Future air pollution in the Shared Socio-economic Pathways
Shilpa Rao
a,b,*, Zbigniew Klimont
a, Steven J. Smith
c,d, Rita Van Dingenen
e, Frank Dentener
e, Lex Bouwman
f,g, Keywan Riahi
a,h, Markus Amann
a,
Benjamin Leon Bodirsky
i,j, Detlef P. van Vuuren
f,k, Lara Aleluia Reis
l,m, Katherine Calvin
c, Laurent Drouet
l,m, Oliver Fricko
a, Shinichiro Fujimori
n, David Gernaat
f, Petr Havlik
a, Mathijs Harmsen
f, Tomoko Hasegawa
n, Chris Heyes
a, Jérôme Hilaire
i,o, Gunnar Luderer
i, Toshihiko Masui
n, Elke Stehfest
f, Jessica Stre fl er
i, Sietske van der Sluis
f,
Massimo Tavoni
l,m,paInternationalInstituteforAppliedSystemsAnalysis,Schlossplatz-1,A-2361,Laxenburg,Austria
bNorwegianInstituteofPublicHealth,POBox4404,Nydalen,0403,Oslo,Norway
cJointGlobalChangeResearchInstitute,PacificNorthwestNationalLaboratory,5825UniversityResearchCourt,Suite3500,CollegePark,MD20740,USA
dDepartmentofAtmosphericandOceanicScience,UniversityofMaryland,CollegePark,MD20742,USA
eJointResearchCentre,InstituteforEnvironmentandSustainability,ViaEnricoFermi2749,I 21027,Ispra(VA),Italy
fPBLNetherlandsEnvironmentalAssessmentAgency,Ant.vanLeeuwenhoeklaan9,3721MA,Bilthoven,TheNetherlands
gDepartmentofEarthSciences,FacultyofGeosciences,UtrechtUniversity,POBox80021,3508TA,Utrecht,TheNetherlands
hGrazUniversityofTechnology,Inffeldgasse,A-8010Graz,Austria
iPotsdamInstituteforClimateImpactResearch(PIK),POBox601203,14412Potsdam,Germany
jCommonwealthScientificandIndustrialResearchOrganization,AgricultureFlagship,StLucia,QLD,4067,Australia
kCopernicusInstituteofSustainableDevelopment,UtrechtUniversity,Heidelberglaan2,3584CSUtrecht,TheNetherlands
lFondazioneEniEnricoMattei(FEEM),CorsoMagenta63,20123Milan,Italy
mCentroEuro-MediterraneoperiCambiamentiClimatici(CMCC),viaAugustoImperatore,16-I-73100Lecce,Italy
nNationalInstituteforEnvironmentalStudies,CenterforSocial&EnvironmentalSystemsresearch,16-2Onogawa,Tsukuba,Ibaraki,305-8506,Japan
oMercatorResearchInstituteonGlobalCommonsandClimateChange(MCC),TorgauerStraße12-1510829Berlin,Germany
pPolitecnicodiMilano,PiazzaLeonardodaVinci,32,20133Milan,Italy
ARTICLE INFO Articlehistory:
Received15December2015 Receivedinrevisedform24May2016 Accepted31May2016
Availableonline15July2016 Keywords:
Scenarios Airpollution
Integratedassessmentmodels
ABSTRACT
Emissionsofairpollutantssuchassulfurandnitrogenoxidesandparticulateshavesignificanthealth impactsaswellaseffectsonnaturalandanthropogenicecosystems.Thesesameemissionsalsocan changeatmosphericchemistryandtheplanetaryenergybalance,therebyimpactingglobalandregional climate.Long-termscenariosforairpollutantemissionsareneededasinputstoglobalclimateand chemistrymodels,andforanalysislinkingairpollutantimpactsacrosssectors.Inthispaperwepresent methodologyandresultsforairpollutantemissionsinSharedSocioeconomicPathways(SSP)scenarios.
Wefirstpresentasetofthreeairpollutionnarrativesthatdescribehigh,central,andlowpollution controlambitionsoverthe21stcentury.Thesenarrativesarethentranslatedintoquantitativeguidance foruseinintegratedassessmentmodels.TheresultingpollutantemissiontrajectoriesundertheSSP scenarios cover a wider range than the scenarios used in previous international climate model comparisons. In the SSP3 and SSP4 scenarios, where economic, institutional and technological limitationsslowairqualityimprovements,globalpollutant emissionsoverthe21stcenturycanbe comparabletocurrentlevels.PollutantemissionsintheSSP1scenariosfalltolowlevelsduetothe assumptionoftechnologicaladvancesandsuccessfulglobalactiontocontrolemissions.
ã2016TheAuthors.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1.Introduction
Despite efforts to control atmospheric pollutant emissions, ambientairqualityremainsamajorconcerninmanypartsofthe world.Airpollution hassignificantnegativeimpactsonhuman health(Popeetal.,2002;Dockeryetal.,1993;Jerrettetal.,2009).
*Correspondingauthorat:InternationalInstituteforAppliedSystemsAnalysis, Schlossplatz-1,A-2361,Laxenburg,Austria.
E-mailaddress:[email protected](S.Rao).
http://dx.doi.org/10.1016/j.gloenvcha.2016.05.012
0959-3780/ã2016TheAuthors.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).
ContentslistsavailableatScienceDirect
Global Environmental Change
j o u r n a lh o m e p ag e :w w w . e l s e vi e r . c o m / l o c a t e / g l o en v c h a
Morethan80%oftheworld’spopulationisexposedtopollutant concentrationsexceedingtheWorldHealthOrganization(WHO) recommendedlevels(Braueretal.,2012)andaround3.6million deathscanbeattributedtoambientairpollutionwithanother4 million from household related sources (Lim et al., 2012).
Moreover,air pollution canalter ecosystems, damagebuildings andmonuments,aswellasinfluenceearth’senergybalanceand thereforeclimatechange.
Long-term global scenarios for air pollutant emissionshave been used for atmospheric chemistry and Earth system model simulationsintendedtoexaminefuturechangesinclimate,air,and watersystems.Thesescenariosreflectplausiblefutureemissions basedonsocioeconomic,environmental,andtechnologicaltrends.
Thesescenariosaregenerallyproducedbyintegratedassessment models(IAMs)(Mossetal.,2010),whichprojecteconomicgrowth, population,energyconsumption,land-useandagriculturealong withassociatedGHGand pollutantemissions.Recentexamples includeinparticular, theRepresentativeConcentrationPathway (RCP)scenarios(vanVuurenetal.,2011a),whichwerethemulti- modelglobalscenariosofgreenhousegasesandairpollutantsused intheCoupledModel IntercomparisonProjectphase 5(CMIP5) (Tayloretal.,2011).TheRCPsweredevelopedtospanarangeof climateforcinglevelsandwerenotassociatedwithspecificsocio- economicnarratives.Thesescenariosreflectedtheprevailingview thatairqualitypolicieswillbesuccessfullyimplementedglobally andthatemissionscontroltechnologywillcontinuetoevolveand asaresultshowsignificantdeclinesinparticulatematter(PM)and ozoneprecursoremissionsoverthe21stcenturyatagloballevel (Amann et al., 2013; van Vuuren et al., 2011b). More recent scenarios have included alternative assumptions on pollution control,inanefforttobetterunderstandtheroleofairpollution controlintermsofreferencescenariodevelopmentand theco- benefitsfromclimatepolicies(seeforexampleRogeljetal.,2014;
Raoetal.,2013;Westetal.,2013;Chuwahetal.,2013).While providingawiderrangeofpollutionfutures,theassumptionson airpollutioncontrolinthesescenariosare,however,stilllargely independentofunderlyingscenarionarratives.
Itisgenerallyassumedinlong-termscenarios,implicitly,that pollutantconcentrationgoalswillcontinuetobemoreambitious overtime,onceincomesbecomesufficientlylarge.However,the time,stringency,andenforcementsuccessoffuturetargetsfora particularregioncannotgenerallybeknownandmustideallybe treatedasscenariovariable.Inalong-termscenariocontext,itis furthernecessarythat assumptions onairpollution controlare consistent with the underlying challenges to climate change mitigationandadaptation.Pollutionoutcomesinsuchscenarios can then be expected tobe a cumulative result of a range of variables including socio-economic development, technological change,efficiencyimprovementsandpoliciesdirectedatpollution controlaswellasalternativeconcernsincludingclimatechange, energyaccess,andagriculturalproduction.
TheShared Socio EconomicPathways (SSPs) (Kriegler etal., 2012)areanewgenerationofscenariosandstorylinesprimarily framed within the context of climate change mitigation and adaptation.TheSSP narratives(van Vuurenet al.,2014; O’Neill etal.,2014)compriseatextualdescriptionofhowthefuturemight unfold,includinga descriptionofmajorsocio-economic,demo- graphic, technological, lifestyle, policy, institutional and other trends.Inthispaper,ouroverarchinggoalistodevelopplausible rangesoffutureairpollutantemissiondevelopmentpathwaysin theSSPscenarios,whicharebased oninternallyconsistentand coherentassumptionsonthedegreeandimplementationoffuture airpollutioncontrol.OtherpapersinthisSpecialIssuesummarize parallel efforts in terms of elaboration of developments in the energysystem,landuseandgreenhousegasemissionsintheSSP scenarios(Baueretal.,2017;Poppetal.,2017).
Thestructureofthepaperisasfollows.Wefirstdescribethe developmentofasetofalternativeassumptionsonthedegreeand implementationof‘pollutioncontrol’intheSSPscenarios.These assumptions then reflect historical evidence and prevailing attitudesandprogressonpollutioncontrolandpotentialattitudes tothehealth andenvironmental impactsofairpollution inthe future. We further postulate a link between these alternative development pathwaysfor pollution control and a specific SSP narrative.Wealsodescribequantitativeguidancewithregardsto implementationoftheseassumptionsinIAMs.Finally,thepaper summarizeskeyresultsfromdifferentIAMinterpretationsofthe SSP scenarios interms of air pollutantemissionsand regional ambientairquality.
2.Methodology
In the following sections, we first summarize the overall descriptionoftheSSPscenarios.Wenextdescribethedevelop- mentofasetofqualitativeassumptionsonpollutioncontrolthat can be linked to the overall SSP narratives and present a quantitative proposal for implementationof these assumptions inIAMs.
2.1.DescriptionofSSPscenarios
The SSPs depict five different global futures (SSP1–5) with substantially different socio-economic conditions. Each SSP is describedbyaqualitativenarrative(Kriegleretal.,2012).Fourof the narratives(SSP1,SSP3, SSP4, and SSP5),are defined bythe variouscombinationsofhighorlowsocio-economicchallengesto climatechangeadaptationandmitigation.Afifthnarrative(SSP2) describes medium challenges of both kinds and is intended to representafutureinwhichdevelopmenttrendsarenotextremein any of the dimensions, but rather follow middle-of-the-road pathways.Aspartofthescenariodevelopmentprocess,consistent andharmonizedquantitativeelaborationsofpopulation;urbani- zationandeconomicdevelopmenthavebeendevelopedforallthe SSPs.ThequantitativeelaborationsoftheSSPnarrativesarethen referredtoas‘baseline’scenarios.
TheSSPnarrativesthemselvesdonotincludeexplicitclimate policies.However,additionalclimatemitigationrunshavebeen developedthatincludeforeachSSPbaseline,additionallong-term radiativeforcingtargetsof2.6,4.5and6.0W/m2in2100.Climate mitigation scenarios in the SSP framework further include a numberofadditionalassumptionsonspecificissuesrelatedtothe level of international cooperation;the timingof themitigation effortovertime;andtheextentoffragmentation(particularlyin the short-to medium-term). These are characterized as shared policyassumptions(SPAs)whichdescribeforeachSSPnarrative, the most relevant characteristics of future climate mitigation policies,consistentwiththeoverallSSPnarrativeaswellastheSSP baselinescenariodevelopments.ThemitigationeffortoftheSSP scenariosisthenafunctionofboththestringencyofthetargetand theunderlyingenergyandcarbonintensitiesinthebaselines.This couldresultinsomecasesin infeasibilitiesintermsof meeting mitigationtargets(foracompleteoverviewoftheSSPbaselineand climatemitigationscenarios(seeRiahietal.,2017).
AnumberofIAMsrantheelaborationsofSSPscenarios.These include IMAGE (van Vuuren et al., 2017); MESSAGE-GLOBIOM (Frickoetal.,2017);AIM/CGE(Fujimorietal.,2017);GCAM(Calvin etal.,2017);REMIND-MAgPIE(Kriegleretal.,2017);andWITCH- GLOBIOM(Emmerlingetal.,2016).Detailedinformationonthe modelscanbefoundintheSupplementaryInformation(SI).For simplification,foreachofthefiveSSPs,onemarkerIAMhasbeen identified(representativeofaspecificSSPfromasingleIAM).The selection was guided by consideration of internal consistency
acrossdifferentSSPinterpretationsaswellastheabilityofamodel torepresentthespecificstorylines.Thishelpedtoensurealsothat thedifferencesbetweenmodelswerewellrepresentedinthefinal setofmarkerSSPs.AdditionalreplicationsoftheSSPsfrom‘non marker’ models then provide insights into possible alternative projectionsofthesamestoryline.Themulti-modelapproachwas importantforunderstandingtherobustnessoftheresultsandthe uncertaintiesassociatedwiththedifferentSSPs.
Table1summarizestheSSPscenarioset.
2.2.PollutioncontrolintheSSPnarratives
In thissection,wenowdescribethedevelopmentofa setof assumptionsonpollution controlthatcanbeusedtoguidethe interpretationofSSPnarratives.
Whilethereisnouniquerelationshipbetweeneitherpollutant levelsoremissioncontrolsandincome(Stern,2005;Carson,2010;
Smithetal.,2005),acontinuedtighteningofpollutiontargetscan beconsideredaconsequenceofgrowingattentiongiventohealth outcomeswithincreasingincome,orperhapsalsoasa resultof new research that ties additional morbidity and mortality modalitiestoair pollution.Theadverseimpactsof airpollution are well documented and costs of control technologies have generallydeclinedovertime.Thismeansthatdevelopingcountries canbenefitfrom past experienceand haveoften implemented pollutioncontrolswellinadvance,relativetoincome,ascompared to historical experience in currently more affluent regions.
Countries have, however, different physical, economic and institutional circumstances that impact both the amount and effort needed to achieve pollution goals. Pollutant emission densitiesinthedevelopingworldaresometimesquitehighand, evenwithmoreadvancedtechnology,reachingpollutiontargets maybemore difficult.The samelevel of pollution controlwill resultindifferentconcentrationlevelsindifferentlocations.
Policies to control the adverse impacts of air pollution are numerous and regionally diverse. They are generally aimed at avoidingexceedingspecifiedtargetsforconcentrationlevels(for example,sulfur-di-oxide,ozone,andparticulatematter)butgoals forecosystemprotection(e.g.,fromacidificationandeutrophica- tion)havealsobeenpursuedinseveralregions.Pollutiontargets areperiodicallyrevisedatboththegloballevel(e.g.WHO)andby nationalandregionalbodies.Levelsofpollutioncontrolarealso often different across sectors. Further, in some circumstances, traditional‘end-ofpipe’pollutioncontrolmayhavelessofarolein reducingemissionsthantheeffectsofsocio-economicgrowthand related fuel and technological shifts (Rafaj et al., 2014). Thus
‘pollutioncontrol’itselfcouldrefertoawiderangeofpoliciesand developments.For example, policiesaddressingclimate change
often, as a co-benefit, reduce atmospheric emissions, thus improvingambientairquality(McCollumetal.,2013;vanVuuren et al., 2006; Bollen, 2008) . Conversely, policies targeting air pollutionwillhavealsoclimateimpacts,e.g.,(Carmichael,2008;
Shindelletal.,2012),althoughclimateco-benefitsmaybesmaller thanpreviouslyexpected(Smithand Mizrahi,2013;Stohletal., 2015).Technologicalavailabilitycanalsobeakeyinfluenceonthe degreeofpollutioncontrol,especiallyiffeworonlycostlyoptions areavailable.Inpracticedamagesare,eitherimplicitlyorexplicitly, balancedagainsttheeconomiccostsofpollutioncontrol,forwhich technologycharacteristics,particularlycostsofpollutioncontrolor loweremissionalternativesareakeydriver.
We cannot capture all these complexities within current integratedscenarios.Wefirstsimplifyourapproachbyidentifying threecharacteristicsforairpollutionnarratives:
1.Pollutioncontroltargets(e.g.concentrationstandards),which wespecifyrelativetothoseincurrentOECDcountries.
2.Thespeedatwhichdevelopingcountries‘catchup’withthese levelsandeffectivenessofpoliciesincurrentOECDcountries.
3.Thepathwaysforpollutioncontroltechnologies,includingthe technologicalfrontierthatrepresentsbestpracticevaluesata giventime.
Basedonthesecharacteristics,wedevelopedthreealternative assumptions for future pollution controls (strong,medium and weak),whicharefurthermappedtospecificSSPscenarios.This terminologyfollowsthesameconventionasotherstudiesusedto informtheSSPscenariodesignprocess(KCandLutz,2017;Crespo Cuaresma,2017).
Themediumpollutioncontrolscenario(SSP2)envisionsaworld thatcontinues followingcurrenttrends.Duetothediffusionof technology and knowledge, there is some ‘catch-up’, where countriesachieve levelsof emission controland policyefficacy inadvance,intermsofincomelevels,ofthehistoricalrecordin currentOECDcountries. Pollutionconcentrationtargetsbecome moreambitiousoverthecenturyasincomegrows,thecommit- menttosetandenforcepollutiontargetsbecomingincreasingly effective,and more valueis placed onhealth andenvironment protection. Toreachthesetargets,someregionswill ultimately require implementation of very efficient technologies, some perhaps requiring advances over current technology levels.
Regions with large population densities or adverse physical conditions (e.g. geographic features that lead to frequent high pollutionepisodes)maynotachievetheirdesiredoutcomes.
Thestrongpollutioncontrolscenarios(SSP1andSSP5)assume that increasing health and environmental concerns result in successful achievementof pollutant targets substantially lower Table1
Summaryofscenarios.
Identifier Descriptor MarkerIAM Alsocomputedby(non-marker IAMs)
CentralSPAassumptionsforClimateMitigation
SSP1 Sustainability IMAGE(van Vuurenet al., 2017)
All Earlyaccessionwithglobalcollaborationasof2020
SSP2 Middle-of-the- road
MESSAGE- GLOBIOM (Frickoetal.,2017)
All Somedelaysinestablishingglobalactionwithregionstransitioningtoglobal cooperationbetween2020and2040
SSP3 Regional rivalry
AIM/CGE(Fujimori etal.,2017)
IMAGE,GCAM,MESSAGE- GLOBIOM,WITCH-GLOBIOM
Lateaccession higherincomeregionsjoinglobalregimebetween2020and2040, whilelowerincomeregionsfollowbetween2030and2050
SSP4 Inequality GCAM AIM/CGE,WITCH-GLOBIOM SameasSSP1
SSP5 Fossil-fuelled development
REMIND-MAgPIE AIM/CGE,GCAM,WITCH- GLOBIOM
SameasSSP2
thancurrentlevelsinthemediumtolongterm.Associatedwith this scenario is a faster rate of pollution control technology development,withgreatereffectivenessascomparedtocurrent technologies. The ambitious air quality goals in the strong pollution control scenario would require, in some regions, implementationofcurrentbestavailabletechnology(andperhaps evenbeyond) andassureoverallenforcementof environmental lawssupportedbyefficientlyoperatinginstitutions.
Weakpollutioncontrolscenarios(SSP3andSSP4)assumethat the implementation of pollution controls is delayed and less ambitiousinthelong-termcomparedtothemediumscenario.This maybeduetothelargechallengesseveralregionsface,including, highemissiondensitiesindevelopingcountries’megacities,failure to develop adequate air quality monitoring, and/or weaker institutionsresultinginpoorenforcementofrespectivelegislation.
Theproblemsareaggravatedbytheassumptionthatinternational cooperation is weaker resulting in low ambition or slow developmentofinternationallawsthatalsoleadstoslowerrates of technological improvements and trans-boundary pollution contributestohigherbackgroundconcentrationsinmanyregions.
These pollution control storylines are matched to the SSP scenario narratives as shown in Table 2. The strong pollution controlnarrativeisassumedfortheSSP1andSSP5scenariosdueto theirhigh levelsof development, focus onhuman capital, and reduced inequality. Conversely, we associate the low pollution controlnarrativewiththeSSP3and SSP4scenariosdue totheir lower levels of development and greater inequality. The SSP2 scenarioismappedtothemediumpollutioncontrolnarrative.The speed and absolutevalue towhich countrygroups convergeis differentiated across the SSPs. While we qualify three sets of assumptionsonpollutioncontrolthataremappedtothefiveSSP scenarios,wenotethatevenwithsimilarassumptionsonpollution control,pollutionoutcomesinspecificSSPscenarioswilldifferdue to varying assumptions on economic and population growth, energyconsumptionpatterns,andotherscenariocharacteristics.
2.3.ImplementationinIAMs
For quantitative interpretation of the storylines, there is a further need to bridge the gap between the complexity in estimatingpollution emissionsand theirimpacts,the abilityof availablemeasures,suchasemissioncontrols,tomitigatethese impacts, and the need for simplified representations of these processes in IAMs. Giventhat IAMs do not generally represent explicitairpollutioncontroltechnologiesonadetailedlevel,we detailbelowanapproachwherescenarioparametersarebroadly
representedintermsofchangesinemissionfactorsderivedfroma moredetailedairpollutionmodel.Thisapproachhasbeenusedin a numberof recentstudies(Riahietal.,2012)and allowsfora relativelysimplisticmethodtorepresentquantitatively,concepts relatedtothespeedand degreeof implementationof pollution controldevelopedanddescribedearlier.
We base ourquantitative guidance ona datasetof regional emission factors(i.e.,emissionsper unitof energy)for energy- relatedcombustionandtransformationsectorsuntil2030based on currentpolicies and technologicaloptions derived fromthe GAINSmodel(Amannetal.,2011,Klimontetal.,inPreparation).
This datasetincludesemission factors for 26 worldregions for sulfurdioxide(SO2),nitrogenoxides(NOx),organiccarbon(OC), blackcarbon (BC),carbonmonoxide(CO),non-methanevolatile organiccarbons(NMVOC),and ammonia(NH3)fromall energy combustion and process sources. The detailed emissions factor datawasprocessedtoaccommodatetheaggregatestructureand resolution of the IAMs (see supplementary information (SI) Section1forfurtherdetails).Theemissionfactorsusedinclude:
CLE: ‘current legislation’ These emission factors assume efficientimplementationofexistingenvironmentallegislation.
Itthusdescribesascenarioofpollutioncontrolwherecountries implement all plannedlegislation until 2030withadequate institutional support. The CLE emission factors are “fleet average” valuesthat aretheaggregateemission factorof all agesofequipmentoperatinginthegivenyear.
MTFR: ‘maximum technically feasible reduction’ These emissionfactorsassumefullimplementationof‘bestavailable technology’asitexiststodayby2030independentoftheircosts butconsideringeconomiclifetimeoftechnologiesandselected other constraints that could limit applicability of certain measures in specific regions. While, the full penetration of MTFRmeasuresinthenear-termisnotafeasiblescenario,these values serve rather as ultimately achievable air pollutant emission factors for conventional technologies considered beingavailableatthepresenttime.
Inordertodeveloptrajectoriesforemissionfactorsthatcould beconsistentwiththeSSPstorylines,wedrawonexperienceand results from a number of existing and forthcoming studies including(Raoetal.,2013;Riahietal.,2012)wheresimilarsets ofemissionfactorshavebeenusedinasingleIAMinconjunction withafullscaleatmosphericchemistrymodel,thusprovidingan indicationoftheimplicationofsuchemissionfactordevelopment in terms of resultingatmospheric concentrations of PM2.5 and correspondinghealthimpactsin themedium-term.Weidentify twomaincomponentsintermsofemissionfactordevelopment:
Table2
QualitativeframeworkforpollutioncontrolintheSSPs.
Policy strength
Policytargets Technological
innovation
SSP link
KeyrelevantcharacteristicsofSSPs
HighIncomecountries MediumandLowincomecountries Strong Policiesoverthe21stcenturyaimformuch
lowerpollutantlevelsthancurrenttargetsin ordertominimizeadverseeffectson population,vulnerablegroups,and ecosystems.
Comparativelyquickcatch-upwiththe developedworld(relativetoincome)
Pollutioncontrol technologycostsdrop substantiallywith controlperformance increasing.
SSP1, SSP5
Sustainabilitydriven;rapid developmentofhumancapital, economicgrowthand
technologicalprogress;prioritized healthconcerns
Medium Lowerthancurrenttargets Catch-upwiththedevelopedworldat incomelevelslowerthanwhenOECD countriesbegancontrols(butnotas quickasinthestrongcontrolcase).
Continuedmodest technologyadvances.
SSP2 Middleoftheroadscenario
Weak Regionallyvariedpolicies. Tradebarriersand/orinstitutional limitationssubstantiallyslowprogress inpollutioncontrol.
Lowerlevelsof technologicaladvance overall.
SSP3, SSP4
Fragmentation,inequalities
Until 2030, emission factors assumed in the different SSP scenarios reflectassumptions ontheattitudestohealth and environment and the institutional capacity to implement pollution control in thenear-term. They include full imple- mentationof CLEpollution controlmeasuresinthemedium scenariobutallowforpartialandadditionalcontrolintheweak andstrongpollutioncontrolscenarios.
After2030,thetrajectoriesareassumedtodependontheextent to which economic development implies that lower-income regions catch-uptoOECD levelsin termsof implementation (e.g.emissionfactorreductions)andtheextentoftechnological change, i.e., the progress towards MTFR levels of emission factors.TheMFTRvaluesareassumedtobestaticthemselves anddo notchangewithtimeandwedonotspeculateabout impact of innovation on further improving the reduction efficiencyof the bestmeasures weincluded. Thus, while in some sense, we may beconservative for the pathways and regionswithhighpenetrationofMTFRequivalenttechnology, ontheotherhand,giventhatmostMFTRvaluesherearebased oncurrent small-scale applications,we assumethat techno- logicalprogressinthescenarioswillmaturethesetechnologies andallowforwideapplicationoverthelonger-term.
Fig.1showsaconceptualrepresentationofthedevelopmentof pollutioncontrolpolicyandassociatedemissionfactorchangein thedifferentSSPs.Amoredetailedillustrationofhowtheemission factorsinthedatasetcanbeusedtoemulatetheaboveguidelines ispresentedinsection1.2oftheSI.
TheIAMsusetheemissionfactordataprovidedandquantita- tiveguidelinesdescribedtoindividuallydeveloptheSSPscenarios.
Theemissionfactorsareimplementedin thebaseline scenarios describing the SSP narratives, while the climate mitigation scenariosthendescribetheadditionalimpactsofclimatepolicies on air pollution emissions and air quality, compared to the baselines.Thus, theclimatemitigationscenariosdo notinclude furtherpoliciesonairpollutioncontrolcomparedtothebaseline scenarios.It is important tonotethat themodelsusedifferent inventoriesforthe2000–2010periods,andarenotbenchmarked toasinglesource.Thedifferencesacrossmodelsinthisperiodthen reflecttheuncertaintyininventorydataandtosomeextent,the regional and sector aggregation of the IAMs. For land-use, international shipping, and other sectors not covered in the emissionfactordataset,additionalassumptionsaremade(seeSI [3943]formoredetailsoninventoriesanddriversforemissions acrosstheIAMs.).Theassumptionsformethane(CH4)fromenergy, wasteandland-usesectorsareseparatelydescribedinBaueretal.
(2017)andPoppetal.(2017)andsummarizedintheSI.
3.Results
Inthissection,wesummarizekeyresultsfortheSSPscenarios in termsofair pollutionemissionsand regionalairquality. We describethefullrangeofmarkerandnon-markerrangesforthe SSPscenarios. In termsof climatemitigation,weonly focuson centralSPAcaseforeachSSP.
Results are mainly presented at a global scale and further discussedforfiveaggregateregions:
OECD90countriesand newEUmemberstatesandcandidates (OECD);
reformingeconomiesofEasternEuropeandtheFormerSoviet Union(excludingEUmemberstates)(REF);
countriesoftheMiddleEastandAfrica(MAF);
countriesofLatinAmericaandtheCaribbean(LAM);and Asiancountries(withtheexceptionoftheMiddleEast,Japanand
FormerSovietUnionstates)(ASIA).
3.1.Emissionsofselectedairpollutants
Fig. 2 shows potential emissions futures across the SSP scenarios in the 2005–2100 period for selected pollutants.
Results forremaining pollutantsare summarizedin theSI.We includeemissionrangesfromtheRCPscenariosetaswellasthe entirerangeofscenariosfromtheIPCCFifthAssessmentReport, in order to place the SSP scenarios in context. Differences in historicalemissionsbetweenthemodels(2000–2010)aredueto use of different inventories by IAMs (Table S1 and individual modeldescriptions)andarewithinuncertaintyranges(Granier et al., 2011; Lamarque et al., 2010). For example, for SO2, historical global emissions uncertainty has been estimated at about 10%, with larger uncertainties for some regions (Smith et al., 2010). Uncertainty is much larger for black carbon emissions, estimatedtobea factorof two (Bond etal., 2004).
Beyond uncertainties in activity data and emissions factors, additional aspectsinclude the relatively aggregate representa- tion ofsectorsin IAMsandthelargeuncertainties inland-use and land-use change emissions (see Popp et al., 2017 for full descriptionof land-usesector).
TheSSP3baselineshowsanincreaseinfutureemissionsover theshort-termacrossallpollutantsexaminedhere,duetolarge population growth and relatively slower and heterogeneous economicgrowth.Atagloballevel,emissionscontinueincreas- ingforthenexttwotothreedecadesandby2100showonlya slightdeclinefromcurrentlevels.TheSSP4baseline,whichhas
Fig.1.ProposedPathwaysforAirPollutionPolicyinSSPsovertime.Righthandinsetshowsschematicdevelopmentofemissionfactors.Weusehereidenticaldefinitionsof incomecountrygroups(lowincome(L)countries,middleincome(M)countries,andhighincome(H)countries)asusedintheSSPprocessfordevelopmentofeconomic projections,basedonrecentWorldBankclassifications.https://secure.iiasa.ac.at/web-apps/ene/SspDb/static/download/ssp_suplementary%20text.pdf.
identical assumptions on pollutant controls, shows lower emissionsthan SSP3 for all pollutants as a result of different evolutionoftheenergysystem(seetextbelow).TheSSP2shows a consistent decline in all pollutants throughout the century whileSSP1 and SSP5 exhibit a more rapid decline as a result of more effective pollution control and lower fossil fuel intensities resulting in lowest emissions in the second half of thecentury.
PollutantemissionsintheSSP scenariosspanacrossa much wider range than the RCP scenarios. In general, baseline SSP3 emissionsaresignificantlyhigherthanthelargestRCPvalues,with NOxandBCemissionsintheSSP1baseline caselowerthanthe lowestRCPvalue.Whilescenariodynamicsandassumptionson transportation and access to clean energy for cooking in developingcountriesaremajor driversofemissionoutcomesof NOx and BC, respectively, anotheraspect is the updated setof pollutantcontrolassumptionsandtheemissionfactorsusedinthis study.Resultsforremainingpollutantsshowsimilartrends(see SI).
Theclimate mitigation scenarios (Fig.2 illustrates 4.5W/m2 (45)and2.6W/m2(26)cases)resultinmostcasesinco-benefitsin termsoflowerpollutantemissionsthanthebaselines.Thelargest co-benefits from climate policy occur in the weak pollution control,SSP3scenario,whichalsohasthehighestcorresponding baselineemissions,whiletheSSP1/5scenariosshowmorelimited reductionsinairpollutantsfromclimatepolicies.WhileSO2and NOxemissionsshowthelargestreductionsandthemodelranges within the SSPs are much smaller than in baseline cases, BC emissionsdonotdeclineasmuchasa resultofassumptionson fuel-substitution in the residential sector (see discussion in Section3.3).
3.2.Emissionintensities
Fasteconomicgrowthandhighemissionintensities(emissions perunitofenergyused)inmanyAsiancountrieshaveledtosevere pollutionepisodesacrossthecontinent.Inspiteoftheeffortstocut airpollutantemissionsfromkeysources,theintensitiesremain well abovethose observedinOECDcountries(Fig.3)where air qualitystandardsarepresentlythehighest.Emissionintensitiesin theOECDarethusalreadylow,andplannedlegislationisexpected toreducetheseevenfurtherby2030.
In the SSP baselines, emission intensities in ASIA decline significantlyby2050inallSSPs.Economicgrowthandtheaverage incomeinASIAin2030differssignificantlyacrossSSPs,withalow valueof10billionUS2005$inSSSP3andahighvalueof28billion US2005billion$ in SSP5 (see also (Crespo Cuaresma, 2017) for detailsoneconomicassumptionsinSSPs).Thus,countriescouldbe expected to adopt pollution controls with varied schedules, dependingonindividualinstitutional,financialandtechnological capacities(seepreviousdiscussioninSection2).
Therelativecontributionofpollutantcontrolmeasuresinterms ofactualreductionsinairpollutionwilldependontheSSPbaseline pathway. Major energy transitions in the SSP scenarios occur graduallyandassumptionsforpollutioncontrolcanbeassumedto be particularly important in the first few decades in terms of reducingemissionintensities.Forexample,coalbasedelectricity evolves relatively similarly until 2050 across the SSPs and consequentlythedifferencesindevelopmentofemissionintensi- tiesinASIAwithinthistimeframeisadirectreflectionofpollution control.
Over thelonger term, the scenarios divergesignificantly in termsofenergyandfuelstructures.TheSSP1andSSP5baselines Fig.2.EmissionsofSO2,NOXandBCinSSPmarkerbaselines(Ref)and4.5(labeledas45)and2.6(labeledas26)W/m2climatemitigationcases.Shadedareaindicatesrangeof totalemissionsfromRCPscenariorangefrom(vanVuurenetal.,2011a).AssessmentReport(AR5)rangereferstothefullrangeofscenariosreviewedintheFifthAssessment Report(AR5)ofWorkingGroupIIIoftheIntergovernmentalPanelonClimateChange(IPCC)https://tntcat.iiasa.ac.at/AR5DB/;Historicalvaluesarederivedfrom(Lamarque etal.,2010);Coloredbarsindicatetherangeofallmodels(markersandnon-markers)in2100.
showatransitiontowards lesspollutingfuelsand technologies, andthus result ina rapid and sustainedreduction in emission intensities in ASIA. Conversely in the SSP3 and SSP4 worlds, relatively weaker technological change and higher fossil fuel intensitiesintheenergysystemleadtohigherlevelsofpollutant emissions.TheSSP2scenarioshowslarge-scaleelectrification-for example,electrificationinASIAgrowsrapidlyandby2030hasa similar share of final energy as current OECD levels. In the transportationsector,liquid fuels arethemajorfuel untilmid- centuryinallSSPscenarios.TheSSP1showsonlyaslightdeclinein liquids while, SSP5 shows the largest increase. This reflects alternativenarrativesoffuturemobilityresultingfromdifferences inlifestyles,preferencesandtechnology.
WenotethatforBCemissionsfromtheresidentialsectorin ASIA,emissionintensitiesremainhighthroughoutthecenturyin theSSP3andSSP4baselinescenariosmainlybecauseofcontinued biomassuse.IntheSSP3scenario,forexample,biomassuseinASIA iscloseto20EJin2100,almostthesameastoday’slevels.Inthe SSP1, the assumption of rapidly increasing access to cleaner cookingfuelsmeansthatBCemissionsdeclinesubstantiallyandby 2030emissionintensitiesconvergetoOECDlevels.
Assuming properenforcementofairpollutionpoliciesinthe OECDregion,climatepolicieshaveverylittleimpactintermsof pollutantemissionintensities.InASIA,climatepoliciesdecrease emissionintensitiesforSO2andNOx,withmorelimitedimpacton BC,infact,aslightincreaseisindicatedintheSSP3scenario(see discussiononsectorimpactsofclimatepoliciesandco-benefitsin Section3.3).
3.3.Sectoremissions
The SSP scenarios offer a wide diversity of future growth patterns and how they relate to regional energy demand
convergenceandmodernization ofenergy use(seeBaueret al., 2017fordetails).Inordertounderstandtheimpactsofalternative energy developments, we look at broad developments of pollutantsacrosssectors(Fig.4).
3.3.1.Baselinescenarios
The energy sector emissions are dominated by electricity production,whichcurrentlycontributesamajorshareofSO2and in thedeveloping countries alsoof NOx. Bothemission control assumptionsandtechnologyassumptions,suchasthoseforclean coalornon-fossiltechnologies,canhaveasubstantialimpacton futureemissions.
The industrial sector remains an important source of SO2
emissionsin all SSPbaselines and climatemitigationscenarios throughout the century. Fossil-fuel use in the industrial sector comprisesawiderangeofuses,includingprocessheat,internal combustion engines, and process-specific uses such as steel- makingoverarangeofscales,fromsmallplantsandboilerstolarge manufacturing centers. This sector has significant diversity in regulations on pollutant emissions depending on the type of industry.Experiencesofarhasshownthatindustriallegislation lags behind energy or transportation sector in developed and developing countries. Anotherfactor is that fossil fuels can be difficult to replace in some industrial activities, such as those relatedtohightemperatureprocessheat.Someprocessessuchas steel makingrequire specific fuels like coking coal, which also differinpollutantintensityascomparedtocoal.IntheSSPbaseline cases,SSP2andSSP3showacontinuouslyincreasingcoalusein thissectorwhileitdeclinesinSSP1andSSP5,especiallytowards theendofthecenturyresultinginstrongreductionofemissionsof SO2andNOx.Coaluseinsmallboilers,cokeandbrickproduction industrycanbesignificantsourcesofBC(Bondetal.,2004).Inthe longterm,atransitiontomoreefficientandcleanertechnologies Fig.3.EmissionsintensitiesformajorpollutantsinASIAandOECDinSSPbaselinesand26and45mitigationscenarios(bothmarkerandnon-markerscenariosincluded).
Emissionintensitiesdefineddifferentlyforpollutants;SO2intensityisinreferencetoenergysupply,NOxandBCinreferencetofinalenergyfromrespectivesectors.
willresultindeclineinemissions;intheSSP3scenariothissector hasasignificantshareofBCemissionsuntilmid-century.
The transportation sector is a major source of NOx and BC emissionsthroughatleastmid-centuryinnearlyallSSPscenarios.
Asdiscussedearlier,continueduseofliquidfuelsmeansthatNOX
emissionsfromthe transportsector remainrelativelyhighand onlydeclineinthesecondhalfofthecentury.Thesedifferencesare broadly reflected at theregional level as well (SI). The end of centurydecreaseintheSSP1isduetothewidespreadadoptionof hydrogen-fueledvehicles.Inthenextdecades,however,NOXand BCemissionsstillremainrelativelyhighevenintheSSP1scenario, mainlyduetothelargeincreaseinliquidfueluseoffsettingthe increasingstringencyoflegislation,particularlyinASIA.
TheresidentialsectorisamajorsourceofBCemissionsaswell asotherproductsofincompletecombustionlikeorganiccarbon (OC)and carbon monoxide (CO). Except for SSP1 and SSP5,BC emissions from this sector remain fairly constant until mid- centuryacrossallSSPsbutthendeclinesubstantiallyinthesecond halfofthecenturyexceptintheSSP3andSSP4scenarios.Thelatter scenariosassumesustaineduseoftraditionalbiomassthroughout thecentury.Thissubstantiatesrecentfindingsthatemissionsfrom thebuildingssectoraredrivenmorebyassumptionsaboutenergy accessthanexplicitpollutioncontrols(Raoetal.,2016).
Emissionsfrominternationalshippingreflectassumptionson thelevelofimplementationofproposedinternationalregulations inthenear-termaswellasspecificassumptionsonthechangesin fueluseinthebaselinesandclimatemitigationscenariosoverthe longer-term(seeSIforassumptions).TheInternationalConvention for the Prevention of Pollutionfrom Ships or MarinePollution Convention(MARPOL)AnnexVI(IMO,2006)setslimitsonsulfur contentoffuelsandNOxemissionsfromshipexhaust.Whileto someextenttherearedifferencesacrossSSPsintermsoflevelsof implementationofsuchprotocols,weseethatemissionsinallthe baselines show a downward trend for SO2 emissions (50–80%
declinecomparedto2005in2030).
The land-use sector(including openbiomass burning) is an importantsourceofBCemissions(closeto30%ofBCemissionsin 2005).TheassumptionsmadebyIAMsforthissectorvaryquite substantially in their level of detail (see SI for details). The developmentofairpollutantemissionsfromthissectordoesnot necessarilyfollowtheassumptionsdrivingtheairpollutionpolicy intheSSPsbutrather,landusepracticesrelatedtodeforestation andsavannahburning.Inmostscenariosemissionsfromlandopen burningchangeonlymarginallyinthemid-termwiththelong- termtendencytodecline,especiallyintheSSP1.
3.3.2.Climatemitigationscenarios
The emission responsesto a carbon policycan generallybe linkedtochangesinfuelconsumptionorchangesinunderlying technologies.SeeSIforprimaryandfinalenergydetailsintheSSP scenarios. The intensity of the climate policytarget is also an importantfactor;althoughmorestringentmitigationtargetsasin the26scenariodonotnecessarilyalwaysleadtolargerpollutant reductionscomparedthelessstringent45case.
TheaggregateresponseofSO2emissionstoaclimatepolicyis similarinallSSPs.Thisisduelargelytocoalcombustionbeinga common source of both SO2 and CO2, and a similar relative responsetoaclimatepolicyintheelectricitygenerationsector.SO2
emissions fall in all models as coal-fired electricity production either decreases or shifts to carbon capture and storage (CCS) technologies.Soforexample,SSP4andSSP2showincreasedshares of gas-fired CCSand nuclear power becauseof the highsocial acceptancefortheseoptionsinthosestorylines.Reductionsfroma climate policy are larger in the SSP3 and SSP4 scenarios as compared toSSP1. This canpartly beexplained by theweaker assumptionsonpollutioncontrolintheSSP3/4.Themuchstronger transportationBC emissioncontrolsin theSSP1/5scenario and resulting low emission levels, coupled with substantial use of syntheticfuels,meanthat,inabsoluteterms,thereislessroomfor emissionstofurtherdecreaseasliquidfuelconsumptiondecreases Fig.4.World,Emissionsbysector,BaselinesandClimateMitigationcases.RCPscenariosindicatedforreference.OnlymarkerSSPscenariosrepresented.Valuesfor2005are fromRCP8.5whileerrorbarsshowuncertaintyacrosswholerangeofSSPandRCPscenarios.
underaclimatepolicy.ThelargerbaselinecaseemissionsinSSP3 resultinapotentialforalargerrelativereductionintheclimate policycase.SO2emissionsfrominternationalshippingdropoffby theend ofthecenturyintheclimatemitigationscenarios.This responseismainlyduetotheeffectofhighcarbonpricesinthis sectorandthemovetowardsalternativefuelslikeliquefiednatural gas(LNG)inthissector.
For NOx emissions,we seethatmajorreductionsoccuronly mid-century. Before that, relative inertia in the energy system meansthatliquidfuelsremainanimportantpartofthefuelmixin thissector(closetoormorethan90%).Whilepollutantcontrolsin thissectorarerelativelynumerousandstringentinmanyregions, continued oil use in this sector means that emissions do not declinerapidlyevenintheSSP1/5scenarios.NOxemissioncontrols intheenergysectorareusuallylesseffectivethanSO2controlsand asa result, weobservethat NOxemissions responsefrom this sectorisless thanthat ofSO2 (seeSI for summaryof assumed controls).
TheBCemissionsreductioninresponsetoacarbonpolicyis smallerandwefindthatforCO2emissionreductionsofuptoabout 50%,mid-centuryin the45 and26 scenarios, BCemissionsare generally only reduced by 10–20%. The scenarios show a substantial reduction in BC emissions from the transportation sectorduetoreductionsinliquidfuelconsumptionandshiftto electricity,hydrogen,electricity,andbiomass-basedliquids.There isrelativelysmallresponseintheindustrialsectorBCemissionsto climatepolicy,duetothelimitedscopeforreductionsinthissector, thecontinueduseofliquidfuels,andarequirementforsomelevel of carbonaceous fuels. These differences in response in the industrialsectoraredue, inpart,todifferentrepresentationsof industrial fuel demand in these models. Traditional biomass consumptionintheresidentialsectorisonlymildlyimpactedbya climatepolicyinallofthemodels,withmostoftheshiftsalready occurringinthebaselinesduetootherpoliciesandassumptionson energyaccess.Forexample,intheSSP1 scenariowithrelatively rapidratesofmodernizationindevelopingcountriesandaswitch tocleanerorlesspollutingsourcesforcooking,climatepolicydoes notbringadditionalreductions.Althoughnotexplored indetail here,wenotethatitispossiblethatclimatepolicymaynegatively impactemissionsfromthissectorasaresultofhighcarbonprices whichmayinsomecasesresultinanincreaseinbiomassusefor cookingin developingcountriesintheshort-term(seealsoRao etal.,2016).
4.Rangesforregionalairqualityoutcomes
In orderto gain an initialunderstandingof theregional air qualityoutcomes acrossSSP scenarios, we estimate air quality undertheSSPscenariosusingTM5-FASSTmodel(VanDingenen et al.,2009), a reduced-form global air quality source-receptor model (AQ-SRM). This allows us to provide an approximate estimateofairqualityoutcomes,althoughasnotedbelow,more detailed analysis, for example in CMIP6, is warranted. This approachoflinkingemission outcomesfromIAMstoa reduced formairquality modeland allowsus tocomputemulti-model, multi-scenarioairqualityoutcomes(Raoetal.,2016)(seeSIfor detaileddescriptionoftheFASSTmodelanditsapplicationtothe SSPscenarios).WeestimateannualaveragePM2.5concentrations (fineparticulatematterwithdiameterlessthan2.5
m
m)aswellassix-month average ozone concentrations (Fig. 5). We further provideacomparisonofthefractionofpopulationexposedacross the SSP scenarios to WHO levels defined as recommended maximumexposurelevel orairqualityguideline (AQG)(10
m
g/m3)andtwointermediatelevels(35
m
g/m3and25m
g/m3)(WHO,2006). For this purpose, we use here as a basis, a median populationtrajectory(Riahietal.,2012),whichiscomparableto
the SSP2 and SSP4 population projections in 2050 (see SI for comparison of population across the SSP scenarios). Thus, our resultsaspresentedheredonotreflectthediversityinregional populationgrowth across therange of SSPnarratives and only reflect the differencesin assumptions onpollution controland underlying energy and land-use development. Future analysis usingSSP-specificspatiallyexplicitpopulationestimateswillbe useful in enhancing our understandingof in terms of changes within a region due to major shifts in population distribution patterns.
We find that the range of PM2.5 and ozone levels for the differentSSPscenariosisconsistentwiththeRCPrange(whichwas estimated using the same model and population basis), but displaysalargervariabilityamongtheSSPvariants.Differencesare largestinparticularin ASIA,inlinewiththewiderdiversityin growthpatternsreflectedinthepollutantemissiontrends.Inall regions,thefullrangeofmodeloutcomesfortheweakpollution control scenarios (SSP3/4) showsignificantly higher concentra- tionscomparedtothosewithstrongpollutioncontrol(SSP1/5).We alsofindthat,exceptforASIAandtheMAF,inallregions,morethan 95%ofthepopulationiscurrentlyunderthe25
m
g/m3exposurelevelforallscenarios.By2050,OECDcountriesstronglyimprove underallSSPscenarios,reducingconcentrationsfurtherwith80to 95%ofthepopulationexposedtolevelsbelow10
m
g/m3.IntheMAFregion,mineraldustisresponsibleformostoftheexposure above25
m
g/m3,explainingwhyclimateandairpollutionpolicieshavelittleimpactontheexposedpopulation.CurrentlyinASIA, averageconcentrationsarearound25
m
g/m3,andalmost90%ofthepopulationisexposed tolevelsabove10
m
g/m3 and 45%tolevelsabove25
m
g/m3.HoweverthereisawidevariationacrossdifferentpartsofASIA,withChinahavinganaverageof32
m
g/m3;Indiawithanaverageof30
m
g/m3;otherregionshaveanaveragePM2.5 concentration below 10
m
g/m3 and at least 2/3 of thepopulationexposedto10
m
g/m3orbelow.BecausetheASIAmeanPM2.5concentrationisnear35
m
g/m3,apositiveornegativetrendinPM2.5by2050willbereflectedinpopulationexposuretothis limitlevel.Indeed,thestrongpollutioncontrolscenarios(SSP1and SSP5) decrease the population fractionin the above 35
m
g/m3exposureclass toabout15%,whereas thelowpollution control variants(SSP3and SSP4)increasethefractionwith25 and18%
respectively.
By 2050, climate policy leads to substantial co-benefits on pollutionlevelsinASIA,wherePM2.5levelsdecreaseby5–11
m
g/m3 relativeto thebaseline scenario.For the otherregions, the maximalbenefitisaround2
m
g/m3.Thehighestclimatepolicyco-benefitsareobservedinscenariosSSP3/SSP4directairpollution policieswereassumedtobelesseffective,inparticularforASIA (seealsoSI).
Ozoneprecursorsare,ingeneral,moredifficulttocontroland ozonelevelshavealargerimpactfromremotesourcesaswellas increasing methane concentrations. We find that in the SSP scenarios, regional ozone levels do showclear regional differ- encesby2050.ASIAasawholeisnotabletostabilizeozoneat presentlevelsevenunderstrongairpollutionpolicies(SSP1and SSP5),althoughalsointhiscaselargedifferencesintrendsare foundbetweenindividualcountries.India’sozoneconcentrations areestimatedtoincrease(orstabilize)from63ppbvin2005to 2050valuesof63,70,or80ppbvforthelow,mediumandhigh pollution control variant, respectively, while ozone in China decreasesfrom56ppbvin2005to48,50,or53ppbvrespectively in2050.
5.Discussion
The SSP scenarios were developed to include narratives on futureairpollutioncontrolthatareconsistentwithcurrenttrends
inairqualitypolicies;experienceincontroltechnologyapplication
;andregionaldifferencesinaffluenceanddegreeofcontrol.
Thisnewgenerationofglobalscenariosresultsinamuchwider range of air pollution emission trajectories than the RCPs. The baselinerealizationsofSSP3scenariohaveglobalemissionsator abovethehighest levelin theRCPs, whiletheSSP1 scenariois generally near the lower end or below the RCPs. Pollutant emissionsin climatemitigationcasesarelowerstill,withsome SSPtrajectoriesbelowtheRCPemissionlevels.TheSSPscenarios, thus,provideawiderangeoffutureemissions,foruseinglobaland regionalstudiesofclimateandsustainability.
TheSSP1and SSP5scenarios,which includeassumptions on globallysuccessfulimplementationof strongpollution controls, bringthemostsignificantreductionsinairpollutantemissions;by mid-centuryemissionsdeclinegloballyby30–50%inthebaseline
scenariosandupto70%intheclimatemitigationscenarios.The SSP2,middleoftheroadscenario,generallyachievesreductionsby 2100similartoSSP5.IntheSSP3scenario,wherecurrentpollution controlplansarenotfullyachieved,globalpollutantemissionsdo not substantiallydecline andeven slightlyincrease inthemid- term.Inspiteofimprovingemissionintensityinallregions,the improvements in the developing world are too small to offset growthinfossilfueluseandotheremissiondrivers.Evenbythe end of the century when emission intensities in the highest polluting regions decline to the current OECD levels, global emissionsremainhighin SSP3,barelybelowthecurrentlevels.
Exceptforthestrongestclimatepolicycasesconsidered,theair pollutioncontrolpoliciesinSSP3stillresultinrelativelyhigherair pollutantemissions,althoughtherearesignificantreductionsin SO2 and NOX. The emission trajectories for the SSP4 marker Fig.5.Leftpanel:region-populationweightedmeanPM2.5inmg/m3(leftaxis)frommarkerscenario(bluecolorbars)andaveragefromthe3RCPscenarios(greybar), contributionofnaturalPM2.5(hatchedarea)fortheyear2005(leftmostbar)and2050.Green,orangeandredcoloredmarkersindicatethefractionofthepopulationexposed
to<10,<25and<35mg/m3respectively(rightaxis).Rightpanel:meanozoneconcentration(maximal6-monthlymeanofdailymaximumozone).Forthegroupedscenarios
SSP1/5andSSP3/4theconcentrationrepresentsthemeanoftherespectivemarkerscenarios.Errorbarsshowtheconcentrationrange(min/max)ofregionalaveragesfromall modelsinthe(setof)SSPscenariosshown,includingnon-marker.FortheRCPbars,theerrorbarindicatesthemin/maxrangewithinthesetof3RCP2.6,RCP4.5andRCP8.5 scenarios.(Forinterpretationofthereferencestocolourinthisfigurelegend,thereaderisreferredtothewebversionofthisarticle.)