οο οο 4 2 Ο ( / ) δ Ο ( / )δ (ppbv) CO (ppmv)CH 18 ο 18 ο

Alfred Wegener Institute

for Polar and Marine Research

AWI

Land–atmosphere carbon exchange during abrupt climate change

Peter K¨ohler^AWI,

Fortunat Joos^Bern, Stefan Gerber^Bern,PEI, Reto Knutti^Bern,ETH

ESF Conference OCEAN CONTROLS IN ABRUPT CLIMATE CHANGE, Obergurgl 05/2007

Climate Dynamics (2005), 25: 689-708

Data constraints on abrupt climate changes during past 60 kyr Modelled bipolar seesaw (ECBILT-CLIO)

Terrestrial carbon storage in LPJ-DGVM Case study for preindustrial times (1 kyr BP)

Importance of the background climate

-42 -40 -38 -36 -34

δ18 Ο (ο / οο)

-40 -38 -36 -34

δ18 Ο (ο / οο)

180 200 220 240 260 280

CO 2 (ppmv)

-60000 -50000 -40000 -30000 -20000 -10000 0 Age (years)

300 400 500 600 700

CH 4 (ppbv)

A4 A3 A2 A1

▲ ▲ ▲ ▲ ▲

12 8

17 14

H6 H5 H4 H3 H2 H1YD

▲ ▲

H5a

▲

GISP2

BYRD

TAYLOR DOME

Data constraints on abrupt climate chan- ges in the past 60 kyr

Dansgaard/Oeschger events (e.g. 17, 14, 12, 8)

Heinrich events (H1–H6)

Antarctic warming events (A1–A4)

atmospheric CO₂ (±20 ppmv)

atmospheric CH₄ (±200 ppbv)

Grootes and Stuiver, 1997; Johnsen et al.,1972; Inderm¨uhle et al., 1999, 2000;

Blunier et al., 1998, Brook et al., 1996, 2000, Blunier and Brook, 2001

Modelled bipolar seesaw (ECBILT-CLIO)

0 1000 2000 Simulation time (yr) -10

-5 0 5

Temperature anomaly (K)

0 1000 2000

Simulation time (yr) 0 5 10 15 20

NADW (Sv)

0.0 0.1 0.2 0.3 0.4 0.5

Freshwater (Sv)

A

B

C

Scenario F_❏ Scenario F_∆

NATL SO

Bipolar seesaw in ECBILT-CLIO

Knutti et al., 2004.

Mimicking the bipolar seesaw by freshwater discharge into the North Atlantic (50^◦– 70^◦N).

Global coupled atmosphere–

ocean–sea ice model

ECBILT2: T21 atmosphere (Opsteegh et al., 1998) CLIO: OGCM + sea ice (Goose and Fichefet, 1999)

ECBILT-CLIO — Temperature

Zonally averaged ∆T over ice free land area:

North (70^◦N): cooling by 7K South (50^◦S): warming by 2K.

ECBILT-CLIO — Precipitation

Zonally averaged ∆prec over ice free land area:

Drier conditions in the tropics — wetter conditions in the subtropics

-42 -40 -38 -36 -34

δ18 Ο (ο / οο)

-40 -38 -36 -34

δ18 Ο (ο / οο)

180 200 220 240 260 280

CO 2 (ppmv)

-60000 -50000 -40000 -30000 -20000 -10000 0 Age (years)

300 400 500 600 700

CH 4 (ppbv)

A4 A3 A2 A1

▲ ▲ ▲ ▲ ▲

12 8

17 14

H6 H5 H4 H3 H2 H1YD

▲ ▲

H5a

▲

GISP2

BYRD

TAYLOR DOME

Data constraints on abrupt climate chan- ges in the past 60 kyr

Dansgaard/Oeschger events (e.g. 17, 14, 12, 8)

Heinrich events (H1–H6)

Antarctic warming events (A1–A4)

atmospheric CO₂ (±20 ppmv)

atmospheric CH₄ (±200 ppbv)

Grootes and Stuiver, 1997; Johnsen et al.,1972; Inderm¨uhle et al., 1999, 2000;

Blunier et al., 1998, Brook et al., 1996, 2000, Blunier and Brook, 2001

-42 -40 -38 -36 -34

δ18 Ο (ο / οο)

-40 -38 -36 -34

δ18 Ο (ο / οο)

180 200 220 240 260 280

CO 2 (ppmv)

-60000 -50000 -40000 -30000 -20000 -10000 0 Age (years)

300 400 500 600 700

CH 4 (ppbv)

A4 A3 A2 A1

▲ ▲ ▲ ▲ ▲

12 8

17 14

H6 H5 H4 H3 H2 H1YD

▲ ▲

H5a

▲

GISP2

BYRD

TAYLOR DOME

Data constraints on abrupt climate chan- ges in the past 60 kyr

Dansgaard/Oeschger events (e.g. 17, 14, 12, 8)

Heinrich events (H1–H6)

Antarctic warming events (A1–A4)

atmospheric CO₂ (±20 ppmv)

atmospheric CH₄ (±200 ppbv)

Case studies for 4 time slices:

1, 13, 17, 21 kyr BP PRE, YD, H1, LGM

Terrestrial carbon storage in LPJ-DGVM

Terrestrial carbon storage in LPJ-DGVM

PRE Difference LGM

Vegetation versus soil carbon

Vegetation Soil

Rule of thumb:

Tropics: more than 2/3 of carbon in the vegetation Boreal areas: more than 2/3 of carbon in the soil

Case study for preindustrial times (1 kyr BP)

Vegetation — Tree Cover

Southward shift of the northern treeline = f(temperature) Less trees in Sahel area (Mulitza et al.) = f(precipitation)

Zonally averaged land carbon storage anomalies

Increase (soil respiration) & decrease (northern treeline) in C storage Persisting anomaly 20^◦S (model artefact)

0 1000 2000 3000 4000 5000 Simulation time (yr)

-80 -60 -40 -20 0 20 40 60

Carbon storage anomaly (PgC)

F_❏ with CO

2 ff F_❏ without CO₂ ff ECBILT-ctrl

1 kyr BP experiment background

flux recover

1

2

3 Land carbon anomaly

1. CO₂ fertilization (NPP = f(CO₂)) is dampening the amplitude by factor of two 2. Final offset (20–30 PgC) due to single PFT reorga- nisation (precipitation) in a few grid cells in the tropics 3. Peak-to-peak-amplitude:

∆C(terrestrial) = 70 and 120 PgC

But...

Evidence for CO₂ fertilisation effect on vegetation growth is still poor

K¨orner 2006

-80 -60 -40 -20 0 20 40 60 80

Anomaly (PgC)

0 1000 2000 3000

Simulation time (yr) -80

-60 -40 -20 0 20 40 60 80

total

soil+litter vegetation

1 kyr BP

1

2

3

4

Soil versus vegetation

1. Soil respiration is reduced during cold times

⇒ carbon gain

2. Treeline shifts to South

⇒ carbon loss

3. Warming at the end of the experiment leads to increased soil respiration

⇒ carbon loss

4. Treeline back North / delayed recovery

⇒ carbon gain

Importance of the background climate

Atmospheric CO₂ concentration (ice cores) Land ice sheet extent / sea level (ICE–4G)

Global temperature / precipitation fields (HadSM3 model output)

-20000 -15000 -10000 -5000 0 Time (yr BP)

-80 -60 -40 -20 0 20 40 60 80 100 120

Carbon storage anomaly (PgC)

2000 2100 2200 2300 2400 2500 2600 2700 2800 2900 3000

Terrestrial carbon (PgC)

freshwater release exp.

background without

Impact of

climate change during

Termination I

Size and direction of carbon storage anomaly depends on background climate, mainly background temperature.

Zonally averaged land carbon storage anomalies

1 kyr BP 13 kyr BP

17 kyr BP 21 kyr BP

Increase / decrease in terrestrial carbon storage

-75 -50 -25 0 25 50 75

Anomaly (PgC)

-75 -50 -25 0 25 50 75

total

soil+litter vegetation

-75 -50 -25 0 25 50 75

Anomaly (PgC)

-75 -50 -25 0 25 50 75

Anomaly (PgC)

0 1000 2000

Simulation time (yr) -75

-50 -25 0 25 50 75

0 1000 2000 3000

Simulation time (yr)

0 1000 2000 3000

-75 -50 -25 0 25 50 75

Anomaly (PgC)

1 kyr BP 13 kyr BP

21 kyr BP 17 kyr BP

Impact of

climate change for

different times

without CO₂ fertilization 13 kyr BP: Younger Dryas cold event

17 kyr BP: Heinrich event during partly glaciation

21 kyr BP: Heinrich event during full glaciation

275 280 285 290 295 300

CO 2 (ppmv)

without CO

2 fert.

with CO

2 fert.

230 235 240 245 250 255

CO 2 (ppmv)

0 1000 2000

Simulation time (yr) 190

195 200 205 210 215

CO 2 (ppmv)

0 1000 2000

Simulation time (yr)

175 180 185 190 195 200

CO 2 (ppmv)

1 kyr BP 13 kyr BP

21 kyr BP 17 kyr BP

Impacts on atmospheric CO

LGM amplitude:

∼7–12 ppmv

To be consistent with the ice core record the mari- ne carbon cycle needs to contribute about the same magnitude.

-7.0 -6.9 -6.8 -6.7 -6.6

δ13 C (o / oo)

-7.0 -6.9 -6.8 -6.7 -6,6

δ13 C (o / oo)

0 1000 2000

Simulation time (yr) -7,1

-7.0 -6.9 -6.8 -6.7

δ13 C(o / oo)

with CO

2 fert.

without CO

2 fert.

0 1000 2000 3000

Simulation time (yr)

-7.1 -7.0 -6.9 -6.8 -6.7

δ13 C (o / oo)

1 kyr BP 13 kyr BP

21 kyr BP 17 kyr BP

Impacts on atmospheric δ

C

LGM amplitude:

∼–0.2 to –0.3h

Ice core record not clear on δ¹³C:

H1 & YD:

– 0.2 to –0.4h _excursion

but both events occur during the last glacial/

interglacial transition

-7.0 -6.8 -6.6 -6.4 -6.2

13 C[o /oo]

Interval I II III IV H1 BA YD

180 200 220 240 260 280

pCO2[ppmv]

20 18 16 14 12 10 GISP2 Age [kyr BP]

0 100

200 300 400 500

14 C[o /oo]

Data constraints on δ

C

Taylor Dome (Smith et al., 1999)

EPICA Dome C (Monnin et al., 2001)

INTCAL98 (Stuiver et al., 1998) Cariaco basin (Hughen et al., 2004)

Conclusions

• Abrupt climate changes caused by the bipolar seesaw lead to opposing trends in terrestrial carbon storage: a southward shift of the northern treeline (carbon loss) and a reduction in soil respiration in mid latitudes (carbon gain).

• The overall net effect on carbon storage depends on the global balance of losses and gains and is highly sensitive to the applied background climatology.

• Be careful about the magnitude because of the unknowns in the CO₂ fertilisation feedback.

• The increase of 20 ppmv in atmospheric CO₂ during the Antarctic warming events A1–A4 might by ∼50% caused terrestrial carbon re- lease.

• OUTLOOK Impact of same ECBILT-CLIO scenarios on CH₄ cycle in LPJ-DGVM: see poster by Jed Kaplan and Joe Melton.

LPJ-DGVM

Sitch et al., 2003 GCB.

Drivers:

– monthly mean temperature, precipitation and insolation fields – CO₂

– terrestrial ice free land area Spatial resolution: 3.75^◦× 2.5^◦ Plant functional types (PFT)

—————————————————

Tropical broadleaved evergreen tree Tropical broadleaved raingreen tree Temperate needle-leaved evergreen tree Temperate broadleaved evergreen tree Temperate broadleaved summergreen tree Boreal needle-leaved evergreen tree

Boreal summergreen tree C3 grass

C4 grass

180 200 220 240 260 280 300

CO 2 (ppmv)

180 200 220 240 260 280 300

-10 -5 0 5 10 15 20

∆Area (1012 m2 )

-12 -10 -8 -6 -4 -2 0

∆T (K)

20 15 10 5

Time (kyr BP) -12

-10 -8 -6 -4 -2 0

global land 30^oS-30^oN 30^oN-90^oN

20 15 10 5 0

Time (kyr BP)

-175 -150 -125 -100 -75 -50 -25 0

∆prec (mm yr-1 )

A B

C D

ice retreat

net change

sea level

Background forcing of LPJ-DGVM

same as in Joos et al., 2004 GBC.

Present day mean climatology (Leemans and Cramer 1991) with following pertubations:

A: CO2 (EPICA Dome C, Monnin et al., 2001) GISP2 age scale

B: land area from Peltier 1994

C,D: ∆T and ∆prec from Hadley Centre Unified Model

Preindustrial carbon storage LPJ-DGVM

Glacial carbon storage LPJ-DGVM

Difference in carbon storage LPJ-DGVM (21 – 1 kyr BP)

-60 -40 -20 0 20 40 60 80 Latitude

-0,4 -0,3 -0,2 -0,1 0 0,1 0,2

Anomaly in relative foliar projective cover

tropical trees

temperate trees

grasses

boreal trees 1 kyr BP

Impact of climate change on vegetation

Grasses replace boreal trees north of 50^◦N

Boreal trees replace temperate trees in mid high latitudes

Small reduction in tropical tree cover

-20000 -15000 -10000 -5000 0 Time (yr BP)

-80 -60 -40 -20 0 20 40 60 80 100 120

Carbon storage anomaly (PgC)

2000 2100 2200 2300 2400 2500 2600 2700 2800 2900 3000

Terrestrial carbon (PgC)

freshwater release exp.

background without

Impact of

climate change during

Termination I

Background terrestrial carbon:

ice retreat: + 610 PgC sea level rise: – 190 PgC CO2 fertilization: + 650 PgC climate change: – 250 PgC

total: + 820 PgC

Size and direction of carbon storage anomaly depends on background climate.

Atmospheric carbon records

LPJ anomalies coupled to the HILDA carbon cycle model

0 1000 2000 3000 4000 5000

Simulation time (yr) 270

275 280 285 290 295 300

atmospheric CO 2 (ppmv)

F_❏ with CO

2 ff F_❏ without CO

2 ff.

ECBILT-ctrl

1 kyr BP experiment background

0 1000 2000 3000 4000 5000

Simulation time (yr) -7.1

-7.0 -6.9 -6.8 -6.7 -6.6

atmospheric δ13 C (o / oo)

F_❏with CO₂ ff F_❏without CO₂ ff ECBILT-ctrl

1 kyr BP experiment background

Peak-to-peak-amplitudes:

∆CO₂ = 13 and 21 ppmv ∆δ¹³C = 0.24 and 0.40 h

Sensitivity studies

0 1000 2000 3000 4000

Simulation time (yr) -80

-60 -40 -20 0 20 40 60

Carbon storage anomaly (PgC)

F_❏ F_∆

ECBILT-ctrl

0 1000 2000 3000 4000 5000

Simulation time (yr)

-80 -60 -40 -20 0 20 40 60

Carbon storage anomaly (PgC)

F_❏

F_❏δP-only F_❏δT-only ECBILT-ctrl

1 kyr BP 1 kyr BP

A B

experiment background experiment background

Shape of freshwater discharge Temperature vs. Precipitation

-100 -50 0 50 100

-100 -50 0 50 100

Anomaly (PgC)

-100 -50 0 50 100

Anomaly (PgC)

-100 -50 0 50 100

0 1000 2000

Simulation time (yr) -100

-50 0 50 100

Anomaly (PgC)

CO2 LGM Climate LGM Land LGM

0 1000 2000

Simulation time (yr)

-100 -50 0 50 100

Anomaly (PgC)

Standard CO2 PRE Climate PRE Land PRE

1 kyr BP 13 kyr BP

21 kyr BP 17 kyr BP

Sensitivity analysis

on

boundary conditions

without CO₂ fertilization vary CO2

vary climate vary land area

LPJ is most sensitive to climate variations

-100 -50 0 50 100

-100 -50 0 50 100

Anomaly (PgC)

-100 -50 0 50 100

Anomaly (PgC)

-100 -50 0 50 100

0 1000 2000

Simulation time (yr) -100

-50 0 50 100

Anomaly (PgC)

Standard δΤ − 50%

0 1000 2000 3000

Simulation time (yr)

0 1000 2000 3000

-100 -50 0 50 100

Anomaly (PgC)

δΤ − 25%

δΤ + 25%

1 kyr BP 13 kyr BP

21 kyr BP 17 kyr BP

Sensitivity analysis

on the

magnitude of the

temperature anomaly (∆T)

without CO2 fertilization

∆T – 50%

∆T – 25%

∆T + 25%

Especially during full gla- ciation the magnitude of

∆T constitutes the re- sponse