Personal tools

You are here: Home / Science and society / Science for all / Just published! / The great meridional circulation in the Atlantic: stable or bi-stable?

The great meridional circulation in the Atlantic: stable or bi-stable?

Simulations of the world ocean at an unprecedented resolution show that the increase of the freshwater input in the Atlantic could make the oceanic circulation shift from a stable state to another, a result that climate models fail to reproduce.

Bandeau Deshayes.jpg

One of the main processes of the ocean-atmosphere interaction is the meridional overturning circulation (MOC). This planet-wide loop of ocean currents is sometimes called thermohaline circulation because it partly depends on differences of temperature and salinity (thence of density) between water masses: as two water masses with different density cannot lie at the same depth, the densest must flow downwards below the lightest. The heterogeneity of heat exchanges at the ocean surface (the sun brings more heat at the equator than at the poles) causes instable horizontal contrasts of density. As a result, the MOC transports the warm waters of the equatorial Atlantic to the north, where they mix with the surrounding cold waters, whereas the latter flow at greater depths in the opposite direction, from North Atlantic to the equator and to the other oceanic basins.


Schematic representation of the MOC (surface currents in red, bottom currents in purple, intermediate-depth currents in blue; L: Labrador Sea, N: Norwegian Sea, R: Ross Sea, W: Weddell Sea)



MOC fluctuations are of great interest to scientists because in the past, and particularly for the last 30,000 years, they caused climate alternations between temperate (strong MOC) or glacial (weak MOC) periods. These climatic transitions are related to salinity in the North Atlantic: an excess of freshwater makes cold water masses lighter and compensates for the horizontal temperature gradient, thus slowing the Atlantic MOC (AMOC), and vice versa. There are threshold effects in the relation between salinity and AMOC, especially within a so-called "bi-stable" regime where both AMOC states co-exist; then the transition between strong and weak regimes is abrupt.

All climatic models forecast that the increasing precipitations and the melting of Greenland ice sheet will provide an increasing amount of freshwater into the North Atlantic, which will cause a decrease of the AMOC. But most models assume that the AMOC is in the mono-stable intense regime, thus precluding an abrupt collapse, whereas the retrospective analysis of observations using models suggests that present-day AMOC is in the bistable regime. As a result, climate models have been reported as "overly stable".

Glacier Svalbard.jpg
Arctique glacier (Svalbard) © CNRS/N. Morata

To confirm the present regime of the AMOC, the scientists measured the freshwater balance of the Atlantic at its opening towards the Southern Ocean at 30°S, which is an indicator of the existence of the bi-stable regime. Prior studies have shown that when AMOC is under this regime, it imports (northwards) saltier water and exports (southwards) fresher water. To estimate this transport in conditions as close as possible to the real ocean, they used a high-resolution model whose elementary mesh size is 1/12 degree (i.e. less than 10 km); this makes it possible to describe the dynamics of oceanic eddies, which is largely ignored by lower resolution approaches such as climate models. This unique model was developed through the collaboration of French, German and British partners within the DRAKKAR consortium.

The results show that at 30°S sea water is saltier at the surface (salinity higher than 35 g/l down to about 500 m), with a maximum in the western part of the ocean basin. At this latitude, currents mainly flow northwards, except from the surface to the bottom along the western boundary and below 1000 m throughout the basin.

Salinity along 30°S in a high-resolution ocean model: the pink line shows the contour of reference salinity (35 g/l))



Current velocity along 30°S in a high-resolution ocean model: the white horizontal line shows the depth at which meridional circulation changes direction (northwards above, southwards below)

These features reflect the existence and the movement of different water masses: the salty surface waters (SW) which are involved in the subtropical gyre, the fresher Antarctic intermediate waters (AAIW), the deep waters formed in the North Atlantic (NADW) and the Antarctic bottom waters (AABW) flowing on the ocean floor. As a result of current and salinity contrasts, the vertical profile of freshwater transport changes direction according to depth, but the overall balance is negative in all simulations, which means that in the present conditions the AMOC exports freshwater southwards. However the volume transported is highly variable at different scales, and the flux can even change direction.



Schematic distribution of water masses along a N-S profile of the Atlantic Ocean (SW: Surface Water; AAIW: Antarctic Intermediate Water; NADW: North Atlantic Deep Water; AABW: Antarctic Bottom Water)


Four independent simulations of the same high-resolution model lead to similar results in terms of volume transported and of seasonal and annual variability; this confirms the robustness of the results in spite of the uncertainties of these numerical experiments (mainly about evaporation and precipitation fluxes). The direction of this flux (southwards) suggests that the present MOC is in a bi-stable regime, a necessary condition for an abrupt change to occur if freshwater influx increases in the North Atlantic. The bi-stability of present MOC is consistent with the observations and retrospective ocean analyses, but differs from most climatic models (some components of which could suffer from bias in evaporation/precipitation fluxes or western boundary currents). These models then could overestimate the stability of the MOC and the present climate.

These first high-resolution simulations confirm that decreasing the mesh size enables to improve the realism of the MOC in the model, but could lead to climate projections significantly different from the outputs of present-day climate models. Along with numerical simulations, the international project SAMOC (South Atlantic Meridional Overturning Circulation) will bring new data in this key region, thereby allowing a better validation of the freshwater transport and the MOC in ocean and climate models.


The paper

Deshayes J., Tréguier A.-M., Barnier B., Lecointre A., Le Sommer J., Molines J.-M., Penduff T., Bourdallé-Badie R., Drillet Y., Garric G., Benshila R., Madec G., Biastoch A., Böning C. W., Scheinert M., Coward A. C., Hirschi J. J.-M., 2013. Oceanic hindcast simulations at high resolution suggest that the Atlantic MOC is bistable, Geophysical Research Letters, 40 : 3069–3073.

See the first page


The authors

The 17 co-authors are researchers from French laboratories: Laboratoire de Physique des Océans (IUEM-Plouzané and LMI ICEMASA-South Africa, Laboratoire de Glaciologie et Géophysique de l'Environnement (LGGE, Grenoble), Laboratoire des Ecoulements Géophysiques et Industriels (LEGI, Grenoble),Mercator Océan, Laboratoire d'océanographie et du climat – expérimentations et approches numériques (LOCEAN, Paris) and from oceanographic centres GEOMAR (Kiel, Germany) and NOC (Southampton, Great Britain).


The journal

Geophysical Research Letters is published by Wiley on behaof of the American Geophysical Union; it is one of the most cited scientific journals, in particular in the field of climate change. It covers the whole range of domains and disciplines pertaining to geophysics and its aim is to publish as quickly as possible short papers ("Letters") likely to bring a major contribution to the scientific community.


Authors : browse IUEM staff directory
Communication and outreach service:

Back to publications list

Picture of the month