The marker quanti fi cation of the Shared Socioeconomic Pathway 2: A middle-of-the-road scenario for the 21st century
Oliver Fricko
a,*, Petr Havlik
a,*, Joeri Rogelj
a, Zbigniew Klimont
a, Mykola Gusti
a,b, Nils Johnson
a, Peter Kolp
a, Manfred Strubegger
a, Hugo Valin
a, Markus Amann
a,
Tatiana Ermolieva
a, Nicklas Forsell
a, Mario Herrero
c, Chris Heyes
a, Georg Kindermann
a, Volker Krey
a, David L. McCollum
a, Michael Obersteiner
a, Shonali Pachauri
a, Shilpa Rao
a, Erwin Schmid
d, Wolfgang Schoepp
a, Keywan Riahi
a,eaInternationalInstituteforAppliedSystemAnalysis(IIASA),Schlossplatz1,2361Laxenburg,Austria
bLvivPolytechnicNationalUniversity,12Banderastreet,79013Lviv,Ukraine
cCSIRO,306CarmodyRoad,St.Lucia,Australia
dUniversityofNaturalResourceandLifeSciences,Vienna(BOKU),Gregor-Mendel-Straße33,1180Vienna,Austria
eGrazUniversityofTechnology,Inffeldgasse,8010Graz,Austria
ARTICLE INFO Articlehistory:
Received15December2015 Receivedinrevisedform2June2016 Accepted6June2016
Availableonlinexxx Keywords:
Sharedsocioeconomicpathways SSP
Greenhousegasemissions Climatechange
Integratedassessmentmodeling Mitigation
Adaptation
ABSTRACT
Studiesofglobalenvironmentalchangemakeextensiveuseofscenariostoexplorehowthefuturecan evolveunderaconsistentsetofassumptions.TherecentlydevelopedSharedSocioeconomicPathways (SSPs)createaframeworkforthestudyofclimate-relatedscenariooutcomes.Theirfivenarrativesspana widerangeofworldsthatvaryintheirchallengesforclimatechangemitigationandadaptation.Herewe providebackgroundonthequantificationthathasbeenselectedtoserveasthereference,or‘marker’, implementation forSSP2. The SSP2narrative describes a middle-of-the-road development in the mitigationandadaptationchallengesspace.Weexplainhowthenarrativehasbeentranslatedinto quantitativeassumptionsintheIIASAIntegratedAssessmentModellingFramework.WeshowthatourSSP2 markerimplementationoccupiesacentralpositionforkeymetricsalongthemitigationandadaptation challengedimensions.FormanydimensionstheSSP2markerimplementationalsoreflectsanextension ofthehistoricalexperience,particularlyintermsofcarbonandenergyintensityimprovementsinits baseline.Thisleadstoasteadyemissionsincreaseoverthe21stcentury,withprojectedend-of-century warmingnearing4Crelativetopreindustriallevels.Ontheotherhand,SSP2alsoshowsthatglobal- meantemperatureincreasecanbelimitedtobelow2C,pendingstringentclimatepoliciesthroughout theworld.TheaddedvalueoftheSSP2markerimplementationforthewiderscientificcommunityisthat itcanserveasastartingpointtofurtherexploreintegratedsolutionsforachievingmultiplesocietal objectivesinlightoftheclimateadaptationandmitigationchallengesthatsocietycouldfaceoverthe 21stcentury.
ã2016TheAuthors.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1.Introductionandbackground
Studiesofglobalenvironmental changearecharacterizedby deepuncertainty.Manyinterdependentfactorsinfluencehowour worldcouldevolveover time.Theseinclude policychoicesand societalpreferences.Aswehavenomeanstopredictthefutureina highlypreciseway,scenariosareoftenusedasscientifictoolsto explorewhatfutureswecouldforesee,andwhichdecisionstoday
couldmostrobustlyleadtodesiredoutcomes(Riahietal.,2007).In this sense,scenariosare neitherpredictionsnorforecasts.They have instead been described as “stories that happened in the future”(ArmstrongandGreen,2012)andarecreatedbyprojecting aconsistentsetofassumptionsfromtodayintothefuture.These assumptions determine many of the key characteristics of scenarios:howpopulationgrowsanddevelopsovertime,which levelsof educationareachievedwhen, which technologies and energy sources become available, how food is produced and consumed,whichworldviewsandsocial preferencesdominate, andmuchmore.Thespaceanddimensionsthatcanbeexplored arevast,buttomakesensescientifically,itiscrucialthatasingle
* Correspondingauthors.
E-mailaddresses:fricko@iiasa.ac.at(O.Fricko),havlikpt@iiasa.ac.at(P.Havlik).
http://dx.doi.org/10.1016/j.gloenvcha.2016.06.004
0959-3780/ã2016TheAuthors.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).
xxx–xxx ContentslistsavailableatScienceDirect
Global Environmental Change
j o u r n al h o m ep a g e: w w w . el s e v i e r . c o m / l o c at e / g l o e n vc h a
scenario is based on a set of assumptions that is internally consistent. To this end, overarching storylines are typically developed that sketch the general context of scenarios; then, within this context or narrative, policiesand decisions can be varied.
Recently, narratives have been developed for the Shared Socioeconomic Pathways (SSPs) (O’Neill et al., 2015). These descriptionsofalternativefuturesofsocietaldevelopmentspan arangeofpossibleworldsthatstretchalongtwoclimate-change- relateddimensions:mitigationandadaptationchallenges.Togeth- erwiththeRepresentativeConcentrationPathways(RCPs)froma fewyearsearlier(Mossetal.,2010;vanVuurenetal.,2012),they providea toolkitfortheclimatechangeresearchcommunity to carryoutintegrated,multi-disciplinaryanalysis.TheSSPsreflect fivedifferentdevelopmentsoftheworldthatarecharacterizedby varyinglevelsofglobalchallenges[see(Riahietal.,2016,inpress) foranoverview]:(1)developmentunderagreengrowthparadigm (SSP1:Sustainability—TakingtheGreenRoad)(vanVuurenetal., 2016,inpress);(2)developmentalonghistoricalpatterns(SSP2:
Middleof theRoad, thisstudy); (3)a regionally heterogeneous development (SSP3: Regional rivalry—A rocky road) (Fujimori et al., 2016, in press); (4) a development which breeds both geographical and social inequalities (SSP4: Inequality—A road divided)(Calvinetal.,2016,inpress);and(5)adevelopmentpath thatisdominatedbyextensivefossil-fueluse(SSP5:Fossil-fuelled development—Takingthehighway)(Kriegleretal.,2016,inpress).
Here we provide background and information about the quantification of the middle-of-the-roadscenario (SSP2) in the IntegratedAssessmentModelling(IAM)frameworkoftheInterna- tional Institute for Applied Systems Analysis (IIASA). This quantification has been selected as the reference, or ‘marker’, implementation of SSP2, and its results are made available (togetherwiththoseofothermodellingframeworks)asaresource tothewidercommunityinapublicdatabase(https://secure.iiasa.
ac.at/web-apps/ene/SspDb). In the SSP taxonomy, SSP2 is a
“middle-of-the-road” evolution of future societaldevelopments (Box1).Thismeansthatitsobjectiveistocoverthemiddleground intermsofmitigationandadaptationchallengesbetweenmore extreme SSPs, like SSP1 and SSP3 (see Boxes S1 and S2 in SupplementaryInformation,SI).SSP2isconsistentwithdevelop- mentpatterns(e.g.,finalenergyintensityimprovementrates)that have been observed over the past century, but is not a mere extrapolationofrecent trends.TheSSP2 narrativestipulatesan explicitdynamicpathwayinformedbypasttrends,butinwhich future changes are consistent with middle-of-the-roadexpect- ations rather than falling near the upper or lower bounds of possible outcomes (O’Neill et al., 2015). This follows a long
traditionofdynamics-as-usualscenarios(seeSupplementaryText 1).ThebracketingSSP1depictsasustainablefutureinwhichglobal cooperation,lowpopulationgrowthandhigherincomesresultin low challenges of mitigation and adaptation. SSP3 provide a narrativefortheotherextreme(seeBoxesS1andS2inSI).
Theobjectiveofthispaperistoprovideadetailedexplanation of how theSSP2 narrative was translatedinto thequantitative scenario that serves as the marker implementation for the evolution of the future global energy and land system in an SSP2 world. For this, we use the IIASA IAM framework. This frameworkcomprisesacollectionofseveral,uniquedisciplinary modelscoupledtoeachotherforthedevelopmentofcomprehen- sivescenarios. Wefirstprovideanoverviewofthequantitative assumptionsthatwereselectedtorepresentthemaincharacter- istics of the SSP2 narrative. Then, we introduce theIIASA IAM framework and describehow it hasbeenused totranslate the qualitativenarrativesintoaquantitativescenarioofthefuture.The subsequentsection describesthebaseline developments of the energyandlandsystemswithinthisscenarioinabsenceofclimate change mitigation policy, and also explores the impact of increasing climate policy stringency. Where appropriate, a comparisonismadebetweenSSP2andthebracketingSSP1and SSP3 implementations withinthe IIASA IAM framework, which representthetwoextremesintermsofchallengestomitigation and adaptation. The last section then takes a step back and providesanoverviewandconclusions,presentingthemiddle-of- the-roadresultsforSSP2inthewidercontextoftheSSP1andSSP3 narratives.Thispaperthusdocumentsthenovelimplementation of the SSP2 markerscenario in the IIASA IAM framework, and providesafirstassessmentofhowsocietalassumptionsalongthe SSP dimensions translate in varying mitigation and adaptation challenges.
2.Fromnarrativestoquantifiedscenariocharacteristics
Quantifyingpossibleevolutionsoftheenergyandlandsystem inanSSP2worldrequirestheoverarchingnarrative(Box1)tobe translatedintoquantifiedassumptionsforanalysisandmodelling.
What does “middle-of-the-road” mean exactly, in terms of challengestoadaptationandmitigation?Hereweprovideabrief overviewofSSP20seconomicandpopulationdevelopmentsover the 21st century; these are core drivers of the scenarios, particularly for energy services and food demands. We then continuewithalookattheassumptionsmadefortheenergyand land-usesectors.Itisimportanttonotethatsomeofthesecore driverswillalsobefurtheraffectedbyinteractionswithintheIIASA IAM. For example, economic development will be affected by
Box1.SSP2narrativeofamiddle-of-the-roadworld.
“Theworldfollowsapathinwhichsocial,economic,andtechnologicaltrendsdonotshiftmarkedlyfromhistoricalpatterns.
Developmentandincomegrowthproceedunevenly,withsomecountriesmakingrelativelygoodprogresswhileothersfallshortof expectations. Most economies are politically stable. Globally connected markets function imperfectly. Global and national institutions work toward but make slow progress in achieving sustainable development goals, including improved living conditions and access toeducation, safe water, and healthcare. Technological development proceeds apace, but without fundamentalbreakthroughs.Environmentalsystemsexperiencedegradation,althoughtherearesomeimprovementsandoverall theintensityofresourceandenergyusedeclines.Eventhoughfossilfueldependencydecreasesslowly,thereisnoreluctanceto useunconventionalfossilresources.Globalpopulationgrowthismoderateandlevelsoffinthesecondhalfofthecenturyasa consequenceofcompletionofthedemographictransition.However,educationinvestmentsarenothighenoughtoacceleratethe transitiontolowfertilityratesinlow-incomecountriesandtorapidlyslowpopulationgrowth.Thisgrowth,alongwithincome inequalitythatpersistsorimprovesonlyslowly,continuingsocietalstratification,andlimitedsocialcohesion,maintainchallenges toreducingvulnerabilitytosocietalandenvironmentalchangesandconstrainsignificantadvancesinsustainabledevelopment.
Thesemoderatedevelopmenttrendsleavetheworld,onaverage,facingmoderatechallengestomitigationandadaptation,but withsignificantheterogeneitiesacrossandwithincountries.”(O’Neilletal.,2015).
xxx–xxx
investments in the energy system required to mitigate carbon emissions,and thebaseline energyintensitywillbeaffectedby energypricesandstringencyofclimatemitigationaction.
2.1.Populationandeconomicdevelopment
Populationandeconomicdevelopmentshavestrongimplica- tionsfortheanticipatedmitigationandadaptationchallenges.For example,alarger,poorerpopulationwillhavemoredifficultiesto adapttothedetrimentaleffectsofclimatechange(O’Neilletal., 2014). Understanding how population and economic growth develops in the SSPs therefore already gives a first layer of understanding of the multiple challenges. Population growth evolvesinresponsetohowthefertility,mortality,migration,and educationofvarioussocialstrataareassumedtochangeovertime.
In SSP2, globalpopulation steadilygrows to 9.4 billion people around2070,andslowlydeclinesthereafter(KCandLutz,2015).
GrossDomesticProduct(GDP)followsregionalhistoricaltrends (Dellinketal.,2015).Withglobalaverageincomereachingabout 60 (thousand year-2005 USD/capita, purchasing-power-parity– PPP, i.e., GDP/capita) by the end of the century, SSP2 sees an increase ofglobal averageincome bya factor 6.The SSP2 GDP projectionisthussituatedin-betweentheestimatesforSSP1and SSP3,whichreach2100globalaverageincomelevelsof82and22 (thousandyear-2005USD/capitaPPP),respectively.SSP2depictsa future of global progress where developing countries achieve significanteconomicgrowth.Today,averagepercapitaincomein theglobalNorthisaboutfivetimeshigherthanintheglobalSouth (seeSIforregionaldefinitions).InSSP2,developingcountriesreach today’saverageincomelevelsoftheOECDbyaround2060–2090, depending on the region. However, modest improvements of educational attainment levels result in declines in education-
specific fertility rates, leading to incompleteeconomic conver- genceacrossdifferentworldregions.Thisisparticularlyanissue forAfrica.Overall,boththepopulationandGDPdevelopmentsin SSP2aredesignedtobesituatedinthemiddleoftheroadbetween SSP1andSSP3,seeKCandLutz(2015)andDellinketal.(2015)for details.
2.2.Baselineenergyintensityimprovements
Energyintensityimprovementsareamongthekeydistinguish- ingfeaturesoftheassumptionsofthemodelledSSPscenarios–the quantified interpretations of the qualitatively-described narra- tives. These improvements are driven by advances in energy efficiencyand evolving behavioural/lifestyle preferences,which arenotexplicitlymodelled.Historically,theintensityoftotalfinal energy(FEI;finalenergyattheend-uselevelperdollarofGDP) improved at a rate of about 1.2% globally over the 1971–2010 timeframe. Without this improvement, energy use, and by extensiongreenhousegasemissions,wouldbemuchhighertoday thantheycurrentlyare.Energyintensityimprovementsthushave importantimplicationsfortheanticipatedchallengesformitiga- tion.
Fig.1summarizesthebaselineassumptionsforSSP1,SSP2,and SSP3 in termsof theirenergy intensityevolution over the21st century (globally and for the North and South, respectively), highlightingSSP20smiddle-oftheroadpositionwithinthisset.The narrativeofSSP2prescribesthattechnologicaltrendsdonotshift markedlyfromhistoricalpatterns.IntheSSP2baseline(i.e.,when noclimatemitigationeffortsareassumed),finalenergyintensityis thereforeassumedtocontinuetoimproveatapproximatelythe abovementionedaveragehistoricalrate(i.c.,1.3%,seealsoFig.8, below). In contrast, SSP1 and SSP3 assume more extreme
Fig.1.Historicandfuturefinalenergyintensity(totalfinalenergyuseoverGDPPPP)developmentplottedagainstgrossdomesticproduct(GDPPPP)percapita.Thinlines representannualhistoricaldatafrom1900to2010forselectedcountriesbasedonMaddison(2010).OriginalGDPdatafromthissourcehasbeendeflatedfrom1990to2005 usingaUSGDPdeflatoranda10yearmovingaveragehasbeenappliedtotheenergyintensitynumberstosmoothhigh-frequencyfluctuations.Globalmodeldata(thicksolid lines)isprovidedforSSP1,SSP2andSSP3(green,blue,red,respectively)aswellasregionallyaggregateddatafortheglobalNorthandSouth(thickdashedlines).Historical datafortheseregionsfortheperiod1970–2010originatedfromWorldBank(2012).(Forinterpretationofthereferencestocolourinthisfigurelegend,thereaderisreferredto thewebversionofthisarticle.)
xxx–xxx
evolutionsthatimplylowerandhigherchallengestomitigation.
TheSSP1no-climate-policybaselineassumesglobalFEIimprove- mentsof1.7%annuallywhileintheSSP3baselineFEIisassumedto improvemoreslowly(0.2%annually).
Fig. 1 also presents historical energy intensitydata for key countries,whichshowshowthefutureenergyintensityimprove- mentrates of theglobalNorth and South in theSSP2 baseline comparetohistoricaltrends.Fig.1 alsoillustrateshowregional convergenceintermsofeconomicandtechnologicaldevelopment iseither facilitated (inSSP1)or frustrated(inSSP3), withSSP2 providingamiddleground.
Togetherwitheconomicandpopulationdevelopments,energy intensity improvements translate into varying levels of energy demand(presentedinSection4.1),whicharebothaninputandan outputintotheIIASAIAMframework.Adescriptionofhowenergy demandhasbeenderivedisprovidedinSupplementaryText2.
2.3.Fossilenergyresources
Theavailabilityandcostsoffossilfuelswillalsoinfluencethe future direction of the energy system, and therewith future mitigationchallenges.Understandingthevariationsinfossilfuel
Fig.2.Cumulativeglobalresourcesupplycurvesforcoal(top),oil(middle),andgas(bottom)intheIIASAIAMframework.Greenshadedresourcesaretechnicallyand economicallyextractableinallSSPs,purpleshadedresourcesareadditionallyavailableinSSP1andSSP2andblueshadedresourcesareadditionallyavailableinSSP2.
Colouredverticallinesrepresentthecumulativeuseofeachresourcebetween2010and2100intheSSPbaselines(seetoppanelforcolourcoding),andarethustheresultof thecombinedeffectofourassumptionsonfossilresourceavailabilityandconversiontechnologiesintheSSPbaselines.‘Reserves’aregenerallydefinedasbeingthose quantitiesforwhichgeologicalandengineeringinformationindicatewithreasonablecertaintythattheycanberecoveredinthefuturefromknownreservoirsunderexisting economicandoperatingconditions.‘Resources’aredetectedquantitiesthatcannotbeprofitablyrecoveredwithcurrenttechnology,butmightberecoverableinthefuture,as wellasthosequantitiesthataregeologicallypossible,butyettobefound.Theremainderare‘Undiscoveredresources’and,bydefinition,onecanonlyspeculateontheir existence.DefinitionsarebasedonRogneretal.(2012).(Forinterpretationofthereferencestocolourinthisfigurelegend,thereaderisreferredtothewebversionofthis article.)
xxx–xxx
Table1
StorylineelementsandtheirquantitativetranslationinSSP1,SSP2,andSSP3baselines.Allindicatorsapplyto2010–2100;IntensityimprovementsareinFE/GDPannually.
xxx–xxx
availabilityandtheunderlyingextractioncostassumptionsacross theSSPsishenceuseful.Ourfossilenergyresourceassumptions arederivedfromvarioussources(Rogner,1997;Riahietal.,2012) andarealignedwiththestorylinesoftheindividualSSPs.
While thephysicalresourcebaseisidenticalacrosstheSSPs, considerabledifferencesareassumedregardingthetechnicaland economic availability of overall resources, for example, of unconventional oil and gas. What ultimately determines the attractivenessofaparticulartypeofresourceisnotjustthecostat whichitcanbebroughttothesurface,butthecostatwhichitcan beusedtoprovideenergyservices.Assumptionsonfossilenergy resourcesshouldthusbeconsideredtogetherwiththoseonrelated conversiontechnologies.Inlinewiththenarratives,technological changein fossil fuel extraction and conversion technologies is assumed to be slowest in SSP1, while comparatively faster technologicalchangeoccursinSSP3therebyconsiderablyenlarg- ingtheeconomic potentials ofcoal and unconventional hydro- carbons(Table1andFig.2).However,drivenbytendencytoward regionalfragmentation we assume the focus in SSP3 to be on developingcoaltechnologieswhichinthelongertermleadstoa replacementofoilproductsbysyntheticfuelsbasedoncoal-to- liquidstechnologies.Incontrast,forSSP2weassumeacontinua- tion of recent trends, focusing more on developing extraction technologiesforunconventionalhydrocarbonresources,thereby leadingtohigherpotentialcumulativeoilextractionthaninthe otherSSPs(Fig.2,middlepanel).
The regional distribution of fossil energy resources further contributestodifferencesbetweentheSSPs,for example,oil is concentratedintheMiddleEastandNorthAfricawhileRussiaand formerSovietUnionstatesdominatealargeshareofconventional gasresources(Supplementary Text 3).Theemphasis oncoal in SSP3leadstodifferentregionaltradepatternscomparedtoSSP2 whereoilcontinuestobeadominantfuelsignificantlyintothe future.Alltheseassumptionstogetherresultindifferentportfolios of fossil resources being available and used in each SSP (see TableS2inSI).TheuseoftheseresourcesinthevariousSSPsis discussedlater.
2.4.Non-biomassrenewableandnuclearresources
Renewable energyresources(solar,wind,hydro,geothermal) along with nuclear offer an alternative to fossil energy. The variationofthepotentialandcostfornon-biomassrenewables(in particularsolarandwindenergy)acrossourSSPswillthusstrongly impact the perceived climate mitigation challenge. Regional resourcepotentialsforsolarandwindareclassifiedaccordingto resourcequality(annualcapacityfactor)basedonPietzckeretal.
(2014)andEureketal.(inreview).Regionalresourcepotentialsas implementedintheIIASAIAMareprovidedbyregionandcapacity factorforsolarPV,concentratingsolarpower(CSP),andonshore/
offshorewindinJohnsonetal.(inreview).Thephysicalpotentialof thesesourcesisassumedtobethesameacrossallSSPs.However, thepartoftheresourcethatisuseableateconomicallycompetitive costsisassumedtodifferwidely.Consistentwiththenarratives, we assume that SSP1 makes significant progress towards the exploitationofrenewables,whilethereisonlylittleprogressin SSP3. SSP2 here follows a central path. In our calculations, technologicalprogress is determined by income developments andnarrative-specificassumptions(Table1).ThisresultsforSSP2 incostreductionsfornon-biomassrenewabletechnologies(e.g., solarpanels,windturbines)ofabout18–70%from2010to2100 (rangeacrossalltechnologies).Inthegreen-growth-drivenworld ofSSP1,thesereductionsrangefrom20to90%,whileinSSP3they areverymodestreachingmaximally30%by2100.Toaccountfor theintermittencyofsolarandwindenergy,renewableintegration constraintshavebeenintroducedintotheIIASAIAM(Sullivanetal.,
2013).Theseintegrationconstraintsareintendedtocapturethe additional costs and system changes that are required when deploying large shares of variable renewable energy (VRE), includingtheneedforincreasedgenerationflexibilityandbackup capacitytohandleuncertainandintermittentVREgenerationand increasedstorageand/orhydrogenproductiontoavoidelectricity curtailment.
Toallowforabalanceddevelopmentoftheenergyportfolioin SSP2, a technological learning rate comparable to fossil based technologiesisassumedfornuclearpower(30%costreductionby 2100over2010).Itisfurthermoreassumedthatoperatingtimes improveindevelopingcountries(i.e.,theannualfullloadfactor increases from70% in 2010–85% in 2100), allowingdeveloping countriestograduallycatchupwiththeNorth,basedonpercapita incometrajectories.AcomparisontootherSSPimplementationsis providedinSupplementaryText4.
2.5.Bioenergyresourcesanduse
Biomass energy is another potentially important renewable energyresourceintheIIASAIAM.Thisincludesbothcommercial and non-commercial use. Commercial refers to the use of bioenergy in, for example, power plants or biofuel refineries (seeSupplementaryText5),whilenon-commercialreferstothe useofbioenergyforresidentialheatingandcooking,primarilyin rural households of today’s developing countries. Bioenergy potentials differ across SSPs as a result of different levels of competitionoverlandforfoodandfibre,butultimatelyonlyvary toalimiteddegree(Fig.3).Thedriversunderlyingthiscompetition are different land-use developments in the SSPs, which are determined byagriculturalproductivity and global demandfor foodconsumption.Land-usespecificsaredescribedinmoredetail below(Section2.7).Inshort,SSP1experienceslowcompetition betweendifferentpossibleland-useactivitiescomparedtoSSP3, becauseagriculturalproductivityisassumedtoincreaseatalmost doubletherateinSSP1comparedtoSSP3(0.51–0.66%versus0.35%
peryear).Furthermore,fooddemandisassumedtogrowonlyvery slowlyinSSP1comparedtoSSP3(Table1).Reflectingamedium perspectiveonboththesedrivers,theland-usecompetitionfor bioenergyresourcesandtherewithalsothecommercialbioenergy
Fig.3. AvailabilityofbioenergyatdifferentpricelevelsintheIIASAIAMframework forthethreeSSPs.
*Typicallynon-commercialbiomassisnottradedorsold,howeverinsomecases thereisamarket-pricerangefrom0.1–1.5$/GJ(Pachaurietal.,2013)($equals2005 USD).
xxx–xxx
potentialsinSSP2aresituatedinthemiddle(Fig.3).Ultimately,the differencesinbioenergy potentialsacrossdifferentSSPsremain limited.Thisisontheonehandduetothefactthatsubstantial amountsofbiomassareassumedtobesourcedfromfeedstocks suchastraditionalforestmanagementorsawmillresidues,which donotdirectlycompetewithagriculturalproductionandhenceare relativelyunaffectedbythedifferentiatedstorylinesforthatsector.
Atthesametime,justabouthalfoftheenergyplantationsareais assumedtooccurattheexpenseofagriculturalland.Theotherhalf isprojectedtotakeplaceinotherwisenon-cultivatedland.
Thedemandfornon-commercialbiomassiscloselyrelatedto thedegreeofaccesstomodernandcleanfuelsforpoorpopulations indevelopingcountries,whichisderivedfromthepopulationand GDPprojections(Riahietal.,2012;Pachaurietal.,2013)foreach SSP. Adjusted historic traditional biomass consumption (Riahi etal.,2012)was usedtodeterminecurrentbiomassdemandin ruralareas;thesequantitieswerethenprojectedintothefuture withthehelpofurbanizationtrends(JiangandO’Neill,inpress).In SSP3,largesharesoftheglobalpopulationarenotabletoreach income levels that allow for a switch to clean fuels. These populationsthuscontinuetorelyontraditionalfuelsforcooking andheatingandhencethedemandfornon-commercialbioenergy increasesthroughouttheentirecentury.SSP1,ontheotherhand, seekstospurregionaldevelopmentandconvergence,andseesthe demand for non-commercial bioenergy discontinued by 2040.
SSP2takesupthemiddlespotinthisset,withagradualdecline overthecenturyandaphaseoutofnon-commercialbioenergyby 2080.
2.6.Technologycostdevelopments
Primaryenergyresourceslikecoal,biomass,andwind,among others,needtechnologiesfortheirtransformationintoelectricity orothersecondaryenergyforms,suchasliquidorgaseousfuels.
Assumptions must be made about how the costs of these technologieschangeovertime,andtheseassumptionsarecritical astheystronglyinfluencethenatureanddirectionofthebaseline evolutionoftheenergysystem(RoehrlandRiahi,2000).Moreover, the quantitative assumptions should be consistent with the overarching qualitative SSP narrative (Table 1). In SSP1, for instance,whosegreen-growthstorylineismoreconsistentwith asustainabledevelopmentparadigm,higherratesoftechnological progress and learning are assumed for renewables and other advanced technologies that may replace fossil fuels (e.g., the potential for electric mobility is assumed to behigher in SSP1 compared to SSP2 or SSP3). In contrast, SSP3 assumes limited progressacrossahostofadvancedtechnologies,particularlyfor renewables and hydrogen; more optimistic assumptions are instead made for coal-based technologies, not only for power generationbut alsofor liquidfuels production. Meanwhile, the middle-of-the-road SSP2 narrative is characterized by a fairly balancedviewofprogressfor bothconventionalfossilandnon- fossiltechnologies. In this sense,technological developmentin SSP2is notbiasedtowardany particulartechnologygroup.Ifit were,itwouldnotoccupyamiddle-of-the-roadpositionbetween thegreen-growthandfossil-fuelintensiveparadigmsofSSP1and SSP3, respectively. The system-wide, long-term implications of theseassumptions willbecomeclearer in Section4,where the resultsfortheenergysupplymixarediscussed.
Technological costs vary regionally in all SSPs, reflecting markeddifferencesinengineeringandconstructioncostsacross countriesobserved in thereal world.Generally, costsstart out lowerinthedeveloping worldandareassumed toconvergeto those of present-day industrialized countries as the former becomesricherthroughoutthecentury(thus,thecostprojections
considerboth labourandcapital components).Thiscatch-upin costsisassumedtobefastestinSSP1andslowestinSSP3(where differencesremain,evenin2100);SSP2isinbetween.Estimates forpresent-dayandfullylearned-outtechnologycostsarefromthe GlobalEnergyAssessment(Riahietal.,2012)andWorldEnergy Outlook(IEA,2014).Asummaryofthesecostassumptionscanbe foundinSupplementaryFigs.S1–S3.
2.7.Land-usedevelopments
Land-usedevelopmentassumptionsinfluenceprojectedemis- sions and mitigationpotential forthe land-usesectorand thus contributetotheoverallmitigationchallenge.Theydependona multitudeoffactorslikeagriculturalproductivities,feedconver- sion efficiencies, consumption patterns,forest value, and regu- lations,allofwhichplayoutdifferentlyacrossthevariousSSPs.
Agriculturalproductivitygrowth–thekeydriverofland-use requirementsforfoodproduction–isfosteredbyinvestmentsinto new technologies and policies promoting country-to-country knowledgetransfer.Werelateproductivitygrowthtopercapita GDPgrowthineachregion,whichcanbeconsideredaproxyfor thelevelofinvestmentsintoresearchanddevelopment,andatthe sametimeaproxyforthedemandgrowth(Herreroetal.,2014).
Because per capita GDP growth differs across regions, sodoes agriculturalproductivity.In theSSP2narrative,forinstance,the worldremainstoacertaindegreefragmentedeconomically,but cropyieldsgrowrelativelyfasterintheglobalSouththaninthe global North, gradually catching up to the yields in today’s developedcountries(forregionaldefinitions,seeSupplementary TableS4).TheSSP2developmentofcropyieldsissituatedbetween theslightlyfasterandsubstantiallyslowerdevelopmentsofSSP1 andSSP3,respectively.Importantly,innoneoftheSSPs,climate change impacts on food production are taken into account. In addition,differentassumptionsaboutthecharacteroftheyield growtharemadeforeachSSP: e.g.,inSSP2theyieldgrowth is proportional to the growth in input requirements such as fertilizers which has implications for its associated emissions (Valinetal.,2013).
Feedconversionefficiencies(i.e.,thelandproductivityofthe livestocksector)areestimatedbasedonpasttrendscalculatedby (Soussana etal., 2012)andextrapolated forward basedonper capitaGDPgrowth.Thedetailedlivestocksectorrepresentation allows for endogenous intensificationthrough production sys- tems transitions (Havlík et al., 2014).SSP2 assumes moderate flexibilityof thesesystems.Importantbarriersfrustratesystem changesinSSP3,andeducationandotherinfrastructurefacilitate thetransitioninSSP1.Finally,intermsoffoodconsumption,SSP2 occupiesacentralspotbetweenSSP1andSSP3(seeTable1and Supplementary Text 6). Trade assumptions of agricultural commodities(Supplementary Text 5)are notvaried acrossthe SSPs.
3.Implementationframework:SSPscenariodevelopmentcycle
Thelargesetofassumptionsthathavebeenintroducedinthe previoussectionandTable1havetobeassessedinanintegrated wayinordertoproducescenariosthatconsistentlyrepresentall dimensionsoftheSSPs’broadernarratives.Tothisend,theIIASA IAM frameworkis used.The IIASAIAM frameworkconsistsofa combinationoffivedifferentmodelsormoduleswhichcomple- menteach otherandarespecializedindifferentareas.Herewe provideasuccinctoverviewofthecomponentsoftheframework (see Box2), andillustrate theinteractionbetweenthedifferent modelsormodulesinatypicalscenariodevelopmentcycle.
xxx–xxx
3.1.Developmentofbaselinescenarios
AllmodelsandmodulesintroducedinBox2,togetherbuildthe IIASAIAMframework.Theyprovideinputtoanditeratebetween eachotherduringatypicalSSPscenariodevelopmentcycle.Inthe previoussection,wedocumentedtheveryfirststepinsuchanSSP scenariodevelopmentcycle:theselectionofquantitativeassump- tionsforallimportantmodelparameters.Togetherwithinputson GDPandpopulation,theyprovidetheexogenousSSPinputsthat are needed at the start. In the remainder of this section, we
describe which further steps are taken within the IIASA IAM frameworktodevelopanSSPscenario.
MESSAGE represents the core of the IIASA IAM framework (Fig.4)anditsmaintaskistooptimizetheenergysystemsothatit cansatisfyspecifiedenergydemandsatthelowestcosts.MESSAGE carries outthis optimizationin aniterativesetup withMACRO, which provides estimates of the macro-economic demand response that results of energy system and services costs computedbyMESSAGE.Forthesixcommercialend-usedemand categories depicted in MESSAGE (Table S3), MACRO will adjust
Box2.DescriptionofIIASAIntegratedAssessmentModellingframeworkcomponents.
Energysystem
EnergysystemdynamicsaremodelledwiththeMESSAGEmodel(ModelforEnergySupplyStrategyAlternativesandtheirGeneral Environmentalimpacts).MESSAGEisaglobalsystemsengineeringoptimizationmodelusedforlongandmediumtermenergy systemplanning(MessnerandStrubegger,1995;Riahietal.,2012).MESSAGEdividestheworldinto11regions,eachofwhichis characterizedbyadetailedenergysystemrepresentation.Themodel’smainobjectiveistooptimizethecontributionsofvarious energysupplyoptions overtime inordertomeetspecifiedregional energydemandsat thelowestoverall discountedcost.
MESSAGEfeaturesaverybroadportfolioofenergytechnologies,coveringtechnologiesforresourceextraction,fuelconversion, electricityandheatgenerationaswellasendusetechnologies.Thesevarioustechnologiessupplysevendifferentdemands(see SupplementaryTable S3). Finally, MESSAGE also tracks thesources and sinks ofgreenhouse gases (GHG) and estimates anthropogenicGHGemissionsaspartofitsoptimizationprocedure.
Land-usesystem
Land-usedynamicsaremodelledwiththeGLOBIOM(GLobalBIOsphereManagement)model,whichisarecursive-dynamicpartial- equilibriummodel(Havlíketal.,2011).GLOBIOMrepresentsthecompetitionbetweendifferentland-usebasedactivities.Itincludes abottom-uprepresentationoftheagricultural,forestryandbio-energysector,whichallowsfortheinclusionofdetailedgrid-cell information on biophysical constraints and technological costs, as well as a rich set of environmental parameters, incl.
comprehensiveAFOLU(agriculture,forestryandotherlanduse)GHGemission accountsandirrigationwateruse.Itsspatial equilibriummodellingapproachrepresentsbilateraltradebasedoncostcompetitiveness.Forspatiallyexplicitprojectionsofthe changeinafforestation,deforestation,forestmanagement,andtheirrelatedCO2emissions,GLOBIOMiscoupledwiththeG4M (GlobalFORestModel)model(Kindermannetal.,2006;Kindermannetal.,2008;Gusti,2010).ThespatiallyexplicitG4Mmodel comparestheincomeofmanagedforest(differenceofwoodpriceandharvestingcosts,incomebystoringcarboninforests)with incomebyalternativelanduseonthesameplace,anddecidesonafforestation,deforestationoralternativemanagementoptions.
Asoutputs,G4Mprovidesestimatesofforestareachange,carbonuptake andreleasebyforests,andsupply ofbiomassfor bioenergyandtimber.
Airpollution
AirpollutionimplicationsarederivedwiththehelpoftheGAINS(Greenhousegas–AirpollutionINteractionsandSynergies)model.
GAINSallowsforthedevelopmentofcost-effectiveemissioncontrol strategiestomeetenvironmentalobjectivesonclimate, humanhealthandecosystemimpactsuntil2030(Amannetal.,2011).Theseimpactsareconsideredinamulti-pollutantcontext, quantifying thecontributions of sulfurdioxide (SO2),nitrogen oxides (NOx), ammonia(NH3), non-methane volatileorganic compounds(VOC),andprimaryemissions ofparticulatematter(PM),includingfineandcoarsePMaswellascarbonaceous particles(BC,OC).Asastand-alonemodel,italsotracksemissionsofsixgreenhousegasesoftheKyotobasket.TheGAINSmodel has global coverage and holds essential information about key sources of emissions, environmental policies, and further mitigationopportunitiesforabout170country-regions.Themodelrelieson exogenousprojectionsofenergyuse,industrial production,andagriculturalactivityforwhichitdistinguishesallkeyemissionsourcesandseveralhundredcontrolmeasures.
GAINScandevelopfinelyresolvedmid-termairpollutantemissiontrajectorieswithdifferentlevelsofmitigationambition(Cofala etal.,2007;Amannetal.,2013).TheresultsofsuchscenariosareusedasinputtoglobalIAMframeworkstocharacterizeair pollutiontrajectoriesassociatedwithvariouslong-termenergydevelopments(seefurtherbelowand,forexample,Riahietal., 2012;Raoetal.,2013).
Macro-economicsystem
Themacro-economicresponseoftheglobaleconomyintheIIASAIAMframeworkiscapturedbytheMACROmodel.Theformof MACROusedintheIIASAIAMframeworkisderivedfromalong seriesofmodelsbyManneandRichels(1992).Asfurther describedbyMessnerandSchrattenholzer(2000),MACROmaximizestheintertemporalutilityfunctionofasinglerepresentative producer-consumerineachworldregionthroughoptimization.Theresultisasequenceofoptimalsavings,investment,and consumptiondecisions.Themainvariablesofthemodelarethecapitalstock,availablelabor,andenergyinputs,whichtogether determinethetotaloutputofaneconomyaccordingtoanestedproductionfunctionwithconstantelasticityofsubstitution.It considersthesixcommercialenergydemandcategoriesinMESSAGE(seeTableS3).
Climatesystem
Theresponseofthecarbon-cycleandclimatetoanthropogenicclimatedriversismodelledwiththeMAGICCmodel(Modelforthe AssessmentofGreenhouse-gasInducedClimateChange).MAGICCisareduced-complexitycoupledglobalclimateandcarboncycle model which calculates projectionsfor atmospheric concentrationsof GHGsand other atmospheric climate driverslike air pollutants,togetherwithconsistentprojectionsofradiativeforcing,globalannual-meansurfaceairtemperature,andocean-heat uptake(Meinshausenetal.,2011a).MAGICCisanupwelling-diffusion,energy-balancemodel,whichproducesoutputsforglobal- andhemispheric-mean temperature.Here,MAGICC isused in adeterministicsetup (Meinshausenet al., 2011b),but alsoa probabilisticsetup(Meinshausenetal.,2009)hasbeenusedearlierwiththeIIASAIAMframework(Rogeljetal.,2013a;Rogeljetal., 2013b;Rogeljetal.,2015).Climatefeedbacksontheglobalcarboncycleareaccountedforthroughtheinteractivecouplingofthe climatemodelandarangeofgas-cyclemodels.
xxx–xxx
useful energy demands, until the two models have reached equilibrium.Thisiterationreflectsprice-inducedenergyefficiency improvements that can occur when energy prices increase.
GLOBIOMprovides MESSAGE with information on land useand itsimplications,liketheavailabilityandcostofbio-energy,and availability and cost of emission mitigation in the AFOLU (Agriculture,ForestryandLandUse)sector.Toreducecomputa- tional costs, MESSAGE iteratively queries a GLOBIOM emulator which can provide possible land-use outcomes during the optimizationprocessinsteadofrequiringtheGLOBIOMmodelto bererun continuously (see Supplementary Text 7, and Supple- mentaryFigs.S6andS7).OnlyoncetheiterationbetweenMESSAGE andMACROhasconverged,theresultingbioenergydemandsalong withcorrespondingcarbonpricesareusedforaconcludingonline analysiswiththefull-fledgedGLOBIOMmodel.Thisensuresfull consistencyinthemodelledresultsfromMESSAGEandGLOBIOM, andalsoallowstheproductionofamoreextensivesetofreporting variables. Air pollution implications of the energy system are computedinMESSAGEbyapplyingtechnology-specificpollution coefficients from GAINS. In general, cumulative global GHG emissionsfromallsectors areconstrainedat differentlevelsto reachdesiredforcinglevels(cf.right-handsideFig.4).Theclimate constraintsarethustakenupinthecoupledMESSAGE-GLOBIOM optimization,andtheresultingcarbonpriceisfedbacktothefull- fledgedGLOBIOMmodelforfullconsistency.Finally,thecombined results for land use, energy, and industrial emissions from MESSAGEandGLOBIOMaremergedandfedintoMAGICC,aglobal carbon-cycleandclimatemodel,whichthenprovidesestimatesof theclimateimplicationsintermsofatmosphericconcentrations, radiativeforcing,andglobal-meantemperatureincrease.Impor- tantly, climate impacts and impacts of the carbon cycle are currentlynotaccountedforintheIIASAIAMframework.Theentire frameworkislinked toanonline databaseinfrastructurewhich allows straightforward visualisation, analysis, comparison and disseminationofresults.
3.2.Introductionofclimatepolicy
Climate action within the IIASA IAM framework is typically modelled by capping the cumulative amount of CO2 or other
greenhousegasesoverthe21stcentury.Alternativewaystomodel climatepolicyarealsopossible,forexamplebyprescribingcarbon pricesorrenewableenergytargets.Whenapplyingaclimatepolicy toanSSPnarrative,assumptionshavetobemadeabouttheextent and timingofthat policy.To ensureconsistencybetweenthese policyassumptionsandtheSSPnarratives,sharedclimatepolicy storylines(called‘sharedclimatepolicyassumptions’–SPA)have beendeveloped(Kriegleretal.,2014),whicharecomplementaryto the SSP narratives (Table S5). For each SSP, a particular SPA is recommended.ForSSP2,weuseSPA2.SPA2assumesthatclimate policiestargetingemissionsfromfossil-fueluseandindustryare geographically fragmented until 2020 and then converge to a globally uniform carbon price by 2040. For more details see Kriegleretal. (2014)and Riahi etal. (2016,inpress). Land-use emissions are controlled by the same regional carbon prices.
However,inordertocomplywiththespecificationthatSSP2hasto beadynamics-as-usualworld,globalafforestationorelimination of deforestation before 2030 is not allowed to occur. In our implementation,weadjustthenear-termcarbonpriceforland-use emissionsinordertoavoidthis.AlsotheSPAsreflectthegradation inmitigationchallengesbetweenthevariousSSPs.SPA1,whichis appliedtoSSP1,assumesthatfastglobalactionispossible,while SSP3 assumes that a period of fragmented regionalaction will precedeglobalaction(seealso,Kriegleretal.,2014).
4.ResultssummaryfortheSSP2markerscenario
Thissectionillustratessomeofthesalientcharacteristicsofthe markerSSP2implementationintheIIASAIAMframework.Results fortheSSP2baselinearetheinitialfocus;thisscenariodoesnot includeanyclimatepoliciesbeyondthoseinplacetoday.Sucha baseline then provides a reference point against which the effectiveness of climate policies (of varyingstringency) can be measured. We first present how the SSP2 energy intensity improvementscomparetothebracketingSSP1andSSP3scenarios, and then havea closerlookathow globalenergydemand and supply, aswell as land usedevelop over time. Thesubsequent sectionanalysestheevolutionofatmosphericclimate-modifying emissions,includingGHGsandairpollutants.Finally,weexplore transformationsintheenergyand land-usesystemsrequiredto Fig.4.OverviewoftheIIASAIAMframework.Colouredboxesrepresentrespectivespecializeddisciplinarymodelswhichareintegratedforgeneratinginternallyconsistent scenarios.
xxx–xxx
limitclimatechangetotargetedlevelsofanthropogenicradiative forcing over the 21st century. Due to space constraints, this discussionfocussesmainlyonglobaldevelopments,butalsosome informationattheleveloftheglobalNorthandSouthisprovided.
Afullsetofscenarioresultscanbefoundinthepubliclyavailable, onlinedatabase(https://secure.iiasa.ac.at/web-apps/ene/SspDb).
4.1.Baselineenergy-systemcharacteristics 4.1.1.Energydemandandsupply
The varyingenergy intensityimprovementrates oftheSSPs translateintowidelydiverginglevelsofbaselineenergydemandin absenceofclimateaction.Combinedwithalternativeassumptions forfossilresourcesandenergytechnologies(seeearlier),thisleads tomarkedlydifferentenergysystemstructures,bothonthesupply and demand sides. On the demand side in particular, all SSPs exhibitacontinuoustransformationawayfrominconvenientand low quality fuels (mainly solid fuels) toward more flexible, convenientand higher quality carriers, such as electricity. Yet, whilethenatureof thedemand-side transitionmaybesimilar across the SSPs, the pace at which these changes (or this
‘modernization’) occurs is much faster in SSP1 compared to SSP3;SSP2liesinthemiddle(Fig.5,lowerpanels).
FinalenergydemandinSSP2increasessteadilyoverthe21st centuryreachingapproximately640and970EJ/yrby2050and 2100,respectively(Fig.5,topleftpanel).Thelatterisa2.7-fold increasefrom2010.These2100levelsareabout300EJ/yrhigher
thaninSSP1,andabout200EJ/yrlowerthaninSSP3.OurSSP1 implementationthusroughlymanagestostabilizeenergydemand growthinthesecondhalfofthecentury.Fig.5furtherillustrates that the SSP2 marker implementation with the IIASA IAM frameworkis situated roughlyin themiddlebetweentheSSP1 and SSP3 implementations. The differences in baseline energy demandbetweentheSSPsaredominatedbythesurgeintheglobal South(SupplementaryFigs.S10-S12).InSSP2,finalenergydemand for the industry, residential-and-commercial, and transport sectorsincreasesbyapproximately42%by2100over2010levels indevelopedcountries.TheincreaseisevengreaterintheSouth, duetothedrasticincreaseinincomelevels:finalenergydemand quadruplesoverthesameperiodoftime,accountingforaglobal shareof74%by2100comparedtoabout51%in2010.SSP3projects asimilarfinalenergydemandin2100fortheglobalNorthastoday, while SSP1 sees energy demand contract slightly. The bulk of energydemandincreaseinanySSPisthusprojectedtocomefrom developingcountries.
TheresultsfortheprimaryenergymixofSSP2(Fig.5,top-right panel)showthat,muchliketoday,fossilenergycarriersremainthe fuelsofchoiceuntiltheendofthecenturyinanSSP2world.The assumedmoderateinvestmentsinrenewableslimittheirrolein thefuture,despitetheircontinuedgrowth.SSP2seesitsshareof non-fossilenergy intheprimaryenergy mixincrease from17%
today(2010)to23%in2100.InSSP1thisishigherandinSSP3lower (SupplementaryFigs.S13andS14).Thenon-fossilshareinSSP1 increasesto31%in2100becauseofanincreaseinrenewablesother
Fig.5. EnergycharacteristicsandcontextoftheSSP2markerscenario.Topleft:Evolutionoffinalenergydemandoverthe21stcenturyforbaselineSSP1,SSP2,andSSP3 scenariosmodelledbytheIIASAIAMframework(boldgreen,blue,andredlines),comparedtothemulti-modelrangeacrossallrespectiveSSPssubmittedbyothermodelling teams(green,blue,andredshadedareas),aswellastherangefoundintheIPCCAR5ScenarioDatabase(greyrange).ThinlinesindifferentlinetypesrepresentSSP2final energydemandforfourmitigationpathwaysmodelledwiththeIIASAIAMframeworkinlinewithanend-of-centuryradiativeforcingtargetof6.0,4.5,3.4,and2.6W/m2, respectively.Topright:PrimaryenergymixevolutionfortheSSP2marker,modelledbytheIIASAIAMframework.Bottomleft:Shareofelectricityinfinalenergyinthe3SSP baselines.Coloursasinleft-handpanel.Bottomright:Contributionsofsolids(grey),liquids(blue),andgrids(pink)tototalfinalenergyinSSP2.VariationsforSSP1andSSP3 areshownbysolidanddashedlines,respectively.(Forinterpretationofthereferencestocolourinthisfigurelegend,thereaderisreferredtothewebversionofthisarticle.)
xxx–xxx
thanbiomass.Finally,onthedemandside,thethreeSSPsdiffer markedlyintermsoftheirsharesofelectricityinfinalenergyand the relative shares of final energy that are covered by solids, liquids,or grids(seeFig. 5, bottompanels, and Supplementary Figs.S8–S9).
4.2.Baselineland-usecharacteristics
4.2.1.Demandandsupplyofagriculturalcommodities
Population and economic growth, as wellas evolvingsocial preferences(asdiscussedinSection2),driveoveralldemandfor agriculturalproducts, and these demands vary acrossthe SSPs (Poppetal.,inreview).Agriculturalcommoditieslikefoodcrops andlivestockaretradedglobally.ThiscontinuesinSSP2,andin additionagriculturalmarketscontinue theirlong-termtrendof slightlydecreasingagriculturalprices. More rapidtechnological progress and only moderately increasing demand lead toeven moresubstantialpricedecreasesinSSP1,whileslowertechnologi- calprogressandstrongerdemandleadtoincreasingcommodity pricesinSSP3.Alltheseestimatesexcludeanyinfluenceofclimate change.
Asfoodcropdemandrelatestochallengesforadaptation,SSP2 also here provides a middle-of-the-road perspective. In this scenario,humanconsumptionofcrops(globally)isprojectedto increaseby41%until2050andreturntothislevelby2100,after peaking around 2070. Theyear-2100 demand for foodcrops is projectedtobe22%lowerinSSP1and33%higherinSSP3,relative toSSP2.AfurthercharacteristicoftheSSP2baselineisthegrowing leveloflivestockconsumption,whichisconsideredaluxurygood andisthereforeassociatedwithhigherincomes(Supplementary Text 8). The moderate increase in population together with sustained income growth makes the SSP2 scenario the largest livestockproductconsumer.InSSP1,thepartialshifttolessmeat- intensivedietsintheNorthandtheslowlygrowingpopulationin theSouthleadtolivestockproductconsumptionthatisabouta thirdlower.InSSP3,thedecreasingpopulationintheNorthand slowlygrowingincomesintheSouthleadalsotolivestockproduct demandwhichis7%lowerthaninSSP2.Overall,increasingfood consumptioncombinedwithcropdemandsforanimalfeedandfor otherusesleadstoaglobalincreaseofcropproductionof84%in 2100inSSP2,relativeto2010.Thiscomparestoaglobalincreaseof 21%and97%inSSP1andSSP3,respectively.
4.2.2.Demandandsupplyofwoodybiomass
Asecondmaintaskoftheland-usesectoristheprovisionof woodybiomass.Twomajorbiomassusesareconsidered,industrial roundwoodandbiomassforenergyproduction,andthesecanbe sourcedfromroundwoodfromtraditionalforestsorfrombiomass from short rotation tree plantations (Supplementary Text 9).
Unlikeforfood,74%ofthedemandforindustrialroundwoodwas located in the North in 2010. In SSP2, the global demand for industrialroundwoodisprojectedtodoubleby2100,butonlyhalf ofitwouldcomefromtheNorth.ThisdemandissimilarinSSP1 (5% lower), and about 20% lower in SSP3. Similar to livestock products,oppositetrendsinpopulationandeconomicgrowthin SSP1andSSP3canceleachotherout,makingthedemandhighest inSSP2.
Biomassdemandforenergyamountstosome55EJofprimary energyin2010and80%ofthisdemandcomesfromtheSouth.This ismainlytraditionalbiomassusedforcookingandheating.InSSP2, thisdemandprogressivelydecreases(seeearlier).Thetotalenergy biomassdemandinSSP2is8%higherin2050comparedto2010.In thesecondhalfofthecentury,theincreasingdemandformodern bioenergyproductionresultsinanetincreaseof19%comparedto 2010 by the end of the century, reaching 66 EJ. Overall, the commercialbiomassdeploymentby2100intheSSP1baselineisof thesamemagnitudeasinSSP3(around74EJ).Themajordifference between these two scenarios consists in the deployment of traditionalbiomass,whichisphasedoutbefore2100inSSP1,while still representing50 EJinSSP3.Thismakestheoverallbaseline energybiomassdemandinSSP374%largercomparedtoSSP1.
4.2.3.Landuseevolution
Landuseiscloselylinkedtoagriculturalandforestproduction (see Supplementary Text 10 and 11), and these influence the naturalenvironmentandecosystemservices,suchasbiodiversity or carbon sequestration. Land use is therefore simultaneously connected to both adaptation and mitigation challenges. Yet, despite the major crop production increases foreseen in SSP2, globalcroplandonlyexpandsby25%relativeto2010(Fig.6).The remainderoftheproductionincreasecomesfromintensificationof land use supported by technological change. This requires a doublingoffertilizer useanda 10%increase inirrigationwater withdrawals. In SSP1, moderate demand increase and fast technologicalprogressleadtoanincreaseincroplandarea(10%
Fig.6.LandusedevelopmentinthemarkerSSP2baselinescenario.Leftpanel:evolutionofgloballandareaovertime.Rightpanel:agriculturalandforestryproductionover timeinunitsofmilliontonnesofdrymatter.Similarfiguresofa2.6W/m2mitigationcaseisprovidedinSupplementaryFigs.S18.
xxx–xxx
by2100from2010levels).Ontheotherhand,theslightlyhigher demand and slower technological progress of SSP3 mean that almost twice as much additional land and irrigation water is required,comparedtoSSP2.Moreover,althoughlivestockproduc- tionalmostdoublesinSSP2,utilizedgrasslandareaisprojectedto expandbyonly6%in2100comparedto2010(seeSupplementary Text 11). Over the same time period, grassland area in SSP3 increasesbymorethan 3%,despitethelowerlivestockproduct demand,whereasinSSP1substantiallylowerdemandandfaster technological change lead to utilized grassland area being 6%
lower.Intermsofpressureonforestsandothernaturalland,SSP2 representsamiddle-of-the-roadscenariowithanetlossof607 millionhectaresby2100(comparedto2010),whileSSP3andSSP1 see a loss of 719 million and a gain of 63 million hectares, respectively.Thesedevelopmentsarecriticalforbothadaptation andmitigationchallenges,becausewhenmorelandisrequiredfor baseline agricultural and forest production, less land remains availabletoaddresspotentialclimateimpactsonagricultureor implementclimatechangemitigationactivities.
4.3.Climatedriversandclimatepolicy
Uptothispoint,wepresentedresultsfortheSSP2baselinein absence of climatepolicy. Each SSP baseline, however, can be combined withvariouslevels ofclimate policy.This leadstoa matrixofpotentialoutcomeswithvariousSSPnarrativesonthe horizontaland various climatemitigation levelsonthevertical axis.Thelevelofclimatemitigationinthismatrixisdefinedasa limitonthetotalanthropogenicradiativeforcingin2100.Studies inthisissuelookatlimitingradiativeforcingto8.5,6.0,4.5,3.4, and 2.6W/m2, and further studies are under way to develop pathwaysthatlimitradiativeforcingto2.0W/m2in2100orthat significantlyexceedthetargetedend-of-centuryradiativeforcing duringearlierdecades.
4.3.1.Baselineemissiontrajectories
EmissiontrajectoriesoftheSSP2baselinearethestartingpoint forourclimatepolicyanalysis.Fig.7illustratesthatalsointermsof baselinetrajectoriesofGHGandairpollutantemissions,SSP2is fully consistent with its assigned role of a middle-of-the-road scenario(bold linesin Fig.7).Themarker SSP2 trajectoriesare
Fig.7.Globaldevelopmentsforvariousgreenhousegasesandairpollutants.Theevolutionoverthe21stcenturyforbaselineSSP1,SSP2,andSSP3scenariosmodelledbythe IIASAIAMframeworkisprovidedinboldgreen,blue,andredlines.Thesearecomparedtothemulti-modelrangeacrossallrespectiveSSPssubmittedbyothermodelling teams(green,blue,andredshadedareas,discussedintheoverviewpaperbyRiahietal.,2016,inpress),aswellastherangefoundintheIPCCAR5ScenarioDatabase(grey range).ThinlinesindifferentlinetypesrepresentSSP2emissionsforfourmitigationpathwaysmodelledwiththeIIASAIAMframeworkinlinewithanend-of-century radiativeforcingtargetof6.0,4.5,3.4,and2.6W/m2,respectively.(Forinterpretationofthereferencestocolourinthisfigurelegend,thereaderisreferredtothewebversion ofthisarticle.)
xxx–xxx
