1. Impacts of changing U.S. NOx emissions on ozone pollution: Insights from satellites, ground-based measurements and air quality models
Arlene M. Fiore
ExxonMobil Research and Engineering
NOx Controls Workshop
Bridgewater, NJ
January 12, 2016

2. Attaining the U.S. National Ambient Air Quality Standards (NAAQS) for ozone remains a challenge
One or more NAAQS
142.2
Ozone (8-hour)
133.2
PM2.5 (annual/24-hr)
28.2
PM10 (24-hr)
PM10 (24hr)
16.1
SO2 (1-hr)
15.1
Lead (3-month)
8.1
NO2 (annual/1-hr)
2012
CO (8-hr)
Millions of people living in counties with air quality concentrations above the level of the U.S. National Ambient Air Quality Standards
EPA, 2014:

3. Today’s topics
Introduction to atmospheric chemistry of ozone formation and destruction
Influence of U.S. NOx controls on ozone air quality
Overview of air quality modeling
Current research directions and challenges
-- Role of natural NOx sources
-- Accuracy of emission inventories
-- Source attribution for ozone (anthropogenic, biogenic, background components)
-- Changing climate

4. Introduction to ozone pollution chemistry

5. Ground-level O3 is produced photochemically in the atmosphere, from both natural and anthropogenic sources
Raises background ozone levels
CH4
NMVOC +
Fuel local-to-regional ozone
pollution episodes
NOx O3
Observed surface ozone includes background plus any ozone produced from local-to-regional emissions

6. Tropospheric O3 formation & “Background” contributions
stratosphere
lightning
“Background” ozone
Natural
sources
Continent X hn
intercontinental
transport
NO2 NO OH
HO2
VOC, CH4, CO
Human
activity
Fires
Land
Biosphere
(NMVOC,
NOx)
Ocean
Continent

7. Regional average loss pathways for ozone in (polluted) U. S
Regional average loss pathways for ozone in (polluted) U.S. boundary layer
Percentage (%) of total ozone (Ox) lost via each pathway,
estimated with the GEOS-Chem chemical transport model for summer 1995
Fiore et al., Journal of Geophysical Research, 2002
deposition dominates
O3 photolysis followed by OH production and O3 + HOx also important
In high NOx settings (urban, power plant plumes) O3 titration by NO also occurs

8. Regional average loss pathways for ozone in (polluted) U. S
Regional average loss pathways for ozone in (polluted) U.S. boundary layer
Percentage (%) of total ozone (Ox) lost via each pathway,
estimated with the GEOS-Chem chemical transport model for summer 1995
WESTERN U.S.A.
EASTERN U.S.A.
Based on Table 4 from Fiore et al., Journal of Geophysical Research, 2002
Deposition dominates, with O3 photolysis followed by OH production and O3 reaction with odd hydrogen radicals (OH, HO2) also important (relatively moreso in humid eastern U.S.A.)
In high NOx settings (urban, power plant plumes) O3 titration by NO also occurs

9. OZONE CONCENTRATIONS vs
OZONE CONCENTRATIONS vs. NOx AND VOC EMISSIONS Air pollution model calculation for a typical urban airshed
NOx-limited
Ridge
NOx-
saturated

10. Influence of U.S. NOx controls on ozone air quality

11. Cleaner U.S. air is visible from space
Satellite (OMI) tropospheric NO2 columns
c/o Lok Lamsal & Bryan Duncan, NASA GSFC
New OMI NO2 website: airquality.gsfc.nasa.gov

12. Consistent trends derived from satellite (OMI) vs
Consistent trends derived from satellite (OMI) vs. ground-based (AQS) measurements
c/o Bryan Duncan & Lok Lamsal, NASA GSFC

13. Satellite (OMI NO2) data indicate 20-40% decreases from 2005 to 2014 over most of the U.S.A.
c/o Bryan Duncan & Lok Lamsal, NASA GSFC

14. Trends in summer daytime (11am-4pm) 95% ozone at rural U. S
Trends in summer daytime (11am-4pm) 95% ozone at rural U.S. monitoring sites (CASTNet): 1990 to 2010
significant
not significant
Cooper et al., JGR, 2012
Decreases in EUS attributed in observations and models to NOx emission controls [e.g., Frost et al., 2006; Hudman et al., 2007; van der A. et al., 2008; Stavrakou et al., 2008; Bloomer et al., 2009, 2010; Fang et al., 2010]

15. CASTNet site: Penn Station, PA
Extreme value theory methods enable derivation of “return levels” for JJA MDA8 O3 within a given time period (from GPD fit)
Return level = Probability of observing a value x (level) within a time window T (period)
CASTNet site: Penn Station, PA
Sharp decline in return levels
between early and later periods
(NOx SIP call)
Consistent with prior work [e.g., Frost et al., 2006; Bloomer et al., 2009, 2010]
Translates air pollution changes into
probabilistic language
Apply methods to all EUS CASTNet sites to derive
1-year and 5-year return levels
Rieder et al., ERL 2013

16. Highest ozone events decrease over EUS following NOx emission controls
1-year Return Levels for Summertime MDA8 Ozone
1-yr return level decreases by 2-16 ppb
1-year levels remain above the NAAQS ozone level (70 ppb) across much of EUS
Rieder et al., ERL 2013

17. Can we exploit weekday-weekend patterns in NOx emissions (e. g
Can we exploit weekday-weekend patterns in NOx emissions (e.g., diesel) to assess impact of NOx reductions on ozone?
WEEKDAY
WEEKEND
Satellite (OMI)
Tropospheric
NO2 columns
Mid-2000s
NO2 columns decline on weekends; following NOx controls, weekday tropospheric NO2 columns look like past weekends
Early 2010s
Luke Valin
LDEO, in prep.

18. Implies regional NOx-sensitive ozone production
Similar patterns occur in surface ozone 90th% summer afternoon surface ozone (U.S. EPA AQS)
WEEKDAY
WEEKEND
Mid-
2000s
Early
2010s
Luke Valin
LDEO,
in prep.
 Additional diesel NOx controls would lower weekday surface ozone (and tropospheric NO2 columns)
Implies regional NOx-sensitive ozone production

19. What is an air quality model?
See accompanying animation
c/o Lee Murray (LDEO/NASA GISS)

20. CMAQ Overview Biogenic Emissions Anthropogenic Emissions (LADCO)
Aerosol Thermodynamics
Dispersion,
Advection
Meteorology
(WRF)
Other Processes
eularian grid-based model
Takes in my inputs
Processes them with these things
36km grid, 112 rows, 148 columns, 27 layers, 500k grid boxes
(If I were to decrease resolution to 12km number of grid boxes increases by more than 70 times)
Cloud- Aqueous Processes
Gas-phase chemistry
Boundary Conditions
(MOZART)
c/o Alex Karambelas, U Wisconsin-Madison

21. Current research directions and ongoing challenges

22. Natural vs. anthropogenic sources: Nitrogen oxides (NOx)
Relative importance
varies by day, season, year!
Approx ~43 Tg N year in 2010.
NOx = NO + NO2
(Precursor to tropospheric O3)

23. Summertime NOx Sources over Contiguous U.S.A.
Estimates from Table 1 of Hudman et al., J. Geophys Res. Atmos., 2004
for July 1 – August

24. Uncertainties exist in the sectoral breakdown of U. S
Uncertainties exist in the sectoral breakdown of U.S. anthropogenic NOx
NOy (NOx + oxidation products) too high, as is NOy/CO
Power plant NOx is well known (CEMS)
 Implies mobile sources are overestimated (by >50%) in the most recent U.S. EPA inventory (NEI2011)
Modeled (CMAQ) using U.S. EPA NEI2011 inventory
observed
Anderson et al., Atmos. Environ., 2014
Figure c/o Russ Dickerson (U MD)

25. How important are ‘natural’ NOx sources?
Model estimates of soil and lightning NOx influence on monthly mean maximum daily
8-hour (MDA8) average surface ozone
Soil NOx
June
June 2006
Hudman et al.,
Atmos. Chem. Phys., 2010
Lightning NOx
July 2004
Lightning and soil NOx vary daily, seasonally, and inter-annually.
Their impacts on ozone are not well quantified, but are typically
smaller than those from U.S. anthropogenic NOx
Fang et al., J. Geophys. Res., 2010
[ppbv]

26. Wildfires are a ‘hot’ topic: How much ozone do they produce?
Example (July )of how satellite, in situ measurements and models can be combined to detect and attribute exceptional events
Fiore et al., EM 2014 (NASA AQAST special issue;
Figures c/o R. Dickerson (U MD) excerpted from Taubman et al., 2004; Colarco et al., 2004)
Ongoing research strives to understand the impact of fire effluents on high ozone levels

27. An air quality mgmt challenge: SOURCE ATTRIBUTION How much is transported? Background? International?
GEOS-Chem model “zero-out” simulations: one day during 2012 EUS heat wave

28. c/o Kelly Chance, Harvard SAO, AQAST7 Meeting, June 2014
New insights into mid-latitude pollution expected from instruments aboard geostationary platforms with hourly, continuous coverage (~ launch dates)
c/o Kelly Chance, Harvard SAO, AQAST7 Meeting, June 2014

29. Figure 6a of Fiore, Naik, Leibensperger, JAWMA, 2015
An Air Quality Management Challenge: CLIMATE CHANGE Will warmer temperatures worsen O3 pollution?
Observations at U.S. EPA CASTNet site Penn State, PA 41N, 78W, 378m
July mean MDA8 O3 and July mean daily maximum temperature
G. Milly
Figure 6a of Fiore, Naik, Leibensperger, JAWMA, 2015
Co-variance between ozone and temperature implies that climate warming will degrade air quality in polluted regions (well established,
e.g. see Jacob & Winner Atmos. Environ. review, 2009)
Downward ozone trend as EUS NOx emission controls are implemented

30. Decreasing NOx emissions reduces sensitivity of O3 to temperature; helps to guard against any “climate penalty” [e.g., Bloomer et al., 2009; Rasmussen et al., 2012; Brown-Steiner et al., 2015]
: 4.1 ppb/C
: 2.4 ppb/C
July mean MDA8 ozone (ppb)
G. Milly
July mean maximum daily temperature (°C)
Figure 6b of Fiore, Naik, Leibensperger, JAWMA, 2015

31. Our Team at Lamont-Doherty Earth Observatory of Columbia University
Luke Valin
EPA)
Arlene
Fiore (PI)
George Milly
Lee Murray
Dan
Westervelt
Gus
Correa
Olivia
Clifton
Jean
Guo
Xiaomeng
Jin
Nora
Mascioli
Not shown but contributed to NASA AQAST projects: Yuxing Ma (CU, Masters), Cynthia Zucker (Barnard ugrad), Melissa Seto (CU ugrad), Jacob Oberman (U WI ugrad)