The Pattern Effect: Sea Surface Temperature modulation of radiative - - PowerPoint PPT Presentation

“The Pattern Effect”: Sea Surface Temperature modulation

• f radiative damping and climate sensitivity

Cristian Proistosescu

JISAO, University of Washington

Equilibrium Climate Sensitivity: First estimate

“doubling of the percentage of CO2 in the air would

raise the temperature of the earth's surface by 5° “

Svante Arrhenius (1896) “doubling of the percentage of CO2 in the air would

raise the temperature of the earth's surface by 5° “

Equilibrium Climate Sensitivity: First estimate

How have we been doing?

Charney Report likely (1 )

σ

ECS (oC)

Uncertainty has been recalcitrant

Charney Report likely (1 )

σ

ECS (oC)

Some small measure of progress

Charney Report likely (1 )

σ

ECS (oC)

Or not…

Charney Report likely (1 )

σ

ECS (oC)

ECS (oC)

Charney Report likely (1 )

σ

Numerical models (GCMs)

Hypotheses:

• models are too sensitive
• inadequate treatment of observations
• they are not measuring the same process

“No best estimate for ECS is given because of a lack of agreement across lines of evidence“ IPCC 2014

What are they measuring?

• Models are run to equilibrium
• Instrumental: We make an inference from

present day energy budget

Lines of evidence disagree

Changes in Energy Content

Energy Input (radiative forcing) Radiative damping

Estimates from the instrumental record Consider the energy budget at equilibrium

ΔT2×CO2

ΔR

:black body radiation + atmospheric feedbacks (clouds, water-vapor, lapse rate, surface albedo)

ΔR2×CO2 ΔF2×CO2

ECS = ΔF2×CO2 λ

Estimates from the instrumental record Key physics: radiative feedback & Climate Sensitivity

Radiative feedback (efficiency of radiative damping)

λ = ΔR ΔT = ΔF2×CO2 ΔT2×CO2

Energy Input (radiative forcing) Radiative damping

ΔR2×CO2 = ΔF2×CO2

Energy balance

ΔT2×CO2 ΔR2×CO2 ΔF2×CO2

Energy balance Radiative feedback (inferred) (efficiency of radiative damping) Energy Input (radiative forcing) Radiative damping

ΔQ

Ocean heat uptake

ΔF ΔR ΔT

Estimates from the instrumental record Inferred Climate Sensitivity

λ = ΔR ΔT = ΔF − ΔQ ΔT ΔR = ΔF − ΔQ

Inferred ECS:

ECS = ΔF2×CO2 ΔT ΔF − ΔQ

Energy Input (radiative forcing) Radiative damping

ΔF ΔR ΔT

ΔQ

Ocean heat uptake

Estimates from the instrumental record Present day energy budget

ΔT = 0.91 ± 0.1o C

2000:2017 - 1850:1870

ΔT

HadCRUT4

• C

Inferred ECS:

ECS = ΔF2×CO2 ΔT ΔF − ΔQ

Energy Input (radiative forcing) Radiative damping

ΔQ

Ocean heat uptake

ΔF ΔT ΔR

ΔQ = 0.61 ± 0.1 W/m2

Estimates from the instrumental record Present day energy budget

ΔT = 0.91 ± 0.1o C

Inferred ECS:

ECS = ΔF2×CO2 ΔT ΔF − ΔQ

Energy Input (radiative forcing) Radiative damping

ΔF ΔR ΔT

Anthropogenic Aerosols

Natural GHGs Total Anthrop.

ΔQ

Ocean heat uptake

ΔF = 2.33 ± 0.8 W/m2 ΔQ = 0.61 ± 0.1 W/m2 ΔF 2011-1850

(IPCC AR5)

W/m2

Estimates from the instrumental record Present day energy budget

ΔT = 0.91 ± 0.1o C ΔF2×CO2 = 3.7 W/m2

Inferred ECS:

ECS = ΔF2×CO2 ΔT ΔF − ΔQ

Count / PDF Count / PDF Count / PDF

ECS (oC)

ΔF2×CO2 = 3.7 W/m2 ΔF = 2.33 ± 0.8 W/m2 ΔQ = 0.61 ± 0.1 W/m2

Estimates from the instrumental record Inferred Climate Sensitivity

ΔT = 0.91 ± 0.1o C

Inferred ECS:

ECS = ΔF2×CO2 ΔT ΔF − ΔQ

Count / PDF

Estimated from Transient Estimated at Equilibrium

Equilibrium Climate Sensitivity: Apples to oranges

Observations Models

ECS (oC)

Equilibrium Climate Sensitivity: Time-dependent climate-sensitivity in models: tied to SSTs

True ECS Inferred from transient

CMIP5 abrupt4xCO2 mean from Proistosescu & Huybers 2017

Time since CO2 quadrupling (years)

ECS = F2×CO2 λ(t) 3

• 3

Murphy 1995, Senior and Mitchell 2000, Winton et al 2009, Held et al 2010, Armour et al 2013, Andrews et al 2015

Model: year 150 Model: year 20

CCSM4

50 100 150 1 2 4 3 ECS (oC) ΔSST (oC)

Equilibrium Climate Sensitivity: State of the knowledge

True ECS Inferred from transient

CMIP5 abrupt4xCO2 mean from Proistosescu & Huybers 2017

Time since CO2 quadrupling (years)

ECS = F2×CO2 λ(t)

50 100 150 1 2 4 3 ECS (oC)

• ECS determined by the net radiative feedback
• Disagreement between numerical models and
• bservations of the Earth’s energy budget
• Radiative feedback has time-dependency tied to

SST changes Questions:

• Can SST-driven changes in feedback account

for the discrepancy?

• What is the underlying physics?
• How do we empirically constrain radiative

feedbacks?

Count / PDF Count / PDF

CAM 4 simulations

Comparing apples to apples SSTs are different

• We can’t get a true equilibrium from obs
• But we can get a transient from the models

Observations Model (coupled)

1.5

• 1.5

3

HadSST

Observations: 2000-1850

CCSM4

• 3

Model: year 150 ECS (oC) ΔSST (oC) ΔSST (oC)

Count / PDF Count / PDF

Model w/ SSTobs Model (coupled) Observations

1.5

• 1.5

3

HadSST

Observations: 2000-1850

• 3

Model: year 150

CCSM4

CAM 4 simulations

ΔSST (oC) ΔSST (oC)

Proistosescu & Huybers 2017

ECS (oC)

Takeaway: Low sensitivity now is consistent with high ECS in the future

• SST-driven changes in feedback explains discrepancy
• Models are consistent with observations
• Models are not too sensitive. Sensitivity increase.
• Delayed warming in regions of heat uptake

Proistosescu & Huybers 2017 in prep: w/ Kyle Armour, Malte Stuecker, Yue Dong, Tim Andrews, Jonathan Gregory, Thorsten Mauritsen, Levi Silvers & David Paynter

Count / PDF

T2×CO2 RGCM(Tobs) RGCM(TGCM(∞)) Robs(Tobs)

Count / PDF

CAM4, CAM5, HadGEM2, HadAM3, ECHAM6, AM2.1, AM3, AM4 Model w/ SSTobs Models (coupled) Observations

1.5

• 1.5

3

HadSST

Observations: 2000-1850

• 3

Model: year 150

CCSM4

ECS (oC) ΔSST (oC) ΔSST (oC)

Takeaway: Low sensitivity now is consistent with high ECS in the future

• SST-driven changes in feedback explains discrepancy
• Models are consistent with observations
• Models are not too sensitive. Sensitivity increase.
• Delayed warming in regions of heat uptake

ΔR = ΔR (ΔT(x, t)) ΔR = ∂R ∂T(x)ΔT(x, t) + 𝒫(ΔT2)

Feedback depends on spatial and temporal changes in temperature

λ = ΔR ΔT

Key physics: radiative feedback

How does radiation depend on the pattern of warming? Key physics: radiative feedback

Proistosescu & Huybers 2017 Theory and modal decomposition Approximations:

• Radiation depends only on spatial pattern of SSTs
• System is linear

R(y, t) ≈ ∂R(y) ∂T(x) ∑ ψn(x)ϕn(t)

EOFs eigenvectors of state space Principle components Exponential decay

ΔR = ΔR (ΔT(x, t)) ΔR = ∂R ∂T(x)ΔT(x, t) + 𝒫(ΔT2)

Feedback depends on spatial and temporal changes in temperature Approximations:

• Radiation depends only on spatial pattern of SSTs
• System is linear

λ = ΔR ΔT

Key physics: radiative feedback

How does radiation depend on the pattern of warming? Key physics: radiative feedback

Dong, Proistosescu, Armour, Battisti, in review General question: how does radiation depend on local warming

R(y, t) ≈ ∂R(y) ∂T(x) ∑ ψn(x)ϕn(t)

EOFs eigenvectors of state space Principle components Exponential decay

How does radiation depend on the pattern of warming? Compute response to local warming

Warm SSTs in a single patch. Keep SST and SIC fixed everywhere else

Dong, Proistosescu, Armour, Battisti, in review Zhou, Zelinka, Klein, 2017

137 fixedSST runs on CAM4

∂R ∂T(x)

How does radiation depend on the pattern of warming? Radiative response to local warming

Dong, Proistosescu, Armour, Battisti, in review

Global radiation response to

• 1oC of warming in West Pacific: +30 W/m2
• 1oC of warming in East Pacific: -10 W/m2

Outgoing Radiation

W/m2/oC

∂R ∂T(x)

Outgoing Radiation — Full model — Reconstruction 1950 1980 2010

• 1

1 2

∂R ∂T(x) ⋆ Tobs(x, t)

ΔR (W/m2)

How does radiation depend on the pattern of warming? Radiative response to local warming (sanity check) ∂R ∂T(x)

W/m2/oC

How does radiation depend on the pattern of warming? Radiative response to local warming tied to SST climatology

Outgoing Radiation

W/m2/oC

Global radiation response to

• 1K of warming in West Pacific: 30 W/m2
• 1K of warming in East Pacific: 10 W/m2

SST climatology

• C

∂R ∂T(x)

Warm Pool Subtropics Klein and Hartman 1993, Wood & Bretherton 2006 , Bretherton & Blossey 2014

Low cloud amount depends

• n strength of inversion and

local temperature

W/m2/oC

• C

Controlled by East Pacific SST +EP SST - LCC

Warm Pool Subtropics

+SST

W/m2/oC

• C

+T +T

Controlled by West Pacific SST +WP SST + LCC

Warm Pool Subtropics

+SST

W/m2/oC

• C

Near-Surface air temperature change to single patch of SST warming Fixed: Sea Surface Temp and Sea Ice Concentration Temperature changes over land and Sea Ice

How does radiation depend on the pattern of warming?: Local warming

Dong, Proistosescu, Armour, Battisti, in review

Near-Surface air temperature change to single patch of SST warming

How does radiation depend on the pattern of warming? Vertical structure of temperature

Zonal view: SST in ascent regions control upper troposphere temperature

Dong, Proistosescu, Armour, Battisti, in review

Dong, Proistosescu, Armour, Battisti, in review

How does radiation depend on the pattern of warming? Local & non-local cloud response

Near-Surface air temperature change to single patch of SST warming Warming in West Pacific: global increase in low clouds Warming in East Pacific: local decrease in low clouds

Takeaway: Atmospheric radiation interacts with evolving ocean temperatures

Weak feedback over historical period is consistent with strong feedback (high ECS) in the future,

• Atmospheric radiation interacts w/ evolving
• cean temperatures
• Historical warming: focused on West

Pacific, leading to increases in low clouds (+ outgoing radiation)

• Future warming: regions of ocean upwelling

(- outgoing radiation)

λ(t) = ΔR ΔT

ECS = F2×CO2 λ(t)

Proistosescu & Huybers 2017 Dong, Proistosescu, Armour, Battisti, in review WCRP report, IPCC AR6

RGCM(Tobs) RGCM(TGCM(∞))

Count / PDF ECS (oC)

1.5

• 1.5

Observations

∂R ∂T(x)

Model: year 150 Observations: 2000-1850

• 20

20 Model w/ SSTobs Models (coupled)

W/m2/oC

3 ΔSST (oC)

• 3

ΔSST (oC)

Takeaway: Atmospheric radiation interacts with evolving ocean temperatures

RGCM(Tobs) RGCM(TGCM(∞))

Count / PDF ECS (oC)

Observations

∂R ∂T(x)

• 20

20 Model w/ SSTobs Models (coupled)

W/m2/oC

• Atmospheric radiation interacts w/ evolving
• cean temperatures
• Historical warming: focused on West

Pacific, leading to increases in low clouds (+ outgoing radiation)

• Future warming: regions of ocean upwelling

(- outgoing radiation)

λ(t) = ΔR ΔT

ECS = F2×CO2 λ(t)

Proistosescu & Huybers 2017 Dong, Proistosescu, Armour, Battisti, in review WCRP report, IPCC AR6

Historical Warming / Equilibrium Warming

0.3 0.6

Weak feedback over historical period is consistent with strong feedback (high ECS) in the future,