Understanding oceansSustaining our future

Enhanced storm activities triggered the North Pacific deep convection during the Younger Dryas event

Xiaopei LinPhysical Oceanography Lab, Ocean University of China

Qingdao National Lab for Marine Science and Technology

Cunjie Zhang, Baolan Wu

Jian Zhao, Gerrit Lohmann, Xun Gong, Xu Zhang, Haijun Yang, Zhengyu Liu, Ping Chang, Min-Te Chen…

Trenberth and Caron, 2001

Ocean Meridional Heat Transport

AMOCThermohaline Circulation

PMOCWind Driven Circulation

fttp://global.britannica.com/science/thermohalinecirculation http://www4.ncsu.edu/eos/users/c/ceknowle/public/chapter07/part1.html

Yang et al. 2015

Indo-Pacific MOC and Potential TemperaturePMOC and AMOC under modern conditions

Sea Surface Salinity

Stommel two-box model，1961

AMOC and Potential Temperature

contours: PTshading: MOC

(positive: clockwise)

contours: PTshading : MOC

(positive: clockwise)

Change of ICE Sheet

Data source ICE-6G

AMOC Strength

Ritz et al，20135 Sv 7 Sv

H1 YD

The modes of AMOC and PMOC might change during stadials in the last deglaciation.

Kleman et.al 2007

Most melt waters discharged into the

North Atlantic because of the topography

Zhang and Delworth, 2005, JCCold and Dry in the North Pacific

Stronger and southward shift Westerly

Sun*…Lin et al., 2011, Nature Geoscience.

Arctic Ice

Temperature

Antarctic Ice

Temperature

Kuroshio

Temperature

Kuroshio S

AMOC Shutdown

Kuroshio IncreaseImplying a Sea Saw

between AMOC and PMOC

Chen* & Lin et al., 2010, GRL

SST anomaly

Besides the Kuroshio, there is a warm belt in the central PacificSubtropical Counter Current——Part of PMOC

[Oka and Qiu, 2011]

[Kobashi et al., 2006]

Mode Water——Subtropical Counter Current

This warm belt is due to more mode water, stronger STCC by cold event and deep convection

Zhang&Lin et al., 2018, CD

YD H1

stre

ngth

ened

North Pacific Ventilation Strength

North Pacific water age anomaliesin a collapsed AMOC state

Deep water extending 2500-3000 m was formed in the

North Pacific during H1. (following articles prefer 2000 m).

Previous explanations：saltier surface North Pacific

Kiefer, 2010 Science

North Pacific salinity proxy

YD H1

Limits： Salinity proxy：only one core in the western

subtropical gyre. Applied to H1，but not to YD.

Okazaki et,al 2010 Science

Locations of paleo records ：

Number：22

0.06

0.07

0.08

0.09OC

E326

-GG

C5

231

Pa/

230

Th

YD H1

Vent

ilatio

n

500

1000

1500

2000Ve

ntila

tion

Age

[yr]

Vent

ilatio

n

500

1000

1500

2000

2500

3000Vent

ilatio

n Ag

e [y

r]

Vent

ilatio

n

CH84-14

GH02-1030

KT89-18-P4

PC4/PC5

MD01-2420

-0.6

-0.2

0.2

0.6

1

13C

[‰]

SO17

8-13

-6

Vent

ilatio

n

-0.4

-0.3

-0.2

-0.1

SO20

1-2-

85KL

13C

[‰]

33

33.5

34

34.5

35

Rec

onst

ruct

ed

SSS

[psu

]

10 11 12 13 14 15 16 17 18 19 20

Calendar Age [ka B.P.]

0

0.5

1

18O

W [

‰]

SSS

A) N Atlantic +

-

B) Okhotsk and Bering Sea +

-

C) NW Pacific +

-

D) Okhotsk and Bering Sea +

-

E) NW Pacific

F) GH02-1030 +

-

6

8

10

PM

OC

[Sv]

YD H1

4

8

12

16

AM

OC

[Sv]

500

600

700

Nor

thw

est P

acifi

cId

eal A

ge [y

r]

200

400

600

Nor

th A

tlant

icId

eal A

ge [y

r]

33.4

33.6

33.8

34

Nor

thw

est P

acifi

c

SS

S [p

su]

3

3.1

3.2

3.3

NH

Atm

osph

ere

MH

T [P

W]

0.6

0.7

0.8

0.9

NH

Oce

anM

HT

[PW

]

0.7

0.8

0.9

Nor

thw

est P

acifi

c

Edd

y A

MH

T [P

W]

10 11 12 13 14 15 16 17 18 19 20

Calendar Age [ka B.P.]

0.4

0.5

Pac

ific

MH

T [P

W]

0.1

0.2

0.3

0.4

0.5

Atla

ntic

MH

T [P

W]

A

B

C

D

E

F

21K CCSM3 Simulation Liu et al., 2013

Paleo Record

Observed deep convection in the North Atlantic

is related to storm activities in the westerly

Deep convection in 2008

2008年的强湍热（潜热感热）失热

Vage et al., 2009, Nature Geoscience M. Femke de Jong,2012,JC

What affects deep convection besides salinity?

turbulent heat loss

1-d MLD model（PWP model）

Without storms

With storms

Sv

Positive : Clockwise

A) control

EQ 10°N 20°N 30°N 40°N 50°N 60°N0

5

10

15

20

25

30

Pote

ntia

l Tem

pera

ture

-27

-18

-9

0

9

18

27

Sv

Positive : Clockwise

B) hosing

EQ 10°N 20°N 30°N 40°N 50°N 60°N0

5

10

15

20

25

30

Pote

ntia

l Tem

pera

ture

-27

-18

-9

0

9

18

27

Sv

Positive : Clockwise

C) hosing - control

EQ 10°N 20°N 30°N 40°N 50°N 60°N0

5

10

15

20

25

30

Pote

ntia

l Tem

pera

ture

-27

-18

-9

0

9

18

27

20 o N 20 o N 120 o E 120 o E 150 o W 150 o W 150 o E 180 o W 120 o W

30 o N

40 o N

50 o N

60 o N

70 o N A) SST Anomaly

150 o E 180 o W 120 o W

30 o N

40 o N

50 o N

60 o N

70 o N

°C-3

-2

-1

0

1

2

3

20 o N 20 o N 120 o E 120 o E 150 o W 150 o W 150 o E 180 o W 120 o W

30 o N

40 o N

50 o N

60 o N

70 o N B) SSS Anomaly

150 o E 180 o W 120 o W

30 o N

40 o N

50 o N

60 o N

70 o N

psu-4-3-2-101234

20 o N 20 o N 120 o E 120 o E 150 o W 150 o W 150 o E 180 o W 120 o W

30 o N

40 o N

50 o N

60 o N

70 o N C) SSD Anomaly

150 o E 180 o W 120 o W

30 o N

40 o N

50 o N

60 o N

70 o N

kg·m3

-3

-2

-1

0

1

2

3

CESM Water Hosing Experiments：Fresh water input in the North Atlantic and shutdown the AMOC

Storm is dominant for the ocean heat loss and SST change

Ma et al., 2015, Sci. Rep

Water Hosing Experiment-YD simulation：

MLD Change Turbulent Heat Flux

Storm Days

Heat FluxVertical Temperature and MLD

MLD

Strong Westerly, Strong Storm, Increased Convection, Strong PMOC

Shutdown of AMOC-Cold SST in North Atlantic-Atmospheric TeleconnectionStrong, Southward Westerly and Storm-Deep Convection-Strong PMOC

Wu et al., 2008,JCStorm Change after AMOC Shutdown

From Yang et al., 2012

Why AMOC Shutdown Increase Westerly and Strom?

0=∆+∆=∆ OHTAHTHTtotalBjerknes Compensation

Meridional Heat Transport: Total, Atmosphere and Ocean

OceanAtlanticPac-Ind.

Sea Saw of AMOC and PMOC

Total OHT ⇓ 0.2 PWAtlantic OHT ⇓ 0.3 PW, Pacific-Indian OHT ⇑ 0.1 PW

Chen & Tung, 2018, Nature

AMOC+ More Convection More Heat Down Global Hiatus

Wu et al., 2012,NCC

Global Warming

Hiatus

Stefan et al., 2015

Warming Period ：AMOC ⇓ , KC ⇑Hiatus Period ：AMOC ⇑, KC ⇓

AMOC Strength

Wei et al., 2013

Kuroshio Transport Kosaka&Xie, et al., 2013,Nature

AMOC-PMOC Sea Saw in modern climate

Lozier…Lin et al., 2017, BAMS

Deep Convection in Labrador Sea is not importantOverflow between Subpolar Atlantic and Nordic Sea is Dominant

Lozier*…Lin et al., 2019, Science

AMOC from Labrador Sea

AMOC from Nordic Sea

Meridional Heat Transport By Eddies

Zhao*…Lin* et al., 2018, Nature CommunicationsZhao*…Lin* et al., 2018, JGR

Mesoscale Eddies contribute a lot to AMOC

Summary

• There is a Sea-Saw between AMOC and PMOC.

• Besides salinity, the storm induced deep-convection is also important to explain the observed Sea-Saw, especially in the YD event.

• The modern observation imply that the deep-convection may be not important for the formation and change of AMOC, but eddies play some roles on the AMOC variability.

Thank You!

Convection depth in the Okhotsk Sea and Bering sea:• 1400-2400 m during the H1. By Jaccard and Galbraith [2013]• 1750-2100 m during the H1 and YD. By Max et al., [2014]

Select: <2000-m depth only

Core ID Longitude Latitude Water Depth (m) Area Used in Figure Reference

B34-91 150.34°E 49.14°N 1227 Okhotsk Sea Fig. 2-3.B Keigwin (2002), Okazaki et al. (2010), Cook et al. (2015)

GGC27 150.18°E 49.60°N 995 Okhotsk Sea Fig. 2-3.B Keigwin (2002), Okazaki et al. (2010), Okazaki et al. (2010), Cook et al. (2015)

LV27-2-4 144.75°E 54.50°N 1305 Okhotsk Sea Fig. 2-3.B Gorbarenko et al. (2010), Okazaki et al. (2014)

936 148.31°E 51.02°N 1305 Okhotsk Sea Fig. 2-3.B Gorbarenko et al. (2004), Okazaki et al. (2014)

GGC15 150.4°E 48.6°N 1980 Okhotsk Sea Fig. 2-3.B Keigwin (2002), Okazaki et al. (2010)

GGC18 150.4°E 48.8°N 1700 Okhotsk Sea Fig. 2-3.B Keigwin (2002), Okazaki et al. (2010)

GGC20 150.4°E 48.9°N 1510 Okhotsk Sea Fig. 2-3.B Keigwin (2002), Okazaki et al. (2010)

SO178-13-6 144.70°E 52.72°N 713 Okhotsk Sea Fig. 2-3.B; Fig. 2-3.D Max, L et al. (2014)

LV29-114-3 152.88°E 49.37°N 1765 Okhotsk Sea Fig. 2-3.B Max, L et al. (2014)

SO201-2-101KL 170.68°E 58.87°N 630 Bering Sea Fig. 2-3.B Max, L et al. (2014)

SO201-2-85KL 170.40°E 57.50°N 968 Bering Sea Fig. 2-3.B; Fig. 2-3.D Max, L et al. (2014)

SO202-18-6 179.43°W 60.12°N 1100 Bering Sea Fig. 2-3.B Max, L et al. (2014)

Core ID Longitude LatitudeWater Depth

(m)Area Reference

MD01-2416 167.73°E 51.27°N 2317 Emperor Seamounts Sarnthein el al. (2006), Okazaki et al. (2010) > 2000 m

ODP883 167.77°E 51.20°N 2385 Emperor Seamounts Sarnthein el al. (2006), Okazaki et al. (2010) > 2000 m

SO201-2-12KL 162.37°E 53.98°N 2145 Bering Sea Max, L et al. (2014) > 2000 m

SO201-2-77KL 170.68°E 56.32°N 2135 Bering Sea Max, L et al. (2014) > 2000 m

RNDB 10PC 167.665°E 51.296°N 2364 Emperor Seamounts Keigwin et al. (2015) > 2000 m

RNDB 11PC 167.975°E 51.073°N 3225 Emperor Seamounts Keigwin et al. (2015) > 2000 m

V34-98 153.20°E 50.11°N 1175 Okhotsk Sea Gorbarenko et al. (2002), Okazaki et al. (2014) No data during 10-20 Ka

Adapted

Abandoned

Ventilation Age： Compare 14C age difference between planktonic and benthic foraminifers to estimate the age of bottom waters. Affected by：Ventilation strength + Water depth

0.06

0.07

0.08

0.09OC

E32

6-G

GC

523

1P

a/23

0Th

YD H1

Ven

tilat

ion

500

1000

1500

2000V

entil

atio

n A

ge [y

r]

Ven

tilat

ion

500

1000

1500

2000

2500

3000Ven

tilat

ion

Age

[yr]

Ven

tilat

ion

CH84-14

GH02-1030

KT89-18-P4

PC4/PC5

MD01-2420

-0.6

-0.2

0.2

0.6

1

13C

[‰]

SO

178-

13-6

Ven

tilat

ion

-0.4

-0.3

-0.2

-0.1

SO

201-

2-85

KL

13C

[‰]

33

33.5

34

34.5

35

Rec

onst

ruct

ed

SS

S [p

su]

10 11 12 13 14 15 16 17 18 19 20

Calendar Age [ka B.P.]

0

0.5

1

18O

W [

‰]

SS

S

A) N Atlantic +

-

B) Okhotsk and Bering Sea +

-

C) NW Pacific +

-

D) Okhotsk and Bering Sea +

-

E) NW Pacific

F) GH02-1030 +

-

AMOC strength

Based on： 31Pa/230Th ratio

Ventilation Age for the Okhotsk Sea and Bering Sea based on 14C

Select：< 2000 m depth only

Ventilation Age for the western North Pacific subtropical gyre based on 14C

Select： At least one value for the H1 and YD period

Reconstructed SSS

SSS proxy

Cores： KT90-9, 21；KT90-9, 5；KH94-3,LM-8；MD01-2421；CH84-04

Core：GH02-1030

Fig. 2-3

6

8

10

PMO

C [S

v]

YD H1

4

8

12

16

AMO

C [S

v]

500

600

700N

orth

wes

t Pac

ific

Idea

l Age

[yr]

200

400

600

Nor

th A

tlant

icId

eal A

ge [y

r]

33.4

33.6

33.8

34

Nor

thw

est P

acifi

c

SSS

[psu

]

3

3.1

3.2

3.3

NH

Atm

osph

ere

MH

T [P

W]

0.6

0.7

0.8

0.9

NH

Oce

anM

HT

[PW

]

0.7

0.8

0.9

Nor

thw

est P

acifi

c

Eddy

AM

HT

[PW

]

10 11 12 13 14 15 16 17 18 19 20

Calendar Age [ka B.P.]

0.4

0.5

Paci

fic M

HT

[PW

]

0.1

0.2

0.3

0.4

0.5

Atla

ntic

MH

T [P

W]

A

B

C

D

E

F

Water Age at 1000-m for NA and NW P

Seesaw of AMOC and PMOC

Northwest Pacific SSS

Atmospheric and Oceanic MHT in the northern hemisphere

Pacific and Atlantic MHT in the northern hemisphere

Bjerknes Compensation

Storm induced atmospheric MHT in the Northwest Pacific

21K CCSM3 Simulation, Liu et al., 2013