New insights from models

New insights from models

NCAR T-shirt design, 1988

Anne Marie Treguier (LPO, Brest, France)Claus Böning (IFM Kiel, Germany)

Models yesterday...1988: WOCE Community Model Experiment (Bryan et al)- North Atlantic, 1/3°- 220x242x30 grid points.

… and todayPOP global 1/10° model (Maltrud and McClean)3600x2400x40 grid points.

OCCAM global 1/12° model(Coward and Webb)4320x1735x66 grid points

Plan

1 - Knowledge of the ocean circulation

Discover and quantify Models/observations synergy

Understand processes Sensitivity studies

2 - Mechanisms of heat transport variability

Seasonal to interannual time scales: wind forcing

Decadal scales : buoyancy fluxes come into play

What about eddies?

3 - Perspectives

Discoveries: North Brazil Undercurrent

The NBUC in the CME model (Schott and Böning, 1991).

The NBUC with ADCP data (Stramma, Fisher, Reppin 1995)

Discovery of the North Queensland current

R.D. Hughes, D. Webb

NQc

The Zapiola Anticyclone in the Argentine Basin : An circulation driven by eddy-topography interactions

Marvor and Alace floats (Ollitrault, Davis...)

Bottom Topography

Mean circulation, 350 mSPEM 1/3° model, coordinate

(de Miranda et al., 1999)

3 models, same behavior3 models, same behavior

MICOM

ORCA

OCCAM

Inverse model Inverse model

AAIW

SAMW

Tasman leakage: a new route in the global conveyor belt

(Speich et al, 2002)

Plan

1 - Knowledge of the ocean circulation

Discover and quantify Model/observations synergy

Understand processes Sensitivity studies

2 - Mechanisms of heat transport variability

Seasonal to interannual time scales

Decadal scales

What about eddies?

3 - Perspectives

The dynamical origin of the Azores current

Azores current Strait of

Gibraltar

The med outflow and the Azores

current

With representation of med outflow: strong AZc

Without representation of med outflow: weak AzcOzgökmen et al, 2001: MICOM 1/12° model

Jia, 1999: DYNAMO intercomparison project.

Zonal current in the Brazil basin

Float displacements:Hogg and Owens, 1999

1/6° CLIPPER Atlantic model, Treguier et al.

0.28° global POP model Maltrud et al.

Displacements over 2 years at 2400m (North Atlantic Deep Water).

A possible dynamical origin of the zonal flows: wind forcing

Nakano and Suginohara, 2002

Zonal flow at 140°W in a 1° model of the Pacific

(bottom: contour interval 0.1 cm/s)

Zonal flow in a shallow water model (20 modes) forced by a zonal wind

Dependency on model resolution

Zonal velocity (mean over 10 years) at 2000m (westward in green. Contour interval 0.4 cm/s)

1° model 1/6° model

Zonal jets in the subtropical North Atlantic

Mean zonal velocity (color)

and mean velocity vectors

at 1400m depth in 3 models

High resolution models improve our

knowledge of the ocean circulation

- Models begin to look like reality

- High resolution basin scale models can be used to run sensitivity studies

- Many key processes affecting the large scale circulation have very small scale

Plan

1 - Knowledge of the ocean circulation

Discoveries and quantification

Understanding processes

2 - Mechanisms of heat transport variability

Seasonal to interannual time scales

Decadal scales

What about eddies?

3 - Perspectives

Heat transport and overturning

Robust relationship at 25°N

CMIP2

Vertical overturning and horizontal gyres

Total heat transport overturning

gyre

POP 1/10° North Atlantic model (Smith et al, 2000)

Ekman transport drives the seasonal variability in the Atlantic

Böning and Hermann, 1994Seasonal cycle at 25N:

0.2Pw = 3Sv Also at 6-8 days period:

2.5 Pw, 25 Sv

Robustness: DYNAMO models Böning et al, 2001

Seasonal variability of heat transport in the world ocean

Jayne and Marotzke (2001)

Total heat transport

Ekman contribution

Ekman transport and SAMW variability Rintoul and England, 2002

Variability of temperature and salinity at 45°S, 145°E (winter mixed layer) in a coupled model

Strong negative correlation between Ekman velocity and SST

Interannual variability of the meridional overturning at 45°N in the Atlantic

Beismann et al, 2002

Amplitude: 3 Sv

Robustness across models

Link with the wind stress field.

Full forcing ---- NAO-related forcing

NAO forcing of the meridional overturning

Time series of meridional heat transport at 48°N Eden and Willebrand,2002

Long time series of overturning at 48°N Eden and Jung,2002

heat flux only ---- all surface fluxes

Plan

1 - Knowledge of the ocean circulation

Discover and quantify

Understand processes

2 - Mechanisms of heat transport variability

Seasonal to interannual time scales

Decadal scales

What about eddies?

3 - Perspectives

Eddies in the meridional heat transport

30°S: 20-25% 45°N: 20-25%Tropics: mean/eddy compensation

Warm water flux at 30°S

Eddy contribution at 30°S

- The « classical » eddy flux v’h’ or v’T’: 20-25% of total no parameterization

- Even when the « classical » eddy flux is zero, translating eddies have an effect :modification of mean flow properties.

Oceanic heat transport variability: model insights

- Models reveal the importance of wind-forced variability:

Ekman transport on seasonal scales Robustness at interannual scales

- On decadal time scales, more complex processes importance of buoyancy fluxes

- Even with models, the effect of eddies is difficult to quantify.

Model development: topography, parameterizations…

More horizontal resolution

More vertical resolution

Future challenges

Spatial resolution of models (1)

Modelling the western boundary currents and deep reciculations.

1/6° model recirculation cell: 2Sv

Currentmeter data: a recirculation of more than 20Sv?(Weatherly et al, 2000)

Spatial resolution of models (2)

Increased spatial resolution has a strong influence on the pattern of anthropogenic CO2 uptake in the North Atlantic subpolar gyre

Anthropogenic CO2 flux in the subpolar North Atlantic (december 1989, in mol/m²/year). Arne Biastoch, 2002.

4/3° (100 km) FLAME model 1/3° (20 km) FLAME model

Models and climate change

Bi et al, GRL, 2002: increase of the ACC transport in a transient warming scenario

Banks and Woods, 2001: Possible use of SAMW variability to detect climate change

What resolution in the oceanto model and predict

anthropogenic climate change?