Upload
truongkhuong
View
217
Download
0
Embed Size (px)
Citation preview
Evalua&on,applica&onanddevelopmentofESMinChina
BinWang1,2
Contributors:
1.LASG,Ins&tuteofAtmosphericPhysics,CAS
2.CESS,TsinghuaUniversity
3.BeijingNormalUniversity
4.BeijingClimateCenter
5.FirstIns&tuteofOceanography
16thWGCM,24‐26Sep,2012,Hamburg,Germany 1
Outline
Preliminaryevalua&onsonCMIP5
Importantapplica&ons
Progressincouplingtechniques
FuturedevelopmentforAR6
2
GroupName Ins&tu&on ModelName
LASG‐CESSIns&tuteofAtmosphericPhysics,ChineseAcademyofSciences;TsinghuaUniversity
FGOALS‐g2
LASGIns&tuteofAtmosphericPhysics,ChineseAcademyofSciences
FGOALS‐s2
BNU BeijingNormalUniversity BNU‐ESM
BCC BeijingClimateCenter BCC‐CSM1.1
FIO FirstIns&tuteofOceanography FIO‐ESM1.0
CMIP5modelsinChina
3
Main Variables EvaluatedVariablename Component Evalua&onMethod
Precipita&on Globe RootMeanSquareError
GlobalmeanSATanomaly(SATA)
DetrendofSATATaylorskill(Taylor,2001)
TropicSSTAannualcycle
Pacific PaSernCorrela&on
Warmextremes1980‐1999,21regions
Rela&veRMS
AsianMonsoonJJASmean,annualcycleindex
PaSernCorrela&on
ENSOSD,varianceannualcycle,amplitude,zonalwindstress
Directcomparisonorcorrela&onwithobserva&on
4
Globalarea‐weightedmeanRMSEofPrecipita&on(1980‐2005)(OBS:GPCP)
Mean: 1.31; Max: 1.72; Min: 1.09 (unit: mm/day)
5
20thCenturySATVaria&on(1870‐2005)
Mean: 0.52; Max: 0.64; Min: 0.396
TropicPacificAnnualCycle(2⁰S‐2⁰N,120⁰E‐80⁰W)
Mean: 0.63; Max: 0.82; Min: 0.267
CMIP3CMIP5best worst
CRUCPC
Rela&veRMS
Performance of AR5 models on
presenting warm extremes for
1980-1999 at 21 regions
(Zhao et al, 2012)
8
SkillsofCMIPModelsonJJASAsianPrecipita&onandWind
(OBS:ERA40)
(OBS
:GPC
P)
FGOALS‐g2
BNU‐ESM
FGOALS‐s2BCC‐CSM1.1
CNRM
‐CM5
MIROC‐ESM
(Zhouetal.2012;Sperberetal.2012)
9
5mm/day
季风爆发 季风撤退
5mm/day
Onset Withdraw
Rela&verainfallrateintheBengalBayregion
(Sperberetal.2012)
Summermonsoonannualcycleindex
Dura&on
peak
10
CMIP5MMM
CMIP5MMMCMIP3MMM
CMIP3MMM
ChinaMMM
ChinaMMM
FGOALS‐g2
FGOALS‐g2
BNU‐ESM
BNU‐ESM
FGOALS‐s2
FGOALS‐s2
BCC‐CSM1.1
BCC‐CSM1.1
Onset
Withd
raw
Dura&on
Peak
Asiansummermonsoon1) Comparing with CMIP3
models, CMIP5 modelsimprove the Asian summermonsoon significantly;
2) China MMM performs betterthan CMIP3 MMM on onsetand duration, but worse onwithdraw and peak;
3) FGOALS-g2 presentsbetter onset and durationbut worse withdraw andpeak than both CMIP3MMM and CMIP5 MMM;
4) FGOALS-g2 presents betteronset, withdraw andduration but worse peakthan China MMM.
(Zhouetal.2012;Sperberetal.2012)
(OBS:GPCP) 11
FGOALS‐g2 BNU‐ESM
FIO‐ESM1.0FGOALS‐s2BCC‐CSM1.1
HadISSTHurrell2008
INMCM3.0
GFDL‐CM2p1
HadGEM2‐ES
IPSL‐CM5A‐LRIPSL‐CM5A‐MR
Othermodels:CNRM‐CM5,INMCM4.0,MIROC5,MIROC‐ESM,MPI‐ESM‐LR,NorESM1‐M
Nino3.4region
Taylorskillsofv
ariancean
nualcycle
Standarddevia&on
(Huangetal,2012) 12
BNU‐ESM
FIO‐ESM1.0
FGOALS‐s2
BCC‐CSM1.1HadISST+ERA40Hurrell2008+ERA40
FGOALS
‐g2
INMCM3.0
MIROC5
NorESM1‐M
INMCM4.0
Widthofzonalwindstressanomalies
Othermodels:CNRM‐CM5,HadGEM2‐ES,IPSL‐CM5A‐LR,IPSL‐CM5A‐MR,MIROC‐ESM,MPI‐ESM‐LR
Amplitud
eofpow
erspe
ctrum
Nino3.4region
(Huangetal,2012)13
Outline
Preliminaryevalua&onsonCMIP5
Importantapplica&ons
Progressincouplingtechniques
FuturedevelopmentforAR6
14
Afribu&on,CenturialProjec&onandDecadalPredic&on
HistoricalSimula&ons
FutureProjec&ons
CMIP5models
CESM
BNU‐ESM
FGOALS
FIO‐ESM
BCC‐CSM
Developed
Countries
Developing
Countries
AllCountries
Ems‐Hist(1850‐2005)
DevelopedCountries
DevelopingCountries
AllCountries
CancunPledgeandfuturereduc&on(2005‐2100)
Modeling scheme courtesy of Dong, 2010hSp://114.255.218.74/index.jsp
15
ESM‐Histfrom10CMIP5models
CO2concentra&on:
Developed:61%and71%;
Developing:39%and29%;
CO2radia&veforcing:
Developed:53%and62%;
Developing:47%and38%;
Airtemperature:
Developed:60%and64%
Developing:40%and36%
ALL = All countries’ emission= =AX1+NX1AX1= developed countries’ emissions ;NX1= developing countries’ emissions;
Developedanddevelopingworldresponsibili&esforhistoricalclimatechange
Wei et al., 2012, PNAS16
Developedanddevelopingworldresponsibili&esforCO2mi&ga&on
Contribu&onstoreduc&on
ofglobalwarming:
Developed:33%and35%
Developing:67%and
65%
Mi&ga&ongoal:
Currentmi&ga&on
cannotcontrolrising
temperaturewithin2℃
in2100.Itisnecessaryto
fillthegapwithmore
ambi&ousmi&ga&on
efforts.
ABNB=allcountriesnon‐
mi&ga&on;
ACNB=onlydeveloped
countriesfollowCancun
(2006‐2020)+80%
reduc&onsby2050and
zeroemissionsby2100;
ABNC=onlydeveloping
countriesfollowCancun
(2006‐2020)+50%
reduc&onsby2050and
zeroemissionsby2100;
ACNC=allcountries
mi&ga&onWei et al., 2012, PNAS
18
SurfaceAirTemperatureAnomaliesbyFGOALS‐g2(2ndsampleofhistoricalrun)
FGOALS-g2
HadCRUT3
Cor:0.84
Surface air temperature anomalies (SATA)
1900192019401960198020002020
0.60
0.30
0.00
‐0.30
‐0.60
1966
-199
0
1986
-200
5
FGOALS-g2 (Cor=0.678) FGOALS-g2 (Cor=0.786)
HadCRUT3 HadCRUT3
19
30-year predictions of SATA by FGOALS-g2
2011-2035(Forecast)
1966
-199
019
86-2
005
After detrendH
indc
ast
Outline
Preliminaryevalua&onsonCMIP5
Importantapplica&ons
Progressincouplingtechniques
FuturedevelopmentforAR6
20
Lackofdirectfluxexchangeinavailablecouplers
21
Centralizedfluxcouplerwithoutdirectfluxexchange
Decentralizednon‐fluxcouplerwithoutdirectfluxexchange
NCARCoupler
OASISCoupler
(Liuetal,2011)
C‐Coupler:toovercometheimperfec&on(thefirstcouplerdevelopedbyChinesescien&sts)
22
Centralizedfluxcouplerformodulariza&onandstandardiza&on
Directfluxexchangeforbeferparallelefficiencyofhigh‐resolu&onmodels
(Liuetal,2011)
Onlineensembletoolforcoupledclimatemodel
MultiICEnsemble
MultiModelEnsembleDifferen
tscena
riosfo
ron
lineen
sembles
MotivationHowweathernoise/atmosphericmodeluncertaintyimpactstheclimateresponse?
Ensemblecouplingsolwareframework
(Xueetal,2011) 23
ENSO
OnlineMul&‐ini&alcondi&onensemble
OnlineMul&‐modelensemble
OfflineMul&‐modelensemble
Singleini&alcondi&on
(Xueetal,2011) 24
(Xueetal,2012)
Outline
Preliminaryevalua&onsonCMIP5
Importantapplica&ons
Progressincouplingtechniques
FuturedevelopmentforAR6
25
Furtherevalua&onofChinaCMIP5modelsandotherCMIP5modelsforunderstandingofuncertain&esinthesemodels;
Developmentofhigh‐resolu&onAGCMandOGCM(0.5⁰‐0.1⁰ inhorizontaland50layersinver&calwithin5years);
Developmentofearthsystemmodelbasedonourowncoupler:C‐Coupler;
Improvementsofonlineensemblecouplingframeworkandensemblecouplingstrategies;
Improvementsofdynamicalcores,physicalparameteriza&ons,andcarboncycleprocesses.
FuturePlanforAR6
26
Thank you
27