Northern and southern influences on Northern and southern influences on the MOCthe MOC

Claus BöningClaus Böning (IFM-GEOMAR, Kiel)(IFM-GEOMAR, Kiel)withwith

Arne Biastoch, Markus Scheinert, Erik BehrensArne Biastoch, Markus Scheinert, Erik Behrens

Causes of MOC variability: Causes of MOC variability: what have we learned what have we learned from previous ocean from previous ocean modelling studies?modelling studies?

26°N

Obs

1/12°-mod. 1/3°-mod.

Relation between MOC and heat transport?Relation between MOC and heat transport? Biastoch, Böning, Getzlaff, et al. (J. Climate, 2006)Biastoch, Böning, Getzlaff, et al. (J. Climate, 2006)

Role of Nordic Seas (Overflows) Role of Nordic Seas (Overflows) vsvs. Labrador Sea . Labrador Sea (Convection)? (Convection)?

Schweckendiek and Willebrand (J. Climate, 2005)Schweckendiek and Willebrand (J. Climate, 2005)

Latif, Böning, Willebrand, et al. (J. Climate, 2006)Latif, Böning, Willebrand, et al. (J. Climate, 2006)

Relation convection - MOC?Relation convection - MOC? Böning, Scheinert, Dengg, et al. (GRL, 2006)Böning, Scheinert, Dengg, et al. (GRL, 2006)

Changes in deep water formation Changes in deep water formation vs.vs. wind stress? wind stress? Biastoch, Böning, Getzlaff, et al. (J. Climate, 2006)Biastoch, Böning, Getzlaff, et al. (J. Climate, 2006)

---------------------------------------------- Effects of Agulhas Leakage (I): wave mechanismEffects of Agulhas Leakage (I): wave mechanism Biastoch, Böning, Lutjeharms (Nature, 27 Nov 2008)Biastoch, Böning, Lutjeharms (Nature, 27 Nov 2008)

Effects of Agulhas Leakage (II): advection Effects of Agulhas Leakage (II): advection Biastoch, Böning, Schwarzkopf, Lutjeharms (Nature, 26 Nov 2009)Biastoch, Böning, Schwarzkopf, Lutjeharms (Nature, 26 Nov 2009)

26°N

1.6

Heat transport (PW)

0.6

20

MOC (Sv)

10

MOC and heat transport

Model simulations:

Close correspondence between MOC and heat transport in the subtropical North Atlantic

1.6

Heat transport (PW) 0.6

MOC (in depth-coordinates) is a useful index only south of ~43°N

20°S 0° 20° 40° 60°N

1

r

0

MOC and heat transport

… but caution:

Correlation

based on monthly and 2-yr-filtered time series time series

Changes in deep water formation: effect on MOC?

(I) Labrador Sea:

deep winter mixing

(II)Outflow from the

Nordic Seas:

Role of changes in overflow vs. convection

North Atlantic model, northern boundary: 70°N forced by anomalies from IPCC climate runs

Greenland-Scotland sill OGCM

Climate Model (GFDL)

MOC (40°N)

OGCM study by Schweckendiek & Willebrand (J. Clim., 2005)

south 66° 70°N 1960 2000 2040 2080

 

Role of changes in overflow vs. convection

North Atlantic model: northern boundary: 70°N forced by anomalies from IPCC climate runs

Greenland-Scotland sill OGCM

Climate Model (GFDL)

MOC (40°N)

OGCM study by Schweckendiek & Willebrand (J. Clim., 2005)

south 66° 70°N 1960 2000 2040 2080

(1) The combination of surface forcing and overflow density changes (prescribed in sponge layer 66°-70°N)

gives an almost perfect reconstruction of the coupled experiments

 

Exps. (2) and (3) show: trend related to changing overflow density!

2

3

relation between overflow density and MOC

… as shown by a host of model cases with prescribed changes in the overflow density:

Latif, Böning, Willebrand, et al. (J. Climate, 2006)

Labrador Sea deep winter mixing

Snapshot ofmixed layer depth

in March(1/12°-model)

important factors: - Surface heat flux (destabilising)

- Eddy-flux of fresh water from West Greenland Current (stabilising)

1800

m

1000

200

Greenland

Case study: Permanent shut-down of deep convection

Hüttl and Böning, 2006

Evolution of MOC anomalies (Hovmoeller diagram)

(1/3°-model)

Hindcast simulation with NCEP-heat flux

MOC (43°N) follows the LSW formation rate

with a lag of ~ 2 years (r = 0.71)

6 – 8 Sv change in LSW formation gives ~2 Sv change in MOC

MOC LSW formation

Böning et al., GRL 2006

MOC response to NCEP heat flux variability

Years ofintense LSWproduction

MOCanomalies (in Sv) in exp. „Heat“

C.I. = 0.2 Sv

1-1

Range of decadal MOC variability: 10-15% of mean transport

but these decadal MOC changes are masked…

MOC anomaliesrelated to

LSW changes

in isolation

…and superimposedby wind-driven

variability

… by high-frequency wind-driven variability and eddy effects

Model caseincluding

wind stressvariability

Biastoch et al. (J. Clim., 2008)

Northern vs. Southern InfluencesNorthern vs. Southern Influences

Upper Branch

North Atlantic Circulation [schematic by G. Holloway]

deep-water formation

areas

ORCA05: 50 km grid resolutionORCA05: 50 km grid resolution

specifically: specifically: Role of the Agulhas System?Role of the Agulhas System?

ORCA025: 25 km grid resolutionORCA025: 25 km grid resolutionAG01: 10 km grid resolutionAG01: 10 km grid resolution

Agulhas ringsAgulhas rings

MozambiqMozambique eddiesue eddies

Agulhas CurrentAgulhas Current

AfricaAfrica

Approach: two-way nesting of

high- resolution regional model

(1/10°) in global model domain

(1/2°)

Temperature and velocity at 450m depth

(I) Decadal variability signal induced in Agulhas (I) Decadal variability signal induced in Agulhas regime:regime:

propagation by wavespropagation by waves

Boundary WavesBoundary Waves

Rossby Waves / Agulhas Rings

Rossby Waves / Agulhas Rings

Colour: eddy kinetic energy

[Biastoch, Böning, Lutjeharms; Nature, 2008]

Comparison of model runs Comparison of model runs with with and and without nestwithout nest::

MOC changes manifested in the NBC at 6°S

Biastoch, Böning, Lutjeharms (Nature, 2008)

Agulhas-induced MOC-variability of +/- 1.5 Sv, rapidly propagating to the North Atlantic

Hovmoeller-

plot:

MOC-

difference

Agulhas-induced MOC variability vs. effect of LSW Agulhas-induced MOC variability vs. effect of LSW formation:formation:

Complete Forcing

Standard deviation of interannual MOC strength

Agulhas influence reaches into the North Atlantic

… in tropics comparable to effect of subarctic deepwater formation

[Biastoch, Böning, Lutjeharms; Nature, 2008]

Effect of LSW

Effect of Agulhas variability

(II) Changes in Agulhas leakage: advective effects

Large-scale Circulation Changes South of AfricaLarge-scale Circulation Changes South of Africa

…also seen in SSH Latitude of zero SSH:

model and Aviso satellite altimetry

Trends 1970s – 2000s:

streamfunction wind stress curl

Gyre extended poleward due to the shift of the westerlies

Biastoch, Böning, Schwarzkopf, Lutjeharms: Nature, 26 Nov 2009

Increase in Agulhas LeakageIncrease in Agulhas Leakage

…during the recent decades (1.2 Sv/decade)

GoodHope

Biastoch, Böning, Schwarzkopf, Lutjeharms: Nature,26 Nov 2009

i.e., Lagrangian transport fraction across the Good Hope section:

REF

CLIM

Pathways of Agulhas LeakagePathways of Agulhas Leakage

Example trajectories of virtual floats released along the GoodHope section (T≥10°C)

Pathway to the North Atlantic (upper limb of the MOC):

- No change in volume transport

- but: 25% decrease in freshwater flux [Biastoch et al., Nature, 26 Nov 2009]

Observed salinities in the North Brazil CurrentObserved salinities in the North Brazil Current

Analysis of historic profiles in NBC core off South

America

… invasion of salty Indian Ocean waters

increasing salinity in the NBC near the equator

Biastoch, Böning, Schwarzkopf, Lutjeharms: Nature, 26 Nov 2009

Thermocline Changes in the Southern HemisphereThermocline Changes in the Southern Hemisphere

2000-2004 minus 1968-1972 temperatures

Upper ocean (0-200m) Zonally averaged over SW Indian Ocean (30°-50°E, isolines show mean

temperatures)

Biastoch, Böning, Schwarzkopf, Lutjeharms: Nature, 26 Nov 2009

Nested Agulhas modelNested Agulhas model

High-resolution nest simulates all salient features of the Agulhas Current system

The two-way nesting scheme allows to study the feedback of the Agulhas region on the global circulation

Agulhas leakage dynamics affects decadal variability in Atlantic MOC (wave Agulhas leakage dynamics affects decadal variability in Atlantic MOC (wave process)process)

Mesoscale Agulhas variability has no effect on mean MOC

… but decadal MOC variations of ±1.5 Sv

… quickly reaches into northern hemisphere, with similar magnitude as sub-arctic deepwater formation events

Agulhas leakage change affects Atlantic THC (advective process)Agulhas leakage change affects Atlantic THC (advective process)

Super-gyre has extended due to poleward shift of the westerlies

The Agulhas leakage has increased

… with no effect on time-mean MOC

… but a 25% increase of the salt export towards the North Atlantic

ConclusionsConclusions AaaaAaaa

BbbbBbbb