When are we committed to crossing critical (1.5 or 2 °C) temperature thresholds?
Cristian Proistosescu1 Kyle Armour1 Gerard Roe1 Peter Huybers2
1University of Washington
2Harvard University
AGU Fall Meeting 2017 Courtesy of NASA’s Earth Observatory Two questions
1. When will we cross 1.5 or 2 °C global warming thresholds (e.g., following high or low emission scenarios) – subject to constraints from the observed global energy budget?
2. When will we be geophysically committed to crossing 1.5 or 2 °C global
warming thresholds? What do CMIP5 models say?
Final DraftFinal (7 JuneDraftFinal 2013) (7 June Draft 2013) (7 June 2013) Chapter 12 Chapter 12 Chapter 12 IPCC WGI IPCC Fifth WGI Assessment IPCC Fifth WGI Assessment Report Fifth Assessment Report Report
CMIP5 projections (IPCC AR5)
2 °C
Figure 12.5:Figure Time 12.5:Figure series Time of12.5: globalseries Time ofannual globalseries mean ofannual global surface mean annual air surface temper mean air aturesurface temper anomalies airature temper anomalies (relativeature anomalies to(relative 1986–2005) to(relative 1986–2005) from to CMIP51986–2005) from CMIP5 from CMIP5 concentration-drivenconcentration-drivenconcentration-driven experime experiments. Projections nts.experime Projections arents. shown Projections are for shown each are RCPfor shown each for theRCP for multi each for themodelRCP multi for mean themodel multi(solid mean model lines) (solid andmean lines) the (solid and lines) the and the 5–95% range5–95% (±1.64 range5–95% standard (±1.64 range deviation)standard (±1.64 deviation)standard across the deviation) across distributi the across ondistributi of individualthe ondistributi of individual modelson of individual(shading). models (shading). Discontinuitiesmodels (shading). Discontinuities at 2100Discontinuities at 2100 at 2100 are due toare different due toare differentnumbers due to numbersofdifferent models ofnumbers performing models of performing models the extension performing the extension runs thebeyond extension runs the beyond 21st runs century the beyond 21st and century the have 21st andno century physical have andno physical have no physical meaning.meaning. Only one meaning.Only ensemble one Onlyensemble member one ensemble ismember used from ismember used each from modelis used each and from model numbers each and model numbersin the and figu innumbersre the indicate figu inre the theindicate figunumberre the indicate of number the of number of different differentmodels contributing modelsdifferent contributing models to the contributingdifferent to the differenttime to periods.the timedifferent Noperiods. ranges time No periods. are ranges given No are for ranges giventhe RCP6.0 arefor giventhe projections RCP6.0 for the projections RCP6.0 beyond projections 2100 beyond 2100 beyond 2100 as only twoas onlymodels twoas are onlymodels available. two are models available. are available.
Do Not Cite,Do Not Quote Cite,Do or Not Quote Distribute Cite, or Quote Distribute or Distribute 12-129 12-129 12-129 Total pages:Total 175 pages: Total 175 pages: 175 What do CMIP5 models say?
Final DraftFinal (7 JuneDraftFinal 2013) (7 June Draft 2013) (7 June 2013) Chapter 12 Chapter 12 Chapter 12 IPCC WGI IPCC Fifth WGI Assessment IPCC Fifth WGI Assessment Report Fifth Assessment Report§ Models Report may not agree with observed CMIP5 projections (IPCC AR5) global warming and energy budget constraints
§ Models may not span full range of plausible future warming
§ Computationally expensive to run different emissions scenarios, so can’t ask questions like, when are we 2 °C geophysically committed to 2 °C?
§ Not clear which physical factors are contributing to uncertainty in projected warming
Figure 12.5:Figure Time 12.5:Figure series Time of12.5: globalseries Time ofannual globalseries mean ofannual global surface mean annual air surface temper mean air aturesurface temper anomalies airature temper anomalies (relativeature anomalies to(relative 1986–2005) to(relative 1986–2005) from to CMIP51986–2005) from CMIP5 from CMIP5 concentration-drivenconcentration-drivenconcentration-driven experime experiments. Projections nts.experime Projections arents. shown Projections are for shown each are RCPfor shown each for theRCP for multi each for themodelRCP multi for mean themodel multi(solid mean model lines) (solid andmean lines) the (solid and lines) the and the 5–95% range5–95% (±1.64 range5–95% standard (±1.64 range deviation)standard (±1.64 deviation)standard across the deviation) across distributi the across ondistributi of individualthe ondistributi of individual modelson of individual(shading). models (shading). Discontinuitiesmodels (shading). Discontinuities at 2100Discontinuities at 2100 at 2100 are due toare different due toare differentnumbers due to numbersofdifferent models ofnumbers performing models of performing models the extension performing the extension runs thebeyond extension runs the beyond 21st runs century the beyond 21st and century the have 21st andno century physical have andno physical have no physical meaning.meaning. Only one meaning.Only ensemble one Onlyensemble member one ensemble ismember used from ismember used each from modelis used each and from model numbers each and model numbersin the and figu innumbersre the indicate figu inre the theindicate figunumberre the indicate of number the of number of different differentmodels contributing modelsdifferent contributing models to the contributingdifferent to the differenttime to periods.the timedifferent Noperiods. ranges time No periods. are ranges given No are for ranges giventhe RCP6.0 arefor giventhe projections RCP6.0 for the projections RCP6.0 beyond projections 2100 beyond 2100 beyond 2100 as only twoas onlymodels twoas are onlymodels available. two are models available. are available.
Do Not Cite,Do Not Quote Cite,Do or Not Quote Distribute Cite, or Quote Distribute or Distribute 12-129 12-129 12-129 Total pages:Total 175 pages: Total 175 pages: 175 ARMOUR ET AL.: SEA ICEARMOUR REVERSIBILITY ET AL.: SEA ICE REVERSIBILITYX-5 X-5
Global radiative forcingGlobal (F )changesapproximatelylinearlywithtimeovertheCO radiative forcing (F )changesapproximatelylinearlywithtimeovertheCO2 2 ARMOUR ET AL.: SEA ICE REVERSIBILITY ARMOURX-5 ET AL.: SEA ICE REVERSIBILITY X-5 2 2 rampings, by about 3.7rampings, Wm per by 70 about yr, which 3.7 Wm is the periodper 70 ofyr, CO which2 doubling is the period or halving of CO2 doubling or halving Global radiative forcing (F )changesapproximatelylinearlywithtimeovertheCOARMOUR ET AL.: SEA ICEGlobal REVERSIBILITY radiative forcing (F )changesapproximatelylinearlywithtimeovertheCO2 X-5 2 [Myhre et al., 1998]. The[Myhre o↵set et in al. Figure, 1998]. 1 Thebetween o↵set warming in Figure (red) 1 between and cooling warming (blue) (red) and cooling (blue) Global radiative forcing2 (F )changesapproximatelylinearlywithtimeovertheCO2 2 rampings, by about 3.7 Wm per 70 yr, which is therampings, period of by CO about2 doubling 3.7 Wm or halvingper 70 yr, which is the period of CO2 doubling or halving trajectories implies a laggedtrajectories response implies of hemispheric-mean a lagged response annual-mean of hemispheric-mean surface tempera- annual-mean surface tempera- 2 [Myhrerampings, et al., 1998]. by about The 3 o.7↵set Wm in Figureper 70 1 yr, between which[Myhre is warming the et period al. (red), 1998]. of andCO2 The coolingdoubling o↵set (blue) or in halving Figure 1 between warming (red) and cooling (blue)
ture anomalies ( TNH andture anomaliesTSH), as expected ( TNH and from deepTSH), ocean as expected heat storage from deep [e.g., oceanHeld heat et storage [e.g., Held et trajectories[Myhre implies et al., a 1998]. lagged The response o↵set of in hemispheric-mean Figure 1 betweentrajectories annual-mean warming implies (red) a surface lagged and tempera- response cooling (blue) of hemispheric-mean annual-mean surface tempera- al., 2010]. In order to approximatelyal., 2010]. In order account to forapproximately this lag, we account consider for the this evolution lag, we of consider ice the evolution of ice trajectories implies a lagged response of hemispheric-mean annual-mean surface tempera- ture anomalies ( TNH and TSH), as expected from deepture anomalies ocean heat ( storageTNH and [e.g., TSHHeld), as et expected from deep ocean heat storage [e.g., Held et area as a function of hemisphericarea as a function temperature of hemispheric rather than temperature time. A justification rather than for time. this A justification for this al., 2010].ture In anomalies order to ( approximately TNH and TSH account), as expected for thisal. from lag,, 2010]. wedeep consider In ocean order heat the to evolution approximately storage [e.g., of iceHeld account et for this lag, we consider the evolution of ice treatment is that annual-meantreatment Arctic is that sea annual-mean ice area has been Arctic found sea ice to declinearea has linearly been found with to decline linearly with area asal. a, 2010].function In of order hemispheric to approximately temperature account rather forarea than this as lag, a time. function we A consider justification of hemispheric the evolution for this temperature of ice rather than time. A justification for this increasing global-meanincreasing temperature global-mean across a range temperature of GCMs, across emissions a range scenarios, of GCMs, and emissions scenarios, and treatmentarea is as that a function annual-mean of hemispheric Arctic sea temperature ice area hastreatment rather been found than is that totime. decline annual-mean A justification linearly Arctic with for sea this ice area has been found to decline linearly with climates [Gregory et al.,climates 2002; Ridley [Gregory et al. et, 2008; al., 2002;WintonRidley, 2006, et al. 2008,, 2008; 2011].Winton Specifically,, 2006, 2008, 2011]. Specifically, increasingtreatment global-mean is that annual-mean temperature Arcticacross sea a range ice areaincreasing of GCMs, has been global-mean emissions found to scenarios, decline temperature linearly and across with a range of GCMs, emissions scenarios, and we extend the argumentswe of extendWinton the[2011], arguments relating of Winton hemispheric[2011], ice relating cover to hemispheric global forcing ice cover to global forcing climatesincreasing [Gregory global-mean et al., 2002; temperatureRidley et al., across 2008; Winton aclimates range, 2006, of [Gregory GCMs, 2008, et emissions 2011]. al., 2002; Specifically, scenarios,Ridley et and al., 2008; Winton, 2006, 2008, 2011]. Specifically, through through climates [Gregory et al., 2002; Ridley et al., 2008; Winton, 2006, 2008, 2011]. Specifically, we extend the arguments of Winton [2011], relating hemisphericwe extendOur the ice arguments cover approach to global of Winton forcing [2011], relating hemispheric ice cover to global forcing throughwe extend the arguments of Winton [2011], relatingthrough hemispheric ice cover to global forcing § Use a 2-layer ocean model (Held et al. 2010; dTu radiative radiativedTu cu =forcing Tu + F + " (cTresponseud Tu=)(1) Tu + F + " (Td Tu)(1) Armour 2017) that includes the essential physics dt dt through governing global-mean surface warming: Ocean heat uptake dT dTu efficacy u cu = Tu + F + " (Td Tu)(1)cu = Tu + F + " (Td Tu)(1) dt dT dt dT d dTd u cd =upper (T oceanu T dT)(2)cd = (Tu Td)(2) cu = Tu + F + " (Td Tu)(1)dt u dt dt
deep ocean Td dTd dTd cd = (Tu Td)(2)cd = (Tu Td)(2) dt ~⌧ zˆ dt ⌧ x x dT ~⌧ zˆ ⌧ d !Te = ⇥ = yˆ !Te = ⇥ = yˆ (3) (3) cd = (Tu Td)(2)⇢f ⇢f dt ⇢f ⇢f ~⌧ zˆ ⌧ x ~⌧ zˆ ⌧ x !Te = ⇥ = yˆ !(3)Te = ⇥ = yˆ (3) ⇢f ⇢f 1 ~⌧ ⇢f ⇢f ~⌧ zˆ ⌧ x 1 ~⌧ we = zˆ we = zˆ (4) (4) !Te = ⇥ = yˆ ⇢ · r⇥(3)f ⇢f ⇢f ⇢ · r⇥f 1 ~⌧ 1 ~⌧ we = zˆ (4)we = zˆ (4) ⇢ · r⇥f @w @v @u @@ww⇢ · @r⇥v @fv @@uu @w @v @u 1 ~⌧ ~u = xˆ + ~u = yˆ + xˆ + zˆ yˆ + (5) zˆ (5) we = zˆ r⇥ @y @z r⇥@z @@(4)xy @z @x @@yz @x @x @y ⇢ · r⇥f ✓ ◆ ✓ ✓ ◆ ◆✓ ✓ ◆ ◆ ✓ ◆ @w @v @u @w @v @u @w @v @u @w @v @u ~u = xˆ + yˆ + ~uzˆ = (5) xˆ + yˆ + zˆ (5) r⇥ @y @z DRAFT@z @x @x @r⇥y December14,2017,5:39pm@y @z @z @x @x @ DRAFTy ✓ @w◆ @v ✓ @u ◆ @w ✓ DRAFT@v◆ @u ✓ ◆ ✓ December14,2017,5:39pm◆ ✓ ◆ DRAFT ~u = xˆ + yˆ + zˆ (5) r⇥ @y @z @z @x @x @y ✓ ◆ ✓ ◆ ✓ ◆ DRAFT December14,2017,5:39pmDRAFT DRAFT December14,2017,5:39pm DRAFT DRAFT December14,2017,5:39pm DRAFT ARMOUR ET AL.: SEA ICE REVERSIBILITY X-5
Global radiative forcing (F )changesapproximatelylinearlywithtimeovertheCOARMOUR ET AL.: SEA ICE REVERSIBILITY 2 X-5
Global radiative forcing2 (F )changesapproximatelylinearlywithtimeovertheCO2 rampings, by about 3.7 Wm per 70 yr, which is the period of CO2 doubling or halving
2 [Myhrerampings, et al., 1998]. by about The 3 o.7↵set Wm in Figureper 70 1 yr, between which is warming the period (red) of andCO2 coolingdoubling (blue) or halving trajectories[Myhre implies et al., a 1998]. lagged The response o↵set of in hemispheric-mean Figure 1 between annual-mean warming (red) surface and tempera- cooling (blue) trajectories implies a lagged response of hemispheric-mean annual-mean surface tempera- ture anomalies ( TNH and TSH), as expected from deep ocean heat storage [e.g., Held et al., 2010].ture In anomalies order to ( approximately TNH and TSH account), as expected for this from lag, wedeep consider ocean heat the evolution storage [e.g., of iceHeld et area asal. a, 2010].function In of order hemispheric to approximately temperature account rather for than this lag, time. we A consider justification the evolution for this of ice treatmentarea is as that a function annual-mean of hemispheric Arctic sea temperature ice area has rather been found than totime. decline A justification linearly with for this increasingtreatment global-mean is that annual-mean temperature Arcticacross sea a range ice area of GCMs, has been emissions found to scenarios, decline linearly and with climatesincreasing [Gregory global-mean et al., 2002; temperatureRidley et al., across 2008; Winton a range, 2006, of GCMs, 2008, emissions 2011]. Specifically, scenarios, and climates [Gregory et al., 2002; Ridley et al., 2008; Winton, 2006, 2008, 2011]. Specifically, we extend the arguments of Winton [2011], relating hemisphericOur ice cover approach to global forcing throughwe extend the arguments of Winton [2011], relating hemispheric ice cover to global forcing § Use a 2-layer ocean model (Held et al. 2010; 4 Global surface temperature response Armour 2017) that includes the essential physics to abrupt CO quadrupling through governing global-mean surface warming: 2 dT 63 c u = T + F + " (T T )(1) u dt u d u dTu c = T + F + " (T T )(1)4 u dt u d u 2 Slow warming on timescale of the dTd deep ocean cd = (Tu Td)(2)21 dt Fast warming on dTd timescale of the cd = (Tu Td)(2)surface dt (°C) T change Temperature 0 0 50 100 150 x 0 50 100 150 ~⌧ zˆ ⌧ Year after CO quadrupling !T = ⇥ = yˆ (3) 2 e ⇢f ⇢f ~⌧ zˆ ⌧ x !T = ⇥ = yˆ (3) e ⇢f ⇢f 1 ~⌧ w = zˆ (4) e ⇢ · r⇥f 1 ~⌧ w = zˆ (4) e ⇢ · r⇥f @w @v @u @w @v @u ~u = xˆ + yˆ + zˆ (5) r⇥ @y @z @z @x @x @y ✓ @w◆ @v ✓ @u ◆ @w ✓ @v◆ @u ~u = xˆ + yˆ + zˆ (5) r⇥ @y @z @z @x @x @y ✓ ◆ ✓ ◆ ✓ ◆ DRAFT December14,2017,5:39pm DRAFT DRAFT December14,2017,5:39pm DRAFT ARMOUR ET AL.: SEA ICE REVERSIBILITY X-5
Global radiative forcing (F )changesapproximatelylinearlywithtimeovertheCOARMOUR ET AL.: SEA ICE REVERSIBILITY 2 X-5
Global radiative forcing2 (F )changesapproximatelylinearlywithtimeovertheCO2 rampings, by about 3.7 Wm per 70 yr, which is the period of CO2 doubling or halving
2 [Myhrerampings, et al., 1998]. by about The 3 o.7↵set Wm in Figureper 70 1 yr, between which is warming the period (red) of andCO2 coolingdoubling (blue) or halving trajectories[Myhre implies et al., a 1998]. lagged The response o↵set of in hemispheric-mean Figure 1 between annual-mean warming (red) surface and tempera- cooling (blue) trajectories implies a lagged response of hemispheric-mean annual-mean surface tempera- ture anomalies ( TNH and TSH), as expected from deep ocean heat storage [e.g., Held et al., 2010].ture In anomalies order to ( approximately TNH and TSH account), as expected for this from lag, wedeep consider ocean heat the evolution storage [e.g., of iceHeld et area asal. a, 2010].function In of order hemispheric to approximately temperature account rather for than this lag, time. we A consider justification the evolution for this of ice treatmentarea is as that a function annual-mean of hemispheric Arctic sea temperature ice area has rather been found than totime. decline A justification linearly with for this increasingtreatment global-mean is that annual-mean temperature Arcticacross sea a range ice area of GCMs, has been emissions found to scenarios, decline linearly and with climatesincreasing [Gregory global-mean et al., 2002; temperatureRidley et al., across 2008; Winton a range, 2006, of GCMs, 2008, emissions 2011]. Specifically, scenarios, and climates [Gregory et al., 2002; Ridley et al., 2008; Winton, 2006, 2008, 2011]. Specifically, we extend the arguments of Winton [2011], relating hemisphericOur ice cover approach to global forcing throughwe extend the arguments of Winton [2011], relating hemispheric ice cover to global forcing § Use a 2-layer ocean model (Held et al. 2010; 4 Global surface temperature response Armour 2017) that includes the essential physics to abrupt CO quadrupling through governing global-mean surface warming: 2 dT 63 c u = T + F + " (T T )(1) u dt u d u dTu c = T + F + " (T T )(1)4 u dt u d u 2
dTd cd = (Tu Td)(2)21 dt dTd cd = (Tu Td)(2) dt (°C) T change Temperature 0 0 50 100 150 x 0 50 100 150 ~⌧ zˆ ⌧ Year after CO quadrupling !T = ⇥ = yˆ (3) 2 e ⇢f ⇢f ~⌧ zˆ ⌧ x !T = ⇥ = yˆ (3) e ⇢f ⇢f 1 ~⌧ w = zˆ (4) e ⇢ · r⇥f 1 ~⌧ w = zˆ (4) e ⇢ · r⇥f @w @v @u @w @v @u ~u = xˆ + yˆ + zˆ (5) r⇥ @y @z @z @x @x @y ✓ @w◆ @v ✓ @u ◆ @w ✓ @v◆ @u ~u = xˆ + yˆ + zˆ (5) r⇥ @y @z @z @x @x @y ✓ ◆ ✓ ◆ ✓ ◆ DRAFT December14,2017,5:39pm DRAFT DRAFT December14,2017,5:39pm DRAFT ARMOURARMOUR ET AL.: ET AL.: SEA SEAICE REVERSIBILITY ICE REVERSIBILITY X-5X-5
GlobalGlobal radiative radiative forcing forcing (F )changesapproximatelylinearlywithtimeovertheCO (F )changesapproximatelylinearlywithtimeovertheCO2 2
ARMOUR ET2 AL.:ARMOUR2 SEAARMOUR ICE ET REVERSIBILITY AL.: ET AL.: SEA ICE SEA REVERSIBILITY ICE REVERSIBILITY X-5 X-5X-5 rampings,rampings, by about by about 3.7 Wm 3.7 Wm per 70per yr, 70 which yr, which is the is periodthe period of CO of2 COdoubling2 doubling or halving or halving ARMOUR ET AL.: SEA ICE REVERSIBILITY X-5 Global radiativeGlobal forcingGlobal radiative (F radiative)changesapproximatelylinearlywithtimeovertheCO forcing forcing (F )changesapproximatelylinearlywithtimeovertheCO (F )changesapproximatelylinearlywithtimeovertheCO2 [Myhre[Myhre et al. et, al.1998]., 1998]. The The o↵set o↵ inset Figure in Figure 1 between 1 between warming warming (red) (red) and and cooling cooling2 (blue) (blue) 2 ARMOUR ET AL.: SEA ICE REVERSIBILITY X-5 Global radiative forcing (F )changesapproximatelylinearlywithtimeovertheCO2 2 2 2 rampings, by aboutrampings, 3.7 Wm by aboutper 3 70.7 yr, Wm which per is 70 the yr, period which of is CO the perioddoubling of orCO halvingdoubling or halving trajectoriestrajectoriesrampings, implies implies a lagged by a about lagged response 3 response.7 Wm of hemispheric-mean ofper hemispheric-mean 70 yr, which annual-mean is2 annual-mean the period surface of surface2 CO tempera-2 doubling tempera- or halving Global radiative forcing2 (F )changesapproximatelylinearlywithtimeovertheCO2 rampings, by about 3.7 Wm per 70 yr, which is the[Myhre period et of al. CO,[ 1998].2Myhredoubling[Myhre The et al. et o or,↵ 1998]. al.set halving, in 1998]. Figure The The o↵ 1set between o↵ inset Figure in warming Figure 1 between 1 (red) between warming and warming cooling (red) (blue) (red) and and cooling cooling (blue) (blue) tureture anomalies anomalies ( T (NH TandNH and TSH ),TSH as), expected as expected from from deep deep ocean ocean heat heat storage storage [e.g., [e.g.,HeldHeld et et 2 [Myhrerampings, et al., 1998]. by about The 3 o.7↵set Wm in Figureper 70 1 yr, between which is warming the period (red) of andCO2 coolingdoubling (blue) or halving trajectoriesal., implies 2010].al.trajectories, 2010].trajectories In a lagged order In order implies to response approximately implies to approximately a lagged of a hemispheric-mean lagged response account response account of for hemispheric-mean this for of annual-mean hemispheric-mean this lag, lag, we consider we consider surface annual-mean the annual-mean evolution thetempera- evolution surface of surface ice of tempera- ice tempera- [Myhre et al., 1998]. The o↵set in Figure 1 between warming (red) and cooling (blue) trajectories implies a lagged response of hemispheric-meanture anomalies annual-meanareaarea asture ( a asT functionture anomaliesNH asurface functionand anomalies ofT tempera- hemisphericSH ( of ),T hemispheric asNH ( expectedandTNH and temperatureTSH temperature from),T asSH), expected deep as rather expected ocean rather from than heat than from deeptime. storage time. deep ocean A justification ocean [e.g.,A heat justificationHeld heat storage etfor storage this [e.g.,for this [e.g.,HeldHeld et et trajectories implies a lagged response of hemispheric-meanal., 2010]. In annual-mean orderal., 2010]. to approximately In surface order to tempera- approximately account for this account lag, we for consider this lag, the we evolution consider of the ice evolution of ice ture anomalies ( TNH and TSH), as expected from deep oceantreatment heattreatment storage isal. that, 2010].is [e.g., that annual-mean Inannual-meanHeld order et to Arctic approximately Arctic sea ice sea area ice accountarea has beenhas for been found this found to lag, decline to we decline consider linearly linearly the with evolution with of ice
area as aincreasing functionarea of global-mean as hemispheric a function temperature of temperature hemispheric across rather temperature a range than of time. rather GCMs, A justification than emissions time. scenarios, A for justification this and for this al., 2010].ture In anomalies order to ( approximately TNH and TSH account), as expected for this from lag, wedeep consider oceanincreasing heat thearea evolution storage global-mean as a function [e.g., of ice temperatureHeld of hemispheric et across temperature a range of GCMs,rather than emissions time. scenarios, A justification and for this treatmentclimates isclimates thattreatment [ annual-meanGregorytreatment [Gregory is et that al. is et, Arctic that 2002; annual-meanal., 2002; annual-meanRidley seaRidley ice et area Arctic al. et, Arctichas 2008; al. sea, been 2008;Winton ice sea found areaWinton ice, area has 2006, to, decline been2006, has 2008, been found 2008, linearly2011]. found 2011]. to Specifically, decline with to Specifically, decline linearly linearly with with area asal. a, 2010].function In of order hemispheric to approximately temperature account rather for than this lag, time. we A consider justification the evolution for this of ice increasingwe global-mean extendweincreasing extend theincreasing arguments the temperature global-mean arguments global-mean of Winton of across temperatureWinton temperature[2011], a range[2011], relating across of relating GCMs, across hemispheric a range hemispheric a emissions range of ice GCMs, of cover ice scenarios, GCMs, cover toemissions global to emissions globaland forcing scenarios, forcing scenarios, and and treatmentarea is as that a function annual-mean of hemispheric Arctic sea temperature ice area has rather been found than totime. decline A justification linearly with for this climatesthrough [Gregorythroughclimates etclimates al., [Gregory 2002; [GregoryRidley et al. et et, 2002; al. al.,, 2002; 2008;RidleyRidleyWinton et al. et,, 2008; 2006, al., 2008;Winton 2008,Winton 2011]., 2006,, Specifically, 2006, 2008, 2008, 2011]. 2011]. Specifically, Specifically, increasingtreatment global-mean is that annual-mean temperature Arcticacross sea a range ice area of GCMs, has been emissions found to scenarios, decline linearly and with we extend the argumentswe extendwe extend the of Winton arguments the arguments[2011], of Winton relating of Winton[2011], hemispheric[2011], relating relating ice hemispheric cover hemispheric to global ice cover forcing ice cover to global to global forcing forcing increasing global-mean temperature across a range of GCMs, emissions scenarios, and climates [Gregory et al., 2002; Ridley et al., 2008; Winton, 2006, 2008, 2011]. Specifically,dTu dTu cu c = T=u +TF++F" +(T" d (TTu)(1)T )(1) through throughthrough dtu dt u d u climates [Gregory et al., 2002; Ridley et al., 2008; Winton, 2006, 2008, 2011]. Specifically, we extend the arguments of Winton [2011], relating hemisphericOur ice cover approach to global forcing we extend the arguments of Winton [2011], relating hemispheric ice cover to global forcing through dTu dTdTd udTdTd u § Use a 2-layer ocean model (Held et al. 2010; Step 1:cu Draw =priors Tu of+c dc uF , c +d c ,=u " =, ( ( T= andTdTu= u( + T T T ufromuFdu)(1))(2)++T dfitsF" )(2) +(T" d (TTdu)(1)Tu)(1) Armour 2017) that includes the essential physics of 2-layerdt model to CMIP5dtdt modeldtdt response to through governing global-mean surface warming: CO2 forcing (Geoffroy et al. 2013)
dTu ~⌧ zˆ ⌧ x x dTd ~⌧ dTdzˆdT ⌧ cu = Tu + F + " (Td Tu)(1) !Te = !T =⇥ ⇥= =d yˆ yˆ (3) (3) dt cd = (Teu cdTd)(2)cd = (=Tu (TTud)(2)Td)(2) dT dt ⇢ f ⇢dtf dt⇢f ⇢f c u = T + F + " (T T )(1) u dt u d u
1 x1 ~⌧ ~⌧ x dTd x ~⌧ wzˆe =we =zˆ⌧ z~⌧ˆ z~⌧ˆ zˆ ⌧ ⌧ (4) (4) cd = (Tu Td)(2) !Te = ⇥ = !T⇢e =·⇢ !r⇥Tyˆe =⇥f ⇥=f = yˆ yˆ (3) (3) (3) dt ⇢f ⇢f ⇢· fr⇥⇢f ⇢f ⇢f dT c d = (T T )(2) d dt u d @w @w@v @v @u @u@w @w @v @v@u @u ~⌧ zˆ ⌧ x ~u =~u = 1 xˆ + xˆ +~⌧ 1 1yˆ + y~⌧ˆ + ~⌧ zˆ zˆ (5) (5) r⇥r⇥ @ywe @=y@z zˆ@z w@ze = @wze@zˆ =x @xzˆ @x @x@y @y (4) (4) (4) !Te = ⇥ = yˆ ✓ (3) ◆ ✓ ◆ ✓ ◆ ⇢f ⇢f ✓ ⇢ · r⇥◆ f ✓ ⇢ · r⇥⇢ ◆· r⇥f ✓f ◆ ~⌧ zˆ ⌧ x !Te = ⇥ = yˆ (3) ⇢f ⇢f DRAFTDRAFT December14,2017,5:39pm December14,2017,5:39pm DRAFT DRAFT @w @v @w@u@w@v@w@v @u@v@u@w@u@w @v @v@u @u 1 ~⌧ ~u = ~u =xˆ~u+= xˆ +yˆxˆ++ yˆ +zˆ yˆ + (5)zˆ zˆ (5) (5) w = zˆ r⇥ @y r⇥ @zr⇥(4) @y@ z@y @z @x@z @z@ x@z @x @y@x @x @x@y @y e ⇢ · r⇥f ✓ ◆ ✓ ✓✓ ◆ ◆◆✓ ✓✓ ◆ ◆◆✓ ✓ ◆ ◆ 1 ~⌧ w = zˆ (4) e ⇢ · r⇥f DRAFTDRAFTDRAFT December14,2017,5:39pm December14,2017,5:39pm December14,2017,5:39pm DRAFT DRAFT DRAFT @w @v @u @w @v @u ~u = xˆ + yˆ + zˆ (5) r⇥ @y @z @z @x @x @y ✓ @w◆ @v ✓ @u ◆ @w ✓ @v◆ @u ~u = xˆ + yˆ + zˆ (5) r⇥ @y @z @z @x @x @y ✓ ◆ ✓ ◆ ✓ ◆ DRAFT December14,2017,5:39pm DRAFT DRAFT December14,2017,5:39pm DRAFT ARMOURARMOUR ET AL.: ET AL.: SEA SEAICE REVERSIBILITY ICE REVERSIBILITY X-5X-5
GlobalGlobal radiative radiative forcing forcing (F )changesapproximatelylinearlywithtimeovertheCO (F )changesapproximatelylinearlywithtimeovertheCO2 2
ARMOUR ET2 AL.:ARMOUR2 SEAARMOUR ICE ET REVERSIBILITY AL.: ET AL.: SEA ICE SEA REVERSIBILITY ICE REVERSIBILITY X-5 X-5X-5 rampings,rampings, by about by about 3.7 Wm 3.7 Wm per 70per yr, 70 which yr, which is the is periodthe period of CO of2 COdoubling2 doubling or halving or halving ARMOUR ET AL.: SEA ICE REVERSIBILITY X-5 Global radiativeGlobal forcingGlobal radiative (F radiative)changesapproximatelylinearlywithtimeovertheCO forcing forcing (F )changesapproximatelylinearlywithtimeovertheCO (F )changesapproximatelylinearlywithtimeovertheCO2 [Myhre[Myhre et al. et, al.1998]., 1998]. The The o↵set o↵ inset Figure in Figure 1 between 1 between warming warming (red) (red) and and cooling cooling2 (blue) (blue) 2 ARMOUR ET AL.: SEA ICE REVERSIBILITY X-5 Global radiative forcing (F )changesapproximatelylinearlywithtimeovertheCO2 2 2 2 rampings, by aboutrampings, 3.7 Wm by aboutper 3 70.7 yr, Wm which per is 70 the yr, period which of is CO the perioddoubling of orCO halvingdoubling or halving trajectoriestrajectoriesrampings, implies implies a lagged by a about lagged response 3 response.7 Wm of hemispheric-mean ofper hemispheric-mean 70 yr, which annual-mean is2 annual-mean the period surface of surface2 CO tempera-2 doubling tempera- or halving Global radiative forcing2 (F )changesapproximatelylinearlywithtimeovertheCO2 rampings, by about 3.7 Wm per 70 yr, which is the[Myhre period et of al. CO,[ 1998].2Myhredoubling[Myhre The et al. et o or,↵ 1998]. al.set halving, in 1998]. Figure The The o↵ 1set between o↵ inset Figure in warming Figure 1 between 1 (red) between warming and warming cooling (red) (blue) (red) and and cooling cooling (blue) (blue) tureture anomalies anomalies ( T (NH TandNH and TSH ),TSH as), expected as expected from from deep deep ocean ocean heat heat storage storage [e.g., [e.g.,HeldHeld et et 2 [Myhrerampings, et al., 1998]. by about The 3 o.7↵set Wm in Figureper 70 1 yr, between which is warming the period (red) of andCO2 coolingdoubling (blue) or halving trajectoriesal., implies 2010].al.trajectories, 2010].trajectories In a lagged order In order implies to response approximately implies to approximately a lagged of a hemispheric-mean lagged response account response account of for hemispheric-mean this for of annual-mean hemispheric-mean this lag, lag, we consider we consider surface annual-mean the annual-mean evolution thetempera- evolution surface of surface ice of tempera- ice tempera- [Myhre et al., 1998]. The o↵set in Figure 1 between warming (red) and cooling (blue) trajectories implies a lagged response of hemispheric-meanture anomalies annual-meanareaarea asture ( a asT functionture anomaliesNH asurface functionand anomalies ofT tempera- hemisphericSH ( of ),T hemispheric asNH ( expectedandTNH and temperatureTSH temperature from),T asSH), expected deep as rather expected ocean rather from than heat than from deeptime. storage time. deep ocean A justification ocean [e.g.,A heat justificationHeld heat storage etfor storage this [e.g.,for this [e.g.,HeldHeld et et trajectories implies a lagged response of hemispheric-meanal., 2010]. In annual-mean orderal., 2010]. to approximately In surface order to tempera- approximately account for this account lag, we for consider this lag, the we evolution consider of the ice evolution of ice ture anomalies ( TNH and TSH), as expected from deep oceantreatment heattreatment storage isal. that, 2010].is [e.g., that annual-mean Inannual-meanHeld order et to Arctic approximately Arctic sea ice sea area ice accountarea has beenhas for been found this found to lag, decline to we decline consider linearly linearly the with evolution with of ice
area as aincreasing functionarea of global-mean as hemispheric a function temperature of temperature hemispheric across rather temperature a range than of time. rather GCMs, A justification than emissions time. scenarios, A for justification this and for this al., 2010].ture In anomalies order to ( approximately TNH and TSH account), as expected for this from lag, wedeep consider oceanincreasing heat thearea evolution storage global-mean as a function [e.g., of ice temperatureHeld of hemispheric et across temperature a range of GCMs,rather than emissions time. scenarios, A justification and for this treatmentclimates isclimates thattreatment [ annual-meanGregorytreatment [Gregory is et that al. is et, Arctic that 2002; annual-meanal., 2002; annual-meanRidley seaRidley ice et area Arctic al. et, Arctichas 2008; al. sea, been 2008;Winton ice sea found areaWinton ice, area has 2006, to, decline been2006, has 2008, been found 2008, linearly2011]. found 2011]. to Specifically, decline with to Specifically, decline linearly linearly with with area asal. a, 2010].function In of order hemispheric to approximately temperature account rather for than this lag, time. we A consider justification the evolution for this of ice increasingwe global-mean extendweincreasing extend theincreasing arguments the temperature global-mean arguments global-mean of Winton of across temperatureWinton temperature[2011], a range[2011], relating across of relating GCMs, across hemispheric a range hemispheric a emissions range of ice GCMs, of cover ice scenarios, GCMs, cover toemissions global to emissions globaland forcing scenarios, forcing scenarios, and and treatmentarea is as that a function annual-mean of hemispheric Arctic sea temperature ice area has rather been found than totime. decline A justification linearly with for this climatesthrough [Gregorythroughclimates etclimates al., [Gregory 2002; [GregoryRidley et al. et et, 2002; al. al.,, 2002; 2008;RidleyRidleyWinton et al. et,, 2008; 2006, al., 2008;Winton 2008,Winton 2011]., 2006,, Specifically, 2006, 2008, 2008, 2011]. 2011]. Specifically, Specifically, increasingtreatment global-mean is that annual-mean temperature Arcticacross sea a range ice area of GCMs, has been emissions found to scenarios, decline linearly and with we extend the argumentswe extendwe extend the of Winton arguments the arguments[2011], of Winton relating of Winton[2011], hemispheric[2011], relating relating ice hemispheric cover hemispheric to global ice cover forcing ice cover to global to global forcing forcing increasing global-mean temperature across a range of GCMs, emissions scenarios, and climates [Gregory et al., 2002; Ridley et al., 2008; Winton, 2006, 2008, 2011]. Specifically,dTu dTu cu c = T=u +TF++F" +(T" d (TTu)(1)T )(1) through throughthrough dtu dt u d u climates [Gregory et al., 2002; Ridley et al., 2008; Winton, 2006, 2008, 2011]. Specifically, we extend the arguments of Winton [2011], relating hemisphericOur ice cover approach to global forcing we extend the arguments of Winton [2011], relating hemispheric ice cover to global forcing through dTu dTdTd udTdTd u § Use a 2-layer ocean model (Held et al. 2010; Step 1:cu Draw =priors Tu of+c dc uF , c +d c ,=u " =, ( ( T= andTdTu= u( + T T T ufromuFdu)(1))(2)++T dfitsF" )(2) +(T" d (TTdu)(1)Tu)(1) Armour 2017) that includes the essential physics of 2-layerdt model to CMIP5dtdt modeldtdt response to through governing global-mean surface warming: CO2 forcing (Geoffroy et al. 2013)
dTu Step 2: Drive model with timeseries~⌧ ~⌧zˆ z ˆof ⌧historicalx ⌧ x cu = Tu + F + " (Td Tu)(1)radiative forcingdT (dMeinshausen !T = ⇥ dT et=d al.dT 2011)d yˆ , with (3) c =e ( !TTe =cdT⇥)(2)c == (=Tu (yˆTTd)(2)T )(2)(3) dt priors drawn fromd forcing rangeu⇢f ⇢ dinf dIPCC ⇢f AR5⇢f u d dTu dt dt dt cu = Tu + F + " (Td Tu)(1) dt
1 x1 ~⌧ ~⌧ x dTd x ~⌧ wzˆe =we =zˆ⌧ z~⌧ˆ z~⌧ˆ zˆ ⌧ ⌧ (4) (4) cd = (Tu Td)(2) !Te = ⇥ = !T⇢e =·⇢ !r⇥Tyˆe =⇥f ⇥=f = yˆ yˆ (3) (3) (3) dt ⇢f ⇢f ⇢· fr⇥⇢f ⇢f ⇢f dT c d = (T T )(2) d dt u d @w @w@v @v @u @u@w @w @v @v@u @u ~⌧ zˆ ⌧ x ~u =~u = 1 xˆ + xˆ +~⌧ 1 1yˆ + y~⌧ˆ + ~⌧ zˆ zˆ (5) (5) r⇥r⇥ @ywe @=y@z zˆ@z w@ze = @wze@zˆ =x @xzˆ @x @x@y @y (4) (4) (4) !Te = ⇥ = yˆ ✓ (3) ◆ ✓ ◆ ✓ ◆ ⇢f ⇢f ✓ ⇢ · r⇥◆ f ✓ ⇢ · r⇥⇢ ◆· r⇥f ✓f ◆ ~⌧ zˆ ⌧ x !Te = ⇥ = yˆ (3) ⇢f ⇢f DRAFTDRAFT December14,2017,5:39pm December14,2017,5:39pm DRAFT DRAFT @w @v @w@u@w@v@w@v @u@v@u@w@u@w @v @v@u @u 1 ~⌧ ~u = ~u =xˆ~u+= xˆ +yˆxˆ++ yˆ +zˆ yˆ + (5)zˆ zˆ (5) (5) w = zˆ r⇥ @y r⇥ @zr⇥(4) @y@ z@y @z @x@z @z@ x@z @x @y@x @x @x@y @y e ⇢ · r⇥f ✓ ◆ ✓ ✓✓ ◆ ◆◆✓ ✓✓ ◆ ◆◆✓ ✓ ◆ ◆ 1 ~⌧ w = zˆ (4) e ⇢ · r⇥f DRAFTDRAFTDRAFT December14,2017,5:39pm December14,2017,5:39pm December14,2017,5:39pm DRAFT DRAFT DRAFT @w @v @u @w @v @u ~u = xˆ + yˆ + zˆ (5) r⇥ @y @z @z @x @x @y ✓ @w◆ @v ✓ @u ◆ @w ✓ @v◆ @u ~u = xˆ + yˆ + zˆ (5) r⇥ @y @z @z @x @x @y ✓ ◆ ✓ ◆ ✓ ◆ DRAFT December14,2017,5:39pm DRAFT DRAFT December14,2017,5:39pm DRAFT ARMOURARMOUR ET AL.: ET AL.: SEA SEAICE REVERSIBILITY ICE REVERSIBILITY X-5X-5
GlobalGlobal radiative radiative forcing forcing (F )changesapproximatelylinearlywithtimeovertheCO (F )changesapproximatelylinearlywithtimeovertheCO2 2
ARMOUR ET2 AL.:ARMOUR2 SEAARMOUR ICE ET REVERSIBILITY AL.: ET AL.: SEA ICE SEA REVERSIBILITY ICE REVERSIBILITY X-5 X-5X-5 rampings,rampings, by about by about 3.7 Wm 3.7 Wm per 70per yr, 70 which yr, which is the is periodthe period of CO of2 COdoubling2 doubling or halving or halving ARMOUR ET AL.: SEA ICE REVERSIBILITY X-5 Global radiativeGlobal forcingGlobal radiative (F radiative)changesapproximatelylinearlywithtimeovertheCO forcing forcing (F )changesapproximatelylinearlywithtimeovertheCO (F )changesapproximatelylinearlywithtimeovertheCO2 [Myhre[Myhre et al. et, al.1998]., 1998]. The The o↵set o↵ inset Figure in Figure 1 between 1 between warming warming (red) (red) and and cooling cooling2 (blue) (blue) 2 ARMOUR ET AL.: SEA ICE REVERSIBILITY X-5 Global radiative forcing (F )changesapproximatelylinearlywithtimeovertheCO2 2 2 2 rampings, by aboutrampings, 3.7 Wm by aboutper 3 70.7 yr, Wm which per is 70 the yr, period which of is CO the perioddoubling of orCO halvingdoubling or halving trajectoriestrajectoriesrampings, implies implies a lagged by a about lagged response 3 response.7 Wm of hemispheric-mean ofper hemispheric-mean 70 yr, which annual-mean is2 annual-mean the period surface of surface2 CO tempera-2 doubling tempera- or halving Global radiative forcing2 (F )changesapproximatelylinearlywithtimeovertheCO2 rampings, by about 3.7 Wm per 70 yr, which is the[Myhre period et of al. CO,[ 1998].2Myhredoubling[Myhre The et al. et o or,↵ 1998]. al.set halving, in 1998]. Figure The The o↵ 1set between o↵ inset Figure in warming Figure 1 between 1 (red) between warming and warming cooling (red) (blue) (red) and and cooling cooling (blue) (blue) tureture anomalies anomalies ( T (NH TandNH and TSH ),TSH as), expected as expected from from deep deep ocean ocean heat heat storage storage [e.g., [e.g.,HeldHeld et et 2 [Myhrerampings, et al., 1998]. by about The 3 o.7↵set Wm in Figureper 70 1 yr, between which is warming the period (red) of andCO2 coolingdoubling (blue) or halving trajectoriesal., implies 2010].al.trajectories, 2010].trajectories In a lagged order In order implies to response approximately implies to approximately a lagged of a hemispheric-mean lagged response account response account of for hemispheric-mean this for of annual-mean hemispheric-mean this lag, lag, we consider we consider surface annual-mean the annual-mean evolution thetempera- evolution surface of surface ice of tempera- ice tempera- [Myhre et al., 1998]. The o↵set in Figure 1 between warming (red) and cooling (blue) trajectories implies a lagged response of hemispheric-meanture anomalies annual-meanareaarea asture ( a asT functionture anomaliesNH asurface functionand anomalies ofT tempera- hemisphericSH ( of ),T hemispheric asNH ( expectedandTNH and temperatureTSH temperature from),T asSH), expected deep as rather expected ocean rather from than heat than from deeptime. storage time. deep ocean A justification ocean [e.g.,A heat justificationHeld heat storage etfor storage this [e.g.,for this [e.g.,HeldHeld et et trajectories implies a lagged response of hemispheric-meanal., 2010]. In annual-mean orderal., 2010]. to approximately In surface order to tempera- approximately account for this account lag, we for consider this lag, the we evolution consider of the ice evolution of ice ture anomalies ( TNH and TSH), as expected from deep oceantreatment heattreatment storage isal. that, 2010].is [e.g., that annual-mean Inannual-meanHeld order et to Arctic approximately Arctic sea ice sea area ice accountarea has beenhas for been found this found to lag, decline to we decline consider linearly linearly the with evolution with of ice
area as aincreasing functionarea of global-mean as hemispheric a function temperature of temperature hemispheric across rather temperature a range than of time. rather GCMs, A justification than emissions time. scenarios, A for justification this and for this al., 2010].ture In anomalies order to ( approximately TNH and TSH account), as expected for this from lag, wedeep consider oceanincreasing heat thearea evolution storage global-mean as a function [e.g., of ice temperatureHeld of hemispheric et across temperature a range of GCMs,rather than emissions time. scenarios, A justification and for this treatmentclimates isclimates thattreatment [ annual-meanGregorytreatment [Gregory is et that al. is et, Arctic that 2002; annual-meanal., 2002; annual-meanRidley seaRidley ice et area Arctic al. et, Arctichas 2008; al. sea, been 2008;Winton ice sea found areaWinton ice, area has 2006, to, decline been2006, has 2008, been found 2008, linearly2011]. found 2011]. to Specifically, decline with to Specifically, decline linearly linearly with with area asal. a, 2010].function In of order hemispheric to approximately temperature account rather for than this lag, time. we A consider justification the evolution for this of ice increasingwe global-mean extendweincreasing extend theincreasing arguments the temperature global-mean arguments global-mean of Winton of across temperatureWinton temperature[2011], a range[2011], relating across of relating GCMs, across hemispheric a range hemispheric a emissions range of ice GCMs, of cover ice scenarios, GCMs, cover toemissions global to emissions globaland forcing scenarios, forcing scenarios, and and treatmentarea is as that a function annual-mean of hemispheric Arctic sea temperature ice area has rather been found than totime. decline A justification linearly with for this climatesthrough [Gregorythroughclimates etclimates al., [Gregory 2002; [GregoryRidley et al. et et, 2002; al. al.,, 2002; 2008;RidleyRidleyWinton et al. et,, 2008; 2006, al., 2008;Winton 2008,Winton 2011]., 2006,, Specifically, 2006, 2008, 2008, 2011]. 2011]. Specifically, Specifically, increasingtreatment global-mean is that annual-mean temperature Arcticacross sea a range ice area of GCMs, has been emissions found to scenarios, decline linearly and with we extend the argumentswe extendwe extend the of Winton arguments the arguments[2011], of Winton relating of Winton[2011], hemispheric[2011], relating relating ice hemispheric cover hemispheric to global ice cover forcing ice cover to global to global forcing forcing increasing global-mean temperature across a range of GCMs, emissions scenarios, and climates [Gregory et al., 2002; Ridley et al., 2008; Winton, 2006, 2008, 2011]. Specifically,dTu dTu cu c = T=u +TF++F" +(T" d (TTu)(1)T )(1) through throughthrough dtu dt u d u climates [Gregory et al., 2002; Ridley et al., 2008; Winton, 2006, 2008, 2011]. Specifically, we extend the arguments of Winton [2011], relating hemisphericOur ice cover approach to global forcing we extend the arguments of Winton [2011], relating hemispheric ice cover to global forcing through dTu dTdTd udTdTd u § Use a 2-layer ocean model (Held et al. 2010; Step 1:cu Draw =priors Tu of+c dc uF , c +d c ,=u " =, ( ( T= andTdTu= u( + T T T ufromuFdu)(1))(2)++T dfitsF" )(2) +(T" d (TTdu)(1)Tu)(1) Armour 2017) that includes the essential physics of 2-layerdt model to CMIP5dtdt modeldtdt response to through governing global-mean surface warming: CO2 forcing (Geoffroy et al. 2013) X - 12 ARMOUR ET AL.: SEA ICE REVERSIBILITY X - 12 ARMOUR ET AL.: SEA ICE REVERSIBILITY dTu Step 2: Drive model with timeseries~⌧ ~⌧zˆ z ˆof ⌧historicalx ⌧ x cu = Tu + F + " (Td Tu)(1)radiative forcingdT (dMeinshausen !T = ⇥ dT et=d al.dT 2011)d yˆ , with (3) X - 12 ARMOUR ET AL.: SEA ICE REVERSIBILITYc =e ( !TTe =cdT⇥)(2)c == (=Tu (yˆTTd)(2)T )(2)(3) dt priors drawn fromd forcing rangeu⇢f ⇢ dinf dIPCC ⇢f AR5⇢f u d dTu dt dt dt Tobs cu = Tu + F + " ECS(Td = TRu2)(1)/ obs = R2 Tobs (63) dt ECS = R2 /⇥ obs = R2 Step⇥ R obs3: UseH Bayesianobs inference to estimate(63) ⇥ ⇥ Robs Hobs posterior parameters/Tobs forcings based on 1 x1 ~⌧ ~⌧ x dTd ECS = R2 / obsobserved= R2 warming~⌧ andzˆ energy⌧ budget~⌧ z~⌧ˆ (seezˆ (63)⌧ ⌧ x ⇥ ⇥ R H we =we =zˆ zˆ (4) (4) cd = (Tu Td)(2)F2 also: Forestobs et !Te al.= 2002,obs⇥ 2006;= !T⇢ Stotte =·⇢ !r⇥Ty ˆ&e ·=Forest⇥r⇥f ⇥= f2007)= yˆ yˆ (3) (3) (3) dt ECS = F2 ⇥ ⇢f ⇢f ⇢f ⇢f ⇢f ⇢f dT ECS = ⇥ d (64) cd = (Tu Td)(2) F2 TobsF2 = 0.75 ± 0.2 °C (64) dt =ECSF2 =T⇥obs ⇥ = F ⇥ Q @w @w@v @v (Otto@ etu al.@ 2013;u@w @w @v @v@u @u F obs Q obs -2 ~⌧ zˆ ⌧ x obs obs =~u 0.65=~u ±= 0.27 Wm1 xˆ +2000-2009xˆ +~⌧ relative1 1 toyˆ +(64)y~⌧ˆ + ~⌧ zˆ zˆ (5) (5) r⇥F2r⇥Tobs @ywe @=y@z zˆ@z w@ze = @wze@zˆ =x @xzˆ @x @x@y @y (4) (4) (4) !Te = ⇥ = yˆ = ⇥ ✓ (3) ◆ 1860-1879)✓ ◆ ✓ ◆ ✓ -2 ⇢ · r⇥◆ f ✓ ⇢ · r⇥⇢ ◆· r⇥f ✓f ◆ ⇢f ⇢f x Fobs2 = 2.3Qobs ± 1 Wm ~⌧ zˆ ⌧ R 4 Wm2 (65) !T = ⇥ = yˆ R ⇡4 Wm (3) (65) e ⇢f ⇢f ⇡ DRAFTDRAFT2 December14,2017,5:39pm December14,2017,5:39pm DRAFT DRAFT R 4 Wm@w @v @w@u@w@v@w@v @u@v@(65)u@w@u@w @v @v@u @u 1 ~⌧ ~u⇡= ~u =xˆ~u+= xˆ +yˆxˆ++ yˆ +zˆ yˆ + (5)zˆ zˆ (5) (5) r⇥ @y r⇥ @zr⇥ @y@ z@y @z @x@z @z@ x@z @x @y@x @x @x@y @y we = zˆ ✓ 2 (4)◆ ✓ ✓ ◆ ◆ ✓ ✓ ◆ ◆ ✓ ◆ ⇢ · r⇥f H 0.7 Wm2 ✓ ◆ (66)✓ ◆ ✓ ◆ 1 ~⌧ H ⇡0.7 Wm (66) w = zˆ ⇡ (4) e 2 ⇢ · r⇥f H 0.7 Wm (66) DRAFTDRAFTDRAFT⇡ December14,2017,5:39pm December14,2017,5:39pm December14,2017,5:39pm DRAFT DRAFT DRAFT @w @v @u @w @v @u ~u = xˆ + yˆ + T 0.8 C=0zˆ .8K(5) (67) r⇥ @y @z @z @x @xT ⇡[email protected]