Re-analysis of tropospheric sulfate aerosol and ozone for the period ...

Atmos Chem Phys 11 9563ndash9594 2011wwwatmos-chem-physnet1195632011doi105194acp-11-9563-2011copy Author(s) 2011 CC Attribution 30 License

AtmosphericChemistry

and Physics

Re-analysis of tropospheric sulfate aerosol and ozone for the period1980ndash2005 using the aerosol-chemistry-climate modelECHAM5-HAMMOZ

L Pozzoli1 G Janssens-Maenhout1 T Diehl23 I Bey4 M G Schultz5 J Feichter6 E Vignati1 and F Dentener1

1European Commission Joint Research Centre Institute for Environment and Sustainability Ispra Italy2NASA Goddard Space Flight Center Greenbelt Maryland USA3University of Maryland Baltimore County Baltimore Maryland USA4Center for Climate Systems Modeling and Institute of Atmospheric and Climate Science ETH Zurich Zurich Switzerland5Forschungszentrum Julich Germany6Max Planck Institute for Meteorology Hamburg Germany now at Eurasia Institute of Earth Sciences Istanbul Technical University Turkey

Received 10 March 2011 ndash Published in Atmos Chem Phys Discuss 30 March 2011Revised 12 August 2011 ndash Accepted 29 August 2011 ndash Published 16 September 2011

Abstract Understanding historical trends of trace gas andaerosol distributions in the troposphere is essential to evalu-ate the efficiency of existing strategies to reduce air pollutionand to design more efficient future air quality and climatepolicies We performed coupled photochemistry and aerosolmicrophysics simulations for the period 1980ndash2005 usingthe aerosol-chemistry-climate model ECHAM5-HAMMOZto assess our understanding of long-term changes and inter-annual variability of the chemical composition of the tro-posphere and in particular of ozone and sulfate concentra-tions for which long-term surface observations are avail-able In order to separate the impact of the anthropogenicemissions and natural variability on atmospheric chemistrywe compare two model experiments driven by the sameECMWF re-analysis data but with varying and constant an-thropogenic emissions respectively Our model analysis in-dicates an increase of ca 1 ppbv (0055plusmn 0002 ppbv yrminus1)in global average surface O3 concentrations due to anthro-pogenic emissions but this trend is largely masked by thelarger O3 anomalies due to the variability of meteorology andnatural emissions The changes in meteorology (not includ-ing stratospheric variations) and natural emissions account

Correspondence toF Dentener(frankdentenerjrceceuropaeu)

for the 75 of the total variability of global average sur-face O3 concentrations Regionally annual mean surfaceO3 concentrations increased by 13 and 16 ppbv over Eu-rope and North America respectively despite the large an-thropogenic emission reductions between 1980 and 2005A comparison of winter and summer O3 trends with mea-surements shows a qualitative agreement except in NorthAmerica where our model erroneously computed a posi-tive trend Simulated O3 increases of more than 4 ppbv inEast Asia and 5 ppbv in South Asia can not be corroboratedwith long-term observations Global average sulfate surfaceconcentrations are largely controlled by anthropogenic emis-sions Globally natural emissions are an important driver de-termining AOD variations Regionally AOD decreased by28 over Europe while it increased by 19 and 26 inEast and South Asia The global radiative perturbation cal-culated in our model for the period 1980ndash2005 was rathersmall (005 W mminus2 for O3 and 002 W mminus2 for total aerosoldirect effect) but larger perturbations ranging fromminus054to 126 W mminus2 are estimated in those regions where anthro-pogenic emissions largely varied

Published by Copernicus Publications on behalf of the European Geosciences Union

9564 L Pozzoli et al Re-analysis 1980ndash2005 ECHAM5-HAMMOZ

1 Introduction

Air quality is determined by the emission of primary pol-lutants into the atmosphere by chemical production of sec-ondary pollutants and by meteorological conditions Thetwo air pollutants of most concern for public health ozone(O3) and particulate matter (PM) have also strong impactson climate In the fourth Assessment Report of the Intergov-ernmental Panel on Climate Change (IPCC AR4)Solomonet al (2007) estimate a Radiative Forcing (RF) from tropo-spheric ozone of +035 [minus01 +03] W mminus2 which corre-sponds to the third largest contribution to the total RF aftercarbon dioxide (CO2) and methane (CH4) IPCC AR4 alsoprovides estimates for radiative forcing from aerosols Thedirect RF (scattering and absorption of solar and infrared ra-diation) amounts tominus05 [plusmn04] W mminus2 and the RF throughindirect changes in cloud properties is estimated tominus070[minus11+04] W mminus2

Trends in global radiation and visibility measurements in-deed suggest an important role of aerosols Solar radia-tion measurements showed a consistent and worldwide de-crease at the Earthrsquos surface (an effect dubbed ldquodimmingrdquo)from the 1960s This trend reversed into ldquobrighteningrdquoin the late 1990s in the US Europe and parts of Korea(Wild 2009) Similarly an analysis of visibility measure-ments from 1973ndash2007 byWang et al(2009) suggests aglobal increase of AOD worldwide except in Europe SinceSO2minus

4 is one of the main aerosol components that determinethe aerosol optical depth (Streets et al 2009) the SO2minus

4concentration reductions over Europe and US after the im-plementation of air quality policies may partly explain thedimming-brightening transition observed in the 1990s in Eu-rope whereas in emerging economies such as China andIndia the emission of air pollutants rapidly increased since1990

Inter-annual meteorological variability in the last decadesalso strongly determined the variations of the concentrationsand geographical distribution of air pollutants For exam-ple the El Nino event in 1997ndash1998 and the period after theMt Pinatubo volcanic eruption in 1991 can explain much ofthe past chemical inter-annual variability of tropospheric O3CH4 and OH (Fiore et al 2009 Dentener et al 2003 Hessand Mahowald 2009) In Europe the infamous summer of2003 led to a strong positive anomaly of solar surface radia-tion (Wild 2009) and exacerbated ozone pollution at ground-level and throughout the troposphere (Solberg et al 2008Tressol et al 2008)

Meteorological variability and changes in the precursoremissions of O3 and SO2minus

4 are often concurrent processesand their impact on surface concentrations is difficult to un-derstand from measurements alone (Vautard et al 2006Berglen et al 2007) Therefore there have been severalefforts to re-analyze these trends using tropospheric chem-istry and transport models For example the Europeanproject REanalysis of the TROpospheric chemical composi-

tion (RETROSchultz et al 2007) employed three differentglobal models to simulate tropospheric ozone changes be-tween 1960 and 2000 RecentlyHess and Mahowald(2009)analyzed the role of meteorology in inter-annual variabilityof tropospheric ozone chemistry

In this paper we extend these analyses by using a coupledaerosol-chemistry-climate simulation and discuss our resultsin the light of the previous studies We analyze the chemi-cal variability due to changes in meteorology (ie transportchemistry) and natural emissions and separate them from an-thropogenic emissions induced variability The period 1980ndash2005 was chosen because the advent of satellite observationsin the 1970s introduced a discontinuity in the re-analysisof meteorological datasets (Hess and Mahowald 2009 vanNoije et al 2006) and as mentioned above the consideredperiod includes large meteorological anomalies Particularattention will be given to four regions of the world (NorthAmerica Europe East Asia and South Asia) where signifi-cant changes in terms of the absolute amount of emitted tracegases and aerosol precursors occurred in the last decades Wewill focus on past changes of O3 and SO2minus

4 because of theirimportance for air quality and climate Long measurementrecords are available since the 1980s and they will be usedto evaluate our model results and the simulated trends Wefurther analyze changes in AOD radiative perturbation (RP)and OH radical associated with changes in emissions and me-teorology

The paper is organized as follows In Sect2 the modeldescription and experiment setup are outlined In Sect3 an-thropogenic and natural emissions are described Sections4and5 present an analysis of global and regional variabilityand trends of O3 and SO2minus

4 from 1980 to 2005 The regionalanalysis focuses on Europe (EU) North America (NA) EastAsia (EA) and South Asia (SA) (Fig1 the region bound-aries were defined as in the HTAP studyFiore et al 2009)In Sect6 we will describe the anthropogenic radiative per-turbation due to changes in O3 and SO2minus

4 In Sects7 and8 we will summarize the main findings and further discussimplications of our study for air quality-climate interactions

2 Model and simulation descriptions

ECHAM5-HAMMOZ is a fully coupled aerosol-chemistry-climate model composed of the general circulation model(GCM) ECHAM5 the tropospheric chemistry module MOZand the aerosol module HAM The ECHAM5-HAMMOZmodel is described in detail inPozzoli et al(2008a) Themodel has been extensively evaluated in previous studies(Stier et al 2005 Pozzoli et al 2008ab Auvray et al2007 Rast et al 2011) with comparisons to several mea-surements and within model inter-comparison studies Wefurther remark a substantial overestimate of our computedozone compared to measurements a problem that ECHAM5-

Atmos Chem Phys 11 9563ndash9594 2011 wwwatmos-chem-physnet1195632011

L Pozzoli et al Re-analysis 1980ndash2005 ECHAM5-HAMMOZ 9565

Fig 1 Map of the selected regions for the analysis and measurement stations with long records of O3 and SO2minus

4 surface concentrationsNorth America (NA) [15 Nndash55 N 60 Wndash125 W] Europe (EU) [25 Nndash65 N 10 Wndash50 E] East Asia (EA) [15 Nndash50 N 95 Endash160 E)] and South Asia (SA) [5 Nndash35 N 50 Endash95 E] Triangles show the location of EMEP stations squares of WDCGG stations anddiamonds of CASTNET stations The stations are grouped in sub-regions Northern Europe (NEU) Central Europe (CEU) Western Europe(WEU) Eastern Europe (EEU) Southern Europe (SEU) Western US (WUS) North-Eastern US (NEUS) Mid-Atlantic US (MAUS) Greatlakes US (GLUS) Southern US (SUS)

HAMMOZ shares with many other global models and werefer for a further discussion toEllingsen et al(2008)

In this study a triangular truncation at wavenumber 42(T42) resolution was used for the computation of the generalcirculation Physical variables are computed on an associatedGaussian grid with a horizontal resolution of ca 28

times 28

degrees The model has 31 vertical levels from the surfaceup to 10 hPa and a time resolution for dynamics and chem-istry of 20 min We simulated the period 1979ndash2005 (thefirst year is discarded from the analysis as spin-up) Me-teorology was taken from the ECMWF ERA-40 re-analysis(Uppala et al 2005) until 2000 and from operational analy-ses (IFS cycle-32r2) for the remaining period (2001ndash2005)Even though we do not find any evidence for this we mustnote that this discontinuity may have an impact on the mete-orological variables and therefore on our analysis in the lastyears of the simulated period Such discontinuities may alsoarise within a re-analysis data set due to the inclusion of dif-ferent data sets in the assimilation procedure ECHAM5 vor-ticity divergence sea surface temperature and surface pres-sure are relaxed towards the re-analysis data every time stepwith a relaxation time scale of 1 day for surface pressureand temperature 2 days for divergence and 6 h for vortic-ity (Jeuken et al 1996) The relaxation technique forces thelarge scale dynamic state of the atmosphere as close as pos-sible to the re-analysis data thus the model is in a consis-tent physical state at each time step but it calculates its ownphysics eg for aerosols and clouds The concentrations ofCO2 and other GHGs used to calculate the radiative budgetwere set according to the specifications given in Appendix IIof the IPCC TAR report (Nakicenovic et al 2000) In Ap-pendixA we provide a detailed description of the chemicaland microphysical parameterizations included in ECHAM5-HAMMOZ A detailed description of the ECHAM5 modelcan be found inRoeckner et al(2003)

Two transient simulations were conducted

ndash SREF reference simulation for 1980ndash2005 where me-teorology and emissions are changing on an hourly-to-monthly basis

ndash SFIX simulation for 1980ndash2005 with anthropogenicemissions fixed at year 1980 while meteorology nat-ural and wildfire emissions change as in SREF

3 Emissions

31 Anthropogenic emissions

The anthropogenic emissions of CO NOx and VOCs forthe period 1980ndash2000 are taken from the RETRO inven-tory (httpretroenesorg) (Schultz et al 2007 Endresenet al 2003 Schultz et al 2008) which provides monthlyaverage emission fields interpolated to the model resolu-tion of 28 times 28 In order to prevent a possible driftin CH4 concentrations with consequences for the simula-tion of OH and O3 we prescribed monthly zonal meanCH4 concentrations in the boundary layer obtained fromthe interpolation of surface measurements (Schultz et al2007) The prescribed CH4 concentrations vary annuallyand range from 1520ndash1650 ppbv (Southern and NorthernHemisphere respectively SH and NH) in 1980 to 1720ndash1860 ppbv in 2005 (SH and NH) NOx aircraft emissionsare based onGrewe et al(2001) and distributed accord-ing to prescribed height profiles The AeroCom hind-cast aerosol emission inventory (httpdataipslipsljussieufrAEROCOMemissionshtml) was used for the annual totalanthropogenic emissions of primary black carbon (BC) or-ganic carbon (OC) aerosols and sulfur dioxide (SO2) Specif-ically the detailed global inventory of primary BC and OCemissions byBond et al(2004) was modified byStreets et al(2004 2006) to include additional technologies and new fuelattributes to calculate SO2 emissions using the same energy

wwwatmos-chem-physnet1195632011 Atmos Chem Phys 11 9563ndash9594 2011


drivers as for BC and OC and extended to the period 1980ndash2006 (Streets et al 2009) using annual fuel-use trends andeconomic growth parameters included in the IMAGE model(National Institute for Public Health and the Environment2001) Except for biomass burning the BC OC and SO2anthropogenic emissions were provided as annual averagesPrimary emissions of SO2minus

4 are calculated as constant frac-tion (25 ) of the anthropogenic sulfur emissions The SO2and primary SO2minus

4 emissions from international ship trafficfor the years 1970 1980 1995 and 2001 were based on theEDGAR 2000 FT inventory (van Aardenne et al 2001) andlinearly interpolated in time using total emission estimatesfrom Eyring et al(2005ab) The emissions of CO NOxand VOCs were available only until the year 2000 There-fore we used for 2001ndash2005 the 2000 emissions except forthe regions where significant changes were expected between2000 and 2005 such as US Europe East Asia and SouthEast Asia (defined as in Fig1) for which derived emissiontrends of CO NOx and VOCs were applied to year 2000The emission ratios between the period 2001ndash2005 and theyear 2000 from the USEPA (httpwwwepagovttnchieftrends) EMEP (httpwwwemepint) and REAS (httpwwwjamstecgojpfrcgcresearchp3emissionhtm) emis-sion inventories were applied to year 2000 emissions usedfor this study over the US Europe and Asia

Table1 summarizes the total annual anthropogenic emis-sions for the years 1980 1985 1990 1995 2000 and 2005Global emissions of CO VOCs and BC were relatively con-stant during the last decades with small increases between1980 and 1990 decreases in the 1990s and renewed increasebetween 2000 and 2005 During these 25 yr global NOx andOC emissions increased up to 10 while sulfur emissionsdecreased by 10 However these global numbers maskthat the global distribution of the emission largely changedwith reductions over North America and Europe balancedby strong increases in the economically emerging countriessuch as China and India Figure2 shows the relative trends ofanthropogenic emissions used during this study over the 4 se-lected world regions In Europe and North America there is ageneral decrease of emissions of all pollutants from 1990 onIn East Asia and South Asia anthropogenic emissions gen-erally increase although for EA there is a decrease around2000

For the period 1980ndash2000 our global amounts of emissionfrom anthropogenic sources which are based on RETROare lower by more than 10 for CO and NOx and theyare higher by more than 40 for VOCs when comparedto the new emission inventory prepared byLamarque et al(2010) in support of the Intergovernmental Panel on ClimateChange (IPCC) Fifth Assessment Report (AR5) For the pe-riod 2001ndash2005 several studies have recently shown signifi-cant changes in regional emissions especially in Asia (egRichter et al(2005) Zhang et al(2009) Klimont et al(2009)) In our study compared to the projection for year2005 of Lamarque et al(2010) which includes the refer-

Table 1 Global anthropogenic emissions of CO [Tg yrminus1]NOx [Tg(N) yrminus1] VOCs [Tg(C) yrminus1] SO2 [Tg(S) yrminus1] SO4

2minus

[Tg(S) yrminus1] OC [Tg yrminus1] and BC [Tg yrminus1]

1980 1985 1990 1995 2000 2005

CO 6737 6801 7138 6853 6554 6786NOx 342 337 361 364 367 372VOCs 846 850 879 848 805 843SO2 671 672 662 610 588 590SO4

2minus 18 18 18 17 16 16OC 78 82 83 87 85 87BC 49 49 49 48 47 49

ences cited above we applied significantly lower emissionreductions between 1980 and 2005 for CO and SO2 in EUlarger reductions for BC and OC in both EU and NA loweremission changes except for CO in EA similar emissionchanges except for NOx and SO2 in SA (Fig2)

32 Natural and biomass burning emissions

Some O3 and SO2minus

4 precursors are emitted by natural pro-cesses which exhibit inter-annual variability due to chang-ing meteorological parameters (temperature wind solar ra-diation clouds and precipitation) For example VOC emis-sions from vegetation are influenced by surface temperatureand short wavelength radiation NOx is produced by light-ning and associated with convective activity dimethyl sulfide(DMS) emissions depend on the phytoplankton blooms in theoceans and wind speed and some aerosol species such asmineral dust and sea salt are strongly dependent on surfacewind speed Inter-annual variability of weather will thereforeinfluence the concentrations of tropospheric O3 and SO2minus

4and may also affect the radiative budget The emissionsfrom vegetation of CO and VOCs (isoprene and terpenes)were calculated interactively using the Model of Emissionsof Gases and Aerosols from Nature (MEGAN) (Guentheret al 2006) The total annual natural emissions in Tg(C)of CO and biogenic VOCs (BVOCs) for the period 1980ndash2005 range from 840 to 960 Tg(C) yrminus1 with a standard de-viation of 26 Tg(C) yrminus1 (Fig 3a) which corresponds to 3 of the annual mean natural VOC emissions for the consid-ered period 80 of these total biogenic emissions occur inthe tropics

Lightning NOx emissions (Fig3b) are calculated fol-lowing the parameterization ofGrewe et al(2001) Wecalculated a 5 variability for NOx emissions from light-ning from 355 to 425 Tg(N) yrminus1 with a decreasingtrend of 0017 Tg(N) yrminus1 (R2 of 053 95 confidenceboundsplusmn 0007) We note here that there are consider-able uncertainties in the parameterization of NOx emissions(eg Grewe et al 2001 Tost et al 2007 Schumann andHuntrieser 2007) and different parameterizations simulated



opposite trends over the same period (Schultz et al 2007)An annual constant contribution of 93 Tg(N) yrminus1 is in-cluded in our simulations for NOx emissions from microbialactivity in soils (Schultz et al 2007)

DMS predominantly emitted from the oceans is an SO2minus

4aerosol precursor Its emissions depend on seawater DMSconcentrations associated with phytoplankton blooms (Ket-tle and Andreae 2000) and model surface wind speedwhich determines the DMS sea-air exchange (Nightingaleet al 2000) Terrestrial biogenic DMS emissions followPham et al(1995) The total DMS emissions are rangingfrom 227 to 244 Tg(S) yrminus1 with a standard deviation of03 Tg(S) yrminus1 or 1 of annual mean emissions (Fig3c)The highest DMS emissions correspond to the 1997ndash1998ENSO event Since we have used a climatology of seawaterDMS concentrations instead of annually varying concentra-tions the variability may be misrepresented

The emissions of sea salt are based onSchulz et al(2004)The strong dependency on wind speed results in a range from5000ndash5550 Tg yrminus1 (Fig 3e) with a standard deviation of2

Mineral dust emissions are calculated online using theECHAM5 wind speed hydrological parameters (eg soilmoisture) and soil properties following the work ofTegenet al (2002) and Cheng et al(2008) The total mineraldust emissions have a large inter-annual variability (Fig3d)ranging from 620 to 930 Tg yrminus1 with a standard deviationof 72 Tg yrminus1 almost 10 of the annual mean average Min-eral dust originating from the Sahara and over Asia con-tribute on average 58 and 34 to the global mineral dustemissions respectively

Biomass burning from tropical savannah burning de-forestation fires and mid-and high latitude forest fires arelargely linked to anthropogenic activities but fire severity(and hence emissions) are also controlled by meteorologi-cal factors such as temperature precipitation and wind Weuse the compilation of inter-annual varying biomass burningemissions published bySchultz et al(2008) who used liter-ature data satellite observations and a dynamical vegetationmodel in order to obtain continental-scale emission estimatesand a geographical distribution of fire occurrence We con-sider emissions of the components CO NOx BC OC andSO2 and apply a time-invariant vertical profile of the plumeinjection height for forest and savannah fire emissions Fig-ure 3f shows the inter-annual variability for CO NOx andOC biomass burning emissions with different peaks such asin year 1998 during the strong ENSO episode

Other natural emissions are kept constant during the en-tire simulation period CO emissions from soil and ocean arebased on the Global Emission Inventory Activity (GEIA) asin Horowitz et al(2003) amounting to 160 and 20 Tg yrminus1respectively Natural soil NOx emissions are taken fromthe ORCHIDEE model (Lathiere et al 2006) resulting in9 Tg(N) yrminus1 SO2 volcanic emissions of 14 Tg(S) yrminus1 arefrom Andres and Kasgnoc(1998) andHalmer et al(2002)

Since in the current model version secondary organic aerosol(SOA) formation is not calculated online we applied amonthly varying OC emission (19 Tg yrminus1) to take into ac-count the SOA production from biogenic monoterpenes (afactor of 015 is applied to monoterpene emissions ofGuen-ther et al 1995) as recommended in the AEROCOM project(Dentener et al 2006)

4 Variability and trends of O 3 and OH during1980ndash2005 and their relationship to meteorologicalvariables

In this section we analyze the global and regional variabilityof O3 and OH in relation to selected modeled chemical andmeteorological variables Section41 will focus on globalsurface ozone Sect42will look into more detail to regionaldifferences in O3 and for Europe and the US compare thedata to observations Section43 will describe the variabil-ity of the global ozone budget and Sect44 will focus onthe related changes in OH As explained earlier we will sep-arate the influence of meteorological and natural emissionvariability from anthropogenic emissions by differencing theSREF and SFIX simulations

41 Global surface ozone and relation to meteorologicalvariability

In Fig 4 we show the ECHAM5-HAMMOZ SREF inter-annual monthly anomalies (difference between the monthlyvalue and the 25-yr monthly average) of global mean surfacetemperature water vapor and surface concentrations of O3SO2minus

4 total column AOD and methane-weighted OH tropo-spheric concentrations will be discussed in later sections Tobetter visualize the inter-annual variability over 1980ndash2005in Fig 4 we also display the 12-months running averageof monthly mean anomalies for both SREF (blue) and SFIX(red) simulations

Surface temperature evolution showed large anomaliesin the last decades associated with major natural eventsFor example large volcanic eruptions (El Chichon 1982Mt Pinatubo 1991) generated cooler temperatures due to theemission into the stratosphere of sulfate aerosols The 1997ndash1998 ENSO caused an increase in temperature of ca 04 KThe strong coupling of the hydrological cycle and tempera-ture is reflected in a correlation of 083 between the monthlyanomalies of global surface temperature and water vaporcontent These two meteorological variables influence O3and OH concentrations For example the large 1997ndash1998ENSO event corresponds with a positive ozone anomaly ofmore than 2 ppbv There is also an evident correspondencebetween lower ozone and cooler periods in 1984 and 1988However the correlation between monthly temperature andO3 anomalies (in SFIX) is only moderate (R = 043)



1980 1985 1990 1995 2000 200570

60

50

40

30

20

10

0

10a) Anthropogenic emission changes over EU

CO9563 Tg yr 1

VOC2066 Tg(C) yr 1

NOx753 Tg(N) yr 1

SO22565 Tg(S) yr 1

BC131 Tg yr 1

OC163 Tg yr 1

1980 1985 1990 1995 2000 200560

50

40

30

20

10

0

10b) Anthropogenic emission changes over NA

CO12740 Tg yr 1

VOC2611 Tg(C) yr 1

NOx753 Tg(N) yr 1

SO21289 Tg(S) yr 1

BC068 Tg yr 1

OC087 Tg yr 1

1980 1985 1990 1995 2000 200520

0

20

40

60

80

100

120

140c) Anthropogenic emission changes over EA

CO9817 Tg yr 1

VOC1083 Tg(C) yr 1

NOx296 Tg(N) yr 1

SO21065 Tg(S) yr 1

BC142 Tg yr 1

OC194 Tg yr 1

1980 1985 1990 1995 2000 20050

50

100

150

200

250

300

350d) Anthropogenic emission changes over SA

CO6015 Tg yr 1

VOC553 Tg(C) yr 1

NOx098 Tg(N) yr 1

SO2147 Tg(S) yr 1

BC039 Tg yr 1

OC116 Tg yr 1

Fig 2 Percentage changes in total anthropogenic emissions (CO VOC NOx SO2 BC and OC) from 1980 to 2005 over the four selectedregions as shown in Fig1 (a) Europe EU(b) North America NA(c) East Asia EA(d) South Asia SA In the legend of each regionalplot the total annual emissions for each species are reported for the year 1980 The colored points represent the percentage changes inregional emissions between 1980 and 2005 inLamarque et al(2010) which includes recent assessments in the projections for the year 2005

Fig 3 Total annual natural and biomass burning emissions for the period 1980ndash2005(a) biogenic CO and VOCs emissions from vegetation(b) NOx emissions from lightning(c) DMS emissions from oceans(d) mineral dust aerosol emissions(e)marine sea salt aerosol emissions(f) CO NOx and OC aerosol biomass burning emissions



Table 2 Average standard deviation (SD) and relative standard deviation (RSD standard deviation divided by the mean) of globally averagedvariables in this work for the simulation with changing anthropogenic emission and with fixed anthropogenic emissions (1980ndash2005) Threedimensional variables are density weighted and averaged between the surface and 280 hPa Three dimensional quantities evaluated at thesurface are prefixed with Sfc The standard deviation is calculated as the standard deviation of the monthly anomalies (the monthly valueminus the mean of all years for that month)

SREF SFIX

Average SD RSD Average SD RSD

Sfc O3 (ppbv) 3645 0826 00227 3597 0627 00174O3 (ppbv) 4837 11020 00228 4764 08851 00186CO (ppbv) 0103 0000416 00402 0101 0000529 00519OH (molecules cmminus3

times 106 ) 120 0016 0013 118 0029 0024HNO3 (pptv) 12911 1470 01139 12573 1391 01106

Emi S (Tg yrminus1) 1042 349 00336 10809 047 00043SO2 (pptv) 2317 1502 00648 2469 870 00352Sfc SO4

minus2 (microg mminus3) 112 0071 00640 118 0061 00515SO4

minus2 (microg mminus3) 069 0028 00406 072 0025 00352

Fig 4 Monthly mean anomalies for the period 1980ndash2005 of globally averaged fields for the SREF and SFIX ECHAM5-HAMMOZsimulations Light blue lines are monthly mean anomalies for the SREF simulation with overlaying dark blue giving the 12 month runningaverages Red lines are the 12 month running averages of monthly mean anomalies for the SFIX simulation The grey area representsthe difference between SREF and SFIX The global fields are respectively(a) surface temperature (K)(b) tropospheric specific humidity(g Kgminus1) (c) O3 surface concentrations (ppbv)(d) OH tropospheric concentration weighted by CH4 reaction (molecules cmminus3) (e) SO2minus

4surface concentrations (microg mminus3) (f) total aerosol optical depth



Fig 5 Maps of surface O3 concentrations and the changes due to anthropogenic emissions and natural variability In the first column weshow 5 yr averages (1981ndash1985) of surface O3 concentrations over the selected regions Europe North America East Asia and South AsiaIn the second column(b) we show the effect of anthropogenic emission changes in the period 2001ndash2005 on surface O3 concentrationscalculated as the difference between SREF and SFIX simulations In the third column(c) the natural variability of O3 concentrations isshown which is due to natural emissions and meteorology in the simulated 25 yr calculated as the difference between the 5 yr averageperiods (2001ndash2005) and (1981ndash1985) in the SFIX simulation The combined effect of anthropogenic emissions and natural variability isshown in column(d) and it is expressed as the difference between the 5 yr average period (2001ndash2005)ndash(1981ndash1985) in the SREF simulation

The 25-yr global surface O3 average is 3645 ppbv (Ta-ble 2) and increased by 048 ppbv compared to the SFIXsimulation with year 1980 constant anthropogenic emis-sions About half of this increase is associated with anthro-pogenic emission changes (Fig4c (grey area)) The inter-annual monthly surface ozone concentrations varied by up toplusmn 217 ppbv (1σ = 083 ppbv) of which 75 (063 ppbv inSFIX) was related to natural variations- especially the 1997ndash1998 ENSO event

42 Regional differences in surface ozone trendsvariability and comparison to measurements

Global trends and variability may mask contrasting regionaltrends Therefore we also perform a regional analysis forNorth America (NA) Europe (EU) East Asia (EA) andSouth Asia (SA) (see Fig1) To quantify the impact onsurface concentrations after 25 yr of changing anthropogenicemission we compare the averages of two 5-yr periods1981ndash1985 and 2001ndash2005 We considered 5-yr averagesto reduce the noise due to meteorological variability in thesetwo periods

In Fig 5 we provide maps for the globe Europe NorthAmerica East Asia and South Asia (rows) showing in thefirst column (a) the reference surface concentrations and inthe other 3 columns relative to the period 1981ndash1985 theisolated effect of anthropogenic emission changes (b) mete-orological and natural emission changes (c) and combinedchanges (d) The global maps provide insight into inter-regional influences of concentrations

A comparison to observed trends provides additional in-sight into the accuracy of our calculations (Fig6) This com-parison however is hampered by the lack of observations be-fore 1990 and the lack of long-term observations outside ofEurope and North America We will therefore limit the com-parison to the period 1990ndash2005 separately for winter (DJF)and summer (JJA) acknowledging that some of the largerchanges may have happened before Figure1 displays themeasurement locations we used 53 stations in North Amer-ica and 98 stations in Europe To allow a realistic comparisonwith our coarse-resolution model we grouped the measure-ment in 5 subregions for EU and 5 subregions for NA Foreach subregion we calculated the trend of the median winterand summer anomalies In AppendixB we give an extensive



Fig 6 Trends of the observed (OBS) and calculated (SREF and SIFX) O3 and SO2minus

4 seasonal anomalies (DJF and JJA) averaged overeach group of stations as shown in Fig1 Northern Europe (NEU) Central Europe (CEU) Western Europe (WEU) Eastern Europe (EEU)Southern Europe (SEU) Western US (WUS) North-Eastern US (NEUS) Mid-Atlantic US (MAUS) Great lakes US (GLUS) Southern US(SUS) The vertical bars represent the 95 confidence interval of the trends The number of stations used to calculate the average seasonalanomalies for each subregion is shown in parentheses Further details in AppendixB

description of the data used for these summary figures andtheir statistical comparison with model results

We note here that O3 and SO2minus

4 in ECHAM5-HAMMOZwere extensively evaluated in previous studies (Stier et al2005 Pozzoli et al 2008ab Rast et al 2011) showing ingeneral a good agreement between calculated and observedSO2minus

4 and an overestimation of surface O3 concentrations insome regions We implicitly assume that this model bias doesnot influence the calculated variability and trends (Fig B1 inAppendixB)

421 Europe

In Fig 5e we show the SREF 1981ndash1985 annual mean EUsurface O3 (ie before large emission changes) correspond-ing to an EU average concentration of 469 ppbv High an-nual average O3 concentrations up to 70 ppbv are found overthe Mediterranean basin and lower concentrations between20 to 40 ppbv in Central and Eastern Europe Figure5f showsthe difference in mean O3 concentrations between the SREFand SFIX simulation for the period 2001ndash2005 The de-cline of NOx and VOC anthropogenic emissions was 20 and 25 respectively (Fig2a) Annual averaged surfaceO3 concentration increases by 081 ppbv between 1981ndash1985and 2001ndash2005 The spatial distribution of the calculatedtrends shows a strong variation over Europe the computedO3 increased between 1 and 5 ppbv over Northern and Cen-

tral Europe while it decreased by up to 1ndash5 ppbv over South-ern Europe The O3 responses to emission reductions aredriven by a complex nonlinear photochemistry of the O3NOx and VOC system Indeed we can see very differentO3 winter and summer sensitivities in Figs7a and8 Theincrease (7 compared to year 1980) in annual O3 surfaceconcentration is mainly driven by winter (DJF) values whilein summer (JJA) there is a small 2 decrease between SREFand SFIX This winter NOx titration effect on O3 is particu-larly strong over Europe (and less over other regions) as wasalso shown in eg Fig 4 ofFiore et al(2009) due to the rel-atively high NOx emission density and the mid-to-high lati-tude location of Europe The impact of changing meteorol-ogy and natural emissions over Europe is shown in Fig5gshowing an increase by 037 ppbv between 1981ndash1985 and2001ndash2005 Winter-time variability drives much of the inter-annual variability of surface O3 concentrations (see Figs7aand8a) While it is difficult to attribute the relationship be-tween O3 and meteorological conditions to a single processwe speculate that the European surface temperature increaseof 07 K from 1981ndash1985 to 2001ndash2005 could play a signif-icant role Figure5d 5h and5l show that modeled NorthAtlantic ozone increased during the same period contribut-ing to the increase of the baseline O3 concentrations at thewestern border of EU Figure5f and5g show that both emis-sions and meteorological variability may have synergistically



Fig 7 Winter (DJF) anomalies of surface O3 concentrations averaged over the selected regions (shown in Fig1) On the left y-axisanomalies of O3 surface concentrations for the period 1980ndash2005 are expressed as the ratios between seasonal means for each year and theyear 1980 The blue line represents the SREF simulation (changing meteorology and changing anthropogenic emissions) the red line theSFIX simulation (changing meteorology and fixed anthropogenic emissions at the level of 1980) while the gray area indicates the SREF-SFIX difference On the right y-axis the green and pink dashed lines represent the changes ie the ratio between each year and 1980 of totalannual VOC and NOx emissions respectively

Fig 8 As Fig7 for summer (JJA) anomalies of surface O3 concentrations



caused upward trends in Northern and Central Europe anddownward trends in Southern Europe

The regional responses of surface ozone can be comparedto the HTAP multi-model emission perturbation study ofFiore et al(2009) which considered identical regions andis thus directly comparable to this study They found thatthe 20 reduction over EU of all HTAP emissions deter-mined an increase of O3 by 02 ppbv in winter and an O3decrease byminus17 ppbv in summer (average of 21 models) toa large extent driven by NOx emissions Considering similarannual mean reductions in NOx and VOC emissions over Eu-rope from 1980ndash2005 we found a stronger emission drivenincrease of up to 3 ppbv in winter and a decrease of up to1 ppbv in summer An important difference between the twostudies may be the use of a spatially homogeneous emis-sion reduction in HTAP while the emissions used here gener-ally included larger emission reductions in Northern Europewhile emissions in Mediterranean countries remained moreor less constant

The calculated and observed O3 trends for the period1990ndash2005 are relatively small compared to the inter-annualvariability In winter (DJF) (Fig6) measured trends con-firm increasing ozone in most parts of Europe The observedtrends are substantially larger (03ndash05 ppbv yrminus1) than themodel results (0ndash02 ppbv yrminus1) except in Western Europe(WEU) In summer (JJA) the agreement of calculated andobserved trends is small in the observations they are closeto zero for all European regions with large 95 confidenceintervals while calculated trends show significant decreases(01ndash045 ppbv yrminus1) of O3 Despite seasonal O3 trends notbeing well captured by the model the seasonally averagedmodeled and measured surface ozone concentrations are rea-sonably well correlated for a large number of stations (Ap-pendix B) In winter the simulated inter-annual variabilityseemed to be somewhat underestimated pointing to miss-ing variability coming from eg stratosphere-troposphere orlong-range transport In summer the correlations are gener-ally increasing for Central Europe and decreasing for West-ern Europe

Our results are generally in good agreement with previousestimates of observed O3 trends egJonson et al(2006)Lamarque et al(2010) Cui et al (2011) Wilson et al(2011) In those studies they reported observed annual O3trends of 032ndash040 ppbv yrminus1 for stations in central Europewhich are comparable with the observed trends calculated inour study (Fig B3 in AppendixB) In Mace Head Lamar-que (2010) reported an increasing trend of 018 ppbv yrminus1comparable to our study (Fig B3 in AppendixB) Jonsonet al(2006) reported seasonal trends of O3 ranging between013 and 05 ppbv yrminus1 in winter (JF) and betweenminus059andminus012 ppbv yrminus1 in summer (JJA)Wilson et al(2011)calculated positive annual O3 trends in Central and North-Western Europe (014ndash041 pbbv yrminus1) and significant nega-tive annual trends at 11 of sites mainly located in Easternand South-Western Europe (minus128ndashminus024 pbbv yrminus1)

422 North America

Computed annual mean surface O3 over North America(Fig5i) for 1981ndash1985 was 483 ppbv Higher O3 concentra-tions are found over California and in the continental outflowregions Atlantic and Pacific Oceans and the Gulf of Mex-ico and lower concentrations north of 45 N AnthropogenicNOx in NA decreased by 17 between 1980 and 2005 par-ticularly in the 1990s (minus22 ) (Fig 2b) These emissionreductions produced an annual mean O3 concentration de-crease up to 1 ppbv over all the Eastern US 1ndash2 ppbv overthe Southern US and 1 ppbv in the Western US Changesin anthropogenic emissions from 1980 to 2005 (Fig5j) re-sulted in a small average increase of ozone by 028 ppbvover NA where effects on O3 of emission reductions in theUS and Canada were balanced by higher O3 concentrationsmainly over the tropics (below 25 N) Changes in meteorol-ogy and natural emissions (Fig5k) increased O3 between1 and 5 ppbv over the continent and reduced O3 by up to5 ppbv over Western Mexico and the Atlantic Ocean (aver-age of 089 ppbv over the entire region) Thus natural vari-ability was the largest driver of the over-all average regionalincrease of 158 ppbv in SREF (Fig5l) Over NA natu-ral variability is the main driver of summer (JJA) O3 fluc-tuations fromminus7 to 5 (Fig8b) In winter (Fig7b)the contribution of emissions and chemistry is strongly in-fluencing these relative changes Computed and measuredsummer concentrations were better correlated than those inwinter (AppendixB) However while in winter an analy-sis of the observed trends seems to suggest 0ndash02 ppbv yrminus1

O3 increases the model predicts small O3 decreases insteadthough these differences are not often significant (see alsoAppendixB) In summer modeled upward trends (SREF) arenot confirmed by measurements except for the Western US(WUS) These computed upward trends were strongly deter-mined by the large-scale meteorological variability (SFIX)and the model trends solely based on anthropogenic emis-sion changes (SREF-SFIX) would be more consistent withobservations

Despite the difficulty in comparing trends calculated withdifferent methods and for different periods our observedtrends are qualitatively in good agreement with previousstudies over the Western USOltmans et al(2008) observedpositive trends at some sites but no significant changes atothers Jaffe and Ray(2007) estimated positive O3 trendsof 021ndash062 ppbv yrminus1 in winter and 043ndash050 ppbv yrminus1 insummerParrish et al(2009) found 043plusmn 017 ppbv yrminus1

in winter and 024plusmn 016 ppbv yrminus1 in summer O3 trendsLamarque et al(2010) found annual increasing O3 trend of033 ppbv yrminus1 Chan and Vet(2010) found larger positivedaytime O3 summer and winter trends close to 1 ppbv yrminus1for the period 1997ndash2006 Over the Eastern US our studyqualitatively agrees withChan and Vet(2010) who found de-creasing significant trends over all the Eastern US in summerand mainly no significant trends in winter



423 East Asia

In East Asia we calculated an annual mean surface O3 con-centration of 436 ppbv for 1981ndash1985 Surface O3 is lessthan 30 ppbv over North-Eastern China and influenced bycontinental outflow conditions up to 50 ppbv concentrationsare computed over the Northern Pacific Ocean and Japan(Fig5m) Over EA anthropogenic NOx and VOC emissionsincreased by 125 and 50 respectively from 1980 to2005 Between 1981ndash1985 and 2001ndash2005 O3 is reduced by10 ppbv in North Eastern China due to reaction with freshlyemitted NO In contrast O3 concentrations increase close tothe coast of China by up to 10 ppbv and up to 5 ppbv over theentire north Pacific reaching North America (Fig5b) Forthe entire EA region (Fig5n) we found an increase of annualmean O3 concentrations of 243 ppbv The effect of meteo-rology and natural emissions is generally significantly posi-tive with an EA-wide increase of 16 ppbv and up to 5 ppbvin northern and southern continental EA (Fig5o) The com-bined effect of anthropogenic emissions and meteorology isan increase of 413 ppbv in O3 concentrations (Fig5p) Dur-ing this period the seasonal mean O3 concentrations wereincreasing by 3 and 9 in winter and summer respec-tively (Figs7c and8c) The effect of natural variability onthe seasonal mean O3 concentrations shows opposite effectsin winter and summer a reduction between 0 and 5 in win-ter and an increase between 0 and 10 in summer The 3long-term measurement datasets at our disposal (not shown)indicate large inter-annual variability of O3 and no signifi-cant trend in the time period from 1990 to 2005 thereforethey are not plotted in Fig6

424 South Asia

Of all 4 regions the largest relative change in anthropogenicemissions occurred over SA NOx emissions increased by150 VOC by 60 and sulfur by 220 South AsianO3 inter-annual variability is rather different from EA NAand EU because the SA region is almost completely situ-ated in the tropics Meteorology is highly influenced by theAsian monsoon circulation with the wet season in JunendashAugust We calculate an annual mean surface O3 concen-tration of 482 ppbv with values between 45 and 60 ppbvover the continent (Fig5q) Note that the high concentra-tions at the northern edge of the region may be influencedby the orography of the Himalaya The increasing anthro-pogenic emissions enhanced annual mean surface O3 con-centrations by on average 424 ppbv (Fig5r) and more than5 ppbv over India and the Gulf of Bengal In the NH winter(dry season) the increase in O3 concentrations of up to 10 due to anthropogenic emissions is more pronounced than inthe summer (wet season) with an increase of only 5 in JJA(Figs7d and8d) The effect of meteorology produced an an-nual mean O3 increase of 115 ppbv over the region and morethan 2 ppbv in the Ganges valley and in the southern Gulf of

Bengal (Fig5s) The total variability in seasonal mean O3concentrations is on the order of 5 both in winter and sum-mer (Figs7d and8d) In 25 yr the computed annual meanO3 concentrations increased by 512 ppbv over SA approxi-mately 75 of which are related to increasing anthropogenicemissions (Fig5t) Unfortunately to our knowledge no suchlong-term data of sufficient quality exist in India

43 Variability of the global ozone budget

We will now discuss the changes in global tropospheric O3To put our model results in a multi-model context we showin Fig9 the global tropospheric O3 budget along with budgetterms derived fromStevenson et al(2006) O3 budget termswere calculated using an assumed chemical tropopause witha threshold of 150 ppbv of O3 The annual globally integratedchemical production (P ) loss (L) surface deposition (D)and stratospheric influx (Sinf = L+DminusP ) terms are well inthe range of those reported byStevenson et al(2006) thoughO3 burden and lifetime are at the high end In our study thevariability in production and loss are clearly determined bymeteorological variability with the 1997ndash1998 ENSO eventstanding out The increasing turnover of tropospheric ozonemanifests in gradually decreasing ozone lifetimes (minus1 dayfrom 1980 to 2005) while total tropospheric ozone burdenincreases from 370 to 380 Tg caused by increasing produc-tion and stratospheric influx

Further we compare our work to a re-analysis study byHess and Mahowald(2009) which focused on the relation-ship between meteorological variability and ozoneHess andMahowald(2009) used the chemical transport model (CTM)MOZART2 to conduct two ozone simulations from 1979 to1999 without considering the inter-annual changes in emis-sions (except for lightning emissions) and is thus very com-parable to our SFIX simulation The simulations were drivenby two different re-analysis methodologies the NationalCenter for Environmental PredictionNational Center for At-mospheric Research (NCEPNCAR) re-analysis the outputof the Community Atmosphere Model (CAM3Collins et al2006) driven by observed sea surface temperatures (SNCEPand SCAM inHess and Mahowald(2009) respectively) Thecomparison of our model (in particular the SFIX simulation)driven by the ECMWF ERA-40 re-analysis withHess andMahowald (2009) provides insight in the extent to whichthese different approaches impact the inter-annual variabil-ity of ozone In Table3 we compare the results of SFIX withthe twoHess and Mahowald(2009) model results We ex-cluded the last 5 yr of our SFIX simulation in order to allowdirect statistical comparison with the period 1980ndash2000

431 Hydrological cycle and lightning

The variability of photolysis frequencies of NO2 (JNO2) atthe surface is an indicator for overhead cloud cover fluctua-tions with lower values corresponding to larger cloud cover



Table 3 Average standard deviation (SD) and relative standard deviation (RSD standard deviation divided by the mean) of globallyaveraged variables in this work SCAM and SNCEP(Hess and Mahowald 2009) (1980ndash2000) Three dimensional variables are densityweighted and averaged between the surface and 280 hPa Three dimensional quantities evaluated at the surface are prefixed with Sfc Thestandard deviation is calculated as the standard deviation of the monthly anomalies (the monthly value minus the mean of all years for thatmonth)

ERA-40 (this work SFIX) SCAM (Hess 2009) SNCEP (Hess 2009)

Average SD RSD Average SD RSD Average SD RSD

SfcT (K) 287 0112 0000391 287 0116 0000403 287 0121 000042SfcJNO2 (sminus1

times 10minus3) 213 00121 000567 243 000454 000187 239 000814 000341LNO (TgN yrminus1) 391 0153 00387 471 0118 00251 279 0211 00759PRECT (mm dayminus1) 295 00331 0011 242 00145 0006 24 00389 00162Q (g kgminus1) 472 0060 00127 346 00411 00119 338 00361 00107O3 (ppbv) 4779 0819 001714 46 0192 000418 484 0752 00155Sfc O3 (ppbv) 361 0595 00165 298 0122 00041 312 0468 0015CO (ppbv) 0100 0000450 004482 0083 0000449 000542 00847 0000388 000458OH (mole moleminus1

times 1015 ) 631 1269 002012 735 0707 000962 704 0847 0012HNO3 (pptv) 127 1395 01096 121 122 00101 121 149 00123

Fig 9 Global tropospheric O3 budget calculated for the period 1980ndash2005 for the SREF (blue) and SFIX (red) ECHAM5-HAMMOZsimulations(a) chemical production (P ) (b) chemical loss (L) (c) surface deposition (D) (d) stratospheric influx (Sinf = L + D minus P ) (e)tropospheric burden (BO3) (f) lifetime (τO3 = BO3(L + D)) The black points for the year 2000 represent the meanplusmn standard deviationbudgets as found in the multi model study ofStevenson et al(2006)

JNO2 values are ca 10 lower in our SFIX (ECHAM5) sim-ulation compared to the two simulations reported byHessand Mahowald(2009) This may be due to a different rep-resentation of the cloud impact on photolysis frequencies (inpresence of a cloud layer lower rates at surface and higherrates above) as calculated in our model using Fast-J2 (see

AppendixA) compared to the look-up-tables used byHessand Mahowald(2009) Furthermore we found 22 higheraverage precipitation and 37 higher tropospheric water va-por in ECHAM5 than reported for the NCEP and CAMre-analyses respectively As discussed byHagemann et al(2006) ECHAM5 humidity may be biased high regarding



processes involving the hydrological cycle especially in theNH summer in the tropics The computed production of NOxfrom lightning (LNO) of 391 Tg yrminus1 in our SFIX simu-lation resides between the values found in the NCEP- andCAM-driven simulations ofHess and Mahowald(2009) de-spite the large uncertainties of lightning parameterizations(see Sect32)

432 Variability re-analysis and nudging methods

The global multi annual averages of the meteorological vari-ables and gas concentrations at the surface and in the tropo-sphere are rather similar in the 3 simulations (Table3) O3 isranging from 46 to 484 ppbv while 20 higher values arefound for surface O3 in our SFIX simulation Our global tro-pospheric average of CO concentrations is 15 higher thanin Hess and Mahowald(2009) probably due to different bio-genic CO and VOCs emissions Despite larger amounts ofwater vapor and lower surfaceJNO2 in SFIX our calcu-lated OH tropospheric concentrations are smaller by 15 Remarkable differences compared toHess and Mahowald(2009) are found in the inter-annual variability In generalthe variability calculated in our SFIX simulation is closer tothe SNCEP simulation ofHess and Mahowald(2009) butour model exhibits a larger inter-annual variability of globalO3 CO OH and HNO3 than the CAM-driven simulationanalysis byHess and Mahowald(2009) This likely indi-cates that nudging with different re-analysis datasets (such asECMWF and NCEP) or only prescribing monthly averagedsea-surface temperatures can give significantly different an-swers on the processes that govern inter-annual variations inthe chemical composition of the troposphere The variabilityof OH (calculated as the relative standard deviation RSD)in our SFIX simulation is higher by a factor of 2 than thosein SCAM and SNCEP and CO HNO3 up to a factor of 10We speculate that these differences point to differences in thehydrological cycle among the models which influence OHthrough changes in cloud cover and HNO3 through differentwashout rates A similar conclusion was reached byAuvrayet al (2007) who analyzed ozone formation and loss ratesfrom the ECHAM5-MOZ and GEOS-CHEM models for dif-ferent pollution conditions over the Atlantic Ocean Since themethodology used in our SFIX simulation should be rathercomparable to that used for the NCEP-driven MOZART2simulation we speculate that in addition to the differences inre-analysis (NCEPNCAR and ECMWFERA-40) also dif-ferent nudging methodologies may strongly impact the cal-culated inter-annual variabilities

44 OH variability

The main processes that contribute to the OH variabil-ity are both meteorological and chemicalDentener et al(2003) found that OH variability for the period 1979ndash1993was mainly driven by meteorological processes ie humid-

itytemperature and wet removalprecipitations They foundonly a small total contribution from the changes in chem-ical species like CH4 O3 and emissions of NOx VOCsand CO Differently fromDentener et al(2003) in ourstudy the effects of natural processes (in the SFIX sim-ulation) includes also changes in emissions of CO NOxand VOCs from biomass burning and lightning emissionsFurthermore the variability from stratospheric O3 is notincluded as well as the effect of Mt Pinatubo eruptionTo estimate the changes in the global oxidation capacitywe calculated the global mean tropospheric OH concentra-tion weighted by the reaction coefficient of CH4 follow-ing Lawrence et al(2001) We found a mean value of12plusmn 0016times 106 molecules cmminus3 in the SREF simulation(Table2 within the 106ndash139times 106 molecules cmminus3 rangecalculated byLawrence et al 2001) The mean value ofthe yearly means of CH4 lifetime over the period 1980ndash2005is 104plusmn 014 yr in the SREF simulation This value is justabove the 1-σ interval of CH4 lifetimes reported by the allmodel average inStevenson et al(2006) In Fig 4d weshow the global tropospheric average monthly mean anoma-lies of OH During the period 1980ndash2005 we found a de-creasing OH trend ofminus033times 104 molecules cmminus3 (R2

=

079 due to natural variability) balanced by an oppositetrend of 030times 104 molecules cmminus3 (R2

= 095) due to an-thropogenic emission changes In agreement with an earlierstudy byFiore et al(2006) we found a strong relationship be-tween global lightning and OH inter-annual variability with acorrelation of 078 This correlation drops to 032 for SREFwhich additionally includes the effect of changing anthro-pogenic CO VOC and NOx emissions Lower correlationswere found with water vapor (R = minus048) and photolysisrates at surface (egJNO2 R = minus065 JO1D R = minus056)The resulting OH inter-annual variability of 2 for the im-pact of meteorology (SFIX) was close to the estimate ofHess and Mahowald(2009) (for the period 1980ndash2000 Ta-ble 3) and much smaller than the 10 variability estimatedby Prinn et al(2005) but somewhat larger than the globalinter-annual variability of 15 analyzed byDentener et al(2003) The latter authors however did not include inter-annual varying biomass burning emissionsDentener et al(2003) computed an increasing trend of 024plusmn 006 yrminus1

in OH global mean concentrations for the period 1979ndash1993 using a different model which was mainly causedby meteorological variability For a slightly shorter period(1980ndash1993) we found a decreasing trend ofminus027 yrminus1

in our SFIX run Montzka et al(2011) estimate an inter-annual variability of OH in the order of 2plusmn 18 for theperiod 1985ndash2008 which compares reasonably well withour results of 16 and 24 for the SREF and SFIX runs(1980ndash2005 Table2) respectively The calculated OH de-cline from 2001 to 2005 which was not strongly correlatedto global surface temperature humidity or lightning couldhave implications for the understanding of the stagnation ofatmospheric methane growth during the first part of 2000s



However we do not want to over-interpret this decline sincein this period we used meteorological data from the opera-tional ECMWF analysis instead of ERA-40 (Sect2) Onthe other hand we have no evidence of other discontinuitiesin our analysis and the magnitude of the OH changes wassimilar to earlier changes in the period 1980ndash1995

5 Surface and column SO2minus

4

In this section we analyze global and regional sulfate sur-face concentrations (Sect51) and the global sulfate budget(Sect52) including their variability The regional analysisand comparison with measurements follows the approach ofthe ozone analysis above

51 Global and regional surface sulfate

Global average surface SO2minus

4 concentrations are 112 and118 microg(S) mminus3 for the SREF and SFIX runs respectively(Table 2) Anthropogenic emission changes induce a de-crease of ca 01 microg(S) mminus3 SO2minus

4 between 1980 and 2005 inSREF (Fig4e) Monthly anomalies of SO2minus

4 surface concen-trations range fromminus01 to 02 microg(S) mminus3 The 12-monthrunning averages of monthly anomalies are in the range ofplusmn01 microg(S) mminus3 for SREF and about half of this in the SFIXsimulation The anomalies in global average SO2minus

4 surfaceconcentrations do not show a significant correlation with me-teorological variables on the global scale

511 Europe

For 1981ndash1985 we compute an annual average SO2minus

4 surfaceconcentration of 257 microg(S) mminus3 (Fig 10e) with the largestvalues over the Mediterranean and Eastern Europe Emis-sion controls (Fig2a) reduced SO2minus

4 surface concentrations(Figs10f 11a and12a) by almost 50 in 2001ndash2005 Fig-ures10f and g show that emission reductions and meteoro-logical variability contribute ca 85 and 15 respectivelyto the overall differences between 2001ndash2005 and 1981ndash1985 indicating a small but significant role for meteorolog-ical variability in the SO2minus

4 signal Figure10h shows thatthe largest SO2minus

4 decreases occurred in Southern and North-Eastern Europe In Figs11a and12a we display seasonaldifferences in the SO2minus

4 response to emissions and meteoro-logical changes SFIX European winter (DJF) surface sulfatevariesplusmn30 compared to 1980 while in summer (JJA) con-centrations are between 0 and 20 larger than in 1980 Thedecline of European surface sulfur concentrations in SREF islarger in winter (50 ) than in summer (37 )

Measurements mostly confirm these model findings In-deed inter-annual seasonal anomalies of SO2minus

4 in winter(AppendixB) generally correlate well (R gt 05) at most sta-tions in Europe In summer the modeled inter-annual vari-ability is always underestimated (normalized standard devi-

ation of 03ndash09) most likely indicating an underestimatein the variability of precipitation scavenging in the modelover Europe Modeled and measured SO2minus

4 trends are ingood agreement in most European regions (Fig6) In winterthe observed declines of 002ndash007 microg(S) mminus3 yrminus1 are un-derestimated in NEU and EEU and overestimated in theother regions In summer the observed declines of 002ndash008 microg(S) mminus3 yrminus1 are overestimated in SEU and under-estimated in CEU WEU and EEU

512 North America

The calculated annual mean surface concentration of sul-fate over the NA region for the period 1981ndash1985 is065 microg(S) mminus3 Highest concentrations are found over theEastern and Southern US (Fig10i) NA emissions reduc-tions of 35 (Fig2b) reduced SO2minus

4 concentrations on av-erage by 018 microg(S) mminus3) and up to 1 microg(S) mminus3 over theEastern US (Fig10j) Meteorological variability results in asmall overall increase of 005 microg(S) mminus3 (Fig10k) Changesin emissions and meteorology can almost be combined lin-early (Fig 10l) The total decline is thus 011 microg(S) mminus3minus20 in winter andminus25 in summer between 1980ndash2005indicating a fairly low seasonal dependency

Like in Europe also in North America measured inter-annual seasonal variability is smaller than in our calcu-lations in winter and larger in summer (AppendixB)Observed winter downward trends are in the range of0ndash003 microg(S) mminus3 yrminus1 in reasonable agreement with therange of 001ndash006 microg(S) mminus3 yrminus1 calculated in the SREFsimulation In summer except for the Western US(WUS) observed SO2minus

4 trends range betweenminus008 andminus005 microg(S) mminus3 yrminus1 and the calculated trends decline onlybetween 003ndash004 microg(S) mminus3 yrminus1) (Fig6) We suspect thata poor representation of the seasonality of anthropogenic sul-fur emissions contributes to both the winter overestimate andthe summer underestimate of the SO2minus

4 trends

513 East Asia

Annual mean SO2minus

4 during 1981ndash1985 was 09 microg(S) mminus3

with higher concentrations over Eastern China Korea andin the continental outflow over the Yellow Sea (Fig10m)Growing anthropogenic sulfur emissions (60 over EA)produced an increase in regional annual mean SO2minus

4 concen-trations of 024 microg(S) mminus3 in the 2001ndash2005 period Largestincreases (practically a doubling of concentrations) werefound over Eastern China and Korea (Fig10n) The con-tribution of changing meteorology and natural emissions isrelatively small (001 microg(S) mminus3 or 15 Fig10o) and thecoupled effect of changing emissions and meteorology isdominated by the emissions perturbation (019 microg(S) mminus3Fig 10p)



Fig 10 As Fig5 but for SO2minus

4 surface concentrations

514 South Asia

Annual and regional average SO2minus

4 surface concentrationsof 070 microg(S) mminus3 (1981ndash1985) increased by on average039 microg(S) mminus3 and up to 1 microg(S) mminus3 over India due to the220 increase in anthropogenic sulfur emissions in 25 yrThe alternation of wet and dry seasons is greatly influencingthe seasonal SO2minus

4 concentration changes In winter (dry sea-son) following the emissions the SO2minus

4 concentrations in-crease two-fold In the wet season (JJA) we do not see a sig-nificant increase in SO2minus

4 concentrations and the variability isalmost completely dominated by the meteorology (Figs11dand12d) This indicates that frequent rainfall in the monsooncirculation keeps SO2minus

4 low regardless of increasing emis-sions In winter we calculated a ratio of 045 between SO2minus

4wet deposition and total SO2minus

4 production (both gas and liq-uid phases) while during summer months about 124 timesmore sulfur is deposited in the SA region than is produced

52 Variability of the global SO2minus

4 budget

We now analyze in more detail the processes that contributeto the variability of surface and column sulfate Figure13shows that inter-annual variability of the global SO2minus

4 burdenis largely determined by meteorology in contrast to the sur-

face SO2minus

4 changes discussed above In-cloud SO2 oxidationprocesses with O3 and H2O2 do not change significantly over1980ndash2005 in the SFIX simulation (472plusmn 04 Tg(S) yrminus1)while in the SREF case the trend is very similar to those ofthe emissions Interestingly the H2SO4 (and aerosol) pro-duction resulting from the SO2 reaction with OH (Fig13c)responds differently to meteorological and emission variabil-ity than in-cloud oxidation Globally it increased by 1 Tg(S)between 1980 and late 1990s (SFIX) and then decreasedagain after the year 2000 (see also Fig4e) The increaseof gas phase SO2minus

4 production (Fig13c) in the SREF simu-lation is even more striking in the context of overall declin-ing emissions These contrasting temporal trends can be ex-plained by the changes in the geographical distribution of theglobal emissions and the variation of the relative efficiencyof the oxidation pathways of SO2 in SREF which are givenin Table4 Thus the global increase in SO2minus

4 gaseous phaseproduction is disproportionally depending on the sulfur emis-sions over Asia as noted earlier by egUnger et al(2009)For instance in the EU region the SO2minus

4 burden increases by22times 10minus3 Tg(S) per Tg(S) emitted while this response ismore than a factor of two higher in SA Consequently de-spite a global decrease in sulfur emissions of 8 the globalburden is not significantly changing and the lifetime of SO2minus

4is slightly increasing by 5



Fig 11 Winter (DJF) anomalies of surface SO2minus

4 concentrations averaged over the selected regions (shown in Fig1) On the left y-axis

anomalies of SO2minus

4 surface concentrations for the period 1980ndash2005 are expressed as the ratios between seasonal means for each year andthe year 1980 The blue line represents the SREF simulation (changing meteorology and changing anthropogenic emissions) the red line theSFIX simulation (changing meteorology and fixed anthropogenic emissions at the level of 1980) while the gray area indicates the SREF-SFIX difference On the right y-axis the pink dashed line represents the changes ie the ratio between each year and 1980 of total annualsulfur emissions

Fig 12 As Fig11 for summer (JJA) anomalies of surface SO2minus

4 concentrations



Fig 13 Global tropospheric SO2minus

4 budget calculated for the period 1980ndash2005 for the SREF (blue) and SFIX (red) ECHAM5-HAMMOZ

simulations(a) total sulfur emissions(b) SO2minus

4 liquid phase production(c) SO2minus

4 gaseous phase production(d) surface deposition(e)

SO2minus

4 burden(f) lifetime

6 Variability of AOD and anthropogenic radiativeperturbation of aerosol and O3

The global annual average total aerosol optical depth (AOD)ranges between 0151 and 0167 during the period 1980ndash2005 (SREF) which is slightly higher than the range ofmodelmeasurement values (0127ndash0151) reported byKinneet al (2006) The monthly mean anomalies of total AOD(Fig 4f 1σ = 0007 or 43 ) are determined by variationsof natural aerosol emissions including biomass burning thechanges in anthropogenic SO2 emissions discussed aboveonly cause little differences in global AOD anomalies Theeffect of anthropogenic emissions is more evident at the re-gional scale Figure14a shows the AOD 5-yr average cal-culated for the period 1981ndash1985 with a global average of0155 The changes in anthropogenic emissions (Fig14b)decrease AOD over a large part of the Northern Hemispherein particular over Eastern Europe and they largely increaseAOD over East and South Asia The effect of meteorologyand natural emissions is smaller ranging betweenminus005 and01 (Fig14c) and it is almost linearly adding to the effect ofanthropogenic emissions (Fig14d)

In Fig 15 and Table5 we see that over EU the reduc-tions in anthropogenic emissions produced a 28 decreasein AOD and 14 over NA In EA and SA the increasingemissions and particularly sulfur emissions produced an in-

Table 4 Relationships in Tg(S) per Tg(S) emitted calculated forthe period 1980ndash2005 between sulfur emissions and SO2minus

4 burden

(B) SO2minus

4 production from SO2 in-cloud oxidation (In-cloud) and

SO2minus

4 gaseous phase production (Cond) over Europe (EU) NorthAmerica (NA) East Asia (EA) and South Asia (SA)

EU NA EA SAslope R2 slope R2 slope R2 slope R2

B (times 10minus3) 221 086 079 012 286 079 536 090In-cloud 031 097 038 093 025 079 018 090Cond 008 093 009 070 017 085 037 098

crease of AOD of 19 and 26 respectively The variabil-ity in AOD due to natural aerosol emissions and meteorologyis significant In SFIX the natural variability of AOD is upto 10 over EU 17 over NA 8 over EA and 13 over SA Interestingly over NA the resulting AOD in theSREF simulation does not show a large signal The same re-sults were qualitatively found also from satellite observations(Wang et al 2009) where the AOD decreased only over Eu-rope no significant trend was found for North America andit increased in Asia

In Table 5 we present an analysis of the radiative per-turbations due to aerosol and ozone comparing the periods



Table 5 Globally and regionally (EU NA EA and SA) averaged effect of changing anthropogenic emissions (SREF-SFIX) during the 5-yrperiods 1981ndash1985 and 2001ndash2005 on surface concentrations of O3 (ppbv) SO2minus

4 () and BC () total aerosol optical depth (AOD) ()

and total column O3 (DU) the total anthropogenic aerosol radiative perturbation at top of the atmosphere (RPTOAaer ) at surface (RPSURF

aer )(W mminus2) and in the atmosphere (RPATM

aer = (RPTOAaer minus RPSURF

aer ) the anthropogenic radiative perturbation of O3 (RPO3) correlations calcu-

lated over the entire period 1980ndash2005 between anthropogenic RPTOAaer and1AOD between anthropogenic RPSURF

aer and1AOD betweenRPATM

aer and1AOD between RPATMaer and1BC between anthropogenic RPO3 and1O3 at surface between anthropogenic RPO3 and1O3

column High correlation coefficients are highlighted in bold

GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus

4 [] minus10 minus36 minus25 27 591BC [] 0 minus43 minus34 30 70

1AOD [] 0 minus28 minus14 19 261O3 [DU] 118 135 154 285 299

RPTOAaer [W mminus2] 002 126 039 minus053 minus054

RPSURFaer [W mminus2] minus003 205 071 minus119 minus183

RPATMaer [W mminus2] 005 minus079 minus032 066 129

RPO3 [W mminus2] 005 005 006 012 015

slope R2 slope R2 slope R2 slope R2 slope R2

RPTOAaer [W mminus2] vs 1AOD minus1786 085 minus161 099 minus1699 097 minus1357 099 minus1384 099

RPSURFaer [W mminus2] vs 1AOD minus132 024 minus2671 099 minus2810 097 minus2895 099 minus4757 099

RPATMaer [W mminus2] vs 1AOD minus462 003 1061 093 1111 068 1538 097 3373 098

RPATMaer [W mminus2] vs 1BC [ ] 035 011 173 099 095 099 192 097 174 099

RPO3 [mW mminus2] vs 1O3 [ppbv] 428 091 352 065 513 049 467 094 360 097RPO3 [mW mminus2] vs 1O3 [DU] 408 099 364 099 442 099 441 099 491 099

1981ndash1985 and 2001ndash2005 We define the difference be-tween the instantaneous clear-sky total aerosol and all skyO3 RF of the SREF and SFIX simulations as the totalaerosol and O3 short-wave radiative perturbation due to an-thropogenic emissions and we will refer to them as RPaerand RPO3 respectively

61 Aerosol radiative perturbation

The instantaneous aerosol radiative forcing (RF) inECHAM5-HAMMOZ is diagnostically calculated from thedifference in the net radiative fluxes including and excludingaerosol (Stier et al 2007) For aerosol we focus on clearsky radiative forcing since unfortunately a coding error pre-vents us from evaluating all-sky forcing In Fig15we showtogether with the AOD (see before) the evolution of the nor-malized RPaer at the top-of-the-atmosphere (RPTOA

aer ) and atthe surface (and RPSurf

aer )For Europe the AOD change byminus28 (mainly due to

changes in removal of SO2minus

4 aerosol) leads to an increase ofRPTOA

aer of 126 W mminus2 and RPSurfaer of 205 The 14 AOD

reduction in NA corresponds to a RPTOAaer of 039 W mminus2 and

RPSurfaer of 071 W mminus2 In EA and SA AOD increased by

19 and 26 corresponding to a RPTOAaer of minus053 and

minus054 W mminus2 respectively while at surface we found RPaerof minus119 W mminus2 andminus183 W mminus2 The larger difference

between TOA and surface forcing in South Asia comparedto the other regions indicates a much larger contribution ofBC absorption in South Asia

Globally we found a significant correlation between theanthropogenic aerosol radiative perturbation at the top of theatmosphere with the percentage changes in AOD due to an-thropogenic emissions (Table5 R2

= 085 for RPTOAaer vs

1AOD) while there is no correlation between anthropogenicaerosol radiative perturbation and anthropogenic changes inAOD (1AOD) or BC (1BC) at the surface and in the at-mosphere Within the four selected regions the correlationbetween anthropogenic emission induced changes in AODand RPaer at the top-of-the-atmosphere surface and the at-mosphere is much higher (R2 gt 093 for all regions exceptfor RPATM

aer over NA Table5)We calculate a relatively constant RPTOA

aer betweenminus17 tominus13 W mminus2 per unit AOD around the world and a largerrange ofminus48 tominus26 W mminus2 per unit AOD for RPaer at thesurface The atmospheric absorption by aerosol RPaer (cal-culated from the difference of RP at TOA and RP at the sur-face) is around 10 W mminus2 in EU and NA 15 W mminus2 in EAand 34 W mminus2 in SA showing the importance of absorbingBC aerosols in determining surface and atmospheric forcingas confirmed by the high correlations of RPATM

aer with surfaceblack carbon levels



Fig 14 Maps of total aerosol optical depth (AOD) and the changes due to anthropogenic emissions and natural variability We show(a) the5-yr averages (1981ndash1985) of global AOD(b) the effect of anthropogenic emission changes in the period 2001ndash2005 on AOD calculated asthe difference between SREF and SFIX simulations(c) the natural variability of AOD which is due to natural emissions and meteorologyin the simulated 25 yr calculated as the difference between the 5-yr average periods (2001ndash2005) and (1981ndash1985) in the SFIX simulation(d) the combined effect of anthropogenic emissions and natural variability expressed as the difference between the 5-yr average period(2001ndash2005)ndash(1981ndash1985) in the SREF simulation

62 Ozone radiative perturbation

For convenience and completeness we also present in thissection a calculation of O3 RP diagnosed using ECHAM5-HAMMOZ O3 columns in combination with all-sky radia-tive forcing efficiencies provided by D Stevenson (personalcommunication 2008 for a further discussionGauss et al2006) Table5 and Fig5 show that O3 total column and sur-face concentrations increased by 154 DU (158 ppbv) overNA 135 DU (128 ppbv) over EU 285 (413 ppbv) overEA and 299 DU (512 ppbv) over SA in the period 1980ndash2005 The RPO3 over the different regions reflect the totalcolumn O3 changes 005 W mminus2 over EU 006 W mminus2 overNA 012 W mminus2 over EA and 015 W mminus2 over SA As ex-pected the spatial correlation of O3 columns and radiativeperturbations is nearly 1 also the surface O3 concentrationsin EA and SA correlate nearly as well with the radiative per-turbation The lower correlations in EU and NA suggest thata substantial fraction of the ozone production from emissionsin NA and EU takes place above the boundary layer (Table5)

7 Summary and conclusions

We used the coupled aerosol-chemistry-climate generalcirculation model ECHAM5-HAMMOZ constrained with25 yr of meteorological data from ECMWF and a compila-tion of recent emission inventories to evaluate the responseof atmospheric concentrations aerosol optical depth and ra-diative perturbations to anthropogenic emission changes andnatural variability over the period 1980ndash2005 The focus ofour study was on O3 and SO2minus

4 for which most long-termsurface observations in the period 1980ndash2005 were availableThe main findings are summarized in the following points

ndash We compiled a gridded database of anthropogenic COVOC NOx SO2 BC and OC emissions utilizingreported regional emission trends Globally anthro-pogenic NOx and OC emissions increased by 10 while sulfur emissions decreased by 10 from 1980to 2005 Regional emission changes were larger egall components decreased by 10ndash50 in North Amer-ica and Europe but increased between 40ndash220 in Eastand South Asia

ndash Natural emissions were calculated on-line and de-pended on inter-annual changes in meteorology We



Fig 15 Annual anomalies of total aerosol optical depth (AOD) averaged over the selected regions (shown in Fig1) On the left y-axisanomalies of AOD for the period 1980ndash2005 are expressed as the ratios between annual means for each year and 1980 The blue linerepresents the SREF simulation (changing meteorology and changing anthropogenic emissions) the red line the SFIX simulation (changingmeteorology and fixed anthropogenic emissions at the level of 1980) while the gray area indicates the SREF-SFIX difference On the righty-axis the pink and green dashed lines represent the clear-sky aerosol anthropogenic radiative perturbation at the top of the atmosphere(RPTOA

aer ) and at the surface (RPSurfaer ) respectively

found a rather small global inter-annual variability forbiogenic VOCs emissions (3 of the multi-annual av-erage) DMS (1 ) and sea salt aerosols (2 ) Alarger variability was found for lightning NOx emis-sions (5 ) and mineral dust (10 ) Generally wecould not identify a clear trend for natural emissionsexcept for a small decreasing trend of lightning NOxemissions (0017plusmn 0007 Tg(N) yrminus1)

ndash The impacts of natural variability (including meteorol-ogy biogenic VOC emissions biomass burning emis-sions and lightning) on surface ozone tropospheric OHand AOD are often larger than the impacts due to an-thropogenic emission changes Two important meteo-rological drivers for atmospheric composition change ndashhumidity and temperature ndash are strongly correlated Themoderate correlation (R = 043) of the global mean an-nual surface ozone concentration and surface tempera-ture suggests important contributions of other processesto the tropospheric ozone budget

ndash The set-up of this study does not allow to specifically in-vestigate the contribution of meteorological parametersto changes in the chemical composition except in a fewcases where major events such as the 1997ndash1998 ENSOlead to strong enhancements of surface ozone and AOD

beyond the response that is expected from the changesof natural emissions

ndash Global surface O3 increased in 25 yr on average by048 ppbv due to anthropogenic emissions but 75 ofthe inter-annual variability of the multi-annual monthlysurface ozone was related to natural variations

ndash The seasonally averaged modeled and measured surfaceozone concentrations are reasonably well correlated fora large number of North American and European sta-tions However this is not always an indication thatthe observed trends are correctly reproduced A re-gional analysis suggests that changing anthropogenicemissions increased O3 on average by 08 ppbv in Eu-rope with large differences between southern and otherparts of Europe which is generally a larger responsecompared to the HTAP study (Fiore et al 2009) Mea-surements qualitatively confirm these trends in EuropeHowever the observed trends are up to a factor of 3(03ndash05 ppbv yrminus1) larger than those that are calculatedespecially in winter while in summer the small nega-tive trends simulated be the model were not confirmedby the measurements In North America anthropogenicemissions slightly increased O3 by 03 ppbv on aver-age even though decreases between 1 and 2 ppbv were



calculated over a large fraction of the US Annual av-erages hid some seasonal discrepancies between modelresults and observed trends In East Asia we computedan increase of surface O3 by 41 ppbv 24 ppbv from an-thropogenic emissions and 16 ppbv contribution frommeteorological changes The scarce long-term obser-vational datasets (mostly Japanese stations) do not con-tradict these computed trends In 25 yr annual meanO3 concentrations increased by 51 ppbv over SA withapproximately 75 related to increasing anthropogenicemissions Confidence in the calculations of O3 and O3trends is low since the few available measurements sug-gest much lower O3 over India

ndash The tropospheric O3 budget and variability agree wellwith earlier studies byStevenson et al(2006) andHessand Mahowald(2009) During 1980ndash2005 we calculatean intensification of tropospheric O3 chemistry leadingto an increase of global tropospheric ozone by 3 anda decreasing O3 lifetime by 4 Comparing similar re-analysis studies such as our study andHess and Ma-howald (2009) it is shown that the choice of the re-analysis product and nudging method has strong im-pacts on variability of O3 and other components Forinstance the comparison of our study with an alter-native data assimilation technique presented byHessand Mahowald(2009) ie prescribing sea-surface-temperatures (often used in climate modeling time sliceexperiments) resulted in substantially less agreementEven if the main large-scale meteorological patterns arecaptured by prescribing SSTs in a GCM simulationthe modeled dynamics in a GCM constrained by a fullnudging methodology should be closer to the observedmeteorology Therefore we suggest that it is also moreconsistent when comparing simulated chemical compo-sition with observations However we have not per-formed the simulations with prescribed SSTs ourselvesand the forthcoming ACC-Hindcast and ACC-MIP pro-grams should shed the light on this matter

ndash Global OH which determines the oxidation capacity ofthe atmosphere decreased byminus027 yrminus1 due to nat-ural variability of which lightning was the most im-portant contributor Anthropogenic emissions changescaused an opposite trend of 025 yrminus1 thus nearlybalancing the natural emission trend Calculated inter-annual variability is in the order of 16 in disagree-ment with the earlier study ofPrinn et al(2005) of largeinter-annual fluctuations on the order of 10 but closerto the estimates ofDentener et al(2003) (18 ) andMontzka et al(2011) (2 )

ndash The global inter-annual variability of surface SO2minus

4(10 ) is strongly determined by regional variations ofemissions Comparison of computed trends with mea-surements in Europe and North America showed in gen-

eral good agreement Seasonal trend analysis gave ad-ditional information For instance in Europe mea-surements suggest similar downward trends of 005ndash01 microg(S) mminus3 yrminus1 in both summer and winter whilesimulated surface SO2minus

4 downward trends are somewhatstronger in winter than in summer In North Americain winter the model reproduces the observed SO2minus

4 de-clines well in some but not all regions In summercomputed trends are generally underestimated by up to50 We expect that a misrepresentation of temporalvariations of emissions together with non-linear oxida-tion chemistry could play a role in these winter-summerdifferences In East and South Asia the model resultssuggest increases of surface SO2minus

4 by ca 30 howeverto our knowledge no datasets are available that couldcorroborate these results

ndash Trend and variability of sulfate columns are very dif-ferent from surface SO2minus

4 Despite a global decreaseof SO2minus

4 emissions from 1980 to 2005 global sulfateburdens were not significantly changing due to a south-ward shift of SO2 emissions which determines a moreefficient production and longer lifetime of SO2minus

4

ndash Globally surface SO2minus

4 concentration decreases by ca01 microg(S) mminus3 while the global AOD increases by ca001 (or ca 5 ) the latter driven by variability of dustand sea salt emissions Regionally anthropogenic emis-sions changes are more visible we calculate a signif-icant decline of AOD over Europe (28 ) a relativelyconstant AOD over North America (it decreased onlyby 14 in the last 5 yr) and a strongly increase overEast (19 ) and South Asia (26 ) These results dif-fer substantially fromStreets et al(2009) Since theemission inventory used in this study and the one byStreets et al(2009) are very similar we expect thatour explicit treatment of aerosol chemistry and micro-physics lead to very different results than the scalingof AOD with emission trends used byStreets et al(2009) Our analysis suggests that the impact of anthro-pogenic emission changes on radiative perturbations istypically larger and more regional at the surface thanat the top-of-the atmosphere with a strong atmosphericwarming by BC aerosol specifically over South AsiaThe global top-of-the-atmosphere radiative perturbationfollows more closely aerosol optical depth reflectinglarge-scale SO2minus

4 dispersion patterns Nevertheless ourstudy corroborates an important role for aerosol in ex-plaining the observed changes in surface radiation overEurope (Wild 2009 Wang et al 2009) Our study isalso qualitatively consistent with the reported world-wide visibility decline (Wang et al 2009) In Europeour calculated AOD reductions are consistent with im-proving visibility (Wang et al 2009) and increasing sur-face radiation (Wild 2009) O3 radiative perturbations



(not including feedbacks of CH4) are regionally muchsmaller than aerosol RPs but globally equal to or largerthan aerosol RPs

8 Outlook

Our re-analysis study showed that several of the overallprocesses determining the variability and trend of O3 andaerosols are qualitatively understood- but also that many ofthe details are not well included As such it gives some trustin our ability to predict the future impacts of aerosol and re-active gases on climate but it also shows that many modelparameterizations need further improvement for more reli-able predictions For example the variability due to strato-spheretroposphere O3 fluxes is not included in our study andthe description of lightning and biogenic emissions are char-acterized by large uncertainties We have seen that aerosolvariability and in particular sulfate is mainly driven by emis-sions Improving the description of anthropogenic emissioninventories natural emission parameterizations and the de-scription of secondary formation of aerosols (eg secondaryorganic aerosols (SOA) which are not included in this study)may improve our understanding of aerosol variability It isour feeling that comparisons focussing on 1 or 2 yr of datawhile useful by itself may mask issues with compensatingerrors and wrong sensitivities Re-analysis studies are use-ful tools to unmask these model deficiencies The analysis ofsummer and winter differences in trends and variability wasparticularly insightful in our study since it highlights ourlevel of understanding of the relative importance of chemi-cal and meteorological processes The separate analysis ofthe influence of meteorology and anthropogenic emissionschanges is of direct importance for the understanding and at-tribution of observed trends to emission controls The anal-ysis of differences in regional patterns again highlights ourunderstanding of different processes

The ECHAM5-HAMMOZ model is like most other cli-mate models continuously being improved For instance theoverestimate of surface O3 in many world regions or the poorrepresentation in a tropospheric model of the stratosphere-troposphere exchange (STE) fluxes (as reported by this studyRast et al 2011andSchultz et al 2007) reduces our trustin our trend analysis and should be urgently addressed Par-ticipation in model inter-comparisons and comparison ofmodel results to intensive measurement campaigns of mul-tiple components may help to identify deficiencies in themodel process descriptions Improvement of parameteriza-tions and model resolution will improve in the long run themodel performances Continued efforts are needed to im-prove our knowledge of anthropogenic and natural emissionsin the past decades which will help to understand better re-cent trends and variability of ozone and aerosols New re-analysis products such as the re-analyses from ECMWF andNCEP are frequently becoming available and should give

improved constraints for the meteorological conditions Itis of outmost importance that the few long-term measure-ment datasets are being continued and that long-term com-mitments are also implemented in regions outside of Eu-rope and North America Other datasets such as AOD fromAERONET and various quality controlled satellite datasetsmay in the future become useful for trend analysis

Chemical re-analyses are computationally expensive andtime consuming and cannot be easily performed for ev-ery new model version However an updated chemical re-analysis every couple of years following major model andre-analysis product upgrades seems highly recommendableThese studies should preferentially be performed in closecollaboration with other modeling groups which allow shar-ing data and analysis methods A better understanding ofthe chemical climate of the past is particularly relevant in thelight of the continued effort to abate the negative impacts ofair pollution and the expected impacts of these controls onclimate (Arneth et al 2009 Raes and Seinfeld 2009)

Appendix A

ECHAM5-HAMMOZ model description

A1 The ECHAM5 GCM

ECHAM5 is a spectral GCM developed at the Max PlanckInstitute for Meteorology (Roeckner et al 2003 2006Hagemann et al 2006) based on the numerical weather pre-diction model of the European Centre for Medium-RangeWeather Forecast (ECMWF) The prognostic variables ofthe model are vorticity divergence temperature and surfacepressure and are represented in the spectral space The mul-tidimensional flux-form semi-Lagrangian transport schemefrom Lin and Rood(1996) is used for water vapor cloudrelated variables and chemical tracers Stratiform cloudsare described by a microphysical cloud scheme (Lohmannand Roeckner 1996) with a prognostic statistical cloud coverscheme (Tompkins 2002) Cumulus convection is param-eterized with the mass flux scheme ofTiedtke (1989) withmodifications fromNordeng(1994) The radiative trans-fer calculation considers vertical profiles of the greenhousegases (eg CO2 O3 CH4) aerosols as well as the cloudwater and ice The shortwave radiative transfer followsCagnazzo et al(2007) considering 6 spectral bands Forthis part of the spectrum cloud optical properties are cal-culated on the basis of Mie calculations using idealized sizedistributions for both cloud droplets and ice crystals (Rockelet al 1991) The long-wave radiative transfer scheme is im-plemented according toMlawer et al(1997) andMorcretteet al(1998) and considers 16 spectral bands The cloud op-tical properties in the long-wave spectrum are parameterizedas a function of the effective radius (Roeckner et al 2003Ebert and Curry 1992)



A2 Gas-phase chemistry module MOZ

The MOZ chemical scheme has been adopted from theMOZART-2 model (Horowitz et al 2003) and includes 63transported tracers and 168 reactions to represent the NOx-HOx-hydrocarbons chemistry The sulfur chemistry includesoxidation of SO2 by OH and DMS oxidation by OH and NO3(Feichter et al 1996) Stratospheric O3 concentrations areprescribed as monthly mean zonal climatology derived fromobservations (Logan 1999 Randel et al 1998) These con-centrations are fixed at the topmost two model levels (pres-sures of 30 hPa and above) At other model levels above thetropopause the concentrations are relaxed towards these val-ues with a relaxation time of 10 days followingHorowitzet al(2003) The photolysis frequencies are calculated withthe algorithm Fast-J2 (Bian and Prather 2002) consideringthe calculated optical properties of aerosols and clouds Therates of heterogeneous reactions involving N2O5 NO3 NO2HO2 SO2 HNO3 and O3 are calculated based on the modelcalculated aerosol surface area A more detailed descriptionof the tropospheric chemistry module MOZ and the couplingbetween the gas phase chemistry and the aerosols is given inPozzoli et al(2008a)

A3 Aerosol module HAM

The tropospheric aerosol module HAM (Stier et al 2005)predicts the size distribution and composition of internally-and externally-mixed aerosol particles The microphysicalcore of HAM M7 (Vignati et al 2004) treats the aerosoldynamics and thermodynamics in the framework of modalparticle size distribution the 7 log-normal modes are char-acterized by three moments including median radius num-ber of particles and a fixed standard deviation (159 forfine particles and 200 for coarse particles Four modesare considered as hydrophilic aerosols composed of sulfate(SU) organic (OC) and black carbon (BC) mineral dust(DU) and sea salt (SS) nucleation (NS) (r le 0005 microm)Aitken (KS) (0005 micromlt r le 005 microm) accumulation (AS)(005 micromlt r le 05 microm) and coarse (CS) (r gt 05 microm) (wherer is the number median radius) Note that in HAM the nucle-ation mode is entirely constituted of sulfate aerosols Threeadditional modes are considered as hydrophobic aerosolscomposed of BC and OC in the Aitken mode (KI) andof mineral dust in the accumulation (AI) and coarse (CI)modes Wavelength-dependent aerosol optical properties(single scattering albedo extinction cross section and asym-metry factor) were pre-calculated explicitly using Mie theory(Toon and Ackerman 1981) and archived in a look-up-tablefor a wide range of aerosol size distributions and refractiveindices HAM is directly coupled to the cloud microphysicsscheme allowing consistent calculations of the aerosol indi-rect effects (Lohmann et al 2007)

A4 Gas and aerosol deposition

Gas and aerosol dry deposition follows the scheme ofGanzeveld and Lelieveld(1995) andGanzeveld et al(19982006) coupling the Wesely resistance approach with land-cover data from ECHAM5 Wet deposition is based onStieret al (2005) including scavenging of aerosol particles bystratiform and convective clouds and below cloud scaveng-ing The scavenging parameters for aerosol particles aremode-specific with lower values for hydrophobic (externally-mixed) modes For gases the partitioning between the airand the cloud water is calculated based on Henryrsquos law andcloud water content

Appendix B

O3 and SO2minus

4 measurement comparisons

Figure B1 provides an example of the ECHAM5-HAMMOZcapabilities to reproduce surface O3 variability We com-pared the observed and calculated monthly mean anomaliesof surface O3 concentrations at 6 remote stations Despitethe fact that ECHAM5-HAMMOZ overestimates monthlymean O3 concentrations by up to 15 ppbv and that the ob-served and calculated anomalies are not well correlated forall the considered stations the model can capture the ob-served range of O3 variability at very different locations inthe world The anomalies at these remote stations are drivenby the meteorological and natural emission variability there-fore it is hard to distinguish the effect of anthropogenic emis-sions comparing the SREF and SFIX simulations In Eu-rope the variability at the stations of Mace Head and Ho-henpeissenberg is well captured by the model while loweragreement is found at Zingst The observed O3 trend atMace Head (019 ppbv yrminus1) is not captured by the model(minus004 ppbv yrminus1) mainly because of the largest anomaliescalculated in the 1990s We obtained similar results also atZingst A better agreement is found at Hohenpeissenberg032 ppbv yrminus1 observed and 045 ppbv yrminus1 calculated O3trends At Barrow (Alaska) and Mauna Loa (Hawaii) ex-cluding the 1990s the variability is qualitatively well cap-tured while in the Southern Hemisphere at Samoa the vari-ability is better captured in the 1980s

Long measurement records of O3 and SO2minus

4 are mainlyavailable in Europe North America and at a few stationsin East Asia from the following networks the EuropeanMonitoring and Evaluation Programme (EMEPhttpwwwemepint) the Clean Air Status and Trends Network (CAST-NET httpwwwepagovcastnet) the World Data Centrefor Greenhouse Gases (WDCGGhttpgawkishougojpwdcgg) O3 measures are available starting from the year1990 for EMEP and at a few stations back to 1987 in theCASTNET and WDCGG networks For this reason we se-lected only the stations that have at least 10 yr records in the



1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head

SREFSFIXObservations

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa

Fig B1 Time evolution (1980ndash2005) of surface ozone monthly mean anomalies (12-month running mean) at a variety of remote sitesObservations are in red ECHAM5-HAMMOZ results are in blue (SREF simulation) and green (SFIX simulation)

period 1990ndash2005 for both winter and summer A total of 81stations were selected over Europe from EMEP 48 stationsover North America from CASTNET and 25 stations fromWDCGG The average record length among all selected sta-tions is of 14 yr For SO2minus

4 measures we selected 62 stationsfrom EMEP and 43 stations from CASTNET The averagerecord length among all selected stations is of 13 yr

The location of each selected station is plotted in Fig1 forboth O3 and SO2minus

4 measurements Each symbol represents ameasuring network (EMEP triangle CASTNET diamondWDCGG square) and each color a geographical subregionSimilarly to Fiore et al(2009) we grouped stations in Cen-tral Europe (CEU which includes mainly the stations of Ger-many Austria Switzerland The Netherlands and Belgium)and Southern Europe (SEU stations below 45 N and in theMediterranean basin) but we also included in our study agroup of stations for Northern Europe (NEU which includesthe Scandinavian countries) Eastern Europe (EEU which in-cludes stations east of 17 E) Western Europe (WEU UKand Ireland) Over North America we grouped the sites in4 regions Northeast (NEUS) Great Lakes (GLUS) Mid-Atlantic (MAUS) and Southwest (SUS) based onLehmanet al (2004) representing chemically coherent receptor re-gions for O3 air pollution and an additional region for West-ern US (WUS)

For each station we calculated the seasonal mean anoma-lies (DJF and JJA) by subtracting the multi-year averageseasonal means from each annual seasonal mean Thesame calculations were applied to the values extracted fromECHAM5-HAMMOZ SREF simulation and the observedand calculated records were compared The correlation coef-ficients between observed and calculated O3 DJF anomalies

is larger than 05 for 44 39 and 36 of the selectedEMEP CASTNET and WDCGG stations respectively Theagreement (R ge 05) between observed and calculated O3seasonal anomalies is improving in summer months with52 for EMEP 66 for CASTNET and 52 for WDCGGFor SO2minus

4 the correlation between observed and calculatedanomalies is larger than 05 for 56 (DJF) and 74 (JJA)of the EMEP selected stations In 65 of the selected CAST-NET stations the correlation between observed and calcu-lated anomalies is larger than 05 both in winter and sum-mer

The comparisons between model results and observationsare synthesized in so-called Taylor diagrams displaying theinter-annual correlation and normalized standard deviation(ratio between the standard deviations of the calculated val-ues and of the observations) It can be shown that the dis-tance of each point to the black dot (10) is a measure of theRMS error (Fig B2) Winter (DJF) O3 inter-annual anoma-lies are relatively well represented by the model in WesternEurope (WEU) Central Europe (CEU) and Northern Europe(NEU) with most correlation coefficients between 05ndash09but generally underestimated standard deviations A loweragreement (R lt 05 standard deviationlt 05) is generallyfound for the stations in North America except for the sta-tions over GLUS and MAUS These winter differences indi-cate that some drivers of wintertime anomalies (such as long-range transport- or stratosphere-troposphere exchange of O3)may not be sufficiently strongly included in the model Thesummer (JJA) O3 anomalies are in general better representedby the model The agreement of the modeled standard de-viation with measurements is increasing compared to winteranomalies A significant improvement compared to winter



Fig B2 Taylor diagrams comparing the ECHAM5-HAMMOZ SREF simulation with EMEP CASTNET and WDCGG observations of O3seasonal means(a) DJF and(b) JJA) and SO2minus

4 seasonal means(c) DJF and(d) JJA The black dot is used as a reference to which simulatedfields are compared Continuous grey lines show iso-contours of skill score Stations are grouped by regions as in Fig1 Northern Europe(NEU) Central Europe (CEU) Western Europe (WEU) Eastern Europe (EEU) Southern Europe (SEU) Western US (WUS) North-EasternUS (NEUS) Mid-Atlantic US (MAUS) Great lakes US (GLUS) Southern US (SUS)

anomalies is found over North American stations (SEUSNEUS and MAUS) while the CEU stations have low nor-malized standard deviation even if correlation coefficientsare above 06 Interestingly while the European inter-annualvariability of the summertime ozone remains underestimated(normalized standard deviation around 05) the opposite istrue for North American variability indicating a too largeinter-annual variability of chemical O3 production perhapscaused by too large contributions of natural O3 precursorsSO2minus

4 winter anomalies are in general reasonably well cap-tured in winter with correlationR gt 05 at most stations andthe magnitude of the inter-annual variations In general wefound a much better agreement in summer than in winter be-tween calculated and observed SO2minus

4 variability with inter-annual coefficients generally larger than 05 In strong con-trast with the winter season now the inter-annual variabilityis always underestimated (normalized standard deviation of03ndash09) indicating an underestimate in the model of chemi-cal production variability or removal efficiency It is unlikelythat anthropogenic emissions (the dominant emissions) vari-ability was causing this lack of variability

Tables B1 and B2 list the observed and calculated seasonaltrends of O3 and SO2minus

4 surface concentrations in the Euro-pean and North American regions as defined before In win-ter we found statistically significant increasing O3 trends (p-valuelt 005) in all European regions and in 3 North Amer-ican regions (WUS NEUS and MAUS) see Table B1 TheSREF model simulation could capture significant trends onlyover Central Europe (CEU) and Western Europe (WEU)We did not find significant trends for the SFIX simulationwhich may indicate no O3 trends over Europe due to naturalvariability In 2 North American regions we found signifi-cant decreasing trends (WUS and NEUS) in contrast withthe observed increasing trends We must note that in these2 regions we also observed a decreasing trend due to nat-ural variability (TSFIX) which may be too strongly repre-sented in our model In summer the observed decreasing O3trends over Europe are not significant while the calculatedtrends show a significant decrease (TSREF NEU WEU andEEU) which is partially due to a natural trend (TSFIX NEUand EEU) In North America the observations show an in-creasing trend in WUS and decreasing trends in all other



Table B1 Observed and simulated trends of surface O3 seasonal anomalies for European (EU) and North American (NA) stations groupedas in Fig1 (Northern Europe (NEU) Central Europe (CEU) Western Europe (WEU) Eastern Europe (EEU) Western US (WUS) North-Eastern US (NEUS) Mid-Atlantic US (MAUS) Great lakes US (GLUS) Southern US (SUS)) The number of stations (NSTA) for eachregions is listed in the second column The seasonal trends are listed separately for winter (DJF) and summer (JJA) Seasonal trends(ppbv yrminus1) are calculated for observations (TOBS) SREF and SFIX simulations (TSREFandTSFIX) as linear fitting of the median surfaceO3 anomalies of each group of stations (the 95 confidence interval is also shown for each calculated trend) The statistically significanttrends (p-valuelt 005) are highlighted in bold The correlation coefficient between median observed and simulated (SREF) anomalies arelisted (ROBSSREF) The statistically significant correlation coefficients (p-valuelt 005) are highlighted in bold

O3 Stations Winter anomalies (DJF) Summer anomalies (JJA)

REGION NSTA TOBS TSREF TSFIX ROBSSREF TOBS TSREF TSFIX ROBSSREF

NEU 20 027plusmn 018 005plusmn 023 minus008plusmn 026 049 minus000plusmn 022 minus044plusmn 026 minus029plusmn 027 036CEU 54 044plusmn 015 021plusmn 040 006plusmn 040 046 004plusmn 032 minus011plusmn 030 minus000plusmn 029 073WEU 16 042plusmn 036 040plusmn 055 021plusmn 056 093 minus010plusmn 023 minus015plusmn 028 minus011plusmn 028 062EEU 8 034plusmn 038 006plusmn 027 minus009plusmn 028 007 minus002plusmn 025 minus036plusmn 036 minus022plusmn 034 071

WUS 7 012plusmn 021 minus008plusmn 011 minus012plusmn 012 026 041plusmn 030 011plusmn 021 021plusmn 022 080NEUS 11 022plusmn 013 minus010plusmn 017 minus017plusmn 015 003 minus027plusmn 028 003plusmn 047 014plusmn 049 055MAUS 17 011plusmn 018 minus002plusmn 023 minus007plusmn 026 066 minus040plusmn 037 009plusmn 081 019plusmn 083 068GLUS 14 004plusmn 018 005plusmn 019 minus002plusmn 020 082 minus019plusmn 029 020plusmn 047 034plusmn 049 043SUS 4 008plusmn 020 minus008plusmn 026 minus006plusmn 027 050 minus025plusmn 048 029plusmn 084 040plusmn 083 064

Table B2 Observed and simulated trends of surface SO2minus

4 seasonal anomalies for European (EU) and North American (NA) stationsgrouped as in Fig1 (Northern Europe (NEU) Central Europe (CEU) Western Europe (WEU) Eastern Europe (EEU) Southern Europe(SEU) Western US (WUS) North-Eastern US (NEUS) Mid-Atlantic US (MAUS) Great lakes US (GLUS) Southern US (SUS)) Thenumber of stations (NSTA) for each regions is listed in the second column The seasonal trends are listed separately for winter (DJF) andsummer (JJA) Seasonal trends (microg(S) mminus3 yrminus1) are calculated for observations (TOBS) SREF and SFIX simulations (TSREFandTSFIX) aslinear fitting of the median surface SO2minus

4 anomalies of each group of stations (the 95 confidence interval is also shown for each calculatedtrend) The statistically significant trends (p-valuelt 005) are highlighted in bold The correlation coefficient between median observed andsimulated (SREF) anomalies are listed (ROBSSREF) The statistically significant correlation coefficients (p-valuelt 005) are highlighted inbold

SO2minus

4 Stations Winter anomalies (DJF) Summer anomalies (JJA)


NEU 18 minus003plusmn 002 minus001plusmn 004 002plusmn 008 059 minus003plusmn 001 minus003plusmn 001 minus001plusmn 002 088CEU 18 minus004plusmn 003 minus010plusmn 008 minus006plusmn 014 067 minus006plusmn 002 minus004plusmn 002 000plusmn 002 079WEU 10 minus005plusmn 003 minus008plusmn 005 minus008plusmn 010 083 minus004plusmn 002 minus002plusmn 002 001plusmn 003 075EEU 11 minus007plusmn 004 minus001plusmn 012 005plusmn 018 028 minus008plusmn 002 minus004plusmn 002 minus001plusmn 002 078SEU 5 minus002plusmn 002 minus006plusmn 005 minus003plusmn 008 078 minus002plusmn 002 minus005plusmn 002 minus002plusmn 003 075

WUS 6 minus000plusmn 000 minus001plusmn 001 minus000plusmn 001 052 minus000plusmn 000 minus000plusmn 001 001plusmn 001 minus011NEUS 9 minus003plusmn 001 minus004plusmn 003 minus001plusmn 005 029 minus008plusmn 003 minus003plusmn 002 002plusmn 003 052MAUS 14 minus002plusmn 001 minus006plusmn 002 minus002plusmn 004 057 minus008plusmn 003 minus004plusmn 002 000plusmn 002 073GLUS 10 minus002plusmn 002 minus004plusmn 003 minus001plusmn 006 011 minus008plusmn 004 minus003plusmn 002 001plusmn 002 041SUS 4 minus002plusmn 002 minus003plusmn 002 minus000plusmn 002 minus005 minus005plusmn 004 minus003plusmn 003 000plusmn 004 037

regions All the observed trends are statistically significantThe model could reproduce a significant positive trend onlyover WUS (which seems to be determined by natural vari-ability TSFIX gt TSREF) In all other North American regionswe found increasing trends but not significant Neverthe-less the correlation coefficients between observed and sim-ulated anomalies are better in summer and for both Europeand North America

SO2minus

4 trends are in general better represented by themodel Both in Europe and North America the observationsshow decreasing SO2minus

4 trends (Table B2) both in winter andsummer The trends are statistically significant in all Eu-ropean and North American subregions both in winter andsummer In North America (except WUS) the observed de-creasing trends are slightly higher in summer (fromminus005andminus008 microg(S) mminus3 yrminus1) than in winter (fromminus002 and



Fig B3 Annual trends of O3 and SO2minus

4 surface concentrationsThe trends are calculated as linear fitting of the annual mean sur-face O3 and SO2minus

4 anomalies for each grid box of the model simu-lation SREF over Europe and North America The grid boxes withstatistically not significant (p-valuegt 005) trends are displayed ingrey The colored circles represent the observed trends for EMEPand CASTNET measuring stations over Europe and North Amer-ica respectively Only the stations with statistically significant (p-valuelt005) trends are plotted

minus003 microg(S) mminus3 yrminus1) The calculated trends (TSREF) are ingeneral overestimated in winter and underestimated in sum-mer We must note that a seasonality in anthropogenic sulfuremissions was introduced only over Europe (30 higher inwinter and 30 lower in summer compared to annual mean)while in the rest of the world annual mean sulfur emissionswere provided

In Fig B3 we show a comparison between the observedand calculated (SREF) annual trends of O3 and SO2minus

4 forthe European (EMEP) and the North American (CASTNET)measuring stations The grey areas represent the grid boxesof the model where trends are not statistically significant Wealso excluded from the plot the stations where trends are notsignificant Over Europe the observed O3 annual trends areincreasing while the model does not show a significant O3trend for almost all of Europe In the model statistically sig-nificant decreasing trends are found over the Mediterraneanand in part of Scandinavia and the Baltic Sea In North Amer-ica both the observed and modeled trends are mainly notstatistically significant Decreasing SO2minus

4 annual trends arefound over Europe and North America with a general goodagreement between the observations from single stations andthe model results

AcknowledgementsThis work has received partial funding fromthe European Communityrsquos Seventh Framework Programme(FP7) in the project PEGASOS (grant agreement 265148) Wegreatly acknowledge Sebastian Rast at Max Planck Institute forMeteorology Hamburg for the scientific and technical supportWe would like also to thank the Deutsches Klimarechenzentrum(DKRZ) and the Forschungszentrum Julich for the computingresources and technical support We would like to thank the EMEPWDCGG and CASTNET networks for providing ozone and sulfatemeasurements over Europe and North America

Edited by Y Balkanski

References

Andres R and Kasgnoc A A time-averaged inventory of sub-aerial volcanic sulfur emissions J Geophys Res-Atmos 10325251ndash25261 1998

Arneth A Unger N Kulmala M and Andreae M O Cleanthe Air Heat the Planet Science 326 672ndash673httpwwwsciencemagorgcontent3265953672short 2009

Auvray M Bey I Llull E Schultz M G and Rast S Amodel investigation of tropospheric ozone chemical tendenciesin long-range transported pollution plumes J Geophys Res112 D05304doi1010292006JD007137 2007

Berglen T Myhre G Isaksen I Vestreng V and Smith SSulphate trends in Europe Are we able to model the recentobserved decrease Tellus B 59 773ndash786httpwwwscopuscominwardrecordurleid=2-s20-34547919247amppartnerID=40ampmd5=b70fa1dd6e13861b2282403bf2239728 2007

Bian H and Prather M Fast-J2 Accurate simulation of strato-spheric photolysis in global chemical models J Atmos Chem41 281ndash296 2002

Bond T C Streets D G Yarber K F Nelson S M Woo J-H and Klimont Z A technology-based global inventory ofblack and organic carbon emissions from combustion J Geo-phys Res 109 D14203doi1010292003JD003697 2004

Cagnazzo C Manzini E Giorgetta M A Forster P M De Fand Morcrette J J Impact of an improved shortwave radia-tion scheme in the MAECHAM5 General Circulation Model At-mos Chem Phys 7 2503ndash2515doi105194acp-7-2503-20072007

Chan E and Vet R J Baseline levels and trends of ground levelozone in Canada and the United States Atmos Chem Phys 108629ndash8647doi105194acp-10-8629-2010 2010

Cheng T Peng Y Feichter J and Tegen I An improve-ment on the dust emission scheme in the global aerosol-climatemodel ECHAM5-HAM Atmos Chem Phys 8 1105ndash1117doi105194acp-8-1105-2008 2008

Collins W D Bitz C M Blackmon M L Bonan G BBretherton C S Carton J A Chang P Doney S C HackJ J Henderson T B Kiehl J T Large W G McKennaD S Santer B D and Smith R D The Community ClimateSystem Model Version 3 (CCSM3) J Climate 19 2122ndash2143doi101175JCLI37611 2006

Cui J Pandey Deolal S Sprenger M Henne S StaehelinJ Steinbacher M and Nedelec P Free tropospheric ozonechanges over Europe as observed at Jungfraujoch (19902008)



An analysis based on backward trajectories J Geophys Res116 D10304doi1010292010JD015154 2011

Dentener F Peters W Krol M van Weele M Berga-maschi P and Lelieveld J Interannual variability andtrend of CH4 lifetime as a measure for OH changes in the1979ndash1993 time period J Geophys Res-Atmos 108 4442doi1010292002JD002916 2003

Dentener F Kinne S Bond T Boucher O Cofala J GenerosoS Ginoux P Gong S Hoelzemann J J Ito A Marelli LPenner J E Putaud J-P Textor C Schulz M van der WerfG R and Wilson J Emissions of primary aerosol and pre-cursor gases in the years 2000 and 1750 prescribed data-sets forAeroCom Atmos Chem Phys 6 4321ndash4344doi105194acp-6-4321-2006 2006

Ebert E and Curry J A parameterization of ice-cloud opti-cal properties for climate models J Geophys Res-Atmos 973831ndash3836 1992

Ellingsen K Gauss M van Dingenen R Dentener F J Em-berson L Fiore A M Schultz M G Stevenson D S Ash-more M R Atherton C S Bergmann D J Bey I ButlerT Drevet J Eskes H Hauglustaine D A Isaksen I S AHorowitz L W Krol M Lamarque J F Lawrence M G vanNoije T Pyle J Rast S Rodriguez J Savage N StrahanS Sudo K Szopa S and Wild O Global ozone and air qual-ity a multi-model assessment of risks to human health and cropsAtmos Chem Phys Discuss 8 2163ndash2223doi105194acpd-8-2163-2008 2008

Endresen A Sorgard E Sundet J K Dalsoren S B Isaksen IS A Berglen T F and Gravir G Emission from internationalsea transportation and environmental impact J Geophys Res108 4560doi1010292002JD002898 2003

Eyring V Kohler H W van Aardenne J and Lauer A Emis-sions from international shipping 1 The last 50 years J Geo-phys Res 110 D17305doi1010292004JD005619 2005

Eyring V Kohler H W Lauer A and Lemper B Emis-sions from international shipping 2 Impact of future technolo-gies on scenarios until 2050 J Geophys Res 110 D17306doi1010292004JD005620 2005

Feichter J Kjellstrom E Rodhe H Dentener F Lelieveld Jand Roelofs G Simulation of the tropospheric sulfur cycle in aglobal climate model Atmos Environ 30 1693ndash1707 1996

Fiore A Horowitz L Dlugokencky E and West J Impact ofmeteorology and emissions on methane trends 1990ndash2004 Geo-phys Res Lett 33 L12809doi1010292006GL026199 2006

Fiore A M Dentener F J Wild O Cuvelier C Schultz M GHess P Textor C Schulz M Doherty R M Horowitz L WMacKenzie I A Sanderson M G Shindell D T Steven-son D S Szopa S van Dingenen R Zeng G Atherton CBergmann D Bey I Carmichael G Collins W J DuncanB N Faluvegi G Folberth G Gauss M Gong S Hauglus-taine D Holloway T Isaksen I S A Jacob D J JonsonJ E Kaminski J W Keating T J Lupu A Marmer EMontanaro V Park R J Pitari G Pringle K J Pyle J ASchroeder S Vivanco M G Wind P Wojcik G Wu Sand Zuber A Multimodel estimates of intercontinental source-receptor relationships for ozone pollution J Geophys Res 114D04301doi1010292008JD010816 2009

Ganzeveld L and Lelieveld J Dry deposition parameterizationin a chemistry general circulation model and its influence on

the distribution of reactive trace gases J Geophys Res-Atmos100 20999ndash21012 1995

Ganzeveld L Lelieveld J and Roelofs G A dry deposition pa-rameterization for sulfur oxides in a chemistry and general circu-lation model J Geophys Res-Atmos 103 5679ndash5694 1998

Ganzeveld L N van Aardenne J A Butler T M Lawrence MG Metzger S M Stier P Zimmermann P and Lelieveld JTechnical Note Anthropogenic and natural offline emissions andthe online EMissions and dry DEPosition submodel EMDEP ofthe Modular Earth Submodel system (MESSy) Atmos ChemPhys Discuss 6 5457ndash5483doi105194acpd-6-5457-20062006

Gauss M Myhre G Isaksen I S A Grewe V Pitari G WildO Collins W J Dentener F J Ellingsen K Gohar L KHauglustaine D A Iachetti D Lamarque F Mancini EMickley L J Prather M J Pyle J A Sanderson M GShine K P Stevenson D S Sudo K Szopa S and Zeng GRadiative forcing since preindustrial times due to ozone changein the troposphere and the lower stratosphere Atmos ChemPhys 6 575ndash599doi105194acp-6-575-2006 2006

Grewe V Brunner D Dameris M Grenfell J L Hein R Shin-dell D and Staehelin J Origin and variability of upper tro-pospheric nitrogen oxides and ozone at northern mid-latitudesAtmos Environ 35 3421ndash3433 2001

Guenther A Hewitt C Erickson D Fall R Geron C GraedelT Harley P Klinger L Lerdau M McKay W Pierce TScholes B Steinbrecher R Tallamraju R Taylor J andZimmerman P A global-model of natural volatile organic-compound emissions J Geophys Res-Atmos 100 8873ndash8892 1995

Guenther A Karl T Harley P Wiedinmyer C Palmer P Iand Geron C Estimates of global terrestrial isoprene emissionsusing MEGAN (Model of Emissions of Gases and Aerosols fromNature) Atmos Chem Phys 6 3181ndash3210doi105194acp-6-3181-2006 2006

Hagemann S Arpe K and Roeckner E Evaluation of the Hy-drological Cycle in the ECHAM5 Model J Climate 19 3810ndash3827 2006

Halmer M Schmincke H and Graf H The annual volcanic gasinput into the atmosphere in particular into the stratosphere aglobal data set for the past 100 years J Volcanol GeothermalRes 115 511ndash528 2002

Hess P and Mahowald N Interannual variability in hindcasts ofatmospheric chemistry the role of meteorology Atmos ChemPhys 9 5261ndash5280doi105194acp-9-5261-2009 2009

Horowitz L Walters S Mauzerall D Emmons L Rasch PGranier C Tie X Lamarque J Schultz M Tyndall GOrlando J and Brasseur G A global simulation of tropo-spheric ozone and related tracers Description and evaluationof MOZART version 2 J Geophys Res-Atmos 108 4784doi1010292002JD002853 2003

Jaffe D and Ray J Increase in surface ozone at rural sites inthe western US Atmos Env 41 54525463 ISSN 1352-2310doi101016jatmosenv200702034 2007

Jeuken A Siegmund P Heijboer L Feichter J and BengtssonL On the potential of assimilating meteorological analyses ina global climate model for the purpose of model validation JGeophys Res-Atmos 101 16939ndash16950 1996

Jonson J E Simpson D Fagerli H and Solberg S Can we ex-



plain the trends in European ozone levels Atmos Chem Phys6 51ndash66doi105194acp-6-51-2006 2006

Kettle A and Andreae M Flux of dimethylsulfide from theoceans A comparison of updated data seas and flux models JGeophys Res-Atmos 105 26793ndash26808 2000

Kinne S Schulz M Textor C Guibert S Balkanski Y BauerS E Berntsen T Berglen T F Boucher O Chin M CollinsW Dentener F Diehl T Easter R Feichter J Fillmore DGhan S Ginoux P Gong S Grini A Hendricks J HerzogM Horowitz L Isaksen I Iversen T Kirkevag A KlosterS Koch D Kristjansson J E Krol M Lauer A LamarqueJ F Lesins G Liu X Lohmann U Montanaro V MyhreG Penner J Pitari G Reddy S Seland O Stier P Take-mura T and Tie X An AeroCom initial assessment - opticalproperties in aerosol component modules of global models At-mos Chem Phys 6 1815ndash1834doi105194acp-6-1815-20062006

Klimont Z Cofala J Xing J Wei W Zhang C Wang SKejun J Bhandari P Mathura R Purohit P Rafaj P Cham-bers A Amann M and Hao J Projections of SO2 NOx andcarbonaceous aerosols emissions in Asia Tellus 61B 602ndash617doi101111j1600-0889200900428x 2009

Lamarque J-F Bond T C Eyring V Granier C Heil AKlimont Z Lee D Liousse C Mieville A Owen BSchultz M G Shindell D Smith S J Stehfest E van Aar-denne J Cooper O R Kainuma M Mahowald N Mc-Connell J R Naik V Riahi K and van Vuuren D P His-torical (1850ndash2000) gridded anthropogenic and biomass burningemissions of reactive gases and aerosols methodology and ap-plication Atmos Chem Phys 10 7017ndash7039doi105194acp-10-7017-2010 2010

Lathiere J Hauglustaine D A Friend A D De Noblet-Ducoudre N Viovy N and Folberth G A Impact of climatevariability and land use changes on global biogenic volatile or-ganic compound emissions Atmos Chem Phys 6 2129ndash2146doi105194acp-6-2129-2006 2006

Lawrence M G Jockel P and von Kuhlmann R What does theglobal mean OH concentration tell us Atmos Chem Phys 137ndash49doi105194acp-1-37-2001 2001

Lehman J Swinton K Bortnick S Hamilton C Baldridge EEder B and Cox B Spatio-temporal characterization of tropo-spheric ozone across the eastern United States Atmos Environ38 4357ndash4369 2004

Lin S and Rood R Multidimensional flux-form semi-Lagrangiantransport schemes Mon Weather Rev 124 2046ndash2070 1996

Logan J An analysis of ozonesonde data for the troposphereRecommendations for testing 3-D models and development ofa gridded climatology for tropospheric ozone J Geophys Res-Atmos 104 16115ndash16149 1999

Lohmann U and Roeckner E Design and performance of a newcloud microphysics scheme developed for the ECHAM generalcirculation model Climate Dynamics 12 557ndash572 1996

Lohmann U Stier P Hoose C Ferrachat S Kloster S Roeck-ner E and Zhang J Cloud microphysics and aerosol indi-rect effects in the global climate model ECHAM5-HAM At-mos Chem Phys 7 3425ndash3446doi105194acp-7-3425-20072007

Mlawer E Taubman S Brown P Iacono M and Clough SRadiative transfer for inhomogeneous atmospheres RRTM a

validated correlated-k model for the longwave J Geophys Res-Atmos 102 16663ndash16682 1997

Montzka S A Krol M Dlugokencky E Hall B Jockel Pand Lelieveld J Small Interannual Variability of Global Atmo-spheric Hydroxyl Science 331 67ndash69httpwwwsciencemagorgcontent331601367abstract 2011

Morcrette J Clough S Mlawer E and Iacono M Imapct ofa validated radiative transfer scheme RRTM on the ECMWFmodel climate and 10-day forecasts Tech Rep 252 ECMWFReading UK 1998

Nakicenovic N Alcamo J Davis G de Vries H Fenhann JGaffin S Gregory K Grubler A Jung T Kram T RovereE L Michaelis L Mori S Morita T Papper W Pitcher HPrice L Riahi K Roehrl A Rogner H-H Sankovski ASchlesinger M Shukla P Smith S Swart R van RooijenS Victor N and Dadi Z Special Report on Emissions Sce-narios Intergovernmental Panel on Climate Change CambridgeUniversity Press Cambridge 2000

Nightingale P Malin G Law C Watson A Liss P LiddicoatM Boutin J and Upstill-Goddard R In situ evaluation of air-sea gas exchange parameterizations using novel conservative andvolatile tracers Global Biogeochem Cy 14 373ndash387 2000

Nordeng T Extended versions of the convective parameterizationscheme at ECMWF and their impact on the mean and transientactivity of the model in the tropics Tech Rep 206 ECMWFReading UK 1994

Oltmans S J Lefohn A S Harris J M and Shadwick D Back-ground ozone levels of air entering the west coast of the US andassessment of longer-term changes Atmos Environ 42 6020ndash6038 2008

Parrish D D Millet D B and Goldstein A H Increasingozone in marine boundary layer inflow at the west coasts ofNorth America and Europe Atmos Chem Phys 9 1303ndash1323doi105194acp-9-1303-2009 2009

Pham M Muller J Brasseur G Granier C and Megie G Athree-dimensional study of the tropospheric sulfur cycle J Geo-phys Res-Atmos 100 26061ndash26092 1995

Pozzoli L Bey I Rast S Schultz M G Stier Pand Feichter J Trace gas and aerosol interactions in thefully coupled model of aerosol-chemistry-climate ECHAM5-HAMMOZ 1 Model description and insights from the spring2001 TRACE-P experiment J Geophys Res 113 D07308doi1010292007JD009007 2008a

Pozzoli L Bey I Rast S Schultz M G Stier P and FeichterJ Trace gas and aerosol interactions in the fully coupled modelof aerosol-chemistry-climate ECHAM5-HAMMOZ 2 Impactof heterogeneous chemistry on the global aerosol distributionsJ Geophys Res 113 D07309doi1010292007JD0090082008b

Prinn R G Huang J Weiss R F Cunnold D M FraserP J Simmonds P G McCulloch A Harth C Reimann SSalameh P OrsquoDoherty S Wang R H J Porter L W MillerB R and Krummel P B Evidence for variability of atmo-spheric hydroxyl radicals over the past quarter century GeophysRes Lett 32 L07809doi1010292004GL022228 2005

Raes F and Seinfeld J Climate Change and Air Pollution Abate-ment A Bumpy Road Atmos Environ 43 5132ndash5133 2009

Randel W Wu F Russell J Roche A and Waters J Sea-sonal cycles and QBO variations in stratospheric CH4 and H2O



observed in UARS HALOE data J Atmos Sci 55 163ndash1851998

Rast J Schultz M Aghedo A Bey I Brasseur G Diehl TEsch M Ganzeveld L Kirchner I Kornblueh L Rhodin ARoeckner E Schmidt H Schroede S Schulzweida U StierP and van Noije T Evaluation of the tropospheric chemistrygeneral circulation model ECHAM5ndashMOZ and its application tothe analysis of the inter-annual variability in tropospheric ozonefrom 1960ndash2000 chemical composition of the troposphere for theperiod 1960ndash2000 (RETRO) MPI-Report (Reports on Earth Sys-tem Science) 2011

Richter A Burrows J P Nuss H Granier C and Niemeier UIncrease in tropospheric nitrogen dioxide over China observedfrom space Nature 437(7055) 129ndash132 2005

Rockel B Raschke E and Weyres B A parameterization ofbroad band radiative transfer properties of water ice and mixedclouds Beitr Phys Atmos 64 1ndash12 1991

Roeckner E Bauml G Bonaventura L Brokopf R EschM Giorgetta M Hagemann S Kirchner I KornbluehL Manzini E Rhodin A Schlese U Schulzweida Uand Tompkins A The atmospehric general circulation modelECHAM5 Part 1 Tech Rep 349 Max Planck Institute for Me-teorology Hamburg 2003

Roeckner E Brokopf R Esch M Giorgetta M HagemannS Kornblueh L Manzini E Schlese U and SchulzweidaU Sensitivity of Simulated Climate to Horizontal and VerticalResolution in the ECHAM5 Atmosphere Model J Climate 193771ndash3791 2006

Schultz M Backman L Balkanski Y Bjoerndalsaeter SBrand R Burrows J Dalsoeren S de Vasconcelos MGrodtmann B Hauglustaine D Heil A Hoelzemann JIsaksen I Kaurola J Knorr W Ladstaetter-Weienmayer AMota B Oom D Pacyna J Panasiuk D Pereira J PullesT Pyle J Rast S Richter A Savage N Schnadt C SchulzM Spessa A Staehelin J Sundet J Szopa S ThonickeK van het Bolscher M van Noije T van Velthoven P VikA and Wittrock F REanalysis of the TROpospheric chemicalcomposition over the past 40 years (RETRO) A long-term globalmodeling study of tropospheric chemistry Final Report Techrep Max Planck Institute for Meteorology Hamburg Germany2007

Schultz M G Heil A Hoelzemann J J Spessa A Thon-icke K Goldammer J G Held A C Pereira J M Cand van het Bolscher M Global wildland fire emissionsfrom 1960 to 2000 Global Biogeochem Cy 22 GB2002doi1010292007GB003031 2008

Schulz M de Leeuw G and Balkanski Y Emission Of Atmo-spheric Trace Compounds chap Sea-salt aerosol source func-tions and emissions 333ndash359 Kluwer 2004

Schumann U and Huntrieser H The global lightning-inducednitrogen oxides source Atmos Chem Phys 7 3823ndash3907doi105194acp-7-3823-2007 2007

Solberg S Hov O Sovde A Isaksen I S A Coddeville PDe Backer H Forster C Orsolini Y and Uhse K Europeansurface ozone in the extreme summer 2003 J Geophys Res113 D07307doi1010292007JD009098 2008

Solomon S Qin D Manning M Chen Z Marquis M Av-eryt K Tignor M and Miller H L Contribution of WorkingGroup I to the Fourth Assessment Report of the Intergovernmen-

tal Panel on Climate Change 2007 Cambridge University PressCambridge United Kingdom and New York NY USA 2007

Stevenson D Dentener F Schultz M Ellingsen K van NoijeT Wild O Zeng G Amann M Atherton C Bell NBergmann D Bey I Butler T Cofala J Collins W Der-went R Doherty R Drevet J Eskes H Fiore A Gauss MHauglustaine D Horowitz L Isaksen I Krol M Lamar-que J Lawrence M Montanaro V Muller J Pitari GPrather M Pyle J Rast S Rodriguez J Sanderson MSavage N Shindell D Strahan S Sudo K and Szopa SMultimodel ensemble simulations of present-day and near-futuretropospheric ozone J Geophys Res-Atmos 111 D08301doi1010292005JD006338 2006

Stier P Feichter J Kinne S Kloster S Vignati E WilsonJ Ganzeveld L Tegen I Werner M Balkanski Y SchulzM Boucher O Minikin A and Petzold A The aerosol-climate model ECHAM5-HAM Atmos Chem Phys 5 1125ndash1156doi105194acp-5-1125-2005 2005

Stier P Seinfeld J H Kinne S and Boucher O Aerosol ab-sorption and radiative forcing Atmos Chem Phys 7 5237ndash5261doi105194acp-7-5237-2007 2007

Streets D G Bond T C Lee T and Jang C On the future ofcarbonaceous aerosol emissions J Geophys Res 109 D24212doi1010292004JD004902 2004

Streets D G Wu Y and Chin M Two-decadalaerosol trends as a likely explanation of the global dim-mingbrightening transition Geophys Res Lett 33 L15806doi1010292006GL026471 2006

Streets D G Yan F Chin M Diehl T Mahowald NSchultz M Wild M Wu Y and Yu C Anthropogenicand natural contributions to regional trends in aerosol op-tical depth 1980ndash2006 J Geophys Res 114 D00D18doi1010292008JD011624 2009

Taylor K Summarizing multiple aspects of model performancein a single diagram J Geophys Res-Atmos 106 7183ndash71922001

Tegen I Harrison S Kohfeld K Prentice I Coe M andHeimann M Impact of vegetation and preferential source areason global dust aerosol Results from a model study J GeophysRes-Atmos 107 4576doi1010292001JD000963 2002

Tiedtke M A comprehensive mass flux scheme for cumulus pa-rameterization in large-scale models Mon Weather Rev 1171779ndash1800 1989

Tompkins A A prognostic parameterization for the subgrid-scalevariability of water vapor and clouds in large-scale models andits use to diagnose cloud cover J Atmos Sci 59 1917ndash19422002

Toon O and Ackerman T Algorithms for the calculation of scat-tering by stratified spheres Appl Optics 20 3657ndash3660 1981

Tost H Jockel P and Lelieveld J Lightning and convec-tion parameterisations - uncertainties in global modelling At-mos Chem Phys 7 4553ndash4568doi105194acp-7-4553-20072007

Tressol M Ordonez C Zbinden R Brioude J Thouret VMari C Nedelec P Cammas J-P Smit H Patz H-Wand Volz-Thomas A Air pollution during the 2003 Europeanheat wave as seen by MOZAIC airliners Atmos Chem Phys8 2133ndash2150doi105194acp-8-2133-2008 2008

Unger N Menon S Koch D M and Shindell D T Im-



pacts of aerosol-cloud interactions on past and future changes intropospheric composition Atmos Chem Phys 9 4115ndash4129doi105194acp-9-4115-2009 2009

Uppala S M KAllberg P W Simmons A J Andrae U Bech-told V D C Fiorino M Gibson J K Haseler J HernandezA Kelly G A Li X Onogi K Saarinen S Sokka N Al-lan R P Andersson E Arpe K Balmaseda M A BeljaarsA C M Berg L V D Bidlot J Bormann N Caires SChevallier F Dethof A Dragosavac M Fisher M FuentesM Hagemann S Holm E Hoskins B J Isaksen L JanssenP A E M Jenne R Mcnally A P Mahfouf J-F Mor-crette J-J Rayner N A Saunders R W Simon P SterlA Trenberth K E Untch A Vasiljevic D Viterbo P andWoollen J The ERA-40 re-analysis Q J Roy Meteorol Soc131 2961ndash3012doi101256qj04176 2005

van Aardenne J Dentener F Olivier J Goldewijk C andLelieveld J A 1 1 resolution data set of historical anthro-pogenic trace gas emissions for the period 1890ndash1990 GlobalBiogeochem Cy 15 909ndash928 2001

van Noije T P C Segers A J and van Velthoven P F J Timeseries of the stratosphere-troposphere exchange of ozone simu-lated with reanalyzed and operational forecast data J GeophysRes 111 D03301doi1010292005JD006081 2006

Vautard R Szopa S Beekmann M Menut L Hauglus-taine D A Rouil L and Roemer M Are decadal an-thropogenic emission reductions in Europe consistent with sur-face ozone observations Geophys Res Lett 33 L13810doi1010292006GL026080 2006

Vignati E Wilson J and Stier P M7 An efficient size-resolved aerosol microphysics module for large-scale aerosoltransport models J Geophys Res-Atmos 109 D22202doi1010292003JD004485 2004

Wang K Dickinson R and Liang S Clear sky visibility hasdecreased over land globally from 1973 to 2007 Science 3231468ndash1470 2009

Wild M Global dimming and brightening A review J GeophysRes 114 D00D16doi1010292008JD011470 2009

Wilson R C Fleming Z L Monks P S Clain G HenneS Konovalov I B Szopa S and Menut L Have pri-mary emission reduction measures reduced ozone across Eu-rope An analysis of European rural background ozone trends1996ndash2005 Atmos Chem Phys Discuss 11 18433ndash18485doi105194acpd-11-18433-2011 2011

Zhang Q Streets D G Carmichael G R He K B Huo HKannari A Klimont Z Park I S Reddy S Fu J S ChenD Duan L Lei Y Wang L T and Yao Z L Asian emis-sions in 2006 for the NASA INTEX-B mission Atmos ChemPhys 9 5131ndash5153doi105194acp-9-5131-2009 2009



1 Introduction

Air quality is determined by the emission of primary pol-lutants into the atmosphere by chemical production of sec-ondary pollutants and by meteorological conditions Thetwo air pollutants of most concern for public health ozone(O3) and particulate matter (PM) have also strong impactson climate In the fourth Assessment Report of the Intergov-ernmental Panel on Climate Change (IPCC AR4)Solomonet al (2007) estimate a Radiative Forcing (RF) from tropo-spheric ozone of +035 [minus01 +03] W mminus2 which corre-sponds to the third largest contribution to the total RF aftercarbon dioxide (CO2) and methane (CH4) IPCC AR4 alsoprovides estimates for radiative forcing from aerosols Thedirect RF (scattering and absorption of solar and infrared ra-diation) amounts tominus05 [plusmn04] W mminus2 and the RF throughindirect changes in cloud properties is estimated tominus070[minus11+04] W mminus2

Trends in global radiation and visibility measurements in-deed suggest an important role of aerosols Solar radia-tion measurements showed a consistent and worldwide de-crease at the Earthrsquos surface (an effect dubbed ldquodimmingrdquo)from the 1960s This trend reversed into ldquobrighteningrdquoin the late 1990s in the US Europe and parts of Korea(Wild 2009) Similarly an analysis of visibility measure-ments from 1973ndash2007 byWang et al(2009) suggests aglobal increase of AOD worldwide except in Europe SinceSO2minus

4 is one of the main aerosol components that determinethe aerosol optical depth (Streets et al 2009) the SO2minus

4concentration reductions over Europe and US after the im-plementation of air quality policies may partly explain thedimming-brightening transition observed in the 1990s in Eu-rope whereas in emerging economies such as China andIndia the emission of air pollutants rapidly increased since1990

Inter-annual meteorological variability in the last decadesalso strongly determined the variations of the concentrationsand geographical distribution of air pollutants For exam-ple the El Nino event in 1997ndash1998 and the period after theMt Pinatubo volcanic eruption in 1991 can explain much ofthe past chemical inter-annual variability of tropospheric O3CH4 and OH (Fiore et al 2009 Dentener et al 2003 Hessand Mahowald 2009) In Europe the infamous summer of2003 led to a strong positive anomaly of solar surface radia-tion (Wild 2009) and exacerbated ozone pollution at ground-level and throughout the troposphere (Solberg et al 2008Tressol et al 2008)

Meteorological variability and changes in the precursoremissions of O3 and SO2minus

4 are often concurrent processesand their impact on surface concentrations is difficult to un-derstand from measurements alone (Vautard et al 2006Berglen et al 2007) Therefore there have been severalefforts to re-analyze these trends using tropospheric chem-istry and transport models For example the Europeanproject REanalysis of the TROpospheric chemical composi-

tion (RETROSchultz et al 2007) employed three differentglobal models to simulate tropospheric ozone changes be-tween 1960 and 2000 RecentlyHess and Mahowald(2009)analyzed the role of meteorology in inter-annual variabilityof tropospheric ozone chemistry

In this paper we extend these analyses by using a coupledaerosol-chemistry-climate simulation and discuss our resultsin the light of the previous studies We analyze the chemi-cal variability due to changes in meteorology (ie transportchemistry) and natural emissions and separate them from an-thropogenic emissions induced variability The period 1980ndash2005 was chosen because the advent of satellite observationsin the 1970s introduced a discontinuity in the re-analysisof meteorological datasets (Hess and Mahowald 2009 vanNoije et al 2006) and as mentioned above the consideredperiod includes large meteorological anomalies Particularattention will be given to four regions of the world (NorthAmerica Europe East Asia and South Asia) where signifi-cant changes in terms of the absolute amount of emitted tracegases and aerosol precursors occurred in the last decades Wewill focus on past changes of O3 and SO2minus

4 because of theirimportance for air quality and climate Long measurementrecords are available since the 1980s and they will be usedto evaluate our model results and the simulated trends Wefurther analyze changes in AOD radiative perturbation (RP)and OH radical associated with changes in emissions and me-teorology

The paper is organized as follows In Sect2 the modeldescription and experiment setup are outlined In Sect3 an-thropogenic and natural emissions are described Sections4and5 present an analysis of global and regional variabilityand trends of O3 and SO2minus

4 from 1980 to 2005 The regionalanalysis focuses on Europe (EU) North America (NA) EastAsia (EA) and South Asia (SA) (Fig1 the region bound-aries were defined as in the HTAP studyFiore et al 2009)In Sect6 we will describe the anthropogenic radiative per-turbation due to changes in O3 and SO2minus

4 In Sects7 and8 we will summarize the main findings and further discussimplications of our study for air quality-climate interactions

2 Model and simulation descriptions

ECHAM5-HAMMOZ is a fully coupled aerosol-chemistry-climate model composed of the general circulation model(GCM) ECHAM5 the tropospheric chemistry module MOZand the aerosol module HAM The ECHAM5-HAMMOZmodel is described in detail inPozzoli et al(2008a) Themodel has been extensively evaluated in previous studies(Stier et al 2005 Pozzoli et al 2008ab Auvray et al2007 Rast et al 2011) with comparisons to several mea-surements and within model inter-comparison studies Wefurther remark a substantial overestimate of our computedozone compared to measurements a problem that ECHAM5-







times 28





3 Emissions











2minus


1980 1985 1990 1995 2000 2005

CO 6737 6801 7138 6853 6554 6786NOx 342 337 361 364 367 372VOCs 846 850 879 848 805 843SO2 671 672 662 610 588 590SO4

2minus 18 18 18 17 16 16OC 78 82 83 87 85 87BC 49 49 49 48 47 49

























1980 1985 1990 1995 2000 200570

60

50

40

30

20

10

0


CO9563 Tg yr 1

VOC2066 Tg(C) yr 1

NOx753 Tg(N) yr 1

SO22565 Tg(S) yr 1

BC131 Tg yr 1

OC163 Tg yr 1

1980 1985 1990 1995 2000 200560

50

40

30

20

10

0


CO12740 Tg yr 1

VOC2611 Tg(C) yr 1

NOx753 Tg(N) yr 1

SO21289 Tg(S) yr 1

BC068 Tg yr 1

OC087 Tg yr 1

1980 1985 1990 1995 2000 200520

0

20

40

60

80

100

120


CO9817 Tg yr 1

VOC1083 Tg(C) yr 1

NOx296 Tg(N) yr 1

SO21065 Tg(S) yr 1

BC142 Tg yr 1

OC194 Tg yr 1

1980 1985 1990 1995 2000 20050

50

100

150

200

250

300


CO6015 Tg yr 1

VOC553 Tg(C) yr 1

NOx098 Tg(N) yr 1

SO2147 Tg(S) yr 1

BC039 Tg yr 1

OC116 Tg yr 1






SREF SFIX



times 106 ) 120 0016 0013 118 0029 0024HNO3 (pptv) 12911 1470 01139 12573 1391 01106






















421 Europe













422 North America







423 East Asia


424 South Asia















times 1015 ) 631 1269 002012 735 0707 000962 704 0847 0012HNO3 (pptv) 127 1395 01096 121 122 00101 121 149 00123









44 OH variability



=









4








511 Europe











512 North America





4 trends

513 East Asia









514 South Asia










4 budget



face SO2minus













4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References




















































































































times 28





3 Emissions











2minus


1980 1985 1990 1995 2000 2005

CO 6737 6801 7138 6853 6554 6786NOx 342 337 361 364 367 372VOCs 846 850 879 848 805 843SO2 671 672 662 610 588 590SO4

2minus 18 18 18 17 16 16OC 78 82 83 87 85 87BC 49 49 49 48 47 49

























1980 1985 1990 1995 2000 200570

60

50

40

30

20

10

0


CO9563 Tg yr 1

VOC2066 Tg(C) yr 1

NOx753 Tg(N) yr 1

SO22565 Tg(S) yr 1

BC131 Tg yr 1

OC163 Tg yr 1

1980 1985 1990 1995 2000 200560

50

40

30

20

10

0


CO12740 Tg yr 1

VOC2611 Tg(C) yr 1

NOx753 Tg(N) yr 1

SO21289 Tg(S) yr 1

BC068 Tg yr 1

OC087 Tg yr 1

1980 1985 1990 1995 2000 200520

0

20

40

60

80

100

120


CO9817 Tg yr 1

VOC1083 Tg(C) yr 1

NOx296 Tg(N) yr 1

SO21065 Tg(S) yr 1

BC142 Tg yr 1

OC194 Tg yr 1

1980 1985 1990 1995 2000 20050

50

100

150

200

250

300


CO6015 Tg yr 1

VOC553 Tg(C) yr 1

NOx098 Tg(N) yr 1

SO2147 Tg(S) yr 1

BC039 Tg yr 1

OC116 Tg yr 1






SREF SFIX



times 106 ) 120 0016 0013 118 0029 0024HNO3 (pptv) 12911 1470 01139 12573 1391 01106






















421 Europe













422 North America







423 East Asia


424 South Asia















times 1015 ) 631 1269 002012 735 0707 000962 704 0847 0012HNO3 (pptv) 127 1395 01096 121 122 00101 121 149 00123









44 OH variability



=









4








511 Europe











512 North America





4 trends

513 East Asia









514 South Asia










4 budget



face SO2minus













4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References






















































































































2minus


1980 1985 1990 1995 2000 2005

CO 6737 6801 7138 6853 6554 6786NOx 342 337 361 364 367 372VOCs 846 850 879 848 805 843SO2 671 672 662 610 588 590SO4

2minus 18 18 18 17 16 16OC 78 82 83 87 85 87BC 49 49 49 48 47 49

























1980 1985 1990 1995 2000 200570

60

50

40

30

20

10

0


CO9563 Tg yr 1

VOC2066 Tg(C) yr 1

NOx753 Tg(N) yr 1

SO22565 Tg(S) yr 1

BC131 Tg yr 1

OC163 Tg yr 1

1980 1985 1990 1995 2000 200560

50

40

30

20

10

0


CO12740 Tg yr 1

VOC2611 Tg(C) yr 1

NOx753 Tg(N) yr 1

SO21289 Tg(S) yr 1

BC068 Tg yr 1

OC087 Tg yr 1

1980 1985 1990 1995 2000 200520

0

20

40

60

80

100

120


CO9817 Tg yr 1

VOC1083 Tg(C) yr 1

NOx296 Tg(N) yr 1

SO21065 Tg(S) yr 1

BC142 Tg yr 1

OC194 Tg yr 1

1980 1985 1990 1995 2000 20050

50

100

150

200

250

300


CO6015 Tg yr 1

VOC553 Tg(C) yr 1

NOx098 Tg(N) yr 1

SO2147 Tg(S) yr 1

BC039 Tg yr 1

OC116 Tg yr 1






SREF SFIX



times 106 ) 120 0016 0013 118 0029 0024HNO3 (pptv) 12911 1470 01139 12573 1391 01106






















421 Europe













422 North America







423 East Asia


424 South Asia















times 1015 ) 631 1269 002012 735 0707 000962 704 0847 0012HNO3 (pptv) 127 1395 01096 121 122 00101 121 149 00123









44 OH variability



=









4








511 Europe











512 North America





4 trends

513 East Asia









514 South Asia










4 budget



face SO2minus













4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References
































































































































1980 1985 1990 1995 2000 200570

60

50

40

30

20

10

0


CO9563 Tg yr 1

VOC2066 Tg(C) yr 1

NOx753 Tg(N) yr 1

SO22565 Tg(S) yr 1

BC131 Tg yr 1

OC163 Tg yr 1

1980 1985 1990 1995 2000 200560

50

40

30

20

10

0


CO12740 Tg yr 1

VOC2611 Tg(C) yr 1

NOx753 Tg(N) yr 1

SO21289 Tg(S) yr 1

BC068 Tg yr 1

OC087 Tg yr 1

1980 1985 1990 1995 2000 200520

0

20

40

60

80

100

120


CO9817 Tg yr 1

VOC1083 Tg(C) yr 1

NOx296 Tg(N) yr 1

SO21065 Tg(S) yr 1

BC142 Tg yr 1

OC194 Tg yr 1

1980 1985 1990 1995 2000 20050

50

100

150

200

250

300


CO6015 Tg yr 1

VOC553 Tg(C) yr 1

NOx098 Tg(N) yr 1

SO2147 Tg(S) yr 1

BC039 Tg yr 1

OC116 Tg yr 1






SREF SFIX



times 106 ) 120 0016 0013 118 0029 0024HNO3 (pptv) 12911 1470 01139 12573 1391 01106






















421 Europe













422 North America







423 East Asia


424 South Asia















times 1015 ) 631 1269 002012 735 0707 000962 704 0847 0012HNO3 (pptv) 127 1395 01096 121 122 00101 121 149 00123









44 OH variability



=









4








511 Europe











512 North America





4 trends

513 East Asia









514 South Asia










4 budget



face SO2minus













4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References

















































































































SREF SFIX



times 106 ) 120 0016 0013 118 0029 0024HNO3 (pptv) 12911 1470 01139 12573 1391 01106






















421 Europe













422 North America







423 East Asia


424 South Asia















times 1015 ) 631 1269 002012 735 0707 000962 704 0847 0012HNO3 (pptv) 127 1395 01096 121 122 00101 121 149 00123









44 OH variability



=









4








511 Europe











512 North America





4 trends

513 East Asia









514 South Asia










4 budget



face SO2minus













4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References






























































































































421 Europe













422 North America







423 East Asia


424 South Asia















times 1015 ) 631 1269 002012 735 0707 000962 704 0847 0012HNO3 (pptv) 127 1395 01096 121 122 00101 121 149 00123









44 OH variability



=









4








511 Europe











512 North America





4 trends

513 East Asia









514 South Asia










4 budget



face SO2minus













4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References
























































































































422 North America







423 East Asia


424 South Asia















times 1015 ) 631 1269 002012 735 0707 000962 704 0847 0012HNO3 (pptv) 127 1395 01096 121 122 00101 121 149 00123









44 OH variability



=









4








511 Europe











512 North America





4 trends

513 East Asia









514 South Asia










4 budget



face SO2minus













4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References
















































































































423 East Asia


424 South Asia















times 1015 ) 631 1269 002012 735 0707 000962 704 0847 0012HNO3 (pptv) 127 1395 01096 121 122 00101 121 149 00123









44 OH variability



=









4








511 Europe











512 North America





4 trends

513 East Asia









514 South Asia










4 budget



face SO2minus













4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References





















































































































times 1015 ) 631 1269 002012 735 0707 000962 704 0847 0012HNO3 (pptv) 127 1395 01096 121 122 00101 121 149 00123









44 OH variability



=









4








511 Europe











512 North America





4 trends

513 East Asia









514 South Asia










4 budget



face SO2minus













4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References



















































































































44 OH variability



=









4








511 Europe











512 North America





4 trends

513 East Asia









514 South Asia










4 budget



face SO2minus













4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References


















































































































4








511 Europe











512 North America





4 trends

513 East Asia









514 South Asia










4 budget



face SO2minus













4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References


















































































































514 South Asia










4 budget



face SO2minus













4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References





















































































































4 concentrations








SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References





















































































































SO2minus






4 burden

(B) SO2minus


SO2minus


















GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References


























































































































GLOBAL EU NA EA SA

1O3 [ppb] 098 081 027 244 4251SO2minus






RPO3 [W mminus2] 005 005 006 012 015




























































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References
















































































































































4







8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References

















































































































8 Outlook





Appendix A


A1 The ECHAM5 GCM










Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References






















































































































Appendix B

O3 and SO2minus







1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References
















































































































1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mace Head


1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Hohenpeissenberg

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Zingst

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Barrow

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Mauna Loa

1980 1990 20008

6

4

2

0

2

4

6

8

ppbv

Samoa





























SO2minus






SO2minus














References

































































































































SO2minus






SO2minus














References

























































































































References















































































































Re-analysis of tropospheric sulfate aerosol and ozone for the period ...

Documents

Transcript of Re-analysis of tropospheric sulfate aerosol and ozone for the period ...