Ozone in the Troposphere: Air Quality, Chemical Weather and Climate
Oliver Wild
Centre for Atmospheric Science, Cambridge
Dept. of Environmental Science, University of Lancaster, 5th June 2007
Why are we Interested in Tropospheric Ozone?
• Environmental impacts on local, regional and global scales
• Secondary pollutant: sensitive to many variables– Chemical production can be fast in polluted conditions
– Lifetime is sufficiently long for global-scale transport
Pollution: O3 is an important component of photochemical smog
Climate: Direct: O3 is a greenhouse gasIndirect: O3 governs lifetime of
other GHGs via OHTropospheric oxidation: O3 regulates oxidation through control of OH and controls removal of CH4, VOCs, etc.
Anthropogenic Influence:Surface and Tropospheric O3 is increasing due to human activity
Ozone in the Troposphere
• Intercontinental transport of O3 from industrial sources
– Very long-range transport and the global O3 background
• Regional meteorology and its impacts on O3
– How do physical processes govern chemistry and transport?
• Characterising the uncertainty in current chemistry models– Can we explain the observed trends in O3 and CH4?
– What processes affecting O3 are least well understood?
Underlying themes:1. Development and evaluation of tropospheric chemistry models2. Thorough testing of models against atmospheric measurements3. Application to air quality and climate issues (O3 and CH4)
Processes Controlling Tropospheric O3
NO NO2OH HO2
CO, O3
O3
O3H2O
hν
HO2, RO2
hν
Deposition
Strat.-Trop. Exchange
NMHCs, CH4, CO
EmissionsO3
OH
O3 O3
Processes Controlling Tropospheric O3
NO NO2OH HO2
CO, O3
O3
O3H2O
hν
HO2, RO2
hν
Deposition
Strat.-Trop. Exchange
NMHCs, CH4, CO
EmissionsO3
OH
O3 O3
STE: Governed by meteorological systems, filamentation and mixing
Deposition: dependent on highly variable surface environment
Chemistry: O3 production is non-linear; strongly location-dependent
FRSGC/UCI Global CTM
1000
100
2
800
600
400
200
T42 resolution (2.8°x2.8°); driven with ECMWF-IFS forecast fields
Pre
ssur
e /h
Pa
Emissions
Wild and Prather [2000] Wild and Akimoto [2001] Wild et al., [2003]
Deposition
Tropospheric Chemistry
ASAD, 37 species
Strat. Chemistry: Linoz
PBL Turbulence
Convection: Tiedke
Advection: 2nd oM
Strat-Trop Exchange
Photolysis: Fast-J
Cloud Formation Lightning NOx source
Surface Processes
37 Levels
50
1. Intercontinental Transport of Ozone
• Industrial emission regions located at similar latitudes– Transport times about 1 week; chemical lifetime 3-4 weeks
• How much do major emission regions affect each other?– How much do they contribute to background O3?– Could this affect attainment of air quality standards?
• Explore O3 production and transport with 3-D global CTM– Single-region anthropogenic emission perturbation experiments
Current Industrial/Fossil Fuel NOx Emissions
• Photochemistry active in summer
• Transport most efficient in springWild and Akimoto [2001]
Largest O3 impacts in late spring
Source-Receptor Matrix
East AsianEmissions
USEmissions
EuropeanEmissions
• Major emission regions directly affect each other– Upwind sources contribute 1-2 ppbv to surface background O3
– This is sufficient to affect attainment of air quality standards
– Study now being repeated with many models (HTAP) to inform policy
2. Regional Meteorology and Chemical Weather
Key Questions and Challenges– How are regional and global impacts influenced by meteorology?
• What is the variability in O3 production from a given source?
– How does meteorology govern climate impacts of sources?• How will future changes in meteorology affect climate impacts?
– How well can models simulate the time scales for O3 formation?
Model Approach– Perturb fossil fuel NOx/CO/NMHC emissions over one region for one day
• Follow atmospheric perturbation for 1 month
– Repeat for each day of March 2001 (TRACE-P measurement campaign)
– Look at variability in magnitude and location of O3 production
Ozone Responses
Look at regional and global O3 from a single day’s emissions over Shanghai
March 12
– Sunny, high pressure
– Strong regional P(O3)
March 16
– Heavily overcast
– Little regional P(O3)
Regional production different, Global production similar
– Evolution quite different
– Location of P(O3) different
Meteorological Setting on March 12 and 16, 2001
Column- and latitude-integrated gross O3 production over the first 3 days following 1 day of emissions over Shanghai
L HHL
Ozone Response to Shanghai Emissions
• Effects on O3 burden
– Days with high regional O3 (smog) have a reduced effect on global O3
– Regional meteorology strongly influences climate impacts
Regional Boundary LayerDistant Boundary LayerFree Troposphere
Global Ozone IncreaseRegional Ozone Increase
• P(O3) vs. NOx loss for each day
– O3 production efficiency (OPE) strongly dependent on location
– Good representation of lifting processes is required!
3. Exploring the Uncertainty in Current CTMs
• CTM studies show large differences in O3 burden and lifetime
– Where do these differences originate?
• Perform sensitivity study on key processes in a single CTM– Identify processes contributing to this uncertainty
O3 Burden vs. O3 Lifetime
Diagonals in grey show O3 loss rate (Tg/year)
(τO3 = Burden/Loss)• ACCENT studies• CTM with NMHC• CTM without NMHC
3. Exploring the Uncertainty in Current CTMs
• CTM studies show large differences in O3 burden and lifetime
– Where do these differences originate?
• Perform sensitivity study on key processes in a single CTM– Identify processes contributing to this uncertainty
O3 Burden vs. O3 Lifetime
Diagonals in grey show O3 loss rate (Tg/year)
(τO3 = Burden/Loss)• ACCENT studies• CTM with NMHC• CTM without NMHC
330 Tg/yr
22.4 days
Best estimates from recent model studies
3. Exploring the Uncertainty in Current CTMs
Sensitivity to key variables explains much of the scatter
60 Tg NOx
650 Tg Isop
800 Tg STE
250 Tg STE
−20% H2O
+20% H2O
7.5 Tg NOx lightning
T−5°CT+5°C
0 Tg
0 Tg
460 Tg dep
975 Tg dep
−20%
+20% 30 Tg NOx
O3 Burden vs. O3 Lifetime
Diagonals in grey show O3 loss rate (Tg/year)
3. Exploring the Uncertainty in Current CTMs
• Summary of key sensitivities– NOx emissions: more O3, P(O3), more OH
– Isoprene emissions: more O3, P(O3), less OH
– Lightning NOx: poorly constrained, large impact on O3 and OH
– Meteorology: effects of humidity and STE
• Implications– Current models are not good enough to model trends in O3 and CH4!
Account for 2/3 of model variability
• ACCENT studies• CTM with NMHC• CTM without NMHC
Future Studies
• Modelling atmosphere-vegetation interactions– Important feedbacks between O3, VOC, N-species and plants
– Interaction of anthropogenic and vegetation emissions is very poorly understood and requires spatial disaggregation
– Currently lead the ‘biogenic fluxes’ theme in JULES
Soils Crops
Requires improved treatment of biogenic emissions and deposition.
Involves collaboration with land use and vegetation community and a full Earth System approach
NOx, CO VOC VOC
aerosol
O3
NO
NOy
Climate
Future Studies
• Improved treatment of urban emissions in climate models– Improved simulation of O3 production in coarse-resolution models
– Reduced bias in regional/global O3 important for climate
– Allows better testing against surface observations
– Important for assessing environmental impacts of Megacities
Background
Plume
Mixing zone
These processes function on a range of scales, but their impacts on climate have not been assessed.
Involve strong collaboration with the UK and EU urban & local modelling communityWind Direction
Future Studies
• Modelling the evolution of tropospheric oxidation– Reproducing the observed trends in CH4 and O3
– Important for climate and air quality communities
– Requires improved understanding of tropospheric chemistry
– Need a better characterization of variability in CH4 sources
Need more thorough testing of models vs. observations
Contributes to goals of new international Atmospheric Chemistry and Climate project
Wild and Akimoto [2001]
Annual Mean Impacts on O3
Daily O3 from Source Regions in Springtime
Global Impact
Receptor Region
r2=0.92
OPE=35
TRACE-P Ozonesondes
• Stratospheric intrusion at Cheju, Korea, March 1–2, 2001
• Intercepted by sondes on successive days
– Very different profiles
• CTM captures evolution of features well
– Two layers on March 1
– Background strat. enhancement
– One high layer on March 2
– Residual strat air mixed in
• Suggests mechanisms for STE can be captured, but demonstrates high degree of variability in ozone
Evolution of O3 profile over Cheju, Korea in CTM
Pres
sure
/hPa
March 1, 2001 March 2, 2001
Sonde data: Sam Oltmans, NOAA/CMDL
Tropopause
Net O3 Production Rate
• Instantaneous O3 production in CTM vs. box model constrained by observations
• Mean latitude-altitude profile over all DC8/P3B flights
• Net destruction in tropical marine boundary layer
• Strong production over Japan
• Strong plume activity in outflow region, 23º–32ºN
• Net production in upper trop (underestimated in CTM)
(Box model: Jim Crawford, NASA Langley, Doug Davis, Georgia Tech.)