David Stevenson

Institute for Meteorology

University of Edinburgh

Thanks to:

Colin Johnson, Dick Derwent, Bill Collins (Met. Office)

Modelling global Tropospheric Ozone:

Implications for Future Air Quality and Climate

Talk Structure

• Some background about tropospheric ozone

• Describe the chemistry-climate model

• Model comparisons with observations

• Model predictions

• The future

Tropospheric Ozone (O3)

• Air Pollutant– City and regional-scale photochemical

smogs – Damage to Vegetation– Human health – attacks tissue

• Greenhouse gas– Third most potent after CO2 and CH4

– Strong spatial variation in forcing

Photochemical Smog

Ozone Damage to Vegetation

Human health effects of ozone

Healthylung

Damagedlung

It makes you cry

Observed Ozone trends

European mountain sites

1970-1997 ozone sonde data

IPCC, 2001NH mid-latitudefree troposphere

Radiative forcing 1750-2000

(IPCC, 2001)CO2 1.5 W m-2

CH4 0.5 W m-2

Trop O3 0.35 W m-2

Trop. Ozone radiative forcing1750-2000

W m-2

IPCC, 2001This is a model result

IPCC models O3 2000-2100

Large range, particularly in

tropical UT

IPCC models

STOCHEM• Lagrangian chemistry-transport model• 50,000 air parcels• Coupled 3 hourly to HadAM3/HadCM3• AGCM grid: 3.75° x 2.5° x 58/19 levels• CTM output: 5° x 5° x 22 levels• 70 chemical species

– CH4-CO-NOx-Hydrocarbons– Isoprene, PAN, Acetone, CH3CHO, etc.– 5-minute chemical timestep

STOCHEM Global Chemistry Model Framework

Eulerian gridfrom GCM provides

meteorology

Air parcel centres

Interpolate met. data for eachair parcel

For each air parcel• Advection

– 4th order Runge-Kutta t=1 hr– Plus small random component (=diffusion)

• Emission & deposition fluxes• Integrate chemistry

– Photochemistry– Gas phase chemistry– Aqueous phase chemistry

• Mixing– with surrounding parcels– convective mixing– boundary layer mixing

‘Oddoxygen’

O3 + h → O(3P) + O2

O(3P) + O2 + M → O3

O3 + h → O(1D) + O2

O(1D) + M → O(3P)O3NO2 NO

Stratospheric O3

Dry deposition

O(3P) O(1D)

O3 + NO → NO2 + O2

NO2 + h → O(3P) + NO

HO2

OH

CO CH4 VOC

Anthropogenic& Natural emissions

O3

losses

NOy

losses

Use STOCHEM to look at some of the important factors for future European O3

• European emissions

• Northern hemisphere emissions

• Mix & location of emissions

• Rising levels of methane

• Climate change• Changing stratospheric ozone• Land use change / changing ‘natural’ emissions

Modelling approach

Repeat experiments changing only emissions 1990 (base year) 2030 variants

Experiments changing both emissions and climate

First, comparison with some observations for the 1990s

Anthropogenic NOx emissions 1990

Global total: 24 Tg(N)

(NB excluding biomass burning)

GOME NO2: March 1997NO2 Column Density March 1997 (1015 molecules per cm2)

P. Veefkind, KNMI

EMEP O3 monitoring sites

EMEP/TOR-2 data from NILU (A-G Hjellbrekke & S Solberg)

AOT40

(ppbh) April–September 1999 (daylig

ht

hours).

Harwell monthly mean Ozone

Model – observation comparisonSurface ozone Switzerland

Good agreement at

a rural site

Poor at a nearby

urban site

Model – observation comparison

Surface ozone Scandinavia

Good agreement at 60°N Poor in the Arctic

Observed Julydaytime

mean O3 1990-99

STOCHEM 1800hJuly mean O3

Modelling approach

Repeat experiments changing only emissions 1990 (base year) 2030 (IPCC SRES A2 scenario) 2030, 1990 Europe 2030, 1990 N. America 2030, 1990 Asia

CH4 in 1990: 1745 ppbv (used for all above)

Further 2030 run with CH4 at 2080 ppbv

Biomass burning & natural emissions fixed

Change in Anthropogenic NOx emissions 1990 to 2030

+14.3

+0.3

Global increase: +30.1 Tg(N)

Rest of World +13.1

Based on IPCC SRES A2 scenario

+2.4

IPCC SRES A2 scenario

Changes in other emissions 1990 to 2030

Global Europe N. Amer Asia ROW

NOx +30.1 +0.3 +2.4 +14.3 +13.1

CO +287 -32 -15 +148 +186

NMVOC +26 +1 -0.3 +2.4 +23NOx in Tg(N) CO in Tg(CO) NMVOC in Tg(C)

JAN

OCTJUL

APR

Surface Ozone changes 1990 to 2030 (no CH4 increase)

European spring/summer 0 ppbv in North

up to +8 in S

JAN

OCTJUL

APR

Surface O3 1990 to 2030 – component due to European emissions

European emissions cause

-3 to +6 ppbv

JAN

OCTJUL

APR

Surface Ozone changes 1990 to 2030 – N. American component

N. American emissions cause

0 to +2 ppbv

JAN

OCTJUL

APR

Surface Ozone changes 1990 to 2030 – Asian component

Vertical section 40-45°N

Asian emissions cause

0 to +2 ppbv

Europe

N.America

Asia

Extra O3 due to regional emissions changes

JAN

OCTJUL

APR

Surface Ozone changes 1990 to 2030 (including CH4 increase)

European spring/summer

~ +10 ppbv

JAN

OCTJUL

APR

Surface Ozone changes 1990 to 2030 (excluding CH4 increase)

Climate change effects

• Two mammoth 110-yr coupled chemistry-climate runs (1990-2100) 1. Control climate; SRES A2 emissions

2. SRES A2 climate forcing & emissions

• Johnson et al. (2001 , GRL)

Climate Change effects

Surface Temperature

Methane / ppbv

CH4 lifetime

SRES A2 climate

Control climate

+3.5

K

SRES A2 climate

SRES A2 climate

Control climate

Control climate

Johnson et al. 2001 GRL

+3.5K

SRES A2 climate

Control climate

N. Mid-latitude surface O3 / ppbv

Johnson et al. 2001 GRL

Large negative feedback due to

increases in water vapour and O3

destruction

Ozone chemical production (July)

200 hPa

Surface

Ozone chemical loss (July)

200 hPa

Surface

O3 net chemical production (July)

200 hPa

Surface

Ozone lifetime (July)

200 hPa

Surface

Days

5 10 20 50 100

Conclusions & remaining questions

• UK spring/summer surface O3 up 6 to 10 ppbv by 2030• European emissions: -2 to -4 ppbv

– UK appears to benefit from emissions reductions in E. Europe• N. American emissions: 0 to +2 ppbv• Asian emissions: 0 to +2 ppbv

– Other N. Hem emissions counteract European reductions• Global methane increase: +8 ppbv

– Methane increases appear very important – these are mainly driven by developing world emissions

• Climate change may reduce surface O3 – More water vapour, more O3 destruction

• What about:– Other emissions scenarios ?– Changes in stratospheric ozone ?– Changes in land-use / “natural” emissions ?

Future chemistry-climate modelling

• Higher resolution / nested models– Plume processing– Boundary layer effects – surface & tropopause– Resolved cloud processes – lightning,

convective mixing, aqueous chemistry, washout

• More coupled processes– Biosphere– ENSO – biomass burning, oceanic emissions– Emissions from and deposition to vegetation

O3NO2 NO

Stratospheric O3

Dry deposition

O(3P) O(1D)HO2

OH

CO CH4 VOC

Anthropogenic& Natural emissions

O3

losses

NOy

losses