a hector application: sea-level constraints tighten ......a hector application: sea-level...
TRANSCRIPT
A Hector application: Sea-level constraints tighten climate
sensitivity and temperature projections
Ben Vega-Westhoff1, Ryan Sriver1, Corinne Hartin2, Tony Wong3, Klaus Keller4
1University of Illinois, 2Joint Global Change Research Institute, 3University of Colorado, 4Pennsylvania State University
Jeremy Harbeck/NASA
Minnesota Dept. of Agriculture
Climate impacts/damages closely linked to extreme (low-probability) events
Uncertainty quantification, including the tails, is critical for climate projections
2/13
Pinning the tails on Hector• Updated Hector energy balance (DOECLIM, part of
Hector v2.0)• Enhanced sea-level module available (BRICK,
https://github.com/bvegawe/hector/tree/dev_slr)• Bayesian (MCMC) calibration tools available
(https://github.com/bvegawe/hector_probabilistic)• Default: calibrate 39 parameters
1960 2000 2040 2080
−10
12
34
5T
[K]
ObservationsRCP2.6RCP4.5RCP8.5
0 2 4 6 8 10
0.0
0.4
0.8
Climate sensitivity [K]
Dens
ity
0 1 2 3 4
0.05
0.20
0.35
Ocean heat diffusivity [cm2/s]
Dens
ity
0.0 0.5 1.0 1.5 2.0
0.0
1.0
2.0
Aerosol scaling
Dens
ity
3/13
, 95% CI shaded
PriorPosterior
Sanity check #1: Do calibrated results fit the observations?
1880 1900 1920 1940 1960 1980 2000
−0.4
−0.2
0.0
0.2
0.4
0.6
0.8
1.0
T [K
]
HadCRUT4.4HectorMAGICC6
• Simple model -> can’t capture short-term variability (but accounted for in our probabilistic assessment)
• Similar RMSE to MAGICC6 (0.118K for Hector, 0.119 for MAGICC)• GISTEMP instead of HadCRUT -> same result
Sanity check #2: A perfect model experiment
• Create 2 sets of simulated observations with different climate sensitivities:– 1.5 deg C– 4.5 deg C
• Can the calibration tool distinguish between the two?
(equilibrium ΔT due to a doubling of CO2)
0 2 4 6 8 100.0
0.2
0.4
0.6
Climate sensitivity [K]
a
Dens
ity
CS = 1.5CS = 4.5
0 1 2 3 40.0
0.1
0.2
0.3
Ocean heat diffusivity [cm2/s]
b
0.0 0.5 1.0 1.5 2.00.0
0.5
1.0
1.5
2.0
Aerosol scaling
c
1880 1900 1920 1940 1960 1980 2000−0.5
0.0
0.5
1.0
1.5
T [K
]
aSimulated obs, CS = 1.5Simulated obs, CS = 4.5
1960 1970 1980 1990
−40
−20
0
20
40
60
H [1
0^22
J]
b
1880 1900 1920 1940 1960 1980 2000
−0.2
−0.1
0.0
0.1
SLR
[m]
c
1880 1900 1920 1940 1960 1980 2000−0.5
0.0
0.5
1.0
1.5
T [K
]
aSimulated obs, CS = 1.5Simulated obs, CS = 4.595% CI, CS = 1.595% CI, CS = 4.5
1960 1970 1980 1990
−40
−20
0
20
40
60
H [1
0^22
J]
b
1880 1900 1920 1940 1960 1980 2000
−0.2
−0.1
0.0
0.1
SLR
[m]
c
0 2 4 6 8 100.0
0.2
0.4
0.6
Climate sensitivity [K]
a
Dens
ity
CS = 1.5CS = 4.5
0 1 2 3 40.0
0.1
0.2
0.3
Ocean heat diffusivity [cm2/s]
b
0.0 0.5 1.0 1.5 2.00.0
0.5
1.0
1.5
2.0
Aerosol scaling
c
Scientific application of the Hector calibration tool
• Past studies investigated:• How does longer temperature data affect climate
sensitivity estimates and temperature projections (Urban et al., 2014; Shiogoma et al., 2016)
• What about adding in ocean heat observations (Urban and Keller, 2009)?
Urban et al., 2014
Urban and Keller, 2009
Just temperature
Just ocean heat
Both constraints
Our question:How do sea-level constraints affect climate sensitivity estimates and key climate projections?
Our approach:Two Hector calibrations
Scientific application of the Hector calibration tool
Calibration #1: “Without SLR”
• Only calibrate energy balance parameters (3 physical and 6 statistical)
• Ignore sea level
Calibration #2: “With SLR”
• Calibrate energy balance and BRICK sea-level parameters (39 total)
• Include sea-level contributor constraints
Sea-level constraints sharpen the climate sensitivity estimate
0 2 4 6 8 100.0
0.2
0.4
Climate sensitivity [K]
a
Dens
ity
With SLRWithout SLR
0 1 2 3 40.0
0.1
0.2
0.3
Ocean heat diffusivity [cm2/s]
b
0.0 0.5 1.0 1.5 2.00.0
0.5
1.0
1.5
2.0
Aerosol scaling
c
Especially in the tails
Sea-level constraints have little effect on hindcasts (besides sea-level)
1880 1900 1920 1940 1960 1980 2000
−0.5
0.0
0.5
1.0
T [K
]
a2−sigma range, observations95% CI, With SLRWithout SLR
1960 1970 1980 1990
−40
−20
0
20
40
H [1
0^22
J]
b
1880 1900 1920 1940 1960 1980 2000
−0.2
−0.1
0.0
0.1
SLR
[m]
c
20 40 60 80 100
20
40
60
80
100
Forecast probability [%] (CI)
a Temperature
Obs
erve
d re
lativ
e fre
quen
cy [%
] With SLRWithout SLR
overco
nfide
nt
unde
rconfi
dent
20 40 60 80 100
20
40
60
80
100
Forecast probability [%] (CI)
b Ocean heat
overco
nfide
nt
unde
rconfi
dent
20 40 60 80 100
20
40
60
80
100
Forecast probability [%] (CI)
c SLR
overco
nfide
nt
unde
rconfi
dent
20 40 60 80 100
20
40
60
80
100
Forecast probability [%] (CI)
a Temperature
Obs
erve
d re
lativ
e fre
quen
cy [%
] With SLRWithout SLR
overco
nfide
nt
unde
rconfi
dent
20 40 60 80 100
20
40
60
80
100
Forecast probability [%] (CI)
b Ocean heat
overco
nfide
nt
unde
rconfi
dent
20 40 60 80 100
20
40
60
80
100
Forecast probability [%] (CI)
c SLR
overco
nfide
nt
unde
rconfi
dent
20 40 60 80 100
20
40
60
80
100
Forecast probability [%] (CI)
a Temperature
Obs
erve
d re
lativ
e fre
quen
cy [%
] With SLRWithout SLR
overco
nfide
nt
unde
rconfi
dent
20 40 60 80 100
20
40
60
80
100
Forecast probability [%] (CI)
b Ocean heat
overco
nfide
nt
unde
rconfi
dent
20 40 60 80 100
20
40
60
80
100
Forecast probability [%] (CI)
c SLR
overco
nfide
nt
unde
rconfi
dent
Overconfidence in sea level is bad!(e.g. underestimate flood risks, Sriver et al., 2018)
Sea-level constraints sharpen projections
1960 2000 2040 2080
0
1
2
3
4
5
6T
[K]
aObservationsRCP8.5 With SLRWithout SLR
1960 2000 2040 2080
0.0
0.5
1.0
1.5
2.0
2.5
SLR
[m]
bObservationsRCP8.5 With SLRWithout SLR
2 4 6 8 10Clim. sens. [K]
−3
−2
−1
0a
With SLRWithout SLR
log(
1−C
DF)
0 1 2 3 4Diffusivity [cm2/s]
−3
−2
−1
0b
0.0 0.4 0.8 1.2Aerosol scaling
−3
−2
−1
0c
2 3 4 5 62100 T anom. [K]
−3
−2
−1
0d
log(
1−C
DF)
0 100 200 300 400 5002100 H [10^22 J]
−3
−2
−1
0e
0.5 1.5 2.5 3.52100 SLR [m]
−3
−2
−1
0f
2 4 6 8 10Clim. sens. [K]
−3
−2
−1
0a
With SLRWithout SLR
log(
1−C
DF)
0 1 2 3 4Diffusivity [cm2/s]
−3
−2
−1
0b
0.0 0.4 0.8 1.2Aerosol scaling
−3
−2
−1
0c
2 3 4 5 62100 T anom. [K]
−3
−2
−1
0d
log(
1−C
DF)
0 100 200 300 400 5002100 H [10^22 J]
−3
−2
−1
0e
0.5 1.5 2.5 3.52100 SLR [m]
−3
−2
−1
0f
CMIP5 range
IPCC range
A Hector application: Sea-level constraints tighten climate
sensitivity and temperature projections
• Results robust to observational temperature data set and observational time range
• Just submitted (Sunday) to Environmental Research Letters
1960 2000 2040 2080
0
1
2
3
4
5
6
T [K
]
aObservationsRCP8.5 GISTEMP, with SLRWithout SLR
1960 2000 2040 2080
0.0
0.5
1.0
1.5
2.0
2.5
3.0
SLR
[m]
bObservationsRCP8.5 GISTEMP, with SLRWithout SLR
0 2 4 6 8 100.0
0.2
0.4
0.6
Climate sensitivity [K]
a
Dens
ity
GISTEMP, with SLRWithout SLR
0 1 2 3 40.0
0.1
0.2
0.3
Ocean heat diffusivity [cm2/s]
b
0.0 0.5 1.0 1.5 2.00.0
0.5
1.0
1.5
2.0
Aerosol scaling
c
Future work
• Expand calibration to carbon cycle parameters• Investigate calibration sensitivity to length of
observations (what is the role of the hiatus)• Apply probabilistic projections to downscaling
applications (e.g. future tropical cyclone properties)
Acknowledgements
• Elmar Kriegler - DOECLIM model • Gregory Garner - C++ implementation of
DOECLIM • Nathan Urban, Skip Wishbone, Frank Erickson,
and Irene Schaperdoth - invaluable inputs • Co-supported by – DOE Office of Science, as part of research in Multi-
Sector Dynamics, Earth and Environmental System Modeling Program
– Penn State Center for Climate Risk Management
Summary• Bayesian (MCMC) calibration tools available for Hector
(https://github.com/bvegawe/hector_probabilistic)• Sanity checks performed, including perfect model
experiments• A first application: Sea-level constraints tighten climate
sensitivity and temperature projections
0 2 4 6 8 100.0
0.2
0.4
Climate sensitivity [K]
a
Dens
ity
With SLRWithout SLR
0 1 2 3 40.0
0.1
0.2
0.3
Ocean heat diffusivity [cm2/s]
b
0.0 0.5 1.0 1.5 2.00.0
0.5
1.0
1.5
2.0
Aerosol scaling
c
1960 2000 2040 2080
0
1
2
3
4
5
6
T [K
]
aObservationsRCP8.5 With SLRWithout SLR
1960 2000 2040 2080
0.0
0.5
1.0
1.5
2.0
2.5
SLR
[m]
bObservationsRCP8.5 With SLRWithout SLR
References
• Shiogama, H. et al. Predicting future uncertainty constraints on global warming projections. Sci. Rep. 6, 18903 (2016).
• Sriver, R. L., Lempert, R. J., Wikman-Svahn, P., & Keller, K. Characterizing uncertain sea level rise projections to support investment decisions. PLoS One, 13(2), e0190641 (2018). doi:10.1371/journal.pone.0190641.
• Urban, N. M. & Keller, K. Complementary observational constraints on climate sensitivity. Geophys. Res. Lett. 36, L04708 (2009).
• Urban, N. M., Holden, P. B., Edwards, N. R., Sriver, R. L. & Keller, K. Historical and future learning about climate sensitivity. Geophys. Res. Lett. 41, 2543–2552 (2014).
• Vega-Westhoff, B., Sriver, R., Hartin, C. A., Wong, T., & Keller, K. Impacts of observational sea-level change constraints on estimates of climate sensitivity (Submitted to Env. Res. Lett., 2018).
Supplemental: climate parameter correlations
Clim. sens. [K]
0 1 2 3 4
0.61
12
34
56
78
0.780
12
34 ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Diffusivity [cm2/s]
0.30
1 2 3 4 5 6 7 8
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
● ●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
● ●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
● ●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●● ●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
● ●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●●
●
● ●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
● ●
●
●
●
● ●
●
●●
●
●
●
●
●
●
●●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
Aerosol scaling
Supplemental: partial autocorrelations
0 10 20 30 40 50 600.0
0.2
0.4
0.6
0.8
1.0 n = 3.2e6
MCMC posteriora
Clim
ate
sens
itivi
ty P
ACF
0 10 20 30 40 500.0
0.2
0.4
0.6
0.8
1.0 n = 2e5
Rejection samplesb
0 10 20 30 400.0
0.2
0.4
0.6
0.8
1.0 n = 1e4
Final samplesc
0 10 20 30 40 50 600.0
0.2
0.4
0.6
0.8
1.0
d
Diff
usiv
ity P
ACF
0 10 20 30 40 500.0
0.2
0.4
0.6
0.8
1.0
e
0 10 20 30 400.0
0.2
0.4
0.6
0.8
1.0
f
0 10 20 30 40 50 600.0
0.2
0.4
0.6
0.8
1.0
Lag
g
Aero
sol s
calin
g PA
CF
0 10 20 30 40 500.0
0.2
0.4
0.6
0.8
1.0
Lag
h
0 10 20 30 400.0
0.2
0.4
0.6
0.8
1.0
Lag
i
Supplemental: Calibrations to 2009
0 2 4 6 8 100.0
0.2
0.4
0.6
0.8
Climate sensitivity [K]
a
Dens
ity
To 2009, with SLRWithout SLR
0 1 2 3 40.0
0.1
0.2
0.3
Ocean heat diffusivity [cm2/s]
b
0.0 0.5 1.0 1.5 2.00.0
0.5
1.0
1.5
2.0
Aerosol scaling
c
1960 2000 2040 2080
0
1
2
3
4
5
T [K
]
aObservationsRCP8.5 To 2009, with SLRWithout SLR
1960 2000 2040 2080
0.0
0.5
1.0
1.5
2.0
2.5
SLR
[m]
bObservationsRCP8.5 To 2009, with SLRWithout SLR