Picture: METEOSAT Oct 2000

Tropospheric OTropospheric O33 budget budget

of the of the

South Atlantic regionSouth Atlantic region

B. Sauvage,

R. V. Martin, A. van Donkelaar, I. Folkins, X.Liu, P. Palmer, V. Thouret , A. M. Thompson,

P. Bernath & K. Chance

Outstanding scientific issue in the TropicsOutstanding scientific issue in the Tropics

GOME Seasonal Tropospheric O3 Columns 2000DJF

MAM

JJA

SON

Topic: O3 maximum zonal wave-one (from Fishman et al. 1987… to Wang et al 2006)

Scientific interest:

year-round pattern observed since the 80’s. Situated in MT-UT O3 critical for radiative effect Key role on the oxidizing power of the atmosphere O3 maximum attributed to various anthropogenic and natural sources + dynamics

Goal:

What controls O3 maximum?

(Sources / regions)

data from Liu et al 2005DU

MethodologyMethodology

What controls

the O3 maximum?

High estimation emissions uncertainty

Global chemical transport model GEOS-Chem “Original” simulation

1

Constraint & Evaluation

In-situ & satellite observations

Lightning: spatial distribution scaled to OTD-LIS

Soils: a posteriori inventory of NOx from GOME (Jaeglé et al., 2005)

Biomass burning: top-down constraint on NOx & VOCs from GOME

2

Quantification (sources / regions)

O3 maximum

3Constrained

“standard” simulation

Space-based constraint on emissionsSpace-based constraint on emissions

NO

h

O3

NO2

HNO3Lifetime hoursVOC

OHHCHOh

hoursCO

hours

PBL

NOx VOC

Free Troposphere

NO NO2

O3, HO2

hv

HNO3

NOx lifetime ~ week

O3

lifetime ~ month

O3

lifetime ~ days

Tropospheric NO2 column ~ ENOx

Tropospheric HCHO column ~ EVOC GOME: 320x40 km2

OTD-LIS Lightning flash rates

Effect of satellite constraint in simulated tropospheric column O3Effect of satellite constraint in simulated tropospheric column O3

Δ Tropospheric O3 Columns “constrained” – “original ” simulationsDJF

MAM

JJA

SON

ΔDU

Large influence from lightning and biomass burning constraint

Space-based constraint on Space-based constraint on lightninglightning NO NOxx emissionsemissions

OTD-LIS flashes (1995-2004) local seasonal rescaling of lightning emissions

-Regional differences / oceanic emissions

-Same intensity: 5 Tg N yr-1

109 molec N cm-2 s-1

Original constrained with OTD/LIS

Modeled lightning NOx emissions (DJF)

In-situ OIn-situ O33, data used to evaluate the simulation, data used to evaluate the simulation

1.MOZAIC airborne program (Marenco et al., 1998; Thouret et al. 1998): 1994-2004 landing and taking off phase

2.SHADOZ ozone sonde network (Thompson et al., 2003a; 2003b): 1998-2004

More than 9000 vertical profiles of OMore than 9000 vertical profiles of O33 over the Tropics 30 over the Tropics 30°°N-30N-30°°SS

Highlights of simulation evaluation: sensitivity to Highlights of simulation evaluation: sensitivity to lightninglightning

Rescaling improve middle-upper tropospheric O3 from 5-15 ppbv

Main influence over subsident zone; South America; Middle East; East

Sensitivity to lightning intensity:7Tg N/yr too high; 3Tg N/yr too low; 5±2Tg N/yr gives overall agreement.

In-situ

constrained

GOME Model originalModel constrained

Better agreement during biomass burning season

1015 molec cm-2

Better spatial correlations between GOME and model NO2 columns R2 > 0.86

Space-based constraint on Space-based constraint on biomass burningbiomass burning emissions emissions

GOME NO2 regional top-down constraint of biomass burning NOx emissions

data from Martin et al. 2002

Tropics: 4.8TgN/yr 5.8TgN/yr

DJF

MAM

JJA

SON

Better spatial correlations between GOME and model HCHO columns R2> 0.7

Space-based constraint on Space-based constraint on biomass burningbiomass burning emissions emissions

data from Chance et al. 2000

Better agreement during biomass burning season

GOME HCHO top-down constraint of biomass burning VOC emissions

HCHO and alkenes emissions increased x 2GOME GEOS-Chem constrained GEOS-Chem original

Highlights of simulation evaluation: sensitivity to biomass burningHighlights of simulation evaluation: sensitivity to biomass burning

Top-down improves lower tropospheric O3 from 5-20 ppbv during biomass burning season

Main influence over Africa DJF-JJA; India MAM

Use of constrained simulation Quantify (sources/regions) influencing O3 maximum

What controls the O3 maximum?What controls the O3 maximum?

O3 maximum

?

O3 budget / Sensitivity to sourcesO3 budget / Sensitivity to sourcesSensitivity to decreasing NOx emissions by 1% and 100% for each source

ΔDU

-Lightning downwind; largest influence over the Tropics & South Atlantic

>36% >7% >9%DJF

MAM

JJA

SON

-Surface sources local; half of the lightning NOx influence (but similar source strength)

-Lightning Ozone Production Efficiency = 3 times OPE of each surface source-Tropical background 30%

O3 budget Sensitivity to regionsO3 budget Sensitivity to regions

ΔDU

Sensitivity to decreasing NOx emissions by 1% over regions

>20% >15% >6%DJF

MAM

JJA

SON

The zonal-wave oneThe zonal-wave one

Vertical-zonal seasonal cross section of O3 and O3 flux

DJF MAM

JJA SON

subsidenceAfricaS. America

Dynamical description / annual meanDynamical description / annual mean

O3 ppb NOx ppb

3/O3 buildup during transport and subsidence over South Atlantic high area

Zonal transport

1/Surface emissions of O3 precursors

S. Am. Africa

2/Injection of NOx into the MT-UT with lightning emissions and uplift into ITCZ

Modeled

SHADOZ+ MOZAIC

O3 (ppbv)

4/ Meridional transport

Meridional transport

AFRICAATLANTICS N

O3 (ppbv)

ConclusionsConclusions1/ Spatial distribution of lightning scaled to reproduce OTD-LIS seasonal mean Improve MT-UT O3 by 5 to 15 ppbv

2/ Top-down constraint on emission inventories of NOx from soil and biomass burning, of VOCs from biomass burning improve LT O3 by 5 to 20 ppbv

Lightning source of 5 Tg±2Tg best reproduces versus in-situ MOZAIC & SHADOZ

>6%

>20%

> 21% > 36%

EAST

AFRICA

South America

O3 maximum is driven by convergence and sustained largely by lightning NOx emissions, which present larger OPE

>15%

Surface NOx sources STE ~ 6%(500 Tg/yr)

Picture: METEOSAT Oct 2000

Thanks for attention!Thanks for attention!

Comparison of convective schemesComparison of convective schemes

ITCZO3 min/ CO max/ RH max

GEOS-3 presents weak convective outflow

GEOS-4 low clouds altitude & optical thickness

Weak incidence over the Atlantic

ITCZ

ITCZ

Flight altitude mean over Africa, 300-200hPa, JJA seasonCO RH

O3

CO lower estimated in LT / CO; emissions increased by 2 weak or negative impact on modeled versus in-situ

CO & Relative humidity evaluationCO & Relative humidity evaluation

Space-based constraint on emissionsSpace-based constraint on emissions

Different intensity of NOx emissions:

Biomass burning: 4.8TgN/yr 5.8TgN/yr / Soils 3.5TgN/yr 4.5TgN/yr (Tropics). Larger influence over Africa and India.

109 molec

N.cm-2.s-1

Seasonal NOx biomass burning emissions (DJF)

Original Standard