Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.


Moving windows to the deep ocean

Observations show that the Atlantic Meridional Overturning Circulation (AMOC) is slowing down, and this is predicted to continue in response to climate change. This isn’t the only change expected; tracing ocean circulation within a climate model now shows that the locations where water sinks to the deep ocean to feed the AMOC will also shift in the future.

Ever wondered why the weather in London is so much milder than in New York, even though London is more than 10° of latitude closer to the North Pole? This contrast in climate is thanks to the ocean circulation. Like a conveyor belt, warm currents carry heat north from the tropics across the North Atlantic towards Europe. Here they lose heat to the atmosphere before sinking and returning south. If the strength of this heat conveyor changes, it will have dramatic implications for climate in western Europe and across the globe. Writing in Nature Climate Change, Camille Lique and Matthew Thomas1 use a new method of tracing the ocean circulation to show how ‘windows’ where water sinks to the deep ocean in the North Atlantic might shift under a warming climate.

The Gulf Stream collects heat in the Gulf of Mexico, carries it across the North Atlantic Ocean and delivers it to western Europe. This gives Europe the gift of a relatively mild climate. When this water reaches the far north Atlantic, it becomes colder and saltier, increasing its density. This causes the water to sink to the deep ocean in narrow regions that serve as windows to depth from the surface. The sinking water spreads southwards in the deep ocean before returning to the sea surface. Taken together, these branches of ocean circulation create the AMOC. This oceanic heat pump has an enormous impact on Earth’s climate, but is also impacted by surface changes, as anthropogenic climate change is causing the Arctic to warm and sea ice to melt. Scientists think that this extra fresh water could prevent water from sinking to the deep ocean, causing the AMOC to slow — or even shut down.

Oceanographers have been intensely studying ongoing changes of the AMOC and how it will change in the future. Thanks to an array of moored sensors stretching across the North Atlantic, it is known that the strength of the AMOC has declined over the past two decades2. Records of past climate suggest that it is now the weakest it has been over the past few hundred years3. Looking forwards, climate models project that the circulation will be a further 25–30% weaker by the end of the century4,5.

Although there is general consensus that the AMOC will weaken, a lot is still unknown about how and why the AMOC is changing. One process linked to variations in the AMOC is the location of windows to the deep ocean, and how much water sinks through them. Scientists refer to these sites as AMOC source regions. It is necessary to understand how and where these source regions will change in the future to generate accurate climate predictions.

Lique and Thomas tackle this question using a new method that can determine the exact contribution from different source regions to the AMOC. The method releases particles at a fixed location in an ocean model. The model’s three-dimensional flow field is then used to trace particles backwards in time. The authors can thus identify the last place where the particles had access to the surface before they became part of the AMOC. This allows them to construct a geographical map of the relative contributions of each source region to the total strength of the AMOC. Using this approach, Lique and Thomas track the particles in two climate model simulations — one a control simulation representing pre-industrial climate conditions (before significant human-induced global warming), the other a high-emissions global warming scenario. By comparing the two, the authors are able to see how the source regions shift under a warming climate.

The results from the global warming scenario agree with earlier work4,5 (a weakened AMOC), but also identify a clear shift in the main source regions of the AMOC. In the control run, the main source for the AMOC is in the subpolar regions, consistent with our current understanding from observations. Under strong global warming, the subpolar sources almost disappear (Fig. 1). At the same time, sources emerge further to the north in the Arctic Basin and further south in the subtropics. The weakening of the subpolar source region is explained by an increase in the density difference between water at the sea surface (fresher) and the water below; this traps water at the surface, preventing it from sinking to the deep ocean. The emergence of a source region in the Arctic Ocean is also attributed to changes in the surface ocean circulation in the region.

Fig. 1: Shifts in AMOC source regions.

Present-day (yellow) and projected future (red) locations of deep water production in the North Atlantic.

It is well established that the strength of the AMOC is likely to change in the future. These new results imply that the mechanisms involved in setting the strength and structure of the AMOC will also change. It is a reminder that we must fully consider the ocean circulation in three dimensions to make accurate predictions of future climate. Unfortunately, limited computing power means that compromises must be made to run climate simulations far into the future. Many fundamental processes involved in the three-dimensional ocean circulation (for example, fresh water inputs from the melting Greenland Ice Sheet) aren't captured in the climate model used in this study. However, the model used by Lique and Thomas is amongst the best currently available for this purpose. It remains an open question whether the changes shown here can be reproduced in a higher-resolution model.

If the AMOC continues to change under a warming climate as predicted, it is uncertain how this will impact the ocean’s absorption of both carbon and heat. New source regions for the AMOC, as identified by Lique and Thomas1, would probably be exposed to different heat and carbon exchange rates with the atmosphere. The discovery of these shifting windows to the deep ocean brings with it more questions. We must continue research in this vein to reveal the full implications of these windows to the deep ocean for future climate.


  1. 1.

    Lique, C. & Thomas, M. Nat. Clim. Change (2018).

    Article  Google Scholar 

  2. 2.

    Caesar, L., Rahmstorf, G. F. S., Robinson, A. & Saba, V. Nature 556, 191–196 (2018).

    CAS  Article  Google Scholar 

  3. 3.

    Thornalley, D. J. R. et al. Nature 556, 227–230 (2018).

    CAS  Article  Google Scholar 

  4. 4.

    Schmittner, A., Latif, M. & Schneider, B. Geophys. Res. Lett. 32, L23710 (2005).

    Article  Google Scholar 

  5. 5.

    Cheng, W., Chiang, J. C. & Zhang, D. J. Clim. 26, 7187–7197 (2013).

    Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Veronica Tamsitt.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Tamsitt, V. Moving windows to the deep ocean. Nature Clim Change 8, 941–942 (2018).

Download citation


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing