996–2002, open circles) and SCIAMACHY (2003–2011, closed circles)
measurements and AM3 BASE simulations (orange lines). Triangles indicate
trends in NOx emissions (normalized to 2000) from Lamarque et al. (2010)
with annual interpolation after 2000 to RCP8.5 (red) versus RCP4.5 (blue).
(c–d) Differences in annual mean SCIAMACHY VCDtrop NO2 from
2003–2005 to 2009–2011. The red boxes denote the regions where emissions
vary over time in the IAVASIA simulation (Table 1). Satellite NO2 data
are from www.temis.nl, with the retrieval technique described in
Boersma et al. (2004).
We first examine how well the emission inventories in AM3 BASE represent
changes in regional NOx emissions over recent decades inferred from
satellite measurements of tropospheric vertical column density
(VCDtrop) of NO2. The combined record of GOME and SCIAMACHY
shows that VCDtrop NO2 over the highly polluted region of
eastern China almost tripled during 1996–2011 (Fig. 1a). In contrast,
VCDtrop NO2 over the EUS decreased by ∼ 50 % in
the 2000s (Fig. 1b) due to NOx State Implementation Plans (commonly
known as the NOx SIP Call) and many rules that tighten emission
standards for mobile sources (McDonald et al., 2012). Similar decreases
occurred in WUS cities, resulting from the NOx control programs to
achieve O3 and regional haze planning goals. These trends are consistent
with those reported by a few recent studies (e.g., Hilboll et al., 2013),
including those using OMI NO2 data (Russell et al., 2012; Duncan et al.,
2016). For comparison with satellite data, we sample the model archived every
3 h closest to the time of satellite overpass for the SCIAMACHY and GOME
products we use in Fig. 1 (10:00–10:30 local time). Trends in
VCDtrop NO2 are similar to those in NOx emissions
(orange lines versus red triangles in Fig. 1a–b), indicating that any
changes in NOx chemical lifetime or partitioning have negligible
influence in our model, consistent with NO2 loss against OH being minor
during the morning overpasses of GOME and SCIAMACHY. The emission inventory
used in BASE, from Lamarque et al. (2010) with annual interpolation after
2000 to RCP8.5 (Lamarque et al., 2012), mimics the opposing changes in
NOx emissions over eastern China versus the EUS during 1996–2011,
consistent with changes in VCDtrop NO2 retrieved from the
satellite instruments. For comparison, the RCP4.5 interpolation for
2001–2010 in CMIP5 historical simulations analyzed by Parrish et al. (2014)
underestimates the increase in Chinese NOx emissions by a factor of 2
(Fig. 1a). Recent reductions in Chinese NOx emissions after 2011 (Duncan
et al., 2016) are not represented in the inventories used in AM3.
Our BASE model applies interannually varying monthly mean emissions from
biomass burning based on the RETRO inventory (Schultz et al., 2008) for 1970
to 1996 and GFEDv3 (van der Werf et al., 2010) for 1997 onwards, distributed
vertically as recommend by Dentener et al. (2006). Figure S2 illustrates the
interannual variability of biomass burning CO emissions from the main source
regions of the Northern Hemisphere over the period 1980–2014. Boreal fire
emissions in Eurasia almost doubled from 1980–1995 to 1996–2014, with large
fires occurring more frequently in the recent decade, as found for the WUS
(Dennison et al., 2014; Yang et al., 2015).
Ozone observation records and uncertainties
Long-term surface O3 observation records were obtained at 70 selected
rural monitoring sites with 20 (1995–2014) to 27 (1988–2014) years of
continuous hourly measurements from the US National Park Services, the US
Clean Air Status and Trends Network (CASTNet), and the US EPA Air Quality
System. Cooper et al. (2012) reported trends in daytime
(11:00–16:00)
O3 over 1990–2010 at 53 rural sites. We investigate trends in daily
maximum 8 h averaged (MDA8) O3 and expand the analysis of Cooper et
al. (2012) using additional data to 2014 and including 17 additional sites
with measurements begun in 1991–1995. All sites have at least 20 years of
data. If a site has less than 50 % data availability in any season, then
that particular season is discarded. The trend is calculated separately for
the 5th, 50th and 95th percentiles of daily MDA8 O3 for each season
through ordinary linear least-square regression. Statistics are derived for
the slope of the linear regression in units of ppb yr-1, the range of
the slope with a 95 % confidence limit (not adjusted for sample
autocorrelation), and the p value indicating the statistical significance
of the trend based on a two-tailed t test.
Measurement uncertainties. (a) Comparison of observed
monthly mean MDA8 O3 at WUS CASTNet sites. All sites have more than
90 % data availability in every month shown. The gray shading denotes the period when data at
Yellowstone (red) and Rocky Mountain (black) were inconsistent with the other
sites. (b–c) The 1990–2010 trends of median JJA MDA8 O3 at
Yellowstone and median MAM MDA8 O3 at Rocky Mountain with and without
data in 1990.
A cross-site consistency analysis was performed to determine robust changes
in the time evolution of O3 over the WUS during 1988–2014 (Fig. 2). The
monitor at Yellowstone National Park was moved 1.5 km from the Lake
Yellowstone site to the Water Tank site in 1996. While the local transport
patterns are slightly different for the two sites, using MDA8 data from the
well-mixed midday period minimizes the differences (Jaffe and Ray, 2007).
Observed O3 interannual variations show large-scale similarity across
sites over the Intermountain West except for the earlier period 1989–1990.
During this period, observations at Yellowstone and Rocky Mountain national
parks show low-O3 anomalies that do not appear at other sites, but there
is no change in measurement technique. Jaffe and Ray (2007) suggest this
represents large-scale variations in background O3 that are seen in
common at these two parks. However, analysis of meteorological fields and
model diagnostics does not reveal any obvious transport anomaly influencing
O3 variations at these sites in 1990 (Lin et al., 2015a). Observations
at Pinedale in January–February 1990 are also anomalously low relative to
Grand Canyon (GRC474), Centennial (CNT169), and Gothic (GTH161). These
anomalous data at the beginning of measurement records can substantially
influence trends calculated from short records. For example, Cooper et
al. (2012) found a summer O3 increase of
0.42 ± 0.30 ppb yr-1 at Yellowstone over 1990–2010. Removing
1990, we find a weaker increase of 0.28 ± 0.27 ppb yr-1
(Fig. 2b). Removing 1990 at Rocky Mountain resulted in a weaker springtime
O3 increase of 0.29 ± 0.17 ppb yr-1 compared to
0.43 ± 0.23 ppb yr-1 over 1990–2010 (Fig. 2c). To assess robust
O3 changes, we thus remove these apparently uncertain measurements in
1990 from the subsequent analysis.
Model baseline sampling approach
Influence of baseline sampling. Median spring MDA8 O3 trends
over 1988–2014 at WUS sites from (a) observations,
(b) BASE model sampled at the surface, (c) BASE sampled at
700 hPa and filtered to remove the influence from fresh local pollution (see
Sect. 2.4), (d) BASE sampled at 700 hPa without filtering, and
(e–f) Background (with North American anthropogenic emissions shut
off) sampled at the surface versus at 700 hPa. Note that three low-elevation
(< 1.5 km) sites, Joshua Tree, Big Bend and Glacier national parks,
are always sampled at the surface. Larger circles indicate sites with
statistically significant trends (p < 0.05).
Global distribution of MDA8 O3 trends from AM3 BASE over
1988–2014 for boreal spring (left) and summer (right) for the 95th
percentile at the surface (a–b), median at the
surface (c–d), and median in the free troposphere (700 hPa;
e–f). Stippling indicates areas where the trend is statistically
significant (p < 0.05). The color scale is designed to resolve
regional features rather than extreme values and saturates. The range of the
trends is -1 to +2.5 ppb yr-1.
Springtime O3 observations at WUS high-elevation sites (≥ 1.5 km a.s.l.) typically represent baseline conditions with little
influence from fresh local pollution. In a global model with
∼ 200 × 200 km2 horizontal resolution, however, these
remote sites can reside in the same grid cell that contains urban cities
where NOx emissions decreased over the analysis period. For example,
Rocky Mountain National Park (2.7 km a.s.l.) is less than 100 km from the
Denver metropolitan area in Colorado. This limitation of large-scale models
in resolving urban-to-rural gradients and sharp topography results in an
artificial offset of increased baseline O3 at remote sites by decreased
urban pollution within the same model grid cell. Thus, coarse-resolution
models are often unable to reproduce observed O3 increases at the
high-elevation sites representative of remote baseline conditions (Fig. 3a
versus b), as found in many prior modeling analyses (e.g., Parrish et al.,
2014; Strode et al., 2015, and references therein). This limitation can be
addressed by using a baseline selection procedure to identify conditions for
sampling the model to avoid model artifacts caused by poor spatial
resolution, as described below.
All measurements presented in this study are unfiltered. We implement a set
of regional CO-like tracers (COt), with a 50-day exponential decay lifetime
and surface emissions constant in time from each of four northern
mid-latitude source regions (Lin et al., 2014). We use these COt tracers to
bin modeled O3 according to the dominant influence of different
continental air regimes. To represent observed baseline conditions at WUS
sites, we sample AM3 at 700 hPa (∼ 3 km a.s.l.) and filter the
O3 data in the BASE simulation to remove the influence from fresh local
pollution. Specifically, our filter excludes days when North American COt
(NACOt) exceeds the 67th percentile for each season. This procedure yields
higher calculated baseline O3 increases (Fig. 3c), bringing it closer to
observations (Fig. 3a). When sampled at 700 hPa without filtering (Fig. 3d),
BASE gives statistically significant O3 increases, but the rate of
increase is ∼ 0.1 ppb yr-1 weaker than with filtering. With
North American anthropogenic emissions shut off, the model simulates
significant O3 increases that are similar at the surface (Fig. 3e) and
at 700 hPa (Fig. 3f). This finding indicates that the underestimate of
O3 increases in BASE, when sampled at the surface (Fig. 3b), reflects an
excessive offset from domestic pollution decreases in the model relative to
observed conditions, as opposed to the insufficient mixing of free
tropospheric O3 to the surface. As individual sites display observed
trends falling in between the filtered model, and those sampled at the
surface versus aloft, we can use the model to interpret which sites most
frequently sample baseline versus being influenced by North American
anthropogenic emissions. For consistency, in the subsequent analysis we apply
model baseline filtering to all WUS sites with elevations greater than
1.5 km altitude. In the EUS, where the terrain and monitor elevations are
much lower than in the west and observed O3 trends are largely
controlled by regional emission changes, we always sample the model at the
surface without filtering.
Global distribution of lower tropospheric O3 trendsGlobal O3 burden and distribution of trends
We begin by examining the global distribution of lower tropospheric O3
trends over 1988–2014 from the BASE simulation (Fig. 4) and focus on the
differences between the surface and free troposphere (∼ 700 hPa), with
implications for understanding the impact of trends in hemispheric baseline
O3 on surface air quality. The model indicates that surface MDA8 O3
levels in Asia have increased significantly by 1.5–2.5 ppb yr-1 in
the 95th percentile (Fig. 4a–b) and by 1–2 ppb yr-1 in the median
values (Fig. 4c–d), with the largest increases occurring in southern Asia
during spring and over eastern China during summer. In contrast, there is a
marked decrease in surface MDA8 O3 in WUS cities, throughout the EUS and
in central Europe, particularly at the high percentiles and during summer.
The increase in surface O3 over Asia and decreases over the US and
Europe are consistent with changes in regional emissions of O3
precursors over this period (Fig. 1).
Over Southeast Asia (south of 30∘ N) during spring, earlier
springtime O3 photochemical production at lower latitudes coupled with
active frontal transport (Liu et al., 2002; Carmichael et al., 2003; Lin et
al., 2010) leads to a comparable or even greater increase in O3 in the
free troposphere than at the surface (Fig. 4c versus e). In contrast, over central eastern China during
summer the simulated trends of O3 in the free troposphere are at least a
factor of 3 weaker than in surface air (Fig. 4d versus f), consistent with
the analysis of MOZAIC aircraft data over Beijing in 1995–1999 versus
2003–2005 (Ding et al., 2008). Mean O3 at 700 hPa above parts of North
America and Europe show little change in summer or even increase during
spring in the model, similar to the trends at 500 hPa (Fig. S3), despite the
significant decreases in surface air. The global tropospheric O3 burden
in the BASE simulation increases by approximately 30 Tg over the past
35 years (Fig. 5a), attributed mainly to changes in anthropogenic emissions.
Over the 2004–2015 OMI/MLS satellite era, however, meteorological
variability contributes approximately half to the total simulated decadal
trends of O3 burden (Fig. 5a), indicating that attribution of the
satellite-derived decadal trends of global tropospheric O3 burden
requires consideration of internal climate variability.
Comparison of observed and simulated O3 trends in Asia
(a) Time series of changes in global tropospheric O3
burden relative to the 1981–1990 mean from BASE and FIXEMIS simulations
(Table 1). (b) Time series of 12-month running mean anomalies
(relative to the 2005–2014 mean) of O3 averaged over 900–600 hPa at
Hong Kong from the averages of ozonesonde samples (black circles) and the
BASE model co-sampled on sonde launch days (orange circles) versus the true
average from BASE and IAVFIRE with continuous daily sampling (solid lines).
(c) Same as (b), but for Hanoi.
Surface O3 trends in Asia. (a) Observation sites
superimposed on a map of the 95th percentile summer MDA8 O3 trends over
1995–2014 from AM3 BASE. (b) Comparison of median O3 trends
from AM3 (1995–2014) with observations (see text for periods): in central
eastern China at Mt. Tai (July–August, Sun et al., 2016), Beijing
(May–June–July, Ding et al., 2008) and Shangdianzi (SDZ) (JJA, Ma et al.,
2016); in South China at Hong Kong (HK) (annual average, Wang et al., 2009)
and Taiwan (MAM, Lin et al., 2010); at Mt. Waliguan (WLG) in western China
(MAM, Xu et al., 2016); in South Korea (JJA, Lee et al., 2014) and Mt. Happo
Japan (MAM, Tanimoto, 2009). For Mt. Happo (triangle on map) AM3 is sampled
at 700 hPa and filtered for the influence from Asian continental air – more
representative of observed baseline conditions in spring.
Long-term O3 observations are very sparse in Asia, making it difficult
to evaluate modeled O3 trends. We compile available measurements from
the published literature, including ozonesonde profiles at Hong Kong
(2000–2014; www.woudc.org) and Hanoi (2005–2015; SHADOZ, Thompson et
al., 2007), MOZAIC aircraft profiles collected on summer afternoons in the
boundary layer (below 1250 m altitude) over Beijing for 1995–2005 (Ding et
al., 2008), ground-based measurements at Mt. Tai (1.5 km a.s.l.) in central
eastern China for July–August 2003–2015 (Sun et al., 2016), at the GAW
stations, Shangdianzi north of Beijing for 2004–2014 (Ma et al., 2016) and
Mt. Waliguan (3.8 km a.s.l.) on the Tibetan Plateau for 1994–2013 (Xu et
al., 2016), at Taiwan for 1994–2007 (Lin et al., 2010), South Korea for
1990–2010 (Lee et al., 2014), Mt. Happo (1.9 km a.s.l.) in Japan for
1991–2011 (Tanimoto, 2009; Parrish et al., 2014), and a coastal site at Hong
Kong in southern China for 1994–2007 (Wang et al., 2009).
Linear trends in spring (MAM) MDA8 O3 over 1988–2014 at US
rural sites for the 95th, 50th, and 5th percentiles as observed (left) and
simulated (right) in AM3 BASE. Larger circles indicate sites with
statistically significant trends (p < 0.05). For WUS
high-elevation sites, the model is sampled at 700 hPa and filtered to remove
local influence (see text in Sect. 2.4).
Recently, Zhang et al. (2016) compiled sparse O3 profiles above
Southeast Asia from IAGOS commercial aircraft and ozonesondes from Hanoi for
1994–2004 versus 2005–2014 and found a total springtime O3 increase of
20–25 ppb between the two periods (∼ 2 ppb yr-1). However, our
model indicates an increase of up to 1 ppb yr-1 for free tropospheric
O3 over Southeast Asia in spring (Fig. 4e). We illustrate the possible
influence of sampling deficiencies on the O3 trends inferred from sparse
observations (Fig. 5). The ozonesonde frequency is four profiles per month at
Hong Kong and only one to two profiles per month at Hanoi. To determine the
representativeness of O3 trends derived from these sparse measurements,
we compare observations and model results co-sampled on sonde launch days,
with the “true average” determined from O3 fields archived every 3 h
from the model, as in our prior work for WUS sites (Lin et al., 2015a, b).
Figure 5b and c show the comparisons for the annual trends of O3 over
900–600 hPa. The trends are generally consistent across the sonde data,
model co-sampled and “true average” results for Hong Kong, with an increase
of 0.5 ± 0.1 ppb yr-1 over 2000–2014. Observations at Hanoi
show an apparently rapid O3 increase of 1.1 ± 0.2 ppb yr-1
over 2005–2014. AM3 BASE, when sampled sparsely as in the ozonesondes, captures the observed
variability (r2= 0.7), whereas the “true average” over this period
indicates the trend (0.7 ± 0.1 ppb yr-1) is only 63 % of
that inferred from observations. Moreover, interannual variability of O3
resulting from wildfire emissions and meteorology in IAVFIRE is as large as
the total O3 change in BASE over the short period 2005–2014. We
conclude that measurement sampling artifacts influence the O3 trends
reported by Zhang et al. (2016).
As in Fig. 7, but for summer (JJA). Note that the color scale
saturates at ±0.8.
As in Fig. 7, but for winter (DJF). Large squares
in (a) denote AQS sites with significant O3 decreases in the
95th percentile.
Expanding the comparison to a suite of sites across East Asia (Fig. 6), we
find that AM3 captures the key features of observed O3 trends in Asia,
including their seasonal to regional variations, summertime increases
(1–2 ppb yr-1) in central eastern China where NOx emissions have
approximately tripled since 1990 (Fig. 1a), and springtime increases
(0.5 ppb yr-1) at Taiwan and Mt. Happo that are driven by pollution
outflow from the Asian continent. Note that to place the trends derived from
the short observational records into a broader context, we show the 20-year
trends over 1995–2014 from the model, except for South Korea (1990–2010)
and Happo, Japan (1991–2011). We match the time period in the model with
observations at these two sites because AM3 shows weaker O3 increases
when data for the recent years are included, which likely reflects the
offsetting effects of regional emission reductions in South Korea and Japan.
Parrish et al. (2014) show that three CMIP5-like models underestimate the
observed springtime O3 increase at Mt. Happo by a factor of 4. This
discrepancy may reflect a combination of factors: (1) underestimates of Asian
emission growth in the RCP4.5 interpolation after 2000 used in CMIP5
historical simulations (Fig. 1a); (2) trends driven by interannual
meteorological variability that free-running CMIP5 models are not expected to
reproduce exactly; (3) an excessive offset from Japanese pollution decreases
in the models owing to their coarse resolution and limitation in resolving
observed baseline conditions at Mt. Happo. Sampling our BASE model at
700 hPa above Happo, we find an O3 increase of
0.35 ± 0.13 ppb yr-1. When focusing on days strongly influenced
by outflow from the East Asian continent (Chinese COt ≥ 67th), the
model O3 trend increases to 0.48 ± 0.13 ppb yr-1,
approximating the observed increase of 0.76 ± 0.35 ppb yr-1 at
Mt. Happo (Fig. 6b). The observed and simulated trends are not statistically
different given the overlapping confidence limits. The larger confidence
limit (uncertainty) derived from the Happo observations reflects the
measurement inconsistency before 1998 and instrumental problems after 2007
(Tanimoto et al., 2016). We conclude that GFDL-AM3 captures 65–90 % of
the observed O3 increases in Asia, lending confidence in its application
to assess the global impacts of rising Asian emissions.
Regional and seasonal variability of US surface O3 trends
We next focus our analysis on the US, where dense, high-frequency, long-term,
reliable measurements of surface O3 facilitate process-oriented model
evaluation. Comparisons of surface O3 trends over 1988–2014 at 70 rural
monitoring sites across the US as observed and simulated in AM3 BASE are
shown in Fig. 7 for spring, Fig. 8 for summer, Fig. 9 for winter, and in
Fig. S4 for autumn. The trends are calculated separately for the 5th, 50th
and 95th percentiles of the daily MDA8 O3 concentration distribution,
with larger circles on the maps indicating sites with statistically
significant trends (p < 0.05). We first discuss observations
(Sect. 4.1), followed by model evaluation and trend attribution (Sect. 4.2).
Observations
In spring (Fig. 7), observations indicate spatial heterogeneity in O3
trends across the Intermountain West and the northeastern (north of
38∘ N) and southeastern US. At the 95th percentile (Fig. 7a) the
pattern of observed trends is homogeneous across the northeastern and
southeastern US, with approximately 85 % of the sites having
statistically significant O3 decreases of 0.4–0.8 ppb yr-1 and
no sites showing a significant increase. In contrast, significant increases
occur at 25 % of the sites in the Intermountain West. Only Joshua Tree
National Park located downwind of the Los Angeles Basin shows a significant
decrease at the 95th percentile. At the 50th percentile (Fig. 7b) there are
significant O3 decreases of 0.2–0.4 ppb yr-1 in the southeast
and little overall change in the northeast, while significant increases of
0.2–0.5 ppb yr-1 occur at 50 % of the sites in the Intermountain
West. Significant springtime O3 increases occur at all observed
percentiles at Lassen Volcanic National Park in California, Great Basin
National Park in Nevada, Rocky Mountain National Park and US Air Force
Academy in Colorado. At the 5th percentile (Fig. 7c) significant O3
increases occur at most sites in the northeast, while little change and some
negative trends are found in the southeast. The occurrence of the greatest
observed O3 decreases for the highest percentiles is consistent with
high-temperature O3 production being more NOx-limited (Pusede et
al., 2015) and thus more responsive to decreases in NOx emissions.
The north-to-south gradient in springtime O3 trends over the EUS
reflects the earlier seasonal transition from NOx-saturated to
NOx-sensitive O3 production regimes in the southeast, where
plentiful radiation in spring enhances HOx supply and biogenic isoprene
emissions begin earlier than in the northeast. The different response of
springtime O3 to NOx controls in the southeast versus northeast
noticed in this work is not present in prior analyses for shorter time
periods (1990–2010 in Cooper et al., 2012, and 1998–2013 in Simon et al.,
2015). We find 72 % of the southeastern sites experiencing significant
median O3 decreases in spring over 1988–2014, while Cooper et
al. (2012) found only 8 %. Sites with significant 95th
percentile springtime O3 decreases in the EUS are also much more common
in our study (85 % versus 43 % in Cooper et al. (2012). In
the 5th percentile, 45 % of the northeastern sites in our analysis have
significant spring O3 increases, with only 15 % in Cooper et
al. (2012) Stronger O3 reductions in the southeast than the
northeast also occur during autumn (Fig. S4), reflecting an extension of
biogenic isoprene emissions and NOx-sensitive O3 production in the
southeast to autumn.
In summer (Fig. 8), as radiation intensifies and isoprene emissions peak
seasonally, the O3 production becomes more NOx-limited across both
the southeastern and northeastern US, where NOx emission controls have
led to significant O3 decreases of 0.8–1.8 ppb yr-1 in the 95th
percentile and 0.4–0.8 ppb yr-1 in the median value (Fig. 8a–b). In
the southeast, significant decreases have also occurred at the lowest
percentiles during summer (Fig. 8c), in contrast to the weak response during
spring (Fig. 7c). Many northeastern states in the late 1990s and early 2000s
did not turn on power plant NOx emission controls until the O3
season (May–September), which may contribute to observed differences between
spring and summer O3 trends. Compared to the 1990–2010 trends reported
in Cooper et al. (2012), the EUS summer O3 decreases reported here with
additional data to 2014 are 33 % stronger. Despite reductions in
precursor emissions in the WUS cities (Fig. 1d), there are no significant
summer O3 decreases at the intermountain sites, except in Yosemite and
Joshua Tree national parks for the 95th percentile. Instead, a significant
summer increase of ∼ 0.3 ppb yr-1 occurs across the entire
O3 distribution at Yellowstone. Significant summer increases are found
in the 5th percentile for Lassen, Mesa Verde, and Rocky Mountain national
parks.
In winter (Fig. 9), observed O3 increases are more common than in spring
and summer across the US. The wintertime O3 increases are strongest in
the lowest percentiles over the EUS, indicating the influence from weakened
NOx titration as a result of regional NOx emission controls (see
also Gao et al., 2013; Clifton et al., 2014; Simon et al., 2015). Even during
winter, some decreasing O3 trends are found in the highest percentiles
over the southeast (Fig. 9a), most prominently in Texas (Dallas and Houston),
where tropical climate and year-round active photochemistry makes O3
most responsive to regional NOx emission controls. Despite the greatest
NOx emission reductions over the past decade in the central and
northeastern US regions, observed O3 reductions have been most
pronounced in the southeast, particularly in spring and autumn.
Model evaluation and attribution of observed O3 trends
The BASE simulation with GFDL-AM3 captures the salient features of observed
O3 trends over 1988–2014 at rural sites across the US: (1) the overall
springtime increases and the lack of significant trends in summer over the
Intermountain West; (2) the north-to-south gradients in O3 trends during
spring and the largest decreases in the 95th percentile during summer over
the EUS; (3) wintertime increases in the 5th and 50th percentiles (left
versus right panels in Figs. 7 to 9). AM3 also simulates a median springtime
O3 increase of 0.32 ± 0.11 ppb yr-1 over 1988–2014
(0.64 ± 0.50 ppb yr-1 over 2004–2014) at Mount Bachelor
Observatory in Oregon, consistent with the positive trend
(0.63 ± 0.41 ppb yr-1) observed over the shorter 2004–2015
period (Gratz et al., 2014). These analyses imply that GFDL-AM3 represents
the underlying chemical and physical processes controlling the response of US
surface O3 means and extremes to changes in global-to-regional precursor
emissions and climate, despite mean state biases (Figs. S5–S6).
The filtered model shows greater 95th percentile O3 increases than
observed at some WUS sites (e.g., Yosemite; Grand Canyon; Canyonlands) for
both spring and summer (Figs. 7a, d and 8a, d), reflecting that
observations at these sites sometimes can be influenced by transport of
photochemically aged plumes from nearby urban areas and from southern
California during late spring and summer. When sampled at the surface, AM3
simulates small summertime O3 decreases in the 95th and 50th percentiles
over the Intermountain West (Fig. 4b, d), consistent with observations at
Yosemite, Grand Canyon, and Canyonlands (Fig. 8a, b). As illustrated in Fig. 3
for spring and discussed in Sect. 2.4, individual sites in the west display
observed trends falling in between the filtered model and those sampled at
the surface versus aloft.
Linear trends in the 95th (left) and 50th (right) percentile
springtime MDA8 O3 over 1988–2014 at US rural sites from BASE (top),
Background (middle) and FIXEMIS simulations (bottom). Larger circles indicate
sites with statistically significant trends (p < 0.05). Top panels are
repeated from Fig. 7d, e. Note that the 95th (50th) percentile is sampled
separately from the Background and FIXEMIS simulations without depending on
the times when the BASE simulation is experiencing the 95th (50th)
percentile days.
As in Fig. 10, but for summer. Top panels are repeated from Fig. 8d, e.
We examine how US surface O3 responds to changes in regional
anthropogenic emissions, hemispheric background, and meteorology by comparing
O3 trends in the BASE, Background, and FIXEMIS experiments
(Figs. 10–11). With North American anthropogenic emissions shut off in the
Background simulation, little difference is discernable from the BASE
simulation for WUS O3 trends during spring (first versus second rows in
Fig. 10), indicating the key role of hemispheric background driving increases
in springtime O3 over the WUS. With anthropogenic emissions held
constant in time, FIXEMIS still shows statistically significant spring
O3 increases in the 95th percentile (Fig. 10c), approximately half of
the trends simulated in BASE, for Grand Canyon, Canyonlands, Mesa Verde and
Rocky Mountain national parks. Prior work shows that deep stratospheric
intrusions contribute to the highest observed and simulated surface O3
events at these sites (Langford et al., 2009; Lin et al., 2012a). Strong
year-to-year variability of such intrusion events (Lin et al., 2015a) can
confound the attribution of springtime O3 changes over the WUS to
anthropogenic emission trends, particularly in the highest percentile and
over a short record length. Summer avoids this confounding influence when
stratospheric intrusions are at their seasonal minimum, as evidenced by
little O3 change in FIXEMIS over the WUS (Fig. 11c, f). In contrast to
spring, the model shows larger differences in WUS O3 trends between BASE
and Background for summer when North American pollution peaks seasonally
(Fig. 10a, d versus b, e compared to Fig. 11a, d versus b, e). There are
significant increases of 0.2–0.5 ppb yr-1 in the 95th and 50th
percentile summer background O3 at more than 50 % of the western
sites (Fig. 11b, e), offsetting the O3 decreases resulting from US
NOx reductions and leading to little overall change in total observed
and simulated O3 at WUS rural sites during summer (Fig. 8).
The 1990–2012 trends in (a) model JJA total biogenic
isoprene emissions, (b) model 90th percentile JJA daily maximum
temperature, (c) the warmest daily maximum temperature and
(d) the frequency of warm days (i.e., those above the 90th
percentile for the base period 1961–1990) for August obtained from the
GHCNDEX dataset (Donat et al., 2013; available at
http://www.climdex.org/view download.html). Stippling denotes areas
where the change is statistically significant (p < 0.05). Note
that the trends are calculated for the 1990–2012 period, instead of
1988–2014, to avoid the influence from hot extremes in 1988 and cold
conditions in 2014 (Sect. 6). When these years are included, the trends
in (c) and (d) are swamped by the anomalies. The trends
in (a) and (b) are similar between 1990–2012 and
1988–2014.
Over the EUS, AM3 also simulates background O3 increases, occurring in
both the 95th and 50th percentiles, with a rate of 0.1–0.3 ppb yr-1
during spring (Fig. 10b, e) and 0.2–0.5 ppb yr-1 during summer
(Fig. 11b, e). Based on prior model estimates that springtime background
O3 is greater in the northeast than the southeast (Lin et al., 2012a, b;
Fiore et al., 2014), one might assume that the springtime O3 increases
in the 5th percentile observed over the northeast (Fig. 7c) have been
influenced by a rising background. However, AM3 simulates homogeneous
background O3 trends across the entire EUS (Fig. 10b, e), indicating
that the observed north-to-south gradient in O3 trends reflects an
earlier seasonal onset of NOx-sensitive photochemistry in the southeast,
as opposed to the background influence.
A warming climate is most likely to worsen the highest O3 events in
polluted regions (e.g., Schnell et al., 2016; Shen et al., 2016). With
anthropogenic emissions held constant in time over 1988–2014, FIXEMIS
suggests significant increases of 0.2–0.4 ppb yr-1 in the 95th
percentile summertime O3 over the EUS (Fig. 11c). Using self-organizing
map cluster analysis, Horton et al. (2015) identified robust increases in the
occurrence of summer anticyclonic circulations over eastern North America
since 1990. We find that biogenic isoprene emissions over this period
increased significantly by 1–2 % yr-1 (10 to
20 mg C m-2 summer-1) throughout the EUS in the model,
consistent with simulated increases in the 90th percentile JJA daily maximum
temperature (Fig. 12a–b). Increases in isoprene emissions contribute to
raising EUS background O3 in summer (Fig. 11b, e). Using the Global
Land-Based Datasets for Monitoring Climate Extremes (GHCNDEX; Donat et al.,
2013), we find increases in the number of warm days above the 90th percentile
and maximum temperature over the southeastern US in August (Fig. 12c–d). The
trends in temperature extremes are similar between June and August, but there
is no significant trend in July (not shown). While changes in regional
temperature extremes on 20- to 30-year time series may reflect internal
climate variability (Shepherd, 2015), we suggest that increasing hot extremes
and biogenic isoprene emissions over the last 2 decades may have offset some
of the benefits of regional NOx reductions in the EUS.
Impacts of rising Asian emissions, methane and wildfires on western US
O3Historical western US O3 trends in spring
(a) Time series of median spring MDA8 O3 anomalies
(relative to the 1995–2014 mean) at Great Basin, Rocky Mountain, and US Air
Force Academy as observed (black) and simulated in AM3 BASE filtered for
baseline conditions (red; see Sect. 2.4) and in Background with North
American anthropogenic emissions zeroed out (NAB; green). Presented at the
top of the graph are statistics from the linear fit and correlations between
observations and simulations. Numbers at the bottom of the graph denote the
sample size of observations for each year. Gray dots indicate uncertain
observations that are removed from the linear fit (see
Sect. 2.3).(b) Same as Fig. 13a, but for Yellowstone, Pinedale, and
Mesa Verde over the period 1988–2012.
Further indications of the factors driving baseline O3 changes over the
WUS can be inferred by examining the time series at several high-elevation
sites, which most frequently sample baseline O3 in the free troposphere
during spring (Sect. 2.4). Figure 13 shows the results, both observed and
simulated, for six such monitoring sites: Great Basin National Park in Nevada
(2.1 km a.s.l.), Rocky Mountain National Park (2.7 km a.s.l.) in
Colorado, US Air Force Academy (1.9 km a.s.l.) in Colorado Springs,
Yellowstone National Park (2.4 km a.s.l.) and Pinedale (2.4 km a.s.l.) in
Wyoming, and Mesa Verde National Park (2.2 km a.s.l.) in the Colorado–New
Mexico–Arizona–Utah four-corner region. The observed median values of
springtime MDA8 O3 have increased significantly at a rate of
0.2–0.5 ppb yr-1 over the past 20–27 years at these sites, except
Pinedale, where the increase in background O3 is likely offset by the
O3 decrease due to recent emission control for the large oil and gas
production fields in this area
(http://deq.wyoming.gov/aqd/winter-ozone/resources/technical-documents/).
When filtered to remove the influence from fresh local pollution (Sect. 2.4),
AM3 BASE captures the long-term trends of O3 observed at these sites.
Correlating AM3 Background with observed O3 indicates that most of the
observed variability reflects changes in the background, with fluctuations in
stratospheric influence contributing to anomalies on interannual timescales
(e.g., the 1999 anomaly, Lin et al., 2015a), whereas Asian influence
dominates the decadal trends as discussed below. The O3 reduction
resulting from US anthropogenic emission controls is less than
0.1 ppb yr-1 (BASE minus Background) at these baseline sites. We show
model results for the entire 1980–2014 period for Great Basin, Rocky
Mountain, and the US Air Force Academy to provide context for observed trends
in the 2 most recent decades (Fig. 13a). In the 1980s when Chinese NOx
emissions (∼ 4 Tg yr-1 NO) were much lower than US NOx emissions
(∼ 15 Tg yr-1 NO) (Granier et al., 2011), there was little
overall O3 change over the WUS in the model. From the mid-1990s onwards,
with NOx emissions in China rising steeply (Fig. 1a) and surpassing US
emissions in the 2000s, the O3 trends at remote WUS sites appear to be
dominated by trends of background, reflecting rising emissions outside the
US. The largest spring O3 increases from 1981–1990 to 2003–2012 at
700 hPa extend from Southeast Asia to the subtropical North Pacific Ocean to
the southwestern US (Fig. S7a), consistent with the influence of rising Asian
precursor emissions.
Summary of springtime median MDA8 O3 trends (in ppb yr-1)
over 1988–2012 at WUS sites from observations and AM3 simulations. Trends
with the 95 % confidence intervals and levels of significance
(bold: < 1 %; italic: 1–5 %; plain: ≥ 5 %) were estimated by the
two-tailed t test.
The * mask indicates data filtered to represent baseline
conditions (NACOt ≤ 67th). The EA subscript indicates that data were
filtered to represent transport conditions favoring the import of Asian
pollution (EACOt ≥ 67th).
Table 2 contains a summary of the drivers of O3 trends in the model at
seven CASTNet sites that exhibit a significant spring O3 increase
observed over 1988–2012. Here we focus our attribution analysis on the period
1988–2012 (instead of 1988–2014) because the IAVASIA and IAVCH4
simulations only extend to 2012. Meteorology varies from year to year in all
experiments. Thus, we quantify the contributions of rising Asian emissions in
IAVASIA, global methane in IAVCH4, and wildfire emissions in IAVFIRE by
subtracting out the slope of the linear regression of seasonal O3 means
in FIXEMIS. Simulated O3 with anthropogenic emissions varying in both
South and East Asia but held constant elsewhere shows statistically
significant increases of 0.1–0.2 ppb yr-1 (p≤ 0.01; IAVASIA minus
FIXEMIS in Table 2), consistent with trends of 0.2 ppb yr-1 estimated
by scaling results from HTAP phase 1 multi-model sensitivity experiments with
Asian emissions reduced by 20 % (Reidmiller et al., 2009). This Asian
influence can explain 50–65 % of the modeled background O3 increase
in spring (Table 2).
With only methane varying, the model trends are less than 0.1 ppb yr-1
(IAVCH4 minus FIXEMIS), accounting for an average of 15 % of the
background increase. The contribution from wildfire emissions during spring
is of minor importance (IAVFIRE minus FIXEMIS, Table 2). A stratospheric
O3 tracer (O3Strat) in AM3 (Lin et al., 2012a, 2015a) demonstrates
a positive but insignificant trend in stratospheric O3 transport to the
sites. We examine the trends of lower tropospheric O3 at these sites
when transport conditions favor the import of Asian pollution into western
North America, as diagnosed by the East Asian CO tracer (EACOt) exceeding the
67th percentile for each spring. Similar to the conclusion of Lin et
al. (2015b), we find that the rate of O3 increase in the Background
simulation is greater by 0.05–0.1 ppb yr-1 under strong transport
from Asia than without filtering. Filtering the IAVASIA simulation for Asian
influence also results in greater O3 increases than filtering for
baseline conditions (Table 2).
Rising Asian emissions even influence trends of O3 downwind of the Los
Angeles Basin during spring. O3 measured in Joshua Tree National Park
shows an increase of 0.31 ± 0.25 ppb yr-1 in spring over
1990–2010 (Cooper et al., 2012), despite significant improvements in O3
air quality in the Los Angeles Basin (Warneke et al., 2012). The O3
record extended to 2014 shows a decline in the 95th percentile O3 in
Joshua Tree National Park for both spring and summer (Figs. 7–8), whereas
the 5th percentile continues to increase in spring and there is no
significant trend in the median. Sampling the AM3 Background simulation at
this site indicates a rising background (0.31 ± 0.14 ppb yr-1).
Aircraft measurements in May–June 2010 indicate the presence of Asian
pollution layers 2 km above southern California with distinct sulfate
enhancements coincident with low organic mass (Lin et al., 2012b), supporting
the conclusion that rising Asian emissions can contribute to trends of
O3 observed in this region. Yosemite National Park (1.6 km a.s.l.) and
Chiricahua National Monument (1.5 km a.s.l.) are also influenced by
increases in Asian emissions and concurrent decreases in local pollution in
California. O3 observed at Yosemite shows an increase from 1995 to
around 2012 (0.37 ± 0.32 ppb yr-1; Fig. S8), which the model
attributes primarily to rising Asian emissions (Table 2), but observations
have remained constant since then, reflecting an offset by O3 decreases
in California (Fig. 4).
Projecting western US springtime O3 for the 21st
Century
Future projections. Time series of median springtime O3 changes
relative to 2010 in GFDL AM3 hindcast (orange circles) and CM3 future
simulations for RCP8.5 (red) versus RCP4.5 (blue; shading represents the
range of three ensemble members), sampled at 700 hPa over the WUS
(35–45∘ N, 120–105∘ W). Black circles indicate observed
changes averaged from the Lassen, Great Basin, and Rocky Mountain national
parks.
Under the RCP8.5 scenario, Chinese NOx emissions are projected to peak
in 2020–2030, reflecting an increase of ∼ 50 % from 2010
(Fig. 1a), followed by a sharp decrease, reaching 1990 levels by 2050. Global
methane increases by ∼ 60 % from 2010 to 2050 under RCP8.5
(Fig. S1). Under the RCP4.5 scenario, in contrast, NOx emissions in
China change little over 2010–2030 and global methane remains almost
constant from 2010 to 2050. NOx emissions in the US decrease through
2050 under both scenarios, by ∼ 40 % from 2010. A number of studies
have examined future US O3 changes under the RCPs (e.g., Gao et al.,
2013; Clifton et al., 2014; Pfister et al., 2014; Fiore et al., 2015; Barnes
et al., 2016). However, as discussed earlier, the trends of O3 in the
model when sampled near the surface are overwhelmingly dominated by US
anthropogenic emission trends. Thus, the future O3 changes estimated by
these prior studies do not represent baseline conditions, particularly the
response to rising Asian emissions. In Fig. 14 we show changes in WUS free
tropospheric (700 hPa) O3 relative to 2010 in the CM3 future
simulations under RCP8.5 versus RCP4.5. Historical hindcasts and observations
are also shown for context. Under RCP4.5, springtime O3 over the WUS
shows little overall change over 2010–2050. Under RCP8.5, in contrast,
springtime WUS O3 increases by ∼ 10 ppb from 2010 to 2030 and
remains almost constant from 2030 to 2050, consistent with the projected
trends in Asian emissions and global methane.
Trends and variability of western US O3 in summer
Summertime O3 in Yellowstone National Park. (a) Median
JJA MDA8 O3 trends over 1988–2012 at Yellowstone from observations
(black) and simulations sampled at 700 hPa for BASE without filtering
(pink), BASE filtered for baseline conditions (hatched pink), IAVASIA (solid
purple, baseline), IAVASIA filtered for Asian influence (EACOt ≥ 67th,
hatched purple), IAVCH4 (cyan), IAVFIRE (orange) and FIXEMIS (red).
(b) Time series of anomalies in August median MDA8 O3 at
Yellowstone as observed (black) and simulated by the model sampled at the
surface, with constant (red) and time-varying wildfire emissions (orange).
Trends over 1988–2014 are reported. (c) Interannual correlations of
JJA mean MDA8 O3 observed at Yellowstone with JJA mean daily maximum
temperature from observations (Harris et al., 2014).
Yellowstone National Park is the only site with statistically significant
summer O3 increases observed across all percentiles (Fig. 8a–c). The
1988–2012 trends for the median observed and simulated O3 are
summarized in Fig. 15a. Observations show an increase of
0.32 ± 0.18 ppb yr-1 for JJA, with a greater rate of increase in
June (0.38 ± 0.25 ppb yr-1) than in July–August
(0.26 ± 0.18 ppb yr-1). AM3 BASE sampled at 700 hPa and
filtered for baseline conditions (hatched pink bar in Fig. 15a) captures the
observed increase. Without baseline filtering (solid pink bar), North
American emission reductions offset almost 50 % of the simulated O3
increase at Yellowstone, causing the model to underestimate the observed
O3 trend. The model attributes much of the observed summer O3
increase at Yellowstone to rising Asian emissions, with IAVASIA simulating an
O3 increase of 0.31 ± 0.19 ppb yr-1 under baseline
conditions, increasing to 0.42 ± 0.23 ppb yr-1 under conditions
of Asian influence (EACOt ≥ 67th percentile). The stronger increase
measured in June than in July–August is consistent with the influence of the
Asian summer monsoon producing a surface O3 minimum in July–August in
East Asia (e.g., Lin et al., 2009), as well as the seasonality of
intercontinental pollution transport. Changes in methane, wildfires, and
meteorology over this period are of minor importance for the JJA O3
trends at Yellowstone.
Enhanced wildfire activity in hot and dry weather is thought to be a key
driver of interannual variability of surface O3 in the Intermountain
West in summer (Jaffe et al., 2008; Jaffe, 2011). However, hot and dry
conditions also facilitate the buildup of O3 produced from regional
anthropogenic emissions, which can complicate the unambiguous attribution of
observed O3 enhancements. Using August data at Yellowstone as an
example, we isolate the relative contribution of these two processes to
observed O3 with the IAVFIRE versus FIXEMIS experiments (Fig. 15b). Here
we sample AM3 at the surface to account for any influence of varying boundary
layer mixing depths. Even without interannual variations of wildfire
emissions, FIXEMIS captures much of the observed year-to-year variability of
August mean O3 at Yellowstone (r= 0.67). IAVFIRE with interannually
varying fire emissions only moderately improves the correlations (r= 0.75).
FIXEMIS also captures the observed O3 increase from the early
1990s to around 2002, likely reflecting warmer temperatures and deeper mixing
depths allowing more baseline O3 to mix down to the surface. Over the
entire 1988–2014 (or 1980–2014) period, IAVFIRE gives ∼ 0.1 ppb yr-1 greater O3 increases in August than FIXEMIS,
consistent with an overall increase in boreal wildfire activity (Figs. S2 and
S7b).
Surface MDA8 O3 enhancements from wildfire emissions for
individual months in the years with large biomass burning in boreal regions (1998,
2002, 2003) and over the WUS (2008, 2011, 2012), as diagnosed by the
differences between IAVFIRE and FIXEMIS. The black circle denotes the
location of Yellowstone National Park.
Figure 16 shows year-to-year variability in surface MDA8 O3 enhancements
from wildfires during summer, as diagnosed by the differences between IAVFIRE
and FIXEMIS. The results are shown for individual months since fires are
highly episodic. During the summers of 1998, 2002, and 2003, biomass fires
burned a large area of Siberia and parts of the North American boreal
forests, raising carbon monoxide across the Northern Hemisphere as detected
from space (Yurganov et al., 2005; van der Werf et al., 2010). Long-range
transport of Siberian fire plumes resulted in 2–6 ppb enhancements in
surface MDA8 O3 at the US western coast and in parts of the
Intermountain West in AM3. The model calculates enhancements in monthly mean
MDA8 O3 of up to 8 ppb from the intense wildfire events in northern
California during July 2008 (Huang et al., 2013; Pfister et al., 2013), over
Texas–Mexico during June 2011 (Wang et al., 2015), and in Wyoming–Utah
during August 2012 (Jaffe et al., 2013). The AM3 estimates are roughly
consistent with a previous analysis of boundary layer aircraft data with and
without fire influences (as diagnosed by CH3CN) during June 2008 over
California (Pfister et al., 2013).
While fires during hot and dry summers clearly result in enhanced O3 at
individual sites for some summers, the ability of AM3 with constant fire
emissions to simulate variability of O3 for a high (e.g., 1988, 2002,
2006) versus low (e.g., 1997, 2009) fire year (Fig. 15b) indicates that
biomass burning is not the primary driver of observed O3 interannual
variability. Year-to-year variability of JJA mean MDA8 O3 observed at
Yellowstone is strongly correlated (r > 0.6) with observed
large-scale variations in JJA mean daily maximum temperature across the
Intermountain West (Fig. 15c). Correlations for other ground stations show a
similar large-scale feature. Similar to the conclusion from Zhang et
al. (2014), our analysis indicates that the correlation between O3 and
biomass burning reported by Jaffe et al. (2008) and Jaffe (2011) at rural
sites reflects common underlying correlations with temperature rather than a
causal relationship of fire
with O3.
At remote mountain sites (e.g., Yellowstone), warmer surface temperatures
lead to deeper mixed layers that facilitate mixing of free tropospheric
O3-rich air down to the surface. At sites near sources of air pollution,
hot conditions enhance regional O3 production and orographic lifting of
urban pollution to mountaintop sites during daytime, as occurs at Rocky
Mountain National Park located downwind of the Denver metropolitan area
during summer (Sect. 5.4). Reactive volatile organic compound (VOC) emissions
from fires may enhance O3 production in NOx-rich urban areas (Baker
et al., 2016), although evaluating these impacts needs high-resolution models
and better treatment of sub-grid-scale fire plumes.
Ozone trends in the Denver metropolitan area
Surface O3 trends in Denver. (a) Comparison of
observed trends in annual fourth highest MDA8 O3 at Crestline Los
Angeles (brown) and in Denver (blue, computed from all monitors available in
Denver non-attainment counties). (b) Time series of observed median
MAM MDA8 O3 at Great Basin National Park (red), in comparison with three
monitors in Denver. (c) Time series of observed 95th percentile
July–August MDA8 O3 in Denver, together with statistics (25th, 50th,
75th, 95th) of observed July–August daily maximum temperature at Rocky Flats
(red, right axis).
Efforts to improve air quality have led to a marked decrease in high-O3
events in the Los Angeles Basin as illustrated by the annual 4th highest MDA8
O3 at Crestline – a regionally representative monitor operated
continuously from 1980 to the present (Fig. 17a). In striking contrast, the
4th highest MDA8 O3 in the Denver metropolitan area shows little change
over the past decades, despite significant reductions in NOx (Fig. 1)
and CO emissions (-80 % from 1990 to 2010; Cooper et al., 2012). Recent
field measurements indicate that increased VOC emissions from oil and natural
gas operations are an important source of O3 precursors in the
Denver–Julesberg Basin (Gilman et al., 2013; Halliday et al., 2016; McDuffie
et al., 2016). However, total VOC emissions in Denver may not be increasing
over time due to the marked reductions in VOC emissions from vehicles (Bishop
and Stedman, 2008, 2015). We seek insights into the causes of the lack of
significant O3 responses to emission controls in Denver by separately
analyzing trends in spring and summer (Fig. 17b–c).
The ∼ 200 × 200 km2 AM3 model is not expected to resolve
the urban-to-rural differences between Rocky Mountain National Park and the
Denver metropolitan area. However, if observed O3 variability in Denver
correlates with that at remote sites in the Intermountain West, then model
attribution for the remote sites can be used to infer sources of observed
O3 in Denver. This is demonstrated in Fig. 17b for spring using data at
three representative sites in Denver, Rocky Flats North, National Renewable
Energy Lab (NREL), and Welby, with continuous measurements since the early
1990s. Year-to-year variability of median MDA8 O3 at these sites during
spring correlates strongly with that in Great Basin National Park
(r= 0.7), a fairly remote site in Nevada not influenced by urban emissions
from Denver. Median spring O3 observations in Denver increased
significantly by ∼ 0.3 ppb yr-1, similar to the rate of increase
in Great Basin National Park, which the model attributes to rising background
(Fig. 13a), implying that the tripling of Asian emissions since 1990 also
raised mean springtime O3 in the Denver metropolitan area. Trends in the
95th percentile are statistically insignificant.
During summer, changes in regional emissions and temperature have the
greatest impacts on the highest observed O3 concentrations in polluted
environments. Figure 17c shows times series of July–August 95th percentile
MDA8 O3 in Denver, together with the distribution of daily maximum
temperature. In every year since 1993, the highest summer MDA8 O3
observed at these sites exceeds the 70 ppb NAAQS level. There is a small
negative trend that is swamped by large interannual variability. The summers
with the highest observed O3 coincide with those with the highest
observed temperatures, such as 1998, 2003, 2007, 2011 and 2012. During these
summers, enhancements of MDA8 O3 were also recorded in Rocky Mountain
National Park, reflecting enhanced lifting of pollution from Denver under
warmer conditions (Brodin et al., 2010). Applying quantile regression (e.g.,
Porter et al., 2015) to daily
observations at Rocky Flats North over 1993–2015, we find a
2 ppb ∘C-1 sensitivity of 95th percentile July–August O3
to changes in maximum daily temperature. We suggest that the substantial
increases in extreme heat occurrence over central North America over the last
2 decades, as found by Horton et al. (2015), contribute to raising summer
O3 in Denver, which offsets O3 reductions that otherwise would have
occurred due to emission controls in Denver. Potential shifts in the O3
photochemistry regime can also contribute to trends of summer O3 in
Denver, although advancing this knowledge would require a high-resolution air
quality model.
Impacts of heat waves and droughts on eastern US summer O3
(a) Time series of July mean MDA8 O3 anomalies (relative to
1988–2014) at the Pennsylvania State University (PSU) CASTNET site as
observed (black) and simulated by the GFDL-AM3 model with time-varying
(purple) and constant anthropogenic emissions (red), along with observed
anomalies in July mean daily max temperature (gray lines; right axis). The
green triangle denotes the 1988 O3 anomaly from a sensitivity simulation
using BASE emissions but with 35 % decreases in Vd,O3 (IAVDEP).
(b) Time series of daily MDA8 O3 at PSU from 1 June to 16 July in 1988 from
observations (black), BASE (purple), and IAVDEP simulations (green).
We discuss in this section interannual variability and long-term changes in
summer O3 over the EUS, where air stagnation and high temperatures
typically yield the highest O3 observed in surface air (e.g., Jacob and
Winner, 2009). Evaluating the ability of models to simulate the high-O3
anomalies during historical heat waves and droughts is crucial to
establishing confidence in the model projection of pollution extremes under a
warming climate. Figure 18a shows comparisons of July mean MDA8 O3 at
one regionally representative site, the Pennsylvania State University (PSU)
CASTNet site, from observations and model simulations. With time-varying
emissions, the BASE model simulates an O3 decrease
(-0.45 ± 0.32 ppb yr-1) consistent with observations
(-0.67 ± 0.33 ppb yr-1) and captures the observed July mean
O3 interannual variability (r= 0.82) that is correlated with
large-scale variations in daily maximum temperature (r= 0.57). In
particular, O3 pollution extremes are successfully simulated during the
EUS summer heat waves of 1988, 1995, 1999, 2002, 2011 and 2012 (Leibensperger
et al., 2008; Fiore et al., 2015; Jia et al., 2016). Year-to-year variations
in meteorology can explain 30 % of the total observed O3 variability
(r= 0.55), as inferred by FIXEMIS with constant anthropogenic emissions.
If US anthropogenic emissions remained at 1990s levels (as in FIXEMIS), then
anomalies in July mean MDA8 O3 would have been 10 ppb greater during
the 2011 and 2012 heat waves. Loughner et al. (2014) found that half of the
days in July 2011 would have been classified as O3 exceedance days for
much of the mid-Atlantic region if emissions had not declined.
(a) Comparisons of probability distributions of summertime
MDA8 O3 from 40 EUS CASTNet sites for the pre-NOx SIP Call
(1988–2002; solid black) versus post-NOx SIP Call (2003–2014; dashed gray)
periods and during the extreme heat waves of 1988 (solid purple) versus 2012
(dashed brown). The median (μ) and standard deviation (σ) are
shown (ppb). (b) Same as (a), but from AM3 BASE. Also shown
is the O3 distribution in 1988 from a sensitivity simulation with
35 % decreases in Vd,O3 in drought areas (green).
(c) Standardized soil moisture departures for JJA 1988 (calculated
by dividing anomalies by the 1979–2010 climatological standard deviation,
using data from the NOAA Climate Prediction Center).
Figure 19a compares the probability density functions of MDA8 O3 at 40
EUS surface sites for JJA in the pre-NOx SIP Call (1988–2002) versus
post-NOx SIP Call (2003–2014) periods and during the extreme heat waves
of 1988 versus 2012. Following the NOx SIP Call, the probability
distribution of observed JJA MDA8 O3 over the EUS shifted downward
(solid black versus dotted gray lines in Fig. 19a). The median value declined
by 9 ppb and the largest decreases occurred in the upper tails, leading to
weaker day-to-day O3 variability and a narrower O3 range (standard
deviation σ decreased from 16.4 to 12.9 ppb). These observed O3
changes driven by regional NOx reductions are even more prominent when
comparing the heat waves of 1988 versus 2012 (solid purple versus dotted
brown lines in Fig. 19a): σ= 22.3 versus 13.4 ppb and median value
μ= 68.6 versus 52.2 ppb.
Figure 19b shows the corresponding comparisons using the results from AM3
BASE. Despite the high mean model bias (∼ 20 ppb), AM3 captures the
overall structure of the changes in the surface O3 distributions and
thus the response of surface O3 to the NOx SIP Call, including the
reductions of high-O3 events during the heat wave of 2012 compared to
1988. Nevertheless, there is a noticeable difference between the observations
and simulations in the shape of MDA8 O3 probability distributions for
summer 1988, particularly in the upper tail of the distribution above
110 ppb (purple lines in Fig. 19a versus b). The BASE model also
underestimates the observed July mean O3 anomaly at PSU in 1988 by
∼ 10 ppb (purple versus black dots in Fig. 18a). One possible
explanation for these biases is that drought stress can effectively reduce
the O3 deposition sink to vegetation, leading to an increase in surface
O3 concentrations as found during the 2003 European heat wave (Solberg
et al., 2008), whereas AM3 does not include interannually varying dry
deposition velocities.
The North American drought of 1988 ranks among the worst episodes of drought
in the US (e.g., Seager and Hoerling, 2014), with JJA soil moisture deficits
occurring over the northern Great Plains–Midwest region with magnitudes of
1–2.5 mm standardized departures from the 1979–2010 climatology
(Fig. 19c). Huang et al. (2016) found that monthly mean O3 dry
deposition velocities (Vd,O3) for forests decreased by 33 %
over Texas during the dry summer of 2011. Based on this estimate, we conduct
a sensitivity simulation for 1988 using BASE emissions but decreasing monthly
mean Vd,O3 from May to August by 35 % in the areas over
North America (20–60∘ N) where soil moisture deficits in 1988
exceed -1.0σ mm (Fig. 19c). This experiment (hereafter referred to
as IAVDEP) simulates ∼ 10 ppb higher July mean MDA8 O3 at the PSU
CASTNet site than the BASE model and matches the observed O3 anomaly in
1988 relative to the record mean (green symbol in Fig. 18a). The impact is
largest (up to 15 ppb) on days when observed MDA8 O3 exceeds 100 ppb
(Fig. 18b; Tmax≥ 30 ∘C). Simulated JJA MDA8 O3
at EUS sites in IAVDEP shows an upward shift in the probability distribution,
particularly in the upper tail above 110 ppb (green versus purple lines in
Fig. 19b), bringing it closer to observations in 1988 (Fig. 19a). The O3
standard deviation in IAVDEP (σ= 18 ppb) shifts towards that in
observations (σ= 22 ppb) relative to the BASE model (σ= 16 ppb).
Summary of US surface O3 trends and drivers. Changes in decadal
mean MDA8 O3 from 1981–1990 to 2003–2012 simulated in a suite of
GFDL-AM3 experiments for spring and summer for the western
(32–46∘ N and 123–102∘ W), northeastern
(37–45∘ N and 90–65∘ W) and southeastern
(30–36∘ N and 95–77∘ W) US domains. Observations are not
shown because limited data are available during 1981–1990. Experiments are
color-coded, with the error bars indicating the range of the mean change at
the 95 % confidence level. Filled circles represent the changes under
Background (green) and IAVASIA (purple) when filtered for Asian influence
(EACOt ≥ 67th), while other results are from the unfiltered models.
The text near the bottom of the plot provides the change in NOx
emissions over the same period for each region.
Quantile mapping can be applied to correct systematic distributional biases
in surface O3 compared to observations (Rieder et al., 2015), but this
approach has limitations if there are structural biases in the O3
distribution due to missing physical processes in the model (e.g., variations
of Vd,O3 with droughts). Travis et al. (2016) suggest that the
National Emission Inventory (NEI) for NOx from the US EPA is too high
nationally by 50 %. Decreasing US NOx emissions by this amount
corrects their model bias for boundary layer O3 by 12 ppb in the
southeast for summer 2013, while surface MDA8 O3 in their model is still
biased high by 6 ± 14 ppb, which the authors attribute to excessive
boundary layer mixing. US NOx emissions in the emission inventory used
in AM3 (Sect. 2.2) are approximately 15 % lower than those from the NEI.
The 35 % decrease in NOx emissions from the pre-NOx SIP Call to
the post-NOx SIP Call in the model reduces mean O3 by 8 ppb in the
EUS, implying that the NOx emission bias could correct 40 % of our
model mean bias of ∼ 20 ppb. These estimates support the idea that the
common model biases in simulating surface O3 over the southeastern US
(e.g., Fiore et al., 2009) may partly reflect excessive NOx emissions.
Some of the positive O3 biases could be also due to the averaging over a
deep vertical box in the model surface layer (∼ 60 m in AM3) that can
not resolve near-surface gradients (Travis et al., 2016).
Conclusions and recommendations
Through an observational and modeling analysis of interannual variability and
long-term trends in sources of O3 over the past 35 years, we have
identified the key drivers of O3 pollution over the US. We initially
evaluated the trends of O3 in Asia resulting from rising Asian precursor
emissions (Figs. 4–6). Our synthesis of available observations and
simulations indicates that surface O3 over East Asia has
increased by 1–2 ppb yr-1 since 1990 (i.e., 25–50 ppb over
25 years), with significant implications for regional air quality and global
tropospheric O3 burden. Shifting next to the US, we find
0.2–0.5 ppb yr-1 increases in median springtime MDA8 O3 measured
at 50 % of 16 WUS rural sites, with 25 % of the sites showing
increases across the entire O3 concentration distribution, despite
stringent US domestic emission controls (Fig. 7). While many prior studies
show that global models have difficulty simulating O3 increases observed
at rural baseline sites (e.g., Parrish et al., 2014; Strode et al., 2015), we
reconcile observed and simulated O3 trends in GFDL-AM3 with a novel
baseline sampling approach (Figs. 3 and 13). We suggest that the common
model–observation disagreement in baseline O3 trends reflects
limitations of coarse-resolution global models in resolving observed baseline
conditions. This representativeness problem can be addressed by filtering
model O3 for hemispheric-scale baseline conditions using the
easy-to-implement, low-cost regional CO-like tracers. This approach allows
trends of O3 measured at baseline sites to be compared directly with
multi-decadal global model hindcasts, such as those being conducted for the
Chemistry-Climate Model Initiative (CCMI; Morgenstern et al., 2017).
The ability of the GFDL-AM3 model to reproduce observed US surface O3
trends lends confidence in its application to attribute these observed trends
to specific processes (Figs. 7 to 11). We summarize the overall statistics in
Fig. 20, drawing upon the decadal mean O3 changes from 1981–1990 to
2003–2012 in the BASE and sensitivity simulations. The changes in BASE are
over the WUS 4.3 ± 1.8 ppb for spring and 1.6 ± 1.2 ppb for
summer; over the northeast, -1.8 ± 1.7 ppb for spring and
-6.0 ± 2.0 ppb for summer; and over the southeast,
-3.9 ± 1.4 ppb for spring and -7.5 ± 1.6 ppb for summer.
Increasing O3 in the WUS under BASE coincides with an increase in
background O3 by 6.3 ± 1.9 ppb for spring and
4.2 ± 2.0 ppb for summer. Under conditions of strong transport from
Asia (East Asian COt ≥ 67th), the background trend rose to
7.6 ± 2.2 ppb for spring and 6.0 ± 2.1 ppb for summer (green
dots in Fig. 20). The WUS background O3 increase reflects contributions
from increases in Asian anthropogenic emissions (accounting for 50 % of
background increase in spring; 52 % in summer), rising global methane
(13 % in spring; 23 % in summer), and variability in biomass burning
(6 % in spring; 12 % in summer; excluding the meteorological
influence).
We conclude that the increase in Asian anthropogenic emissions is the major
driver of rising background O3 over the WUS for both spring and summer
in the past decades, with a lesser contribution from methane increases over
this period. The tripling of Asian NOx emissions since 1990 contributes
up to 65 % of modeled springtime background O3 increases
(0.3–0.5 ppb yr-1) over the WUS, outpacing O3 decreases resulting
from 50 % US NOx emission controls (≤ 0.1 ppb yr-1;
Table 2 and Fig. 10). Springtime O3 observed in the Denver metropolitan
area has increased at a rate similar to remote rural sites (Fig. 17b). Mean
springtime O3 above the WUS is projected to increase by ∼ 10 ppb
from 2010 to 2030 under the RCP8.5 global change scenario but to remain
constant throughout 2010 to 2050 under the RCP4.5 scenario (Fig. 14). As
NOx emissions in China continue to decline in response to efforts to
improve air quality (Krotkov et al., 2016; Liu et al., 2016), rising global
methane and NOx emissions in the tropical countries (e.g., India) in
Asia, where O3 production is more efficient, may become more important
in the coming decades. A global perspective is necessary when designing a
strategy to meet US O3 air quality objectives.
During summer, a tripling of Asian anthropogenic emissions from 1988 to 2014
approximately offsets the benefits of 50 % reductions in US domestic
emissions, leading to weak or insignificant O3 trends observed at most
WUS rural sites (Figs. 8 and 11). Rising Asian emissions contribute to
observed summertime O3 increases (0.3 ppb yr-1) at Yellowstone
National Park. Our findings confirm the earliest projection of Jacob et
al. (1999) with a tripling of Asian emissions. While wildfire emissions can
result in 2–8 ppb enhancements to monthly mean O3 at individual sites
in some summers, they are not the primary driver of observed O3
interannual variability over the Intermountain West (Figs. 15 and 16).
Instead, boundary layer depth, high temperatures and the associated buildup
of O3 produced from regional anthropogenic emissions contribute most to
the observed interannual variability of O3 in summer. Summertime O3
measured in Denver during pollution episodes frequently exceeds the 70 ppb
NAAQS level, with little overall trend despite stringent precursor emission
controls (Fig. 17c), likely due to the effects of more frequent occurrences
of hot extremes in the last decade.
In the eastern US, if emissions had not declined, the 95th percentile
summertime O3 would have increased by 0.2–0.4 ppb yr-1 over
1988–2014 (Fig. 11c), due to more frequent hot summer extremes and increases
in biogenic isoprene emissions (1–2 % yr-1) over this period
(Fig. 12). Regional NOx reductions alleviated the O3 buildup during
the recent heat waves of 2011 and 2012 relative to earlier heat waves (e.g.,
1988, 1995, 1999). GFDL-AM3 captures year-to-year variability in monthly mean
O3 enhancements associated with large-scale variations in temperatures
(Figs. 18 and 19). However, there is a need to improve the model
representation of O3 deposition sink to vegetation, in particular its
reduced efficiency under drought stress, as we demonstrated for the severe
North American drought of 1988. Such land–biosphere couplings are poorly
represented in current models and further work is needed to examine their
impacts on O3 pollution extremes in a warming climate.
Following the NOx SIP Call, surface O3 in the eastern US declined
throughout its probability distribution, with the largest decreases occurring
in the highest percentiles during summer (-0.8 to -1.8 ppb yr-1;
Fig. 8). Spatially, historical O3 decreases during non-summer seasons
were more pronounced in the southeast, where the seasonal onset of biogenic
isoprene emissions and NOx-sensitive O3 production occurs earlier
than in the northeast (Figs. 7, 9 and S4). The 95th percentile O3
concentration in the southeast has even decreased during winter. Despite high
mean-state biases, GFDL-AM3 captures the salient features of observed O3
trends over the eastern US, including wintertime increases in the 5th and
50th percentiles in the northeast, greater springtime decreases in the
southeast than the northeast, and summertime decreases throughout the O3
concentration distribution. These results suggest that NOx emission
controls will continue to provide long-term O3 air quality benefits in
the southeastern US during all seasons.
Data availability
All data derived from observations and model simulations used in this study are archived at
NOAA GFDL and are available to the public upon request to Meiyun Lin.
The Supplement related to this article is available online at doi:10.5194/acp-17-2943-2017-supplement.
The authors declare that they have no conflict of interest.
Acknowledgements
This work was supported by funding from NASA grants NNH13ZDA001N-AURAST and
NNX14AR47G to Meiyun Lin. We thank O. Cooper, S. Fan and J. Schnell for
helpful comments on the manuscript. We acknowledge the free use of ozonesonde
data at Hong Kong available on www.woudc.org and GOME-SCIAMACHY
tropospheric NO2 column data available on www.temis.nl.
Arlene M. Fiore acknowledges support under EPA
Assistance Agreement no. 83587801. The views expressed in this document are
solely those of the authors and do not necessarily reflect those of the
agency. Meiyun Lin devotes this article to her father Tianci Lin, who is the
motivation of her life and research career.
Edited by: B. N. Duncan Reviewed by: two anonymous referees
ReferencesAbatzoglou, J. T. and Williams, A. P.: Impact of anthropogenic climate change
on wildfire across western US forests, P. Natl. Acad. Sci. USA, 113,
11770–11775, 10.1073/pnas.1607171113, 2016.Baker, K. R., Woody, M. C., Tonnesen, G. S., Hutzell, W., Pye, H. O. T.,
Beaver, M. R., Pouliot, G., and Pierce, T.: Contribution of regional-scale
fire events to ozone and PM2.5 air quality estimated by photochemical
modeling approaches, Atmos. Environ., 140, 539–554,
10.1016/j.atmosenv.2016.06.032, 2016.Barnes, E. A., Fiore, A. M., and Horowitz, L. W.: Detection of trends in
surface ozone in the presence of climate variability, J. Geophys. Res., 121,
6112–6129, 10.1002/2015jd024397, 2016.Bishop, G. A. and Stedman, D. H.: A decade of on-road emissions measurements,
Environ. Sci. Technol., 42, 1651–1656, 10.1021/es702413b, 2008.Bishop, G. A. and Stedman, D. H.: Reactive Nitrogen Species Emission Trends
in Three Light-/Medium-Duty United States Fleets, Environ. Sci. Technol., 49,
11234–11240, 10.1021/acs.est.5b02392, 2015.Boersma, K. F., Eskes, H. J., and Brinksma, E. J.: Error analysis for
tropospheric NO2 retrieval from space, J. Geophys. Res., 109, D04311,
10.1029/2003jd003962, 2004.Brodin, M., Helmig, D., and Oltmans, S.: Seasonal ozone behavior along an
elevation gradient in the Colorado Front Range Mountains, Atmos. Environ.,
44, 5305–5315, 10.1016/j.atmosenv.2010.06.033, 2010.Brown-Steiner, B. and Hess, P.: Asian influence on surface ozone in the
United States: A comparison of chemistry, seasonality, and transport
mechanisms, J. Geophys. Res., 116, D17309, 10.1029/2011jd015846, 2011.Brown-Steiner, B., Hess, P. G., and Lin, M. Y.: On the capabilities and
limitations of GCCM simulations of summertime regional air quality: A
diagnostic analysis of ozone and temperature simulations in the US using CESM
CAM-Chem, Atmos. Environ., 101, 134–148, 10.1016/j.atmosenv.2014.11.001,
2015.Carmichael, G. R., Tang, Y., Kurata, G., Uno, I., Streets, D., Woo, J. H.,
Huang, H., Yienger, J., Lefer, B., Shetter, R., Blake, D., Atlas, E., Fried,
A., Apel, E., Eisele, F., Cantrell, C., Avery, M., Barrick, J., Sachse, G.,
Brune, W., Sandholm, S., Kondo, Y., Singh, H., Talbot, R., Bandy, A.,
Thorton, D., Clarke, A., and Heikes, B.: Regional-scale chemical transport
modeling in support of the analysis of observations obtained during the
TRACE-P experiment, J. Geophys. Res., 108, 8823, 10.1029/2002jd003117,
2003.Clifton, O. E., Fiore, A. M., Correa, G., Horowitz, L. W., and Naik, V.:
Twenty-first century reversal of the surface ozone seasonal cycle over the
northeastern United States, Geophys. Res. Lett., 41, 7343–7350,
10.1002/2014gl061378, 2014.Cooper, O. R., Parrish, D. D., Stohl, A., Trainer, M., Nedelec, P., Thouret,
V., Cammas, J. P., Oltmans, S. J., Johnson, B. J., Tarasick, D., Leblanc, T.,
McDermid, I. S., Jaffe, D., Gao, R., Stith, J., Ryerson, T., Aikin, K.,
Campos, T., Weinheimer, A., and Avery, M. A.: Increasing springtime ozone
mixing ratios in the free troposphere over western North America, Nature,
463, 344–348, 10.1038/nature08708, 2010.Cooper, O. R., Gao, R.-S., Tarasick, D., Leblanc, T., and Sweeney, C.:
Long-term ozone trends at rural ozone monitoring sites across the United
States, 1990–2010, J. Geophys. Res., 117, D22307, 10.1029/2012JD018261,
2012.Dennison, P. E., Brewer, S. C., Arnold, J. D., and Moritz, M. A.: Large
wildfire trends in the western United States, 1984-2011, Geophys. Res. Lett.,
41, 2928–2933, 10.1002/2014gl059576, 2014.Dentener, F., Kinne, S., Bond, T., Boucher, O., Cofala, J., Generoso, S.,
Ginoux, P., Gong, S., Hoelzemann, J. J., Ito, A., Marelli, L., Penner, J. E.,
Putaud, J.-P., Textor, C., Schulz, M., van der Werf, G. R., and Wilson, J.:
Emissions of primary aerosol and precursor gases in the years 2000 and 1750
prescribed data-sets for AeroCom, Atmos. Chem. Phys., 6, 4321–4344,
10.5194/acp-6-4321-2006, 2006.Ding, A. J., Wang, T., Thouret, V., Cammas, J.-P., and Nédélec, P.:
Tropospheric ozone climatology over Beijing: analysis of aircraft data from
the MOZAIC program, Atmos. Chem. Phys., 8, 1–13, 10.5194/acp-8-1-2008,
2008.Donat, M., Alexander, L., Yang, H., Durre, I., Vose, R., and Caesar, J.:
Global Land-Based Datasets for Monitoring Climatic Extremes, B. Am. Meteorol.
Soc., 94, 997–1006, 10.1175/Bams-D-12-00109.1, 2013.Donner, L. J., Wyman, B. L., Hemler, R. S., et al.: The Dynamical Core, Physical Parameterizations, and
Basic Simulation Characteristics of the Atmospheric Component AM3 of the GFDL
Global Coupled Model CM3, J. Climate, 24, 3484–3519,
10.1175/2011jcli3955.1, 2011.Duncan, B. N., Lamsal, L. N., Thompson, A. M., Yoshida, Y., Lu, Z. F.,
Streets, D. G., Hurwitz, M. M., and Pickering, K. E.: A space-based,
high-resolution view of notable changes in urban NOx pollution around the
world (2005–2014), J. Geophys. Res., 121, 976–996,
10.1002/2015jd024121, 2016.Emberson, L. D., Kitwiroon, N., Beevers, S., Büker, P., and Cinderby, S.:
Scorched Earth: how will changes in the strength of the vegetation sink to
ozone deposition affect human health and ecosystems?, Atmos. Chem. Phys., 13,
6741–6755, 10.5194/acp-13-6741-2013, 2013.Federal Register: US Environmental Protection Agency, National Ambient Air
Quality Standards for Ozone – Final Rule, Federal Register 80, available at:
http://www.gpo.gov/fdsys/pkg/FR-2015-10-26/pdf/2015-26594.pdf,
65292–65468, 2015.Fiore, A. M., Dentener, F. J., Wild, O., et al.: Multimodel estimates of
intercontinental source-receptor relationships for ozone pollution, J.
Geophys. Res., 114, D04301, 10.1029/2008jd010816, 2009.Fiore, A. M., Oberman, J. T., Lin, M. Y., Zhang, L., Clifton, O. E., Jacob,
D. J., Naik, V., Horowitz, L. W., and Pinto, J. P.: Estimating North American
background ozone in U.S. surface air with two independent global models:
Variability, uncertainties, and recommendations Atmos. Environ., 96,
284–300, 10.1016/j.atmosenv.2014.07.045, 2014.Fiore, A. M., Naik, V., and Leibensperger, E. M.: Air Quality and Climate
Connections, J. Air Waste Manage., 65, 645–685,
10.1080/10962247.2015.1040526, 2015.Gao, Y., Fu, J. S., Drake, J. B., Lamarque, J.-F., and Liu, Y.: The impact of
emission and climate change on ozone in the United States under
representative concentration pathways (RCPs), Atmos. Chem. Phys., 13,
9607–9621, 10.5194/acp-13-9607-2013, 2013.Gilman, J. B., Lerner, B. M., Kuster, W. C., and de Gouw, J. A.: Source
Signature of Volatile Organic Compounds from Oil and Natural Gas Operations
in Northeastern Colorado, Environ. Sci. Technol., 47, 1297–1305,
10.1021/es304119a, 2013.Granier, C., Bessagnet, B., Bond, T., D'Angiola, A., van der Gon, H. D.,
Frost, G. J., Heil, A., Kaiser, J. W., Kinne, S., Klimont, Z., Kloster, S.,
Lamarque, J.-F., Liousse, C., Masui, T., Meleux, F., Mieville, A., Ohara, T.,
Raut, J.-C., Riahi, K., Schultz, M. G., Smith, S. J., Thompson, A., van
Aardenne, J., van der Werf, G. R., and van Vuuren, D. P.: Evolution of
anthropogenic and biomass burning emissions of air pollutants at global and
regional scales during the 1980–2010 period, Climatic Change, 109, 163–190,
10.1007/s10584-011-0154-1, 2011.Gratz, L. E., Jaffe, D. A., and Hee, J. R.: Causes of increasing ozone and
decreasing carbon monoxide in springtime at the Mt. Bachelor Observatory from
2004 to 2013, Atmos. Environ., 109, 323–330,
10.1016/j.atmosenv.2014.05.076, 2014.Guenther, A., Karl, T., Harley, P., Wiedinmyer, C., Palmer, P. I., and Geron,
C.: Estimates of global terrestrial isoprene emissions using MEGAN (Model of
Emissions of Gases and Aerosols from Nature), Atmos. Chem. Phys., 6,
3181–3210, 10.5194/acp-6-3181-2006, 2006.Halliday, H. S., Thompson, A. M., Wisthaler, A., Blake, D. R., Hornbrook, R.
S., Mikoviny, T., Muller, M., Eichler, P., Apel, E. C., and Hills, A. J.:
Atmospheric benzene observations from oil and gas production in the
Denver-Julesburg Basin in July and August 2014, J. Geophys. Res.-Atmos., 121,
11055–11074, 10.1002/2016jd025327, 2016.Harris, I., Jones, P. D., Osborn, T. J., and Lister, D. H.: Updated
high-resolution grids of monthly climatic observations – the CRU TS3.10
Dataset, Int. J. Climatol., 34, 623–642, 10.1002/joc.3711, 2014.Hilboll, A., Richter, A., and Burrows, J. P.: Long-term changes of
tropospheric NO2 over megacities derived from multiple satellite
instruments, Atmos. Chem. Phys., 13, 4145-4169, 10.5194/acp-13-4145-2013,
2013.Horton, D. E., Johnson, N. C., Singh, D., Swain, D. L., Rajaratnam, B., and
Diffenbaugh, N. S.: Contribution of changes in atmospheric circulation
patterns to extreme temperature trends, Nature, 522, 465–469,
10.1038/nature14550, 2015.Huang, L., McDonald-Buller, E. C., McGaughey, G., Kimura, Y., and Allen, D.
T.: The impact of drought on ozone dry deposition over eastern Texas, Atmos.
Environ., 127, 176–186, 10.1016/j.atmosenv.2015.12.022, 2016.Huang, M., Carmichael, G. R., Chai, T., Pierce, R. B., Oltmans, S. J., Jaffe,
D. A., Bowman, K. W., Kaduwela, A., Cai, C., Spak, S. N., Weinheimer, A. J.,
Huey, L. G., and Diskin, G. S.: Impacts of transported background pollutants
on summertime western US air quality: model evaluation, sensitivity analysis
and data assimilation, Atmos. Chem. Phys., 13, 359–391,
10.5194/acp-13-359-2013, 2013.Jacob, D. J. and Winner, D. A.: Effect of climate change on air quality,
Atmos. Environ., 43, 51–63, 10.1016/j.atmosenv.2008.09.051, 2009.Jacob, D. J., Logan, J. A., and Murti, P. P.: Effect of rising Asian
emissions on surface ozone in the United States, Geophys. Res. Lett., 26,
2175–2178, 10.1029/1999gl900450, 1999.Jaffe, D.: Relationship between Surface and Free Tropospheric Ozone in the
Western U.S, Environ. Sci. Technol, 45, 432–438, 10.1021/es1028102,
2011.Jaffe, D. and Ray, J.: Increase in surface ozone at rural sites in the
western US, Atmos. Environ., 41, 5452–5463,
10.1016/j.atmosenv.2007.02.034, 2007.Jaffe, D., Chand, D., Hafner, W., Westerling, A., and Spracklen, D.:
Influence of fires on ozone concentrations in the western US, Environ. Sci.
Technol., 42, 5885–5891, 10.1021/es800084k, 2008.Jaffe, D., Wigder, N., Downey, N., Pfister, G., Boynard, A., and Reid, S. B.:
Impact of Wildfires on Ozone Exceptional Events in the Western US, Environ.
Sci. Technol., 47, 11065–11072, 10.1021/es402164f, 2013.Jia, L. W., Vecchi, G. A., Yang, X. S., Gudgel, R. G., Delworth, T. L.,
Stern, W. F., Paffendorf, K., Underwood, S. D., and Zeng, F. R.: The Roles of
Radiative Forcing, Sea Surface Temperatures, and Atmospheric and Land Initial
Conditions in US Summer Warming Episodes, J. Climate, 29,
4121–4135, 10.1175/Jcli-D-15-0471.1, 2016.John, J. G., Fiore, A. M., Naik, V., Horowitz, L. W., and Dunne, J. P.:
Climate versus emission drivers of methane lifetime against loss by
tropospheric OH from 1860–2100, Atmos. Chem. Phys., 12, 12021–12036,
10.5194/acp-12-12021-2012, 2012.Koumoutsaris, S. and Bey, I.: Can a global model reproduce observed trends in
summertime surface ozone levels?, Atmos. Chem. Phys., 12, 6983–6998,
10.5194/acp-12-6983-2012, 2012.Krotkov, N. A., McLinden, C. A., Li, C., Lamsal, L. N., Celarier, E. A.,
Marchenko, S. V., Swartz, W. H., Bucsela, E. J., Joiner, J., Duncan, B. N.,
Boersma, K. F., Veefkind, J. P., Levelt, P. F., Fioletov, V. E., Dickerson,
R. R., He, H., Lu, Z., and Streets, D. G.: Aura OMI observations of regional
SO2 and NO2 pollution changes from 2005 to 2015, Atmos. Chem.
Phys., 16, 4605–4629, 10.5194/acp-16-4605-2016, 2016.Lamarque, J.-F., Bond, T. C., Eyring, V., Granier, C., Heil, A., Klimont, Z.,
Lee, D., Liousse, C., Mieville, A., Owen, B., Schultz, M. G., Shindell, D.,
Smith, S. J., Stehfest, E., Van Aardenne, J., Cooper, O. R., Kainuma, M.,
Mahowald, N., McConnell, J. R., Naik, V., Riahi, K., and van Vuuren, D. P.:
Historical (1850–2000) gridded anthropogenic and biomass burning emissions
of reactive gases and aerosols: methodology and application, Atmos. Chem.
Phys., 10, 7017–7039, 10.5194/acp-10-7017-2010, 2010.Lamarque, J. F., Kyle, G. P., Meinshausen, M., Riahi, K., Smith, S. J., van
Vuuren, D. P., Conley, A. J., and Vitt, F.: Global and regional evolution of
short-lived radiatively-active gases and aerosols in the Representative
Concentration Pathways, Climatic Change 109, 191–212,
10.1007/s10584-011-0155-0, 2012.Langford, A. O., Aikin, K. C., Eubank, C. S., and Williams, E. J.:
Stratospheric contribution to high surface ozone in Colorado during
springtime, Geophys. Res. Lett., 36, L12801, 10.1029/2009gl038367, 2009.Langford, A. O., Senff, C. J., Alvarez II, R. J., Banta, R. M., and Hardesty,
R. M.: Long-range transport of ozone from the Los Angeles Basin: A case
study, Geophys. Res. Lett., 37, L06807, 10.1029/2010gl042507, 2010.Langford, A. O., Senff, C., Alvarez II, R., Brioude, J., Cooper, O.,
Holloway, J., Lin, M., Marchbanks, R., Pierce, R., Sandberg, S., Weickmann,
A., and Williams, E.: An overview of the 2013 Las Vegas Ozone Study (LVOS):
Impact of stratospheric intrusions and long-range transport on surface air
quality, Atmos. Environ., 109, 305–322, 10.1016/j.atmosenv.2014.08.040,
2014.Lee, H.-J., Kim, S.-W., Brioude, J., Cooper, O. R., Frost, G. J., Kim, C.-H.,
Park, R. J., Trainer, M., and Woo, J.-H.: Transport of NOx in East Asia
identified by satellite and in situ measurements and Lagrangian particle
dispersion model simulations, J. Geophys. Res.-Atmos., 119, 2574–2596,
doi:10.1002/2013JD021185, 2014.Leibensperger, E. M., Mickley, L. J., and Jacob, D. J.: Sensitivity of US air
quality to mid-latitude cyclone frequency and implications of 1980–2006
climate change, Atmos. Chem. Phys., 8, 7075–7086,
10.5194/acp-8-7075-2008, 2008.Li, G., Bei, N., Cao, J., Wu, J., Long, X., Feng, T., Dai, W., Liu, S.,
Zhang, Q., and Tie, X.: Widespread and Persistent Ozone Pollution in Eastern
China, Atmos. Chem. Phys. Discuss., 10.5194/acp-2016-864, in review,
2016.Lin, M., Holloway, T., Oki, T., Streets, D. G., and Richter, A.: Multi-scale
model analysis of boundary layer ozone over East Asia, Atmos. Chem. Phys., 9,
3277–3301, 10.5194/acp-9-3277-2009, 2009.Lin, M., Holloway, T., Carmichael, G. R., and Fiore, A. M.: Quantifying
pollution inflow and outflow over East Asia in spring with regional and
global models, Atmos. Chem. Phys., 10, 4221–4239,
10.5194/acp-10-4221-2010, 2010.Lin, M., Fiore, A. M., Cooper, O. R., Horowitz, L. W., Langford, A. O., Levy,
H., Johnson, B. J., Naik, V., Oltmans, S. J., and Senff, C. J.: Springtime
high surface ozone events over the western United States: Quantifying the
role of stratospheric intrusions, J. Geophys. Res., 117, D00V22,
10.1029/2012jd018151, 2012a.Lin, M., Fiore, A. M., Horowitz, L. W., Cooper, O. R., Naik, V., Holloway,
J., Johnson, B. J., Middlebrook, A. M., Oltmans, S. J., Pollack, I. B.,
Ryerson, T. B., Warner, J. X., Wiedinmyer, C., Wilson, J., and Wyman, B.:
Transport of Asian ozone pollution into surface air over the western United
States in spring, J. Geophys. Res., 117, D00V07, 10.1029/2011jd016961,
2012b.Lin, M., Horowitz, L. W., Oltmans, S. J., Fiore, A. M., and Fan, S.:
Tropospheric ozone trends at Mauna Loa Observatory tied to decadal climate
variability, Nat. Geosci., 7, 136–143, 10.1038/ngeo2066, 2014.Lin, M., Fiore, A. M., Horowitz, L. W., Langford, A. O., Oltmans, S. J.,
Tarasick, D., and Rieder, H. E.: Climate variability modulates western U.S.
ozone air quality in spring via deep stratospheric intrusions, Nat. Commun.,
6, 7105, 10.1038/ncomms8105,
2015a.Lin, M., Horowitz, L. W., Cooper, O. R., Tarasick, D., Conley, S., Iraci, L.
T., Johnson, B., Leblanc, T., Petropavlovskikh, I., and Yates, E. L.:
Revisiting the evidence of increasing springtime ozone mixing ratios in the
free troposphere over western North America, Geophys. Res. Lett., 42,
8719–8728, 10.1002/2015GL065311, 2015b.Lin, Y.-K., Lin, T.-H., and Chang, S.-C.: The changes in different ozone
metrics and their implications following precursor reductions over northern
Taiwan from 1994 to 2007, Environ. Monit. Assess., 169, 143–157,
10.1007/s10661-009-1158-4, 2010.Liu, F., Zhang, Q., Ronald, J. V., Zheng, B., Tong, D., Yan, L., Zheng, Y.
X., and He, K. B.: Recent reduction in NOx emissions over China: synthesis of
satellite observations and emission inventories, Environ. Res. Lett., 11,
114002, 10.1088/1748-9326/11/11/114002, 2016.Liu, H. Y., Jacob, D. J., Chan, L. Y., Oltmans, S. J., Bey, I., Yantosca, R.
M., Harris, J. M., Duncan, B. N., and Martin, R. V.: Sources of tropospheric
ozone along the Asian Pacific Rim: An analysis of ozonesonde observations, J.
Geophys. Res., 107, 4573, 10.1029/2001jd002005, 2002.
Loughner, C. P., Duncan, B. N., and Hains, J.: The benefit of historical air
pollution emissions reductions during extreme heat, Environ.
Manage., 9, 34–38, 2014.Ma, Z., Xu, J., Quan, W., Zhang, Z., Lin, W., and Xu, X.: Significant
increase of surface ozone at a rural site, north of eastern China, Atmos.
Chem. Phys., 16, 3969–3977, 10.5194/acp-16-3969-2016, 2016.McDonald, B. C., Dallmann, T. R., Martin, E. W., and Harley, R. A.: Long-term
trends in nitrogen oxide emissions from motor vehicles at national, state,
and air basin scales, J. Geophys. Res., 117, D00V18,
10.1029/2012jd018304, 2012.McDuffie, E. E., Edwards, P. M., Gilman, J. B., Lerner, B. M., Dube, W. P.,
Trainer, M., Wolfe, D. E., Angevine, W. M., deGouw, J., Williams, E. J.,
Tevlin, A. G., Murphy, J. G., Fischer, E. V., McKeen, S., Ryerson, T. B.,
Peischl, J., Holloway, J. S., Aikin, K., Langford, A. O., Senff, C. J.,
Alvarez, R. J., Hall, S. R., Ullmann, K., Lantz, K. O., and Brown, S. S.:
Influence of oil and gas emissions on summertime ozone in the Colorado
Northern Front Range, J. Geophys. Res.-Atmos., 121, 8712–8729,
10.1002/2016jd025265, 2016.Monks, P. S., Archibald, A. T., Colette, A., Cooper, O., Coyle, M., Derwent,
R., Fowler, D., Granier, C., Law, K. S., Mills, G. E., Stevenson, D. S.,
Tarasova, O., Thouret, V., von Schneidemesser, E., Sommariva, R., Wild, O.,
and Williams, M. L.: Tropospheric ozone and its precursors from the urban to
the global scale from air quality to short-lived climate forcer, Atmos. Chem.
Phys., 15, 8889–8973, 10.5194/acp-15-8889-2015, 2015.Morgenstern, O., Hegglin, M. I., Rozanov, E., O'Connor, F. M., Abraham, N.
L., Akiyoshi, H., Archibald, A. T., Bekki, S., Butchart, N., Chipperfield, M.
P., Deushi, M., Dhomse, S. S., Garcia, R. R., Hardiman, S. C., Horowitz, L.
W., Jöckel, P., Josse, B., Kinnison, D., Lin, M., Mancini, E., Manyin, M. E.,
Marchand, M., Marécal, V., Michou, M., Oman, L. D., Pitari, G., Plummer,
D. A., Revell, L. E., Saint-Martin, D., Schofield, R., Stenke, A., Stone, K.,
Sudo, K., Tanaka, T. Y., Tilmes, S., Yamashita, Y., Yoshida, K., and Zeng,
G.: Review of the global models used within phase 1 of the Chemistry-Climate
Model Initiative (CCMI), Geosci. Model Dev., 10, 639–671,
10.5194/gmd-10-639-2017, 2017.Naik, V., Horowitz, L. W., Fiore, A. M., Ginoux, P., Mao, J. Q., Aghedo, A.
M., and Levy, H.: Impact of preindustrial to present-day changes in
short-lived pollutant emissions on atmospheric composition and climate
forcing, J. Geophys. Res., 118, 8086–8110, 10.1002/jgrd.50608, 2013.Parrish, D. D., Lamarque, J.-F., Naik, V., Horowitz, L., Shindell, D. T.,
Staehelin, J., Derwent, R., Cooper, O. R., Tanimoto, H., Volz-Thomas, A.,
Gilge, S., Scheel, H.-E., Steinbacher, M., and Frühlich, M.: Long-term
changes in lower tropospheric baseline ozone concentrations: Comparing
chemistry-climate models and observations at northern midlatitudes, J.
Geophys. Res., 119, 5719–5736, 10.1002/2013JD021435, 2014.Pfister, G. G., Walters, S., Emmons, L. K., Edwards, D. P., and Avise, J.:
Quantifying the contribution of inflow on surface ozone over California
during summer 2008, J. Geophys. Res., 118, 12282–12299,
10.1002/2013jd020336, 2013.Pfister, G. G., Walters, S., Lamarque, J. F., Fast, J., Barth, M. C., Wong,
J., Done, J., Holland, G., and Bruyere, C. L.: Projections of future
summertime ozone over the US, J. Geophys. Res., 119, 5559–5582,
10.1002/2013jd020932, 2014.Porter, W. C., Heald, C. L., Cooley, D., and Russell, B.: Investigating the
observed sensitivities of air-quality extremes to meteorological drivers via
quantile regression, Atmos. Chem. Phys., 15, 10349–10366,
10.5194/acp-15-10349-2015, 2015.Pusede, S. E., Steiner, A. L., and Cohen, R. C.: Temperature and Recent
Trends in the Chemistry of Continental Surface Ozone, Chem. Rev., 115,
3898–3918, 10.1021/cr5006815, 2015.Rasmussen, D. J., Fiore, A. M., Naik, V., Horowitz, L. W., McGinnis, S. J.,
and Schultz, M. G.: Surface ozone-temperature relationships in the eastern
US: A monthly climatology for evaluating chemistry-climate models, Atmos.
Environ., 47, 142–153, 10.1016/j.atmosenv.2011.11.021, 2012.Reidmiller, D. R., Fiore, A. M., Jaffe, D. A., Bergmann, D., Cuvelier, C.,
Dentener, F. J., Duncan, B. N., Folberth, G., Gauss, M., Gong, S., Hess, P.,
Jonson, J. E., Keating, T., Lupu, A., Marmer, E., Park, R., Schultz, M. G.,
Shindell, D. T., Szopa, S., Vivanco, M. G., Wild, O., and Zuber, A.: The
influence of foreign vs. North American emissions on surface ozone in the US,
Atmos. Chem. Phys., 9, 5027–5042, 10.5194/acp-9-5027-2009, 2009.Rieder, H. E., Fiore, A. M., Horowitz, L. W., and Naik, V.: Projecting
policy-relevant metrics for high summertime ozone pollution events over the
eastern United States due to climate and emission changes during the 21st
century, J. Geophys. Res., 120, 784–800, 10.1002/2014jd022303, 2015.Russell, A. R., Valin, L. C., and Cohen, R. C.: Trends in OMI NO2
observations over the United States: effects of emission control technology
and the economic recession, Atmos. Chem. Phys., 12, 12197–12209,
10.5194/acp-12-12197-2012, 2012.Schnell, J. L., Prather, M. J., Josse, B., Naik, V., Horowitz, L. W., Zeng,
G., Shindell, D. T., and Faluvegi, G.: Effect of climate change on surface
ozone over North America, Europe, and East Asia, Geophys. Res. Lett., 43,
3509–3518, 10.1002/2016GL068060, 2016.Schultz, M. G., Heil, A., Hoelzemann, J. J., Spessa, A., Thonicke, K.,
Goldammer, J. G., Held, A. C., Pereira, J. M. C., and van het Bolscher, M.:
Global wildland fire emissions from 1960 to 2000, Global Biogeochem. Cy., 22,
Gb2002, 10.1029/2007gb003031, 2008.Schwietzke, S., Sherwood, O. A., Ruhwiler, L. M. P. B., Miller, J. B.,
Etiope, G., Dlugokencky, E. J., Michel, S. E., Arling, V. A., Vaughn, B. H.,
White, J. W. C., and Tans, P. P.: Upward revision of global fossil fuel
methane emissions based on isotope database, Nature, 538, 88–91,
10.1038/nature19797, 2016.Seager, R. and Hoerling, M.: Atmosphere and Ocean Origins of North American
Droughts, J. Climate, 27, 4581–4606,
10.1175/Jcli-D-13-00329.1, 2014.Shen, L., Mickley, L. J., and Gilleland, E.: Impact of increasing heat waves
on US ozone episodes in the 2050s: Results from a multimodel analysis using
extreme value theory, Geophys. Res. Lett., 43, 4017–4025,
10.1002/2016gl068432, 2016.
Shepherd, T. G.: CLIMATE SCIENCE: The dynamics of temperature extremes,
Nature, 522, 422–424, 2015.Simon, H., Reff, A., Wells, B., Xing, J., and Frank, N.: Ozone Trends Across
the United States over a Period of Decreasing NOx and VOC Emissions,
Environ. Sci. Technol., 49, 186–195, 10.1021/es504514z, 2015.Solberg, S., Hov, O., Sovde, A., Isaksen, I. S. A., Coddeville, P., De
Backer, H., Forster, C., Orsolini, Y., and Uhse, K.: European surface ozone
in the extreme summer 2003, J. Geophys. Res., 113, D07307,
10.1029/2007jd009098, 2008.Strode, S. A., Rodriguez, J. M., Logan, J. A., Cooper, O. R., Witte, J. C.,
Lamsal, L. N., Damon, M., Van Aartsen, B., Steenrod, S. D., and Strahan, S. E.: Trends
and variability in surface ozone over the United States, J. Geophys.
Res.-Atmos., 120, 9020–9042, 10.1002/2014JD022784, 2015.Sun, L., Xue, L., Wang, T., Gao, J., Ding, A., Cooper, O. R., Lin, M., Xu,
P., Wang, Z., Wang, X., Wen, L., Zhu, Y., Chen, T., Yang, L., Wang, Y., Chen,
J., and Wang, W.: Significant increase of summertime ozone at Mount Tai in
Central Eastern China, Atmos. Chem. Phys., 16, 10637–10650,
10.5194/acp-16-10637-2016, 2016.Tanimoto, H.: Increase in springtime tropospheric ozone at a mountainous site
in Japan for the period 1998–2006, Atmos. Environ., 43, 1358–1363,
10.1016/j.atmosenv.2008.12.006, 2009.Tanimoto, H., Zbinden, R. M., Thouret, V., and Nedelec, P.: Consistency of
tropospheric ozone observations made by different platforms and techniques in
the global databases, Tellus B, 67, 27073, 10.3402/tellusb.v67.27073,
2016.Thompson, A. M., Witte, J. C., Smit, H. G. J., Oltmans, S. J., Johnson, B.
J., Kirchhoff, V. W. J. H., and Schmidlin, F. J.: Southern Hemisphere
Additional Ozonesondes (SHADOZ) 1998–2004 tropical ozone climatology: 3.
Instrumentation, station-to-station variability, and evaluation with
simulated flight profiles, J. Geophys. Res., 112, D03304, 10.1029/2005jd007042, 2007.Travis, K. R., Jacob, D. J., Fisher, J. A., Kim, P. S., Marais, E. A., Zhu,
L., Yu, K., Miller, C. C., Yantosca, R. M., Sulprizio, M. P., Thompson, A.
M., Wennberg, P. O., Crounse, J. D., St. Clair, J. M., Cohen, R. C.,
Laughner, J. L., Dibb, J. E., Hall, S. R., Ullmann, K., Wolfe, G. M.,
Pollack, I. B., Peischl, J., Neuman, J. A., and Zhou, X.: Why do models
overestimate surface ozone in the Southeast United States?, Atmos. Chem.
Phys., 16, 13561–13577, 10.5194/acp-16-13561-2016, 2016.van der Werf, G. R., Randerson, J. T., Giglio, L., Collatz, G. J., Mu, M.,
Kasibhatla, P. S., Morton, D. C., DeFries, R. S., Jin, Y., and van Leeuwen,
T. T.: Global fire emissions and the contribution of deforestation, savanna,
forest, agricultural, and peat fires (1997–2009), Atmos. Chem. Phys., 10,
11707–11735, 10.5194/acp-10-11707-2010, 2010.VanCuren, R. and Gustin, M. S.: Identification of sources contributing to
PM2.5 and ozone at elevated sites in the western US by receptor
analysis: Lassen Volcanic National Park, California, and Great Basin National
Park, Nevada, Sci. Total Environ., 530, 505–518,
10.1016/j.scitotenv.2015.03.091, 2015.Verstraeten, W. W., Neu, J. L., Williams, J. E., Bowman, K. W., Worden, J.
R., and Boersma, K. F.: Rapid increases in tropospheric ozone production and
export from China, Nat. Geosci., 8, 690–695, 10.1038/Ngeo2493, 2015.Wang, T., Ding, A. J., Gao, J., and Wu, W. S.: Strong ozone production in
urban plumes from Beijing, China, Geophys. Res. Lett., 33, L21806,
10.1029/2006gl027689, 2006.Wang, T., Wei, X. L., Ding, A. J., Poon, C. N., Lam, K. S., Li, Y. S., Chan,
L. Y., and Anson, M.: Increasing surface ozone concentrations in the
background atmosphere of Southern China, 1994–2007, Atmos. Chem. Phys., 9,
6217–6227, 10.5194/acp-9-6217-2009, 2009.Wang, Y., McElroy, M. B., Munger, J. W., Hao, J., Ma, H., Nielsen, C. P., and
Chen, Y.: Variations of O3 and CO in summertime at a rural site near
Beijing, Atmos. Chem. Phys., 8, 6355–6363, 10.5194/acp-8-6355-2008,
2008.Wang, Y., Xie, Y., Cai, L., Dong, W., Zhang, Q., and Zhang, L.: Impact of the
2011 Southern US Drought on Ground-Level Fine Aerosol Concentration in
Summertime, J. Atmos. Sci., 72, 1075–1093, 10.1175/jas-d-14-0197.1,
2015.Warneke, C., de Gouw, J. A., Holloway, J. S., Peischl, J., Ryerson, T. B.,
Atlas, E., Blake, D., Trainer, M., and Parrish, D. D.: Multiyear trends in
volatile organic compounds in Los Angeles, California: Five decades of
decreasing emissions, J. Geophys. Res., 117, D00v17,
10.1029/2012jd017899, 2012.Xu, W., Lin, W., Xu, X., Tang, J., Huang, J., Wu, H., and Zhang, X.:
Long-term trends of surface ozone and its influencing factors at the Mt
Waliguan GAW station, China – Part 1: Overall trends and characteristics,
Atmos. Chem. Phys., 16, 6191–6205, 10.5194/acp-16-6191-2016, 2016.Yang, J., Tian, H. Q., Tao, B., Ren, W., Pan, S. F., Liu, Y. Q., and Wang, Y.
H.: A growing importance of large fires in conterminous United States during
1984–2012, J. Geophys. Res.-Biogeo., 120, 2625–2640,
10.1002/2015jg002965, 2015.Yurganov, L. N., Duchatelet, P., Dzhola, A. V., Edwards, D. P., Hase, F.,
Kramer, I., Mahieu, E., Mellqvist, J., Notholt, J., Novelli, P. C., Rockmann,
A., Scheel, H. E., Schneider, M., Schulz, A., Strandberg, A., Sussmann, R.,
Tanimoto, H., Velazco, V., Drummond, J. R., and Gille, J. C.: Increased
Northern Hemispheric carbon monoxide burden in the troposphere in 2002 and
2003 detected from the ground and from space, Atmos. Chem. Phys., 5,
563–573, 10.5194/acp-5-563-2005, 2005.
Zhang, L., Jacob, D. J., Boersma, K. F., Jaffe, D. A., Olson, J. R., Bowman,
K. W., Worden, J. R., Thompson, A. M., Avery, M. A., Cohen, R. C., Dibb, J.
E., Flock, F. M., Fuelberg, H. E., Huey, L. G., McMillan, W. W., Singh, H.
B., and Weinheimer, A. J.: Transpacific transport of ozone pollution and the
effect of recent Asian emission increases on air quality in North America: an
integrated analysis using satellite, aircraft, ozonesonde, and surface
observations, Atmos. Chem. Phys., 8, 6117–6136, 10.5194/acp-8-6117-2008,
2008.Zhang, L., Jacob, D. J., Yue, X., Downey, N. V., Wood, D. A., and Blewitt,
D.: Sources contributing to background surface ozone in the US Intermountain
West, Atmos. Chem. Phys., 14, 5295–5309, 10.5194/acp-14-5295-2014, 2014.Zhang, Y. Q., Cooper, O. R., Gaudel, A., Thompson, A. M., Nedelec, P., Ogino,
S. Y., and West, J. J.: Tropospheric ozone change from 1980 to 2010 dominated
by equatorward redistribution of emissions, Nat. Geosci., 9, 875–879,
10.1038/Ngeo2827, 2016.