Glob Change Biol. 2019;00:1–14. wileyonlinelibrary.com/journal/gcb  | 1© 2019 John Wiley & Sons Ltd

Received:11December2018  |  Revised:12March2019  |  Accepted:16March2019DOI:10.1111/gcb.14638

P R I M A R Y R E S E A R C H A R T I C L E

Changes in timing of seasonal peak photosynthetic activity in northern ecosystems

Taejin Park1  | Chi Chen1 | Marc Macias‐Fauria2  | Hans Tømmervik3  | Sungho Choi4 | Alexander Winkler5,6 | Uma S. Bhatt7 | Donald A. Walker8 | Shilong Piao9  | Victor Brovkin5 | Ramakrishna R. Nemani10 | Ranga B. Myneni1

1DepartmentofEarthandEnvironment,BostonUniversity,Boston,Massachusetts2SchoolofGeographyandtheEnvironment,UniversityofOxford,Oxford,UnitedKingdom3NorwegianInstituteforNatureResearch,FRAM–HighNorthResearchCentreforClimateandtheEnvironment,Tromsø,Norway4RhombusPowerInc.,NASAAmesResearchPark,MoffettField,California5Max‐Planck‐InstituteforMeteorology,Hamburg,Germany6InternationalMax‐PlanckResearchSchoolforEarthSystemModeling,Hamburg,Germany7GeophysicalInstitute,UniversityofAlaskaFairbanks,Fairbanks,Alaska8InstituteofArcticBiology,UniversityofAlaska,Fairbanks,Alaska9CollegeofUrbanandEnvironmentalSciences,PekingUniversity,Beijing,China10NASAAmesResearchCenter,MoffettField,California

CorrespondenceTaejinPark,DepartmentofEarthandEnvironment,BostonUniversity,Boston,MA02215,USA.Email:parktj@bu.edu

Funding informationNationalAeronauticsandSpaceAdministration,Grant/AwardNumber:NNX14AI71G,NNX14AP80AandNNX16AO34H;NorgesForskningsråd,Grant/AwardNumber:287402and270992;NaturalEnvironmentResearchCouncil,Grant/AwardNumber:NE/L011859/1

AbstractSeasonality in photosynthetic activity is a critical component of seasonal carbon,water,andenergycyclesintheEarthsystem.Thischaracteristicisaconsequenceofplant'sadaptiveevolutionaryprocessestoagivensetofenvironmentalconditions.Changingclimateinnorthernlands(>30°N)altersthestateofclimaticconstraintsonplantgrowth,andtherefore,changesintheseasonalityandcarbonaccumulationareanticipated.However,howphotosyntheticseasonalityevolvedtoitscurrentstate,andwhatroleclimaticconstraintsandtheirvariabilityplayedinthisprocessandulti‐matelyincarboncycleisstillpoorlyunderstoodduetoitscomplexity.Here,wetakethe“lawsofminimum”asabasisandintroduceanewframeworkwherethetiming(dayofyear)ofpeakphotosyntheticactivity (DOYPmax)actsasaproxy forplant'sadaptivestatetoclimaticconstraintsonitsgrowth.OuranalysesconfirmthatspatialvariationsinDOYPmaxreflectspatialgradientsinclimaticconstraintsaswellassea‐sonalmaximumandtotalproductivity.Wefindawidespreadwarming‐inducedad‐vance in DOYPmax (−1.66±0.30days/decade, p<0.001) across northern lands,indicating a spatiotemporal dynamismof climatic constraints toplant growth.WeshowthattheobservedchangesinDOYPmaxareassociatedwithanincreaseintotalgrossprimaryproductivitythroughenhancedcarbonassimilationearlyinthegrow‐ingseason,which leadstoanearlierphaseshift in land‐atmospherecarbonfluxesandanincreaseintheiramplitude.Suchchangesareexpectedtocontinueinthefu‐turebasedonouranalysisofearthsystemmodelprojections.Ourstudyprovidesasimplified,yetrealisticframeworkbasedonfirstprinciplesforthecomplexmecha‐nisms by which various climatic factors constrain plant growth in northernecosystems.

K E Y W O R D S

carboncycle,climatechange,climateconstraint,earthsystemmodel,eddycovariance,grossprimaryproductivity,lawofminimum,photosyntheticseasonality,remotesensing

www.wileyonlinelibrary.com/journal/gcbmailto:https://orcid.org/0000-0003-0698-6942https://orcid.org/0000-0002-8438-2223https://orcid.org/0000-0001-7273-1695https://orcid.org/0000-0001-8057-2292mailto:parktj@bu.edu

2  |     PARK et Al.

1  | INTRODUC TION

Warmingisgenerallythoughttoeaseclimateconstraintonphotosyn‐theticactivityofvegetationinnorthernlands.Indeed,recentgrow‐ingseasonstudiesbasedongroundobservation(Parmesan&Yohe,2003),eddycovariance(Keenanetal.,2014;Richardsonetal.,2010),remotesensing(Parketal.,2016;Xuetal.,2013),andmodelsimula‐tion(Duveneck&Thompson,2017)haveconcordantlyindicatedthatthegrowingseasondurationfornorthernterrestrialvegetationhassignificantlyextendedoverthepastdecadesduetobothanearlierstartanddelayed termination.Thisprolongedgrowingseasonovernorthern landdrivesa longercarbonassimilationperiodduetotherelaxationof lowtemperature limitsonmetabolism,andinturn, in‐creasedproductivityandcarbonuptakehavebeenobserved(Forkeletal.,2016;Xuetal.,2013).However, longerandwarmergrowingseasons also promote environmental conditions that favor surfacedrying,andthusintensifiedsummerdroughts,treemortality,andwild‐fireshaveresultedinsummerproductivitydecline(Barichivichetal.,2014;D’Orangevilleetal.,2018;Pengetal.,2011).Theseconsequen‐tialdynamicsarehighlyvariableinspaceandovertime,andindicateacomplexinteractionofmultipleclimateconstraintsonplantgrowthanditsdynamism(Garonnaetal.,2018;Nemanietal.,2003;Reichetal.,2018).Toaccuratelyprojecttheresponseofnorthernvegetationtofutureclimate,weneedtobetterunderstandhowclimate–vegeta‐tioninteractionhasevolvedtoitscurrentstate,andwhatroleclimaticconstraintsandtheirvariabilityplayedinthisprocess.

Photosynthetic seasonality is an integrated outcome of howplantsadapttoseasonalvariationsinclimaticconstraints(Chuine& Beaubien, 2001; Eagleson, 2005; Garonna et al., 2018; Jolly,Nemani,&Running,2005),andisthusacriticalindicatorofvege‐tation–climateinteraction.Forinstance,grossprimaryproductiv‐ity (GPP) tracks the seasonal courseof temperature innorthernhigh‐latitudeecosystems,whilethesynchronybetweenGPPandtemperatureisgraduallylostsouthwardstowardwarmeranddrierenvironments(seeFigure1inRotenberg&Yakir,2010).Thelawsofminimum(Blackman,1905;Liebig,1841;Sprengel,1828)explainthese shifts in GPP with respect to varying climatic conditions(Eagleson, 2005). The laws state that although photosyntheticactivityiscontrolledbymultiplefactors(e.g.,radiation,tempera‐ture,wateravailability,etc.),theprevailingrateissetbythemostdeficientofthesefactors(Blackman,1905;Liebig,1841;Sprengel,1828).Thissuggeststhatthetiming(dayofyear)ofpeakphoto‐syntheticrate(DOYPmax)duringtheseasonalcoursecorrespondsto the periodwhen the primary climatic factor controlling plantgrowth is least limiting.Thissimpleyet intuitive indicatorhasanindispensablerolenotonlyindicatingthetimingandmagnitudeofresourceavailability (i.e.,constraint)butalsothecapacityofter‐restrialecosystemproductivity(Xiaetal.,2015;Zhouetal.,2017).Ongoingclimatechangeinthenorthisexpectedtoalterthestateofclimaticconstraintsonplantgrowth,andtherefore,changesinDOYPmaxandproductivity.Previousstudieshaveobservedtrendstowardanearlierpeakof thegrowingseason (Buitenwerf,Rose,& Higgins, 2015; Gonsamo, Chen, & Ooi, 2018). However, the

underlyingmechanismsforspatiallyvaryingrelationsbetweenitschangesand implicationsonseasonal totalproductivityandcar‐boncyclearestilllargelyunknown.

In this study,we take the “laws ofminimum” as a basis and intro‐duceanewframeworkwherethetimingofpeakphotosyntheticactivity(DOYPmax)actsasaproxyforplant'sadaptivestatetoclimaticconstraintson its growth. Two basic principles formulate this new framework(Figure1).First,undernonlimitingclimaticconditions,DOYPmaxwillshowatendencytocoincidewiththeperiodofseasonalpeakradiationloadsoastoresultinmaximumphotosyntheticcapacityconditions(Bauerleetal.,2012;Eagleson,2005;Case1inFigure1).Second,ifaclimaticfactoractsastheprimaryconstrainttophotosyntheticactivity,DOYPmaxshouldshifttowardtheperiodintheseasonalcourseatwhichthatlimitingresourceismoreavailable(Eagleson,2005;Rotenberg&Yakir,2010;Cases2–4inFigure1).Inthisframework,thetimingsofpeakGPP(DOYPmax)andthreeclimatic factors including temperature (DOYTmax), radiation (DOYRmax),andwateravailability(DOYWmax)serveaskeyproxiesforclimateresourceavailability.Weonly introduce these three abiotic controls ofGPPbe‐causeitiswidelyknownthattheyinteracttoprimarilyimposecomplexandvaryinglimitationsonvegetationactivity(Nemanietal.,2003).Duetoreducedwaterlossesduringthecoldseasonovernorthernterrestrialecosystemsandthermalinertia,asequentialorderofthetimingsofpeakclimatic factors (DOYWmax<DOYRmax<DOYTmax) simplifies our frame‐work(FigureS1).Inotherwords,thissuggeststhatpositioningofDOYPmax with respect to DOYRmax (δDOYP,R=DOYPmax−DOYRmax) can indicatetheprimaryclimaticconstraintonecosystems,thatis,water(δDOYP,R<0)ortemperature(δDOYP,R>0).δDOYP,TdefinedasDOYPmax−DOYTmaxisadditionally introducedtosubdividedominanttemperatureconstrainednorthernecosystems.

Ourprimaryobjectivesofthisstudyaretwofold:(a)toexaminetheproposed frameworkusing independentmultipledatasetsandunderstandhownorthernvegetationseasonalityhasbeencharac‐terized;and(b)toinvestigatechangesinDOYPmaxanditsimpactonseasonaltotalproductivityandcarboncycle.Toaccomplishtheob‐jectives,weapplytheproposedframeworktoGPPdynamicsfromthesatelliteobservedvegetationphotosyntheticactivitytoevaluateits validity and changes inDOYPmax. Two independent sources ofvegetation productivity (tower‐measured GPP and satellite‐drivensun‐inducedfluorescence[SIF])areusedtofurthertesttheframe‐work.Weuse the atmosphericCO2 observations atPointBarrow(71.3°N,156.6°W)andtwostate‐of‐the‐artCO2inversionestimatestoinvestigatethepotentialimpactofshiftingDOYPmaxonterrestrialecosystemcarboncycle.Asetofearthsystemmodels(ESMs)isad‐ditionallyintroducedtoevaluatethereproducibilityoftheobservedDOYPmaxchangesand theirconsequencesunderhistoricaland fu‐tureclimatescenarios.

2  | MATERIAL S AND METHODS

2.1 | Study area and bioclimatic zones

Onlynonagriculturalvegetationovernorthof30°Nisconsideredinthis study tominimizehuman‐induced influence.Threebioclimatic

     |  3PARK et Al.

zones includingarctic,boreal,andtemperateregionswereusedtopresentoutcomesofthisstudy.Todiscriminatethebioclimaticzones,wecombinedaterrestrialecoregionscheme(Olsonetal.,2001)oftheWorldWildlifeFund(WWF)andtheModerateResolutionImagingSpectroradiometer (MODIS) International Geosphere‐BiosphereProgramme(IGBP)landcoverdata(Friedletal.,2010;Collection5.1).WefirstusedMODISIGBPtokeeponlynonagriculturalvegetationclasses (Class1–10and16).Then,basedon theWWF'secoregionscheme, tundraandboreal forests/taigaecoregionswereassignedintothearcticandborealbioclimaticzones,respectively.Temperatebroadleafandmixedforests,temperateconiferousforests,temper‐ategrasslands,savannas,andshrublandswereidentifiedasthetem‐peratebioclimaticzone.Wefurtherexcludedthepixelscontainingmorethan25%ofcroplandbasedontheInternationalInstituteforAppliedSystemsAnalysiscroplandfractiondata(Fritzetal.,2015).

2.2 | Data and methods

2.2.1 | Multiscale GPP and its proxy: satellite and tower measurements

Inthisstudy,wemainlyused17year(2000–2016)timeseriesofGPPdata from theMODISaboardNASA'sTerra satellite (Running,Mu,&

Zhao, 2015) to examine the framework and to investigate DOYPmax change innorthern lands.The latestversion (Collection6)ofMODISGPPwith8daytemporalcompositewasspatiallyaggregatedinto0.05degreegrid.Itshightemporalfrequencyisadvantageoustocapturetheseasonalvariationofphotosyntheticactivity.MODISGPPisbasedonaproductionefficiencymodelthatusestheproductoftheabsorbedpho‐tosyntheticallyactiveradiationbyvegetationandalightuseefficiencyfactor.ThequalityofMODISGPPdatasetshasbeencomprehensivelyevaluated against multiple eddy‐covariance tower measurements ofGPPandthroughintercomparisonswithotherGPPproducts(Heinschetal.,2006;Zhao,Heinsch,Nemani,&Running,2005).

We additionally introduced satellite‐driven SIF and eddy‐co‐variance‐based GPP data to verify our framework and resultsfromMODISGPP. The SIF is retrieved near the λ=740nm far‐red peak in chlorophyll fluorescence emission from the GlobalOzone Monitoring Experiment‐2 (GOME‐2) instrument onboardEumetsat's MetOp‐A satellite. The monthly SIF record (version27, level 3) covering 2007–2016 was used in this study (Joiner,Yoshida,Guanter,&Middleton,2016). SIF is anelectromagneticemission inthe650–800nmrangeoriginatingfromplantphoto‐syntheticmachinery,anditistheoreticallylinearlycorrelatedwiththe electron transport rate of photosynthetic activity (Zhang etal.,2014).

F I G U R E 1  ConceptualillustrationoftheproposedDOYPmaxframework.Seasonalcycleoftemperature(T,red),radiation(R,green),wateravailability(W,blue),andgrossprimaryproductivity(GPP)(P,black)overcommonnorthernterrestrialecosystems.Verticallinesindicatewheneachvariablereachesamaximumstate.DOYPmax,DOYTmax,DOYRmax,andDOYWmaxstandforthedayofyearwhenGPP,temperature,radiation,andprecipitationreachrespectivemaximumstateduringeachseasonalcourseoftheyear.Fouridealizedcasesareshowntodemonstratehowphotosyntheticseasonalityoftheecosystemundergivenclimateconstraintdiffersfromeachother:non‐(solidline,Case1),temperature‐(dotdashline,Case2),water‐(longdashline,Case3),andradiation‐(solidline,Case4)constrainedecosystems

4  |     PARK et Al.

The eddy‐covariance tower measurements from theFLUXNET2015database (tier1,Pastorelloet al., 2017)wereusedinthisstudy.FLUXNETisaglobalnetworkofmicrometeorologicaltower sites that use eddy covariancemethods tomeasure the ex‐changesofcarbon,water,andenergybetweenterrestrialecosystemsandtheatmosphere(Baldocchietal.,2001).WeusedGPPestimatesbasedonthefluxpartitioningapproachproposedbyLasslopetal.(2010).Atotalof92sites(thosewithmorethanthreesite‐yearmea‐surements)wereselectedfortheevaluationofourDOYPmaxframe‐workspanningalargeclimaticandbiomegradient(FigureS2a).

2.3 | Multiscale climate data

We used daily climate datasets provided by Global Modeling andAssimilationOffice(GMAO)ReanalysisofNASA(Gelaroetal.,2017).ThecurrentversionofGMAOisanhourlytimestepdatasetgeneratedbytheGoddardEarthObservingSystem‐5(GEOS‐5)dataassimilationsystem.Weaggregatedthenativehourlydata intothedailyscaletoretrievepixel‐wisephasesofclimatevariables.Surfaceairtemperatureanddown‐wellingphotosyntheticallyactiveradiationwereemployedinthisanalysis.DailyclimatedatasetswereusedtocharacterizeDOYTmax andDOYRmax.Wealsoobtainedpotentialevapotranspiration(PET)andactualevapotranspiration(AET)toquantifywateravailabilityonplantgrowthbycalculatingaratioofAETtoPET(RAP)(Prenticeetal.,1992).BothAETandPETwereobtainedbyGlobal LandDataAssimilationSystems (GLDAS,Version2.1;Rodell et al.,2004).WecharacterizedsummerclimateusingmeantemperatureandRAPduringJune–Augustfor investigatinghowDOYPmax positioning varies as functionsof cli‐mate constraints, that is, temperature andwater availability. For thetower‐measuredGPP,theancillarymicroclimatedatasetsincludingairtemperatureandincomingradiation(photosyntheticphotonfluxden‐sity)simultaneouslymeasuredwithGPPwereadditionallyobtained.

2.4 | ESM simulated historical and future GPP

Wealsointroducedasetofthemostrecentclimate‐carbonsimulationsofESMscontributingtothefifthphaseoftheCoupledModelIntercomparisonProject(CMIP5;Taylor,Stouffer,&Meehl,2012).SevenESMs,whichareavailableatCMIP5archive,wereusedinthisstudy:NorESM1‐M,MIROC‐ESM, CanESM2, HadGEM2‐ES, IPSL‐CM5A‐MR, MPI‐ESM‐MR, andCCSM4. The datasets provided monthly GPP output (1980–2099) forsimulationsofbothHistoricalandRepresentativeConcentrationPathway(RCP)4.5(Thomsonetal.,2011).DatafromtheHistoricalandRCP4.5sce‐narioperiodswerecombinedtogeneratecontinuousvariablefieldsfrom1980to2099.Allmodeloutputswereprocessedatthenativespatialreso‐lutionsandaggregatedintoregionalscales(i.e.,arctic,boreal,andtemper‐ateregions)fortrendandcorrelationestimates.

2.5 | Timings of peak seasonal photosynthetic activity and climate

We extracted three metrics indicating a maximal state of sea‐sonal photosynthetic activity (DOYPmax), radiation (DOYRmax),

and temperature (DOYTmax) at two different scales: site and re‐gionalscale.Forbothscales,toreducenoiseandmaintainadis‐tinct seasonal feature ofGPP (or SIF) and climate datasets, thesingularspectrumanalysiswas first implementedatyearlybasis(Vautard, Yiou, &Ghil, 1992). The singular spectrum analysis isanonparametricapproachthatdoesnotneedapriorispecifica‐tionofmodelsoftimeseries,thusit isdata‐adaptive.Itfirstde‐composes a time series into oscillatory components and noisesaccordingtothesingularvaluedecomposition,thereafterrecon‐structsspecificcomponents(i.e.,seasonalsignal)fromtheoriginaltime series. This nonparametric approachhasbeenwidely usedto reconstruct the time series of GPP and other environmentalvariablesbyreducingtheirnoisecomponents(Keenanetal.,2014;Zhouetal.,2017).TimeseriesofGPPandmeteorologicaldata‐setswereusedtoretrieveDOYPmax,DOYRmax,andDOYTmax on a yearly basis. Note thatmultiyear averaged daily GPP, radiation,and temperature time series were used for FLUXNET retriev‐als. For the case ofmonthly data (SIF and CMIP5GPP),we as‐signedmiddleofthemonthasthedayoftheyearforeachmonthandthenimplementedthesameproceduresusedinMODISandFLUXNET.Finally,δDOYP,R(i.e.,DOYPmax−DOYRmax)andδDOYP,T (i.e., DOYPmax−DOYTmax) were also calculated. We additionallyretrievedpixel‐wisegrowingseasonlengthfromMODISGPPbyapplyingafixedthreshold,that is,10%ofthemultiyearaveragemaximumGPP(Zhouetal.,2017).

2.6 | Atmospheric CO2 concentration and fluxes: zero‐crossing date and seasonal amplitude

Daily atmospheric CO2 concentration at Point Barrow (71.3°N,156.6°W)wasobtainedfromtheinsitumeasurementdatasetpro‐vided by the National Oceanic and Atmospheric Administration/EarthSystemResearchLaboratory(NOAA/ESRL).Thespringdown‐ward CO2 zero‐crossing date (DOYzero‐crossing) was extracted byfollowingtheapproachdescribedinPiaoetal.(2008).Wefirstde‐trendedtheinterannualtrendintheatmosphericCO2concentrationwithaquadraticpolynomialcurve, fourharmonics in theseasonalfunction, and time‐filtered residuals. We then used the harmon‐icsplustheresiduals (detrendedCO2seasonalcycle)todefinethedownward CO2 doyzero‐crossing as the day onwhich the detrendedcurvecrossedthezerolinefrompositivetonegative.Allaforemen‐tionedprocesseswereachievedbytheuseofthestandardpackageCCGCRV fromNOAA/ESRL (Thoning,Tans,&Komhyr,1989).Weuseddoyzero‐crossingasanindicatorofproximalDOYPmaxforthreerea‐sons,althoughDOYzero‐crossingisnotanaccuratetermofpeakphoto‐synthesistiming.First,seasonaltrajectoryofGPPstronglygovernschangesinnetbiomeproductivityseasonalityanditstrend(Forkeletal.,2016; Itoetal.,2016).Second,DOYzero‐crossing canbedeter‐minedmoreaccuratelyandit isroughlycorrespondingtothetimeofmaximumcarbonuptakebythebiosphere(Itoetal.,2016).Third,a relativechange in thephaseof thecycle identifiedatonepoint(e.g.,DOYzero‐crossing)willbematchedbyrelativephasechangesatallotherpointssincetheshapeoftheseasonalcycledoesnotchange

     |  5PARK et Al.

significantly (Barichivich, Briffa,Osborn,Melvin, & Caesar, 2012).Wefurtherextractedtheseasonalcycleamplitude(SCA)becauseitschangesreflectvegetationGPPdrivenchangesinnetcarbonuptake(Forkeletal.,2016).

We additionally used two gridded carbon fluxes from atmo‐spheric CO2 inversion products: the Copernicus AtmosphereMonitoring Service (CAMS, version v15r2, 1979–2015; Chevallieret al., 2010) and the Jena CarboScope (JENA, version s81_v3.8,1981–2015; Rödenbeck, Houweling, Gloor, & Heimann, 2003).Atmospheric CO2 inversions estimate net carbon exchange fluxesbetweensurfaceandatmospherebyutilizingCO2concentrationsatmeasurementsites,combinedwithanatmospherictransportmodeland prior information on fossil fuel carbon emissions and carbonexchangebetweentheatmosphereandland(andocean).Weuseddailymeannet fluxestimatesona spatial resolutionof3.75° lati‐tudeand5° longitude (JENA)and1.875° latitudeand3.75° longi‐tude(CAMS)overthevegetatedlandsurface.BothproductswerefirstaggregatedintoregionalscalesthenDOYzero‐crossingandSCAofcarbonfluxeswere,respectively,extracted.Notethatthefluxam‐plitudeisindirectlyrelatedtotheamplitudeintheatmosphericCO2 concentration,astheatmosphericconcentrationisroughlytheinte‐gralofthefluxes(Welpetal.,2016).

2.7 | Analytical approach

BasedontheextractedMODISDOYPmax,wefirsttestedthevalidityofframeworkbyrelatingittosummerclimateconditions(i.e.,tem‐peratureandwateravailability).Thefirstprincipleweformulatedfortheframeworkjustifiesusingsummerseasonasaperiodwhentheprimaryclimateconstraintdictatesvegetationphotosyntheticsea‐sonality,andtherefore,DOYPmax.Bothseasonaltotal(GPPTotal)andmaximumGPP(GPPPmax)werecalculatedto investigatethespatialand temporal relations betweenDOYPmax and vegetation produc‐tivity. Inorder to capture the seasonal distributionofGPPwith asimplemetric,weevaluatedtheratio(GPPRatio)oftotalGPPduringthefirsthalf(January1tothelong‐termmeanDOYPmax)tothatofthewholeyear.Additionally,thelengthofgrowingseasontogetherwithGPPPmaxwasconsidered toexplain theobservedpatternbe‐tween DOYPmax and GPPTotal (e.g., Xia et al., 2015). All exploredrelationshipswereexplainedasfunctionsofδDOYP,R and δDOYP,T. Independent eddy‐covariance tower GPP and GOME‐2 SIF‐basedretrievalswereusedforfurthertestingoftheframework.Notethatwelimitedtheuseoftheseindependentdataonlyforverifyingtheframeworkandnotthechangeanalysisbecauseoflimitedtemporalfrequencyandcoverageofthedata.

For the time series analysis, all trends in time series werecomputed as the slope of linear trends based on ordinary leastsquares regression. The significanceof the trendwas computedby using the nonparametricMann–Kendall trend test. The stan‐darderrorof thetrendslope isalsoreported.Weestimatedthedecadal trend based on the 5yearmoving average approach toreduce thepotential impactof first, last, andoutlierpoints.TheKendall'srankcorrelationcoefficient(r)wasusedtomeasurethe

ordinalassociationbetweengiventwoquantities.Tounderstandhowwarming‐inducedDOYPmax shift has characterizednorthernlandvegetationproductivity,weinvestigatedchangesintempera‐ture,DOYPmax,GPPTotal,andGPPRatio.ThisanalysiswasappliedtobothMODISandESMs‐basedretrievals.AtrendinDOYzero‐crossing of three CO2 data was, respectively, computed and correlationanalysisbetweenannualvariationsinDOYzero‐crossingandSCAwasperformed.

3  | RESULTS

3.1 | Spatial pattern of MODIS DOYPmax and its determinants

Adistinct spatial gradient exists inDOYPmax and in its positioningwith respect to the seasonal course of radiation and temperature(Figure2aandFigureS2a,b).Overall,DOYPmaxinarcticecosystemsis more closely aligned with DOYTmax (δDOYP,T=−9.3±5.5days,mean±1SD)thanDOYRmax(δDOYP,R=29.1±8.5days),whileintheborealecosystemsitshowsamuchcloseralignmentwithpeakradia‐tionlevels(δDOYP,T=−13.3±5.4days,δDOYP,R=12.9±10.5days).Inthetemperateregions,δDOYP,Risnegative(−9.5±27.0days),thatis,DOYPmaxprecedesDOYRmax.Temperatureandwateravailability(i.e.,RAP)limitingphotosyntheticactivityelucidatetheobservedre‐gionalvariationsinDOYPmaxpositioning.Every1°Cincreaseintem‐peratureresultsinaδDOYP,Rchangeof−5.7±0.1days(slope±SE,Figure 2b). In regions with negative δDOYP,R, every 1% decreaseinwateravailability results inaδDOYP,Rchangeof−1.8±0.1days(Figure2c).Theseresultsfollowthetwotenetsofourframework,asoutlinedearliercomplyingwiththelawsofminimum(Blackman,1905;Liebig,1841;Sprengel,1828).This suggests that theuseofDOYPmax and its positioning in relation to DOYRmax and DOYTmax represents a feasible approach to assess plant's adaptive state toclimaticconstraints.

3.2 | Climate constraints, MODIS DOYPmax, and seasonal vegetation productivity

Emerging climatic constraints to plant growth are directly linkedto changes in both GPPTotal (Figure 2d) andGPPPmax (Figure S2c).Regions with large GPPPmax are associated with tight synchronybetweenDOYPmaxandDOYRmax,thatis,bothenergyandwaterac‐cessibilityareleastlimiting(Bauerleetal.,2012).Ecosystemsundereithertemperature‐(δDOYP,R>0)orwater‐limited(δDOYP,R<0)en‐vironments show lowerphotosynthetic capacitybycomplying thegeneralideaofclimaticconstraintstoplantgrowth.Suchinteractionlimitingphotosyntheticactivityisalsotightlyassociatedwithgrow‐ingseasonduration(FigureS2d).ItisinterestingtonotethatinareaswiththelargestGPPTotal (~1.07kgC/m),DOYPmaxslightlyprecedesDOYRmax(δDOYP,R≈−7days)becauseofa jointcontrolbygrowingseason lengthandGPPPmax (Xiaetal.,2015).The longestgrowingseason duration (~6.5months) is foundwhen δDOYP,R is approxi‐matelyequalto−17days.Thisisknownas“phenologicaltrade‐off,”

6  |     PARK et Al.

that is, a longer growing season imposedbywarmer environmentmayresultinahigherGPPTotal,butwarmeranddriersummersmaysuppress GPPPmax, potentially offsetting the increased amount ofGPPTotal(Duveneck&Thompson,2017).

3.3 | Confirmed patterns from two independent data: SIF and eddy‐covariance tower GPP

Fluxtower‐measuredGPPdatafromtheeddy‐covariancenetworkandGOME‐2SIF confirm the abovepatterns observed inMODISGPP products, thus lending further support for the proposedDOYPmaxframework(Figure3andFigureS3).

3.4 | Changes in MODIS DOYPmax during last 17 years

Trend analyses reveal awidespread shift inMODISDOYPmax towardearlierinthegrowingseasondominatingacross60.6%ofthenorthern

vegetatedareaduringlast17years,and32.8%oftheareashowingasig‐nificantnegativetrend(p<0.1,Figure4).Thesechangesareseenacrossallthreebioclimaticzones,thatis,31.9%,38.7%,and26.8%ofthearc‐tic,boreal,andtemperateregions,respectively.Atahemisphericscale,wedetectedasignificanttrendtowardanearlierpeakphotosyntheticrate of −1.66±0.30days/decade (slope±SE, p<0.001) (Figure 5a),with regionallyvaryingdegreeofadvancing trends:a steeperchangein the boreal region (−2.46±0.47days/decade, p<0.001) relative tothe temperate (−1.07±0.26days/decade, p<0.001), and arctic re‐gions(−1.09±0.29days/decade,p<0.001).Thesechangesaremostlyassociatedwithwarminginthelandsnorthof30°N(Figures4and5).ThesensitivityofDOYPmaxtowarmingwasdetectedtobegreater inthe temperate (−4.27±1.50days/°C, p<0.001) than in the arctic(−3.88±1.29days/°C, p<0.001) and boreal (−3.91±1.02days/°C,p<0.001) regions. Note that regionally varying warming rates(TE<AR<BO) lead to adifferentorderof trendand sensitivity esti‐mates.ThesechangesinDOYPmaxareinterpretedasshiftsinδDOYP,R across the arctic (−1.98±7.30days, mean±SD, t test, p<0.001),

F I G U R E 2  Relativepositioningofpeakphotosyntheticactivitytimingwithrespecttotheseasonalcourseoftemperatureandradiation,anditsrelationtoclimaticconstraintsandproductivity.(a)GeographicaldistributionofδDOYP,T(DOYPmax−DOYTmax)andδDOYP,R (DOYPmax−DOYRmax)fornorthernecosystems.RegionaldistributionofδDOYP,T and δDOYP,Roverarctic(AR),boreal(BO),andtemperate(TE)regionsisgivenintheinsetviolinplotwithmeanand1SD(bracket).(b)PositioningofDOYPmaxseenastherelationbetweenδDOYP,R and δDOYP,T,withrespecttotemperature(°C).(c)Sameasbbutforwateravailability(i.e.,RAP).(d)Sameas(b)butforGPPTotal(kgC/m

2).MODIS‐derivedoutcomesareusedforthesepanels

     |  7PARK et Al.

boreal(−3.21±5.83days,p<0.001),andtemperate(−1.28±12.76days,p<0.001) regions (Figure S4a,b).We find that the observed shift inDOYPmaxismainlyresponsibleforthechangesinδDOYP,R(andδDOYP,T)becauseofrelativelystableDOYRmaxandDOYTmaxchanges(FigureS4andTableS1).Accordingtotheprinciplesinourframework,theshiftsresultinganewlyestablishedphotosyntheticseasonalitywithrespecttoseasonalclimatefactorsimplychangesinvegetationresponsetovary‐ingclimaticconstraints,thatis,reducedrelativeimportanceofthermalconstraint in thearctic andboreal vegetationwhileenhanced roleofwateravailabilityinthetemperateregions(Allenetal.,2010;Fuetal.,2015;Garonnaetal.,2018;Piaoetal.,2017;Figures2cand5).NotethatsomeregionstransitioningfrompositivetonegativeδDOYP,Rmightexperienceacritical tippingpointwhere theecosystemsmoves fromtemperature‐towardwater‐limitedecosystems(FigureS5).

3.5 | Implications of changing MODIS DOYPmax on seasonal vegetation productivity

ThechangesinDOYPmaxhaveregionallyvaryingimpactsonGPPTotal. An “earlierpeak–largerproductivity”pattern isdominantover the

arctic (−0.004±0.002kgCm−2 day−1, slope±SE,p<0.05)andbo‐real(−0.006±0.002kgCm−2 day−1,p<0.05)regionsunderawarm‐ingclimate(Figure5c).Theframeworkproposedearlierinformsthatmorefavorablethermalconditionsenablevegetationtoincreaseitssynchronywithseasonalityinincomingradiation,withtheseasonalcourseofphotosyntheticactivitytendingtowardthepeakofradia‐tion.Widelyreportedgrowingseasonextension(likelyinferredfromDOYPmaxshift,FigureS2d)partlyexplainssuch“earlierpeak–largerproductivity”relationacrossthearcticandborealregions(Parketal.,2016;Xuetal.,2013).Warmertemperaturesmightalsoenhanceac‐cesstokeynutrients(e.g.,nitrogen),thusstimulatingphotosyntheticratesoverthecourseoftheentiregrowingseason(Natali,Schuur,&Rubin,2012).Aweaker“earlierpeak–lessproductivity”patterninthetemperateregionsemergesduetocomplexclimate–vegetationinteractions(Figure5c).Here,warmerconditionswithoutmoisture‐stressresultinanearlierDOYPmaxandlargerGPPPmaxandGPPTotal. Inotherparts,wheremoisturestressisstronger,asignificantdeclineinbothGPPPmaxandGPPTotal isseendespiteearlierDOYPmax (e.g.,southwesternEurasia)(Angertetal.,2005).InordertocapturetheseasonaldistributionofGPPwithasimplemetricweevaluatedthe

F I G U R E 3   (a)SameasFigure2abutfortheindependentsatellitesun‐inducedfluorescence.(b)–(d)SameasFigure2b–dbutfortheeddycovariancetowermeasurement.Total92FLUXNETsites(FigureS2a)wereusedandeachdotrepresentsasinglesite

8  |     PARK et Al.

ratio (GPPRatio)of totalGPPduring the firsthalf (January1 to thelong‐termmeanDOYPmax) to thatof thewholeyear.We find thatDOYPmaxoccurringonedayearlierintheseasonincreasesGPPRatio by 0.28±0.07 (temperate, slope±SE, p<0.001) to 0.58±0.08%(boreal,p<0.001),clearlyindicatinganadvanceingrosscarbonas‐similation activity (Figure 5d;Duveneck&Thompson, 2017). Thisis an important indicator, as the photosynthetic activity is tightlylinkedtotheatmosphereviacarbon,water,andenergycycles.Thus,phaseshiftsincarbon,water,andenergycyclescouldbeanticipated(Richardsonetal.,2013).

3.6 | Changes in phase and amplitude of CO2 seasonal cycle

We found that earlier peak photosynthesis andmore carbon as‐similationintheearlypartofthegrowingseasonalteredthesea‐sonal course of atmospheric CO2 concentration. We used CO2 observations from Point Barrow and two state‐of‐the‐art CO2 inversion datasets (i.e., CAMS and JENA). The springtime down‐ward CO2 zero‐crossing date (DOYzero‐crossing) shows trends to‐ward earlier downwardDOYzero‐crossing in the threeCO2 datasets(Figure 6a). The phase of atmospheric CO2 at Point Barrow hasadvancedby1.84±0.20days/decade(slope±SE,p<0.001)since1972. We also observe advancing trends but steeper changesin both CAMS (−2.42±0.21days/decade, p<0.001) and JENA(−3.26±0.21days/decade, p<0.001). This shift corroboratestheadvancingDOYPmaxofgrossphotosyntheticactivityobserved

fromspaceandshowsthepotentialimplicationsofenhancedgrosscarbonassimilationintheearlygrowingseason,thatis,increasedGPPRatio (Barichivichetal.,2012;Randerson,Field,Fung,&Tans,1999; Figure 5a,d). Furthermore, similar towhatwe observed inthe analysis of DOYPmax and GPPTotal (Figure 5c), SCA of threeCO2data isnegativelyassociatedwithDOYzero‐crossing (Figure6b).ThesephaseshiftsintheCO2dataandtheirassociationwiththeenhancedseasonalamplitudesareinaccordancewithseveralob‐servations(Barichivichetal.,2012;Gravenetal.,2013;Randersonetal.,1999)andmodelingstudies(Duveneck&Thompson,2017;Zhao&Zeng,2014)suggestingenhancedpeakphotosyntheticac‐tivityanditsadvancingshift.

3.7 | Changes in ESMs simulated vegetation productivity and DOYPmaxWelastlyaskwhetherstate‐of‐the‐artterrestrialbiospheremod‐elscanreproducetheobservedDOYPmaxchangesandtheircon‐sequencesunderhistoricalandfutureclimatescenarios(Figure7).The ESMsproject an advancingDOYPmax across all northern bi‐oclimatic zones for the period 1980–2030.We see a pattern ofregionalDOYPmaxtrendsfromESMsanalogoustosatelliteobser‐vations, that is, a strong trend for shifting to earlier in the sea‐son over the boreal (−0.94±0.67days/decade, mean±1SD across all ESMs), arctic (−0.86±0.71days/decade), and temper‐ate(−0.58±0.61days/decade)regions.AllmodelsshowatightlylinkednegativerelationbetweenDOYPmaxandGPPTotal,revealing

F I G U R E 4   SpatialpatternofchangesinDOYPmaxandtemperatureduringlast17years(2000–2016).(a)DecadaltrendofMODISbasedDOYPmaxovernorthernlandduringlast17years.(b)Sameas(a)butforsummertemperature(June–August).Thetrendwasderivedbasedonordinaryleastsquaresregression

     |  9PARK et Al.

the “earlierpeak‐largerproductivity” tendencyas incurrent sat‐ellite observations. Particularly, temperature‐constrained arcticandboreal regions have a tighter linkagebetweenDOYPmax and GPPTotalthanthewarmertemperateregions.TheshiftinDOYPmax also increases the GPPRatio, indicating more carbon assimilationin the early part of the growing season than in the later period(Duveneck&Thompson,2017;Zhao&Zeng,2014).Thepaceoffuture(2050–2100)DOYPmaxshiftanditscontributiontoproduc‐tivityisprojectedtocontinue,buttobeslowerandweakerthanatpresent.

4  | DISCUSSION

Our analyses from long‐term satellite records and ESMs reveal awidespread shift inDOYPmax towardearlier in thegrowing season.

ThechangesareassociatedwithdivergentconsequencesonGPPTotal dependingondifferentstatesofclimateconstraintsonplantgrowth.For high latitude arctic ecosystems, the advancement in DOYPmax likelycontinuesinawarmerfutureclimateasseenintheESMsimula‐tions.Ourframeworktranslatesthechangeintoacontinuousrelaxa‐tionoftemperaturelimitonarcticvegetationphotosyntheticactivity.Arecentremotesensingbasedstudysupportsourstudybyidentify‐inga16.4%decline in theareaofvegetated landthat is limitedbytemperature(Keenan&Riley,2018).Yet,ourframeworksuggestsareductionintherelativeimportanceoftemperaturecontrolonplantphotosyntheticactivityratherthanatransitionalstatewhereotherclimate constraints primarily govern the ecosystem (Figure S4a).Indeed, long‐term ground based studies in the Arctic tundra haveshownthattemperatureisaprimarydriverofshrubgrowthanditsexpansioninarcticenvironment,whilesoilmoisturecontrolsthesen‐sitivityofgrowthresponsetowarming(Myers‐Smithetal.,2015).

F I G U R E 5  ChangesinDOYPmaxduringlast17years(2000–2016)andtheirimplicationsonnorthernvegetationproductivity. (a)InterannualvariationofDOYPmaxbyregions(arctic:AR;boreal:BO;temperate:TE;NorthernHemisphere:NH)anditstrendoverlast17years.Thedecadaltrendisestimatedbasedonthe5yearmovingaverageapproachtoreducethepotentialimpactoffirst,last,andoutlierpoints.ThinsolidlinewithmarkersandthicksolidlinerepresentannualDOYPmaxand5yearmovingaverage.Calculatedtrend(slope±SE)basedonordinaryleastsquaresregressionisgivenwithitssignificancelevel(doubleasterisksdenotep<0.001andsingleasterisksdenotep<0.05).ThesignificancewascomputedbyusingthenonparametricMann–Kendalltrendtest.(b)RelationbetweenregionalDOYPmaxandsummertemperature(June–August)anomalies.(c,d)SameasbbutforrespectiverelationbetweenDOYPmax and GPPTotal,andDOYPmaxandGPPRatioanomalies.Thesignificanceoftheslopeestimate(β±SE)isdenotedbydouble(p<0.001)andsingle(p<0.05)asterisks.TheKendall'srankcorrelationcoefficient(r)betweentwovariablesisalsogiven.Darkblue,lightblue,green,andgraystandforAR,BO,TE,andNH,respectively

10  |     PARK et Al.

Some of boreal ecosystems (northwest Russia and southFennoscandia,southandsoutheastCanada)showatransitionfrompositivetonegativeδDOYP,Rduringlasttwodecades(FigureS5).ThistransitiondoesnotnecessarilysignifyadeclineofGPPTotalbecauseof the “phenological trade‐off”mechanism (Figure S2d).However,itiscriticaltomonitortheseecosystemscontinuouslybecauseourframeworksuggests that theremaybeatippingpointwheretheymovefromtemperature‐towardwater‐limitedecosystems.Thatis,continuouswarminganddryingconditionsmayexacerbatemoisturestress, and therefore, productivity reduction in theseecosystems.Interestingly, a recent tree‐ring based study revealed that while

2°Cofwarmingmayincreaseoverallforestproductivity,additionalwarmingcould reverse this trendand lead tosubstantialmoisturestress (D’Orangeville et al., 2018). Also, multiple warming experi‐mentsconfirmthedynamismofclimateconstraintsonplantgrowthinthesouthernborealforestandhighlightthevulnerabilityoftheecosystemtoexcesswarminganddrying(e.g.,Reichetal.,2018).

Warmeranddrierconditionsovertemperatevegetation,wherenegativeδDOYP,Risdominant,generallyresultinadecreaseofplantgrowth.Widespreadincreaseoftreemortalityofthissusceptibleeco‐systemtoworseningmoisturestresshasbeenreported(Allenetal.,2010).Mostepidemicclimate‐inducedtreemortalityeventsoccurs

F I G U R E 6  AnalysisofatmosphericCO2concentrationatPointBarrowandtwoCO2inversionestimates.(a)TimeseriesofDOYzero‐crossing observedatPointBarrowatmosphericobservatoryandtwoindependentCO2inversiondatasets(CAMSandJENA).NotethattheCO2 fluxesforDOYzero‐crossingretrievaloftheinversiondatasetsarebasedonregionallyintegratedfluxesoverthearcticandborealzones,andalltrendestimatesarebasedonthe5yearmovingaverageapproach.Calculatedtrend(slope±SE)basedonordinaryleastsquaresregressionisgivenwithitssignificancelevel(doubleasterisksdenotep<0.001andsingleasterisksdenotep<0.05).ThesignificancewascomputedbyusingthenonparametricMann–Kendalltrendtest.(b)RelationbetweenDOYzero‐crossingandseasonalcycleamplitude(SCA)ofatmosphericCO2concentrationandfluxestimates.SCAanomalywasexpressedaspercentageoflong‐termmean.Thesignificanceoftheslopeestimate(β±SE)isdenotedbydouble(p<0.001)andsingle(p<0.05)asterisks.TheKendallrankcorrelationcoefficient(r)wasusedtomeasurethedegreeofassociation.Red,blue,andgreenstandforCO2datafromPointBarrow,CAMS,andJENA,respectively

F I G U R E 7  AnalysisofmultipleCMIP5Earthsystemmodels(ESMs)duringtwoseparateperiods:(a)1980–2030and(b)2050–2100.DecadaltrendofDOYPmax(left)anditsassociationtoGPPTotal(center)andGPPRatio(right)overnorthernlandsinferredfromthesevenESMs.Barchartswitherrorbarsdepictmean±1SDacrossallESMs.TheKendallrankcorrelationcoefficient(r)wasusedtomeasurethedegreeofassociation.Darkblue,lightblue,green,andgraystandforarctic(AR),boreal(BO),temperate(TE),andNorthernHemisphere(NH),respectively

     |  11PARK et Al.

overtheregionswherewateravailabilityistheprimaryclimatecon‐straintonphotosynthetic activity (i.e.,δDOYP,R<0, seeFigureS5).Itagreeswiththe“earlierpeak–lessproductivity”patterninwarmertemperatevegetationfromMODISdata.However,therelationwasnot reproducedby theESMs.Themodelsprojected thatwarming‐inducedearlierpeakphotosynthesis leadstoanenhancedseasonaltotalproductivity(Figure7a).Recentstudieshaveshownthatcurrentterrestrialcarboncyclemodelssubstantiallyoverestimate(underesti‐mate)positive(negative)effectsassociatedwithwarming(Buermannetal.,2018).Itispossiblybecausethesemodelsinadequatelycapturetheeffectsoftheseasonalbuild‐upofwaterstressonseasonalveg‐etationgrowth.

OuranalysesofDOYzero‐crossingandSCAconfirmtheadvancingandenhancingCO2 seasonal cycle innorthern lands (Barichivichetal.,2012;Forkeletal.,2016;Gravenetal.,2013).Anadditionalremarkmadehereforongoingchangesinbiosphere–atmosphereinteractionisanasymmetricenhancementofterrestrialphotosyn‐thetic activity.We find awidespreadwarming‐inducedDOYPmax advancement and GPPTotal increase across northern lands, andthesechangespossiblyplaya role inongoingshiftandamplifiedatmosphericCO2 seasonalcycle.This isbecausepeakphotosyn‐thesis rate explains about78%of the variationof seasonal totalproductivity and only 21% can be explained by growing seasonchanges(Xiaetal.,2015).Ourresultsconfirmthatalargerbenefi‐cialcarbonuptakefromanextendedgrowingseasonisdominatedby the later part of spring,whenmore fully developed leaf areawithmorefavorablelightandtemperatureisavailableforphoto‐syntheticactivity (Keenanetal.,2014).Togetherwith theseear‐lier studies,our findings suggest thatan intraseasonal scalemayprovideapossiblebutoverlookedmechanism for thechanges inatmosphericCO2seasonalcycle.Furthermore,theobservedshiftintherelative importanceofclimateconstraintsonplantgrowthmaybeapossiblemechanism for the recently reportedweaken‐ingtemperaturecontrolsonspringcarbonuptakeacrossnorthernlands(Piaoetal.,2017).

Furthermore,ourframeworkalsogivesinsightintothechangesingrowingseasondurationanditsimplicationoncarboncycle.Asdescribed in Figure 1, thermal inertia induced decoupling of ra‐diation and temperature characterizes a unique seasonal climateenvironment to local vegetation. For temperature‐constrainedecosystems (see Case 2 in Figure 1), DOYTmax‐ward DOYPmax positioning leads to strong temperature dependence in springphotosynthesis while light availability emerges as an importantcontrollerinautumnalactivity(Garonnaetal.,2018).Thisintrinsicphysical environment indicates contrasting responses of photo‐syntheticactivitytospringversusautumnwarming.Inthiscolden‐vironment,springwarminggenerallystimulatescarbonuptakebyextendingonsetofgrowingseason(Pulliainenetal.,2017).Incon‐trast,autumnalgrowingseasonextensionand itsphotosyntheticcarbon gain will be strongly limited by radiation (Bauerle et al.,2012).Multiplestudieshavereportedthattheincreaseofautumntemperatureresultsinnetcarbonlossindicatingmorerespiratorylossthanphotosyntheticgain innorthern lands (Commaneetal.,

2017;Piaoetal.,2008).Thesecontrastingseasonalresponsesalsopartiallyexplaintheobservedandprojectedasymmetricenhance‐ment of photosynthetic activity and carbon cycles in northernlands.However,furtherstudieswillberequiredtoidentifywhichcasetheautumngrowingseasonextensioncanleadtoincreasedphotosynthesissufficienttobalancethehigherrespirationcarbonloss.

Most of ESMs as well asMODIS GPP estimate used in thisstudy do not include photosynthetic temperature acclimationprocess.Thisphysiological adjustment is commonlyobservedasa shift in the optimum temperature for carbon assimilation ratebymodulatinglocalplant'smetabolism(Yamori,Hikosaka,&Way,2014).We expect that taking the photosynthetic thermal accli‐mationlikelyleadtoaslightlycloseralignmentbetweenDOYPmax andDOYRmaxthantheonewithouttheprocess.Italsomayreducethe observed DOYPmax sensitivity to warming (Smith,Malyshev,Shevliakova, Kattge, & Dukes, 2016). Nevertheless, we believethat theproposedDOYPmax frameworkand itschangesarevalidbecause ofmultiple evidence from independent datasets in thiswork (Figure 3 and Figure S3) and previous studies (Buitenwerfet al., 2015; Gonsamo et al., 2018; Rotenberg & Yakir, 2010).Interestingly, dendrometer‐based intra‐annual tree growth stud‐iesalsosupportourframework(e.g.,Rossietal.,2006).Ongoingefforts for advancements in modeling communities (Rogers etal.,2017)willhelptodeploytemperatureacclimationmodulesinESMsandthusbetterunderstandingsonseasonalphotosynthesisandDOYPmaxchangesareexpected.

Insummary,ourresultshighlightasignificantshiftinterrestrialphotosyntheticactivitynorthof30°N,implyingaconstantlyadapt‐ing state of climatic constraints on plant growth. A consensus ofmultipleEarthobservationsandESMsonthischangeimbuescon‐fidence inour findings.This is a criticaldevelopmentbecause theshiftsinpeakphotosynthesismaycausecascadingperturbationsinEarth system components that include carbon,water, and energybalances(Richardsonetal.,2013),aswellasecologicalinteractions(Walther, 2010). The frameworkproposedhere is oneof the firstattemptstointroducethetimeofpeakphotosynthesisasanindica‐torofaplant'sadaptivestatetoclimaticconstraints,andprovidesasimplified yet realistic framework for the complexmechanismsbywhichvariousclimaticfactorsconstrainplantgrowth.

ACKNOWLEDG EMENTS

WethankfullyacknowledgeFLUXNET,CMIP5,andMODISScienceteammembersfortheirdatamakingavailable.WealsoliketothankChristianRödenbeckandChevallierFrederic forsharing theJENAandCAMSCO2inversiondata.Further,wethankJoannaJoinerforproviding theGOME‐2 SIF data. Thisworkwas funded byNASAEarth Science Directorate (grants NNX16AO34H, NNX14AP80Aand NNX14AI71G) and the Research Council of Norway (grants287402 and 270992). A Natural Environment Research CouncilIndependentResearchFellowship(NE/L011859/1)fundedM.M.‐F.'scontribution.

12  |     PARK et Al.

AUTHOR CONTRIBUTIONS

TPandRBMdesignedtheresearch;TPperformedtheanalysisandwrotethedraft;andalltheauthorscontributedtotheinterpretationoftheresultsandthewritingofthepaper.

ORCID

Taejin Park https://orcid.org/0000‐0003‐0698‐6942

Marc Macias‐Fauria https://orcid.org/0000‐0002‐8438‐2223

Hans Tømmervik https://orcid.org/0000‐0001‐7273‐1695

Shilong Piao https://orcid.org/0000‐0001‐8057‐2292

R E FE R E N C E S

Allen,C.D.,Macalady,A.K.,Chenchouni,H.,Bachelet,D.,McDowell,N.,Vennetier,M.,…Cobb,N.(2010).Aglobaloverviewofdroughtand heat‐induced tree mortality reveals emerging climate changerisks for forests.Forest Ecology and Management,259(4),660–684.https://doi.org/10.1016/j.foreco.2009.09.001

Angert,A.,Biraud,S.,Bonfils,C.,Henning,C.C.,Buermann,W.,Pinzon,J.,…Fung,I. (2005).DriersummerscancelouttheCO2uptakeen‐hancement inducedbywarmer springs.Proceedings of the National Academy of Sciences of the United States of America,102(31),10823–10827.https://doi.org/10.1073/pnas.0501647102

Baldocchi, D., Falge, E., Gu, L., Olson, R., Hollinger, D., Running, S.,… Wofsy, S. (2001). FLUXNET: A new tool to study the tem‐poral and spatial variability of ecosystem–scale carbon di‐oxide, water vapor, and energy flux densities. Bulletin of the American Meteorological Society, 82(11), 2415–2434. https://doi.org/10.1175/1520‐0477(2001)0822.3.CO;2

Barichivich,J.,Briffa,K.,Myneni,R.,Schrier,G.,Dorigo,W.,Tucker,C.,…Melvin,T.(2014).Temperatureandsnow‐mediatedmoisturecon‐trolsofsummerphotosyntheticactivityinnorthernterrestrialeco‐systemsbetween1982and2011.Remote Sensing,6(2),1390–1431.https://doi.org/10.3390/rs6021390

Barichivich, J., Briffa, K. R., Osborn, T. J., Melvin, T. M., & Caesar, J.(2012).Thermalgrowingseasonandtimingofbiosphericcarbonup‐takeacrosstheNorthernHemisphere.Global Biogeochemical Cycles,26(4).https://doi.org/10.1029/2012GB004312

Bauerle,W.L.,Oren,R.,Way,D.A.,Qian,S.S.,Stoy,P.C.,Thornton,P.E.,…Reynolds,R.F.(2012).Photoperiodicregulationoftheseasonalpatternofphotosyntheticcapacityandtheimplicationsforcarboncycling. Proceedings of the National Academy of Sciences of the United States of America, 109(22), 8612–8617. https://doi.org/10.1073/pnas.1119131109

Blackman, F. F. (1905). Optima and limiting factors. Annals of Botany,19(74), 281–295. https://doi.org/10.1093/oxfordjournals.aob.a089000

Buermann, W., Forkel, M., O’Sullivan, M., Sitch, S., Friedlingstein,P., Haverd, V., … Richardson, A. D. (2018). Widespread seasonalcompensation effects of spring warming on northern plant pro‐ductivity. Nature, 562(7725), 110–114. https://doi.org/10.1038/s41586‐018‐0555‐7

Buitenwerf,R.,Rose,L.,&Higgins,S.I.(2015).Threedecadesofmulti‐dimensionalchangeingloballeafphenology.Nature Climate Change,5(4),364–368.https://doi.org/10.1038/nclimate2533

Chevallier,F.,Ciais,P.,Conway,T.,Aalto,T.,Anderson,B.,Bousquet,P.,…Fröhlich,M.(2010).CO2surfacefluxesatgridpointscaleestimatedfrom a global 21 year reanalysis of atmospheric measurements.Journal of Geophysical Research: Atmospheres,115(D21).

Chuine, I.,&Beaubien,E.G. (2001).Phenology isamajordeterminantof tree species range. Ecology Letters, 4(5), 500–510. https://doi.org/10.1046/j.1461‐0248.2001.00261.x

Commane, R., Lindaas, J., Benmergui, J., Luus, K. A., Chang, R.‐W.,Daube,B.C.,…Wofsy, S.C. (2017).Carbondioxide sources fromAlaska driven by increasing early winter respiration from Arctictundra.Proceedings of the National Academy of Sciences of the United States of America, 114(21), 5361–5366. https://doi.org/10.1073/pnas.1618567114

D’Orangeville,L.,Houle,D.,Duchesne,L.,Phillips,R.P.,Bergeron,Y.,&Kneeshaw,D.(2018).Beneficialeffectsofclimatewarmingonborealtreegrowthmaybe transitory.Nature Communications,9(1), 3213.https://doi.org/10.1038/s41467‐018‐05705‐4

Duveneck,M.J.,&Thompson,J.R.(2017).Climatechangeimposesphe‐nological trade‐offs on forest net primary productivity. Journal of Geophysical Research: Biogeosciences,122(9),2298–2313.https://doi.org/10.1002/2017JG004025

Eagleson, P. S. (2005).Ecohydrology: Darwinian expression of vegetation form and function.Cambridge:CambridgeUniversityPress.

Forkel, M., Carvalhais, N., Rödenbeck, C., Keeling, R., Heimann, M.,Thonicke,K.,…Reichstein,M. (2016). Enhanced seasonalCO2 ex‐changecausedbyamplifiedplantproductivity innorthernecosys‐tems.Science,351(6274),696–699.

Friedl,M.A.,Sulla‐Menashe,D.,Tan,B.,Schneider,A.,Ramankutty,N.,Sibley,A.,&Huang,X.(2010).MODISCollection5globallandcover:Algorithmrefinementsandcharacterizationofnewdatasets.Remote Sensing of Environment,114(1), 168–182. https://doi.org/10.1016/j.rse.2009.08.016

Fritz, S., See, L., McCallum, I., You, L., Bun, A., Moltchanova, E., …Obersteiner, M. (2015). Mapping global cropland and field size.Global Change Biology, 21(5), 1980–1992. https://doi.org/10.1111/gcb.12838

Fu,Y.H.,Zhao,H.,Piao,S.,Peaucelle,M.,Peng,S.,Zhou,G.,…Janssens,I. A. (2015). Declining global warming effects on the phenol‐ogy of spring leaf unfolding. Nature, 526(7571), 104. https://doi.org/10.1038/nature15402

Garonna,I.,deJong,R.,Stöckli,R.,Schmid,B.,Schenkel,D.,Schimel,D.,&Schaepman,M.E.(2018).Shiftingrelativeimportanceofclimaticconstraints on land surface phenology. Environmental Research Letters,13(2),024025.https://doi.org/10.1088/1748‐9326/aaa17b

Gelaro,R.,McCarty,W.,Suárez,M.J.,Todling,R.,Molod,A.,Takacs,L.,…Zhao,B.(2017).Themodern‐eraretrospectiveanalysisforresearchand applications, version 2 (MERRA‐2). Journal of Climate, 30(14),5419–5454.https://doi.org/10.1175/JCLI‐D‐16‐0758.1

Gonsamo,A.,Chen,J.M.,&Ooi,Y.W.(2018).Peakseasonplantactiv‐ityshifttowardsspringisreflectedbyincreasingcarbonuptakebyextratropical ecosystems.Global Change Biology,24(5),2117–2128.https://doi.org/10.1111/gcb.14001

Graven,H.D.,Keeling,R. F., Piper, S.C., Patra, P.K., Stephens,B.B.,Wofsy,S.C.,…Bent, J.D. (2013).EnhancedseasonalexchangeofCO2bynorthernecosystemssince1960.Science,341(6150),1085–1089.https://doi.org/10.1126/science.1239207

Heinsch, F. A., Zhao,M., Running, S.W., Kimball, J. S.,Nemani, R. R.,Davis,K.J.,…Flanagan,L.B. (2006).Evaluationofremotesensingbased terrestrial productivity from MODIS using regional towereddy flux network observations. IEEE Transactions on Geoscience and Remote Sensing, 44(7), 1908–1925. https://doi.org/10.1109/TGRS.2005.853936

Ito, A., Inatomi, M., Huntzinger, D. N., Schwalm, C., Michalak, A. M.,Cook,R.,…Post,W.M.(2016).Decadaltrendsintheseasonal‐cycleamplitude of terrestrial CO2 exchange resulting from the ensem‐ble of terrestrial biospheremodels. Tellus B: Chemical and Physical Meteorology,68(1),28968.

Joiner,J.,Yoshida,Y.,Guanter,L.,&Middleton,E.M.(2016).Newmethodsfortheretrievalofchlorophyllredfluorescencefromhyperspectral

https://orcid.org/0000-0003-0698-6942https://orcid.org/0000-0003-0698-6942https://orcid.org/0000-0002-8438-2223https://orcid.org/0000-0002-8438-2223https://orcid.org/0000-0001-7273-1695https://orcid.org/0000-0001-7273-1695https://orcid.org/0000-0001-8057-2292https://orcid.org/0000-0001-8057-2292https://doi.org/10.1016/j.foreco.2009.09.001https://doi.org/10.1073/pnas.0501647102https://doi.org/10.1175/1520-0477(2001)082%3C2415:FANTTS%3E2.3.CO;2https://doi.org/10.1175/1520-0477(2001)082%3C2415:FANTTS%3E2.3.CO;2https://doi.org/10.3390/rs6021390https://doi.org/10.1029/2012GB004312https://doi.org/10.1073/pnas.1119131109https://doi.org/10.1073/pnas.1119131109https://doi.org/10.1093/oxfordjournals.aob.a089000https://doi.org/10.1093/oxfordjournals.aob.a089000https://doi.org/10.1038/s41586-018-0555-7https://doi.org/10.1038/s41586-018-0555-7https://doi.org/10.1038/nclimate2533https://doi.org/10.1046/j.1461-0248.2001.00261.xhttps://doi.org/10.1046/j.1461-0248.2001.00261.xhttps://doi.org/10.1073/pnas.1618567114https://doi.org/10.1073/pnas.1618567114https://doi.org/10.1038/s41467-018-05705-4https://doi.org/10.1002/2017JG004025https://doi.org/10.1002/2017JG004025https://doi.org/10.1016/j.rse.2009.08.016https://doi.org/10.1016/j.rse.2009.08.016https://doi.org/10.1111/gcb.12838https://doi.org/10.1111/gcb.12838https://doi.org/10.1038/nature15402https://doi.org/10.1038/nature15402https://doi.org/10.1088/1748-9326/aaa17bhttps://doi.org/10.1175/JCLI-D-16-0758.1https://doi.org/10.1111/gcb.14001https://doi.org/10.1126/science.1239207https://doi.org/10.1109/TGRS.2005.853936https://doi.org/10.1109/TGRS.2005.853936

     |  13PARK et Al.

satellite instruments: Simulations and application to GOME‐2 andSCIAMACHY.Atmospheric Measurement Techniques,9(8).

Jolly, W. M., Nemani, R., & Running, S. W. (2005). A general‐ized, bioclimatic index to predict foliar phenology in responseto climate. Global Change Biology, 11(4), 619–632. https://doi.org/10.1111/j.1365‐2486.2005.00930.x

Keenan,T.F.,Gray, J.,Friedl,M.A.,Toomey,M.,Bohrer,G.,Hollinger,D.Y.,…Richardson,A.D. (2014).Netcarbonuptakehas increasedthroughwarming‐induced changes in temperate forest phenology.Nature Climate Change, 4(7), 598–604. https://doi.org/10.1038/nclimate2253

Keenan,T.,&Riley,W.(2018).Greeningofthelandsurfaceintheworld'scoldregionsconsistentwithrecentwarming.Nature Climate Change,8(9),825.https://doi.org/10.1038/s41558‐018‐0258‐y

Lasslop,G.,Reichstein,M.,Papale,D.,Richardson,A.D.,Arneth,A.,Barr,A.,…Wohlfahrt,G. (2010).Separationofnetecosystemexchangeinto assimilation and respiration using a light response curve ap‐proach:Critical issuesandglobalevaluation.Global Change Biology,16(1),187–208.https://doi.org/10.1111/j.1365‐2486.2009.02041.x

Liebig, J. (1841).Organic chemistry in its applications to agriculture and physiology.J.Owen.London:TaylorandWalton.

Myers‐Smith,I.H.,Elmendorf,S.C.,Beck,P.S.A.,Wilmking,M.,Hallinger,M.,Blok,D.,…Vellend,M.(2015).Climatesensitivityofshrubgrowthacross the tundrabiome.Nature Climate Change,5(9),887.https://doi.org/10.1038/nclimate2697

Natali, S.M.,Schuur,E.A.,&Rubin,R. L. (2012). Increasedplantpro‐ductivity inAlaskan tundraasa resultofexperimentalwarmingofsoilandpermafrost.Journal of Ecology,100(2),488–498.https://doi.org/10.1111/j.1365‐2745.2011.01925.x

Nemani,R.R.,Keeling,C.D.,Hashimoto,H., Jolly,W.M.,Piper, S.C.,Tucker,C. J.,…Running, S.W. (2003).Climate‐driven increases inglobalterrestrialnetprimaryproductionfrom1982to1999.Science,300(5625),1560–1563.

Olson, D. M., Dinerstein, E., Wikramanayake, E. D., Burgess, N. D.,Powell,G.V.N.,Underwood,E.C.,…Kassem,K.R.(2001).Terrestrialecoregionsoftheworld:Anewmapof lifeonearth:Anewglobalmap of terrestrial ecoregions provides an innovative tool for con‐serving biodiversity. BioScience, 51(11), 933–938. https://doi.org/10.1641/0006‐3568(2001)051[0933:TEOTWA]2.0.CO;2

Park, T., Ganguly, S., Tømmervik, H., Euskirchen, E. S., Høgda, K.‐A.,Karlsen, S. R.,…Myneni, R.B. (2016).Changes in growing seasonduration and productivity of northern vegetation inferred fromlong‐termremotesensingdata.Environmental Research Letters,11(8),084001.https://doi.org/10.1088/1748‐9326/11/8/084001

Parmesan,C.,&Yohe,G. (2003).Agloballycoherentfingerprintofcli‐matechangeimpactsacrossnaturalsystems.Nature,421(6918),37.https://doi.org/10.1038/nature01286

Pastorello,G.,Papale,D.,Chu,H.,Trotta,C.,Agarwal,D.,Canfora,E.,…Torn,M.(2017).Anewdatasettokeepasharpereyeonland‐airex‐changes.Eos, Transactions American Geophysical Union (Online),98(8).https://doi.org/10.1029/2017EO071597

Peng,C.,Ma,Z.,Lei,X.,Zhu,Q.,Chen,H.,Wang,W.,…Zhou,X.(2011).A drought‐induced pervasive increase in tree mortality acrossCanada'sborealforests.Nature Climate Change,1(9),467.https://doi.org/10.1038/nclimate1293

Piao,S.,Ciais,P.,Friedlingstein,P.,Peylin,P.,Reichstein,M.,Luyssaert,S.,…Chen,A.(2008).Netcarbondioxidelossesofnorthernecosystemsinresponsetoautumnwarming.Nature,451(7174),49–52.

Piao,S.,Liu,Z.,Wang,T.,Peng,S.,Ciais,P.,Huang,M.,…Tans,P.P.(2017).Weakening temperature control on the interannual variations ofspringcarbonuptakeacrossnorthernlands.Nature Climate Change,7(5),359–363.https://doi.org/10.1038/nclimate3277

Prentice,I.C.,Cramer,W.,Harrison,S.P.,Leemans,R.,Monserud,R.A.,&Solomon,A.M.(1992).Specialpaper:Aglobalbiomemodelbased

on plant physiology and dominance, soil properties and climate.Journal of Biogeography,117–134.https://doi.org/10.2307/2845499

Pulliainen,J.,Aurela,M.,Laurila,T.,Aalto,T.,Takala,M.,Salminen,M.,… Vesala, T. (2017). Early snowmelt significantly enhances borealspringtime carbon uptake. Proceedings of the National Academy of Sciences of the United States of America, 114(42), 11081–11086.https://doi.org/10.1073/pnas.1707889114

Randerson,J.,Field,C.,Fung,I.,&Tans,P.(1999).Increasesinearlysea‐sonecosystemuptakeexplainrecentchangesintheseasonalcycleofatmosphericCO2athighnorthernlatitudes.Geophysical Research Letters,26(17),2765–2768.

Reich, P. B., Sendall, K.M., Stefanski, A., Rich, R. L., Hobbie, S. E., &Montgomery, R. A. (2018). Effects of climate warming on photo‐synthesis in boreal tree species depend on soil moisture.Nature,562(7726),263.https://doi.org/10.1038/s41586‐018‐0582‐4

Richardson,A.D.,Black,T.A.,Ciais,P.,Delbart,N.,Friedl,M.A.,Gobron,N.,…Migliavacca,M. (2010). Influenceof springandautumnphe‐nologicaltransitionsonforestecosystemproductivity.Philosophical Transactions of the Royal Society of London B: Biological Sciences,365(1555),3227–3246.

Richardson,A.D.,Keenan,T.F.,Migliavacca,M.,Ryu,Y.,Sonnentag,O.,&Toomey,M.(2013).Climatechange,phenology,andphenologicalcontrolofvegetation feedbacks to theclimate system.Agricultural and Forest Meteorology, 169, 156–173. https://doi.org/10.1016/j.agrformet.2012.09.012

Rodell,M.,Houser,P.R.,Jambor,U.,Gottschalck,J.,Mitchell,K.,Meng,C.‐J., … Toll, D. (2004). The global land data assimilation system.Bulletin of the American Meteorological Society, 85(3), 381–394.https://doi.org/10.1175/BAMS‐85‐3‐381

Rödenbeck,C.,Houweling,S.,Gloor,M.,&Heimann,M.(2003).CO2fluxhistory1982–2001inferredfromatmosphericdatausingaglobalin‐versionofatmospherictransport.Atmospheric Chemistry and Physics,3(6),1919–1964.

Rogers,A.,Medlyn, B. E.,Dukes, J. S., Bonan,G., vonCaemmerer, S.,Dietze,M.C.,…Zaehle,S.(2017).Aroadmapforimprovingtherepre‐sentationofphotosynthesisinEarthsystemmodels.New Phytologist,213(1),22–42.https://doi.org/10.1111/nph.14283

Rossi, S., Deslauriers, A., Anfodillo, T., Morin, H., Saracino, A.,Motta, R., … Borghetti, M. (2006). Conifers in cold environ‐ments synchronize maximum growth rate of tree‐ring formationwith day length. New Phytologist, 170(2), 301–310. https://doi.org/10.1111/j.1469‐8137.2006.01660.x

Rotenberg,E.,&Yakir,D.(2010).Contributionofsemi‐aridforeststotheclimatesystem.Science,327(5964),451–454.

Running,S.,Mu,Q.,&Zhao,M.(2015).MOD17A2HMODIS/terragrossprimary productivity 8‐day L4 global 500m SIN grid V006.NASAEOSDISLandProcessesDAAC.

Smith,N.G.,Malyshev,S.L.,Shevliakova,E.,Kattge, J.,&Dukes, J.S.(2016). Foliar temperature acclimation reduces simulated carbonsensitivity to climate.Nature Climate Change,6(4), 407.https://doi.org/10.1038/nclimate2878

Sprengel, C. (1828). Von den Substanzen der Ackerkrume und desUntergrundes[Aboutthesubstancesintheplowlayerandthesub‐soil].Journal for Technische and Okonomische Chemie,2,397–421.(inGerman)

Taylor, K. E., Stouffer, R. J., & Meehl, G. A. (2012). An overviewof CMIP5 and the experiment design. Bulletin of the American Meteorological Society, 93(4), 485–498. https://doi.org/10.1175/BAMS‐D‐11‐00094.1

Thomson,A.M.,Calvin,K.V.,Smith,S.J.,Kyle,G.P.,Volke,A.,Patel,P.,…Clarke,L.E.(2011).RCP4.5:Apathwayforstabilizationofradiativeforcingby2100.Climatic Change,109(1–2),77.

Thoning,K.W.,Tans,P.P.,&Komhyr,W.D.(1989).AtmosphericcarbondioxideatMaunaLoaObservatory:2.AnalysisoftheNOAAGMCC

https://doi.org/10.1111/j.1365-2486.2005.00930.xhttps://doi.org/10.1111/j.1365-2486.2005.00930.xhttps://doi.org/10.1038/nclimate2253https://doi.org/10.1038/nclimate2253https://doi.org/10.1038/s41558-018-0258-yhttps://doi.org/10.1111/j.1365-2486.2009.02041.xhttps://doi.org/10.1038/nclimate2697https://doi.org/10.1038/nclimate2697https://doi.org/10.1111/j.1365-2745.2011.01925.xhttps://doi.org/10.1111/j.1365-2745.2011.01925.xhttps://doi.org/10.1641/0006-3568(2001)051%5B0933:TEOTWA%5D2.0.CO;2https://doi.org/10.1641/0006-3568(2001)051%5B0933:TEOTWA%5D2.0.CO;2https://doi.org/10.1088/1748-9326/11/8/084001https://doi.org/10.1038/nature01286https://doi.org/10.1029/2017EO071597https://doi.org/10.1038/nclimate1293https://doi.org/10.1038/nclimate1293https://doi.org/10.1038/nclimate3277https://doi.org/10.2307/2845499https://doi.org/10.1073/pnas.1707889114https://doi.org/10.1038/s41586-018-0582-4https://doi.org/10.1016/j.agrformet.2012.09.012https://doi.org/10.1016/j.agrformet.2012.09.012https://doi.org/10.1175/BAMS-85-3-381https://doi.org/10.1111/nph.14283https://doi.org/10.1111/j.1469-8137.2006.01660.xhttps://doi.org/10.1111/j.1469-8137.2006.01660.xhttps://doi.org/10.1038/nclimate2878https://doi.org/10.1038/nclimate2878https://doi.org/10.1175/BAMS-D-11-00094.1https://doi.org/10.1175/BAMS-D-11-00094.1

14  |     PARK et Al.

data, 1974–1985. Journal of Geophysical Research: Atmospheres,94(D6),8549–8565.

Vautard,R.,Yiou,P.,&Ghil,M.(1992).Singular‐spectrumanalysis:Atool‐kitforshort,noisychaoticsignals.Physica D: Nonlinear Phenomena,58(1–4),95–126.https://doi.org/10.1016/0167‐2789(92)90103‐T

Walther,G.R.(2010).Communityandecosystemresponsestorecentcli‐matechange.Philosophical Transactions of the Royal Society of London B: Biological Sciences,365(1549),2019–2024.

Welp,L.R.,Patra,P.K.,Rödenbeck,C.,Nemani,R.,Bi, J.,Piper,S.C.,&Keeling,R.F. (2016). Increasing summernetCO2 uptake inhighnorthernecosystemsinferredfromatmosphericinversionsandcom‐parisonstoremote‐sensingNDVI.Atmospheric Chemistry and Physics,16(14),9047–9066.

Xia,J.,Niu,S.,Ciais,P.,Janssens,I.A.,Chen,J.,Ammann,C.,…Luo,Y.(2015).Jointcontrolofterrestrialgrossprimaryproductivitybyplantphenology and physiology. Proceedings of the National Academy of Sciences of the United States of America,112(9),2788–2793.https://doi.org/10.1073/pnas.1413090112

Xu,L.,Myneni,R.B.,ChapinIII,F.S.,Callaghan,T.V.,Pinzon,J.E.,Tucker,C.J.,…Stroeve,J.C.(2013).Temperatureandvegetationseasonal‐ity diminishment over northern lands.Nature Climate Change,3(6),581–586.https://doi.org/10.1038/nclimate1836

Yamori,W.,Hikosaka, K., &Way,D. A. (2014). Temperature responseofphotosynthesisinC3,C4,andCAMplants:Temperatureacclima‐tionandtemperatureadaptation.Photosynthesis Research,119(1–2),101–117.https://doi.org/10.1007/s11120‐013‐9874‐6

Zhang,Y.,Guanter,L.,Berry,J.A.,Joiner,J.,vanderTol,C.,Huete,A.,…Köhler,P. (2014).Estimationofvegetationphotosyntheticcapacity

fromspace‐basedmeasurementsofchlorophyllfluorescenceforter‐restrialbiospheremodels.Global Change Biology,20(12),3727–3742.https://doi.org/10.1111/gcb.12664

Zhao,F.,&Zeng,N.(2014).ContinuedincreaseinatmosphericCO2sea‐sonalamplitude inthe21stcenturyprojectedbytheCMIP5Earthsystemmodels.Earth System Dynamics,5(2),423.

Zhao, M., Heinsch, F. A., Nemani, R. R., & Running, S. W. (2005).ImprovementsoftheMODISterrestrialgrossandnetprimarypro‐ductionglobaldataset.Remote Sensing of Environment,95(2),164–176.https://doi.org/10.1016/j.rse.2004.12.011

Zhou,S.,Zhang,Y.,Ciais,P.,Xiao,X.,Luo,Y.,Caylor,K.K.,…Wang,G.(2017).Dominantroleofplantphysiologyintrendandvariabilityofgross primary productivity in North America. Scientific Reports, 7,41366.https://doi.org/10.1038/srep41366

SUPPORTING INFORMATION

Additional supporting information may be found online in theSupportingInformationsectionattheendofthearticle. 

How to cite this article:ParkT,ChenC,Macias‐FauriaM,etal.Changesintimingofseasonalpeakphotosyntheticactivityinnorthernecosystems.Glob Change Biol. 2019;00:1–14. https://doi.org/10.1111/gcb.14638

https://doi.org/10.1016/0167-2789(92)90103-Thttps://doi.org/10.1073/pnas.1413090112https://doi.org/10.1073/pnas.1413090112https://doi.org/10.1038/nclimate1836https://doi.org/10.1007/s11120-013-9874-6https://doi.org/10.1111/gcb.12664https://doi.org/10.1016/j.rse.2004.12.011https://doi.org/10.1038/srep41366https://doi.org/10.1111/gcb.14638