Discovery: Wind-blown Antarctic sea ice helps drive ocean circulation

Study first to quantify influence of Antarctic sea ice

Ocean waters circulate globally, rising in some regions and sinking in others.CREDIT L.D. Talley et al,, Oceanography, 2013

Ocean waters circulate globally, rising in some regions and sinking in others.CREDIT L.D. Talley et al,, Oceanography, 2013

Antarctic sea ice is constantly on the move as powerful winds blow it away from the coast and out toward the open ocean. A new study shows how that ice migration may be more important for the global ocean circulation than anyone realized.

A team of scientists used a computer model to synthesize millions of ocean and ice observations collected over six years near Antarctica, and estimated, for the first time, the influence of sea ice, glacier ice, precipitation and heating on ocean overturning circulation. Overturning circulation brings deep water and nutrients up to the surface, carries surface water down, and distributes heat and helps store carbon dioxide as it flows through the world’s oceans, making it an important force in the global climate system. The scientists found that freshwater played the most powerful role in changing water density, which drives circulation, and that melting of wind-blown sea ice contributed 10 times more freshwater than melting of land-based glaciers did.

A vital contributor to the process, the scientists discovered, was the seasonal migration of the ice, which is largely driven by winds. If the sea ice were instead forming and melting in the same place, there would be no net effect.

“If you were to turn off these winds and eliminate that pathway for moving sea ice away from Antarctica, you would probably significantly reduce the strength of the overturning circulation,” said lead author Ryan Abernathey, an oceanographer at Columbia University’s Lamont-Doherty Earth Observatory.

The study, published this week in the journal Nature Geoscience, uses a sophisticated approach to examine on the complex problem of what is happening down under the ice, where direct observations are hard to come by. It provides new insight into the basic physics of the ocean that may be critical for answering future questions about climate change, such as how loss of sea ice or changing winds could affect global ocean circulation, said Abernathey.

“Everyone is asking, is sea ice expanding or contracting? We’re coming at it from a different perspective: What does sea ice do to the underlying ocean?” Abernathey said.

When sea ice forms around the edges of Antarctica each winter, the salt in the ocean water doesn’t freeze; it stays behind. That makes the water near the coast much saltier and therefore denser than water off shore. Denser water sinks, and in doing so pushes less dense water up, driving circulation. Meanwhile, as sea ice melts farther out in the open ocean, it deposits its less-dense freshwater, moving denser water down.

Scientists have known for some time that changes in water density, particularly the sinking of cold, saline water, contribute to the ocean’s “abyssal circulation,” the deepest, coldest branch of the ocean conveyor belt, which moves cold Antarctic water northward along the ocean bottom. What has been less well understood is the role salinity might play in the “upper circulation,” which carries mid-depth water up to the surface in the Southern Ocean and eventually toward the tropics.

Using an analysis technique called water-mass transformation, the scientists were able to quantify the rate at which ice freezing and melt contribute to the upper circulation by making water near the coast denser and water in the open ocean lighter.

Ocean circulation is critical to the climate system because it distributes heat and helps store carbon dioxide in the deep ocean. Major climate changes in the past, including glacial periods, are believed to have involved changes in ocean circulation. To understand how circulation may be changing today, the next steps will be to look more closely at how salinity and wind speeds changed in the past, Abernathey said.

“This work shows really clearly that Antarctic sea ice plays a crucial role in the circulation of the world’s oceans,” said coauthor Paul Holland of the British Antarctic Survey. “We have known for many years that the freezing of Antarctic sea ice in winter is responsible for forming the very deepest waters in the world oceans, but this study shows that melting the ice in summer also governs the formation of shallower waters. This advance has only been made possible by the state-of-the-art computer model used in this study, which assimilated millions of ocean observations.”



50 thoughts on “Discovery: Wind-blown Antarctic sea ice helps drive ocean circulation

    • Yes, “which assimilated millions of ocean observations.” Not all models are bad.

      • Yeah, millions sound like a lot right?

        360*180=64800 , a one degree grid for one moment in time.
        let’s ease up a bit and go for a 5×5 degree grid

        10^6 / (64800/5/5) = 386

        so a million observations gets you about a years worth of daily data on 5×5 grid…. at one depth, eg surface.
        So the ” millions of ocean observations” is not going to get you far in characterising a 3D model of global ocean currents.

        The study, published this week in the journal Nature Geoscience, uses a sophisticated approach to examine on the complex problem of what is happening down under the ice, where direct observations are hard to come by. It provides new insight into the basic physics of the ocean

        Hmm does not sound like they have millions of obs to validate the model under ice. Sounds more like they are making it up using non validated models and think that non validated models inform up about “the basic physics of the ocean”.

        .No, they tell us about the the basic physics of the world in the model. .

    • The reliability of a computer model depends on both its construction and the data that is put through it. The climate fear mongers don’t have models that can represent the actual climate at all — and the data they push through them has been given quite a massage beforehand.

    • IMHO we have to be careful to distinguish between a bottom-up “climate” computer model and a useful computer model. I can’t tell if this model is OK, but it could be.

      • Well the computer model should reproduce the actual measured values of observations made at the actual real earth measuring sites. That’s why you make measurements; to fit the model to the observed data.

        Of course if your measured data are not a valid Nyquist samples of the continuous function, then it can only give you total garbage. You have to sample at no more than half the spatial and temporal periods of the highest frequency in your band limited continuous function. Otherwise those samples are not even data; just noise.


    • Ice powered sailboats, instead of carbon fiber sails.

      We might see ice sails on the next Americas Cup hydrofoil sail boats.

      See, it IS the water and NOT the CO2


  1. Many of the ocean currents are illustrated with directions opposite to what was formerly known. For example, the western Pacific Kuroshio Current is part of a clockwise circulation, flowing northward by Japan, then eastward across the north Pacific toward North America. It is illustrated as counterclockwise circulation. The dashed line current off the west coast of South Africa used to be the Benguela Current in my old oceanography text.

    • “the western Pacific Kuroshio Current is part of a clockwise circulation”
      Good eye. The current as drawn looks like it picks up heat from eastern Siberia, flow past Japan, and dumps the heat in the tropical Pacific.

    • Yes, that struck me as odd too.

      I would be very surprised if a major current could go in the opposite direction as that determined by Corelis forces. Sadly the pretty pics which are reproduced here do not include a legend explaining what the colour scheme represents. So much of the information is useless.

    • While I’m always skeptical of maps showing density-driven currents unsupported by observations in situ, it’s apparent that Figure 1 shows not the well-known surface currents mentioned, but the snail-paced subsurface “overturning” circulation.

    • Well it is quite simple to reverse the ocean currents, and make the counter clockwise.

      All you have to do is simply reverse the rotation of the earth, so it goes from East to West instead of verse vicea.

      That’s what causes the NH clockwise currents. Tidal bulges build up on the equator; kinetic energy is converted to potential energy of increased water height, and when it runs into land the accentuated bulge must collapse, and the excess water go poleward to where gravity is higher, and centrifugal force lessened.

      All that ” thermo- haline ” circulation is just so much poppy cock. The earth’s rotation causes the circulation.


      PS when in history did the gulf stream last run down the US coast, instead of up it ??

  2. “This work shows really clearly that Antarctic sea ice plays a crucial role in the circulation of the world’s oceans,” said coauthor Paul Holland of the British Antarctic Survey.

    So there were questions before?

  3. Not sure how they differentiate between cause and effect. Couldn’t it also be that the changes in the deeper ocean circulation drive the flow of Antarctic currents and sea ice? In terms of overall mass it seems they have the tail wagging the dog.

    • Surface currents are mainly wind driven and even the 90% of ice below the water is still “surface” in that context.

      Ocean currents are orders of magnitude slower than wind driven surface currents.

      • The winds in a tornado are an order of magnitude greater than normal winds. I don’t of anyone who thinks they are of “vital importance” to atmospheric air flow.

        All I’m saying is there are much great forces involved here and simply because they built a model, that most likely matches their preconceptions, doesn’t inspire a lot of confidence in me. Maybe the paper is more conservative than the press release.

    • Kind of what I was thinking – it’s the wind driving the bergs. Not so much the ice influencing current as it is the wind.

  4. The take away, is we are making critical decisions with incomplete data and understanding. There is NO emergency current, that justifies making such decisions based on first impressions from incoming data. We need verified models and a much better understanding than we have today before taking our hands out of our collective pockets. Ideology, by itself, will only lead to disaster. GK

  5. What about the cooling effect of putting ice cubes in your tea? It would seem that if more ice is being pushed around by the wind, that some ocean cooling must also take place?

    • If ice moves from land to sea and melts it will cool the sea.

      If it moves around on the sea it will cool sea A for sea B but the net cooling effect will be the same.

  6. “Everyone is asking, is sea ice expanding or contracting? We’re coming at it from a different perspective: What does sea ice do to the underlying ocean?” Abernathey said.

    Why in modern climate papers do they always feel the need to put this sort of phrase in? It’s BS

    • “It’s BS”

      Thanks for such a profound and enlightening rebuttal. I wish I had got there first.

      Maybe you could expand on WHY in YHO it is BS.

      There was much wailing and gnashing of teeth about ASSUMED effects of changes in sea ice coverage in the Arctic, which predicted a ‘death spiral’ and ice free Arctic by 2010, 2012,2013….. still waiting.

      Looks like previous assumptions were wrong. “What does sea ice do to the underlying ocean?” Seems like a damned good thing to be studying to me.

      • Don’t know about BS, but it strikes me as pablum.
        When I’m reading a paper, I really don’t care what “everyone is asking”, what I want to know is what did they do.
        For me such verbiage detracts from the seriousness of the paper.

    • The earth is NOT a sphere. X^2 + y^2 + z^2 = R^2 IS a sphere.

      There are no real spheres, only fictitious mathematical ones.


  7. “melting of wind-blown sea ice contributed 10 times more freshwater than melting of land-based glaciers did”

    That must be an extreme under-estimate since, except for the Antarctic peninsula, land-based glaciers in Antarctica don’t melt at all. They calve into the ocean and melt there together with the sea-ice. But perhaps this is what they actually mean? The people who write the press releases usually don’t have the faintest idea what they are writing aboutet.

    • Found a pre-print of the paper:

      And yes, “ice melt” refers to icebergs melting at sea. I found something rather more interesting though. Apparently this wonderful computer model of theirs completely fails to simulate the generation of Antarctic Bottom Water (AABW) in coastal polynyas. Since AABW is the most important driver of thermohaline circulation in the southern hemisphere I would suggest that their results should be considered extremely dubious.

      • To say nothing of all the mid water currents in between the surface and AABW.

        While it is interesting and another speculation into the question of ocean circulation, once again the mid level currents are completely ignored or standardized in order to study 1 aspect. Until we understand the role of the mid level currents we can run as many computer models with as many surface data points as we wish…but it only refers to the surface. Once below that, it is like the edge of the flat Earth maps…there be monsters here.

  8. The extent to which people in this science just make stuff up is astounding.

    I guess when you have 10,000 scientists each trying to publish and get ahead in this make-stuff-up science, you get a lot of make-stuff-up – some sillier than others..

  9. I’d be wary of the true underlying reason for this paper. As has been seen before, climate scientist will create a paper for the sole purpose of it being used for citation in another paper that predicts some future calamity.

    By saying that wind blown antarctic sea ice helps drive ocean circulation, they are in essence implying that a loss of antarctic sea ice from CAGW will decrease ocean circulation. Some other climate scientist will take this and make it central to the aforementioned calamity.

  10. The sub-heading to the paper should be “Unless you are on the Ship of Fools” as they found the lack of circulation an embarrassment

  11. Only the rich or very well connected people will be on board(the Richard Branson, Tom Styer types). I hope that Al Gore is on board when they strike an iceberg which throws Al overboard where he is saved from drowning by a hungry polar bear

  12. Quick question for all you clever types out there. I’ve always taken sea ice to be seawater that has been frozen. If this is indeed the case then could someone explain how wind-blown sea ice contributes 10 times more freshwater than melting of land-based glaciers. Or have I missed something ?

    • Sea ice is ice on the sea…meaning freshwater ice floating ontop of saltwater. The simplest visual is that salt is kicked out before the water freezes.

    • Sea ice is fresh water. Something called a segregation coefficient determines how much of a dissolved impurity goes into each phase. Both the salts, and the CO2 much prefer to stay in the liquid phase, rather than try to squeeze into the solid phase.

      This phenomenon is commonly used to purify materials by freezing them, sometimes in an iterative freeze and re-melt cycle.

      We made 99.99999 % purity gallium, out of scrap gallium arsenide junk from LED manufacturing, suing a freezing process (along with other chemical means.). We supplied 50% of our total gallium usage by recycling scrap (from sawing and other losses).


      • Thanks Jenn, Phil and George. It still sounds like its misnamed. Wouldn’t it just be easier to call it ice and be done with it!

      • If you think of a crystalline substance (water ice or the other ice known as diamond), the atoms are typically arranged in some organized structure or lattice, with somewhat fixed spacings between the atoms.
        An impurity atom or molecule , say NaCl or CO2 inside such a cage, could try to fit in the space between the crystal atoms (or molecules) often called ” intersticial ” meaning in the interstices or spaces, or it could oust one of the crystal atoms or molecules, and take its place in the structure (or try to). If it’s big enough it can break the crystal locally and cause some sort of “dislocation”.

        Impurities that ” substitute ” in the lattice generally stay put at least at modest temperatures. Interstitial impurities can wander around, and keep changing the properties of the structure.

        In a liquid, the atoms or molecules aren’t quite so tightly bound, and can accommodate impurities somewhat easier. In a gas, it is pretty much a free for all.

        So when water freezes, a whole lot of stuff that may be quite comfortable in the liquid, may just not fit in with the crystalline structure of ice, and that is why they prefer to remain in the liquid (segregate) and the ice is mostly fresh water.

        You have to be carful, because the crystalline ice is very fractured, often because of impurities, so it can form open pockets in which salts or gases could accumulate. So sea ice, if you crush it and melt it, could in fact contain a good amount of brine, that is in pockets, and note really “inside” the ice crystal structure.

        Well this is a somewhat simplified exposition for the lay reader. The Quantum Mechanics can add barnacles all over it to make it look much nastier. Well they get their kicks their way !


  13. This is probably too late for anyone to notice, but I’d look for a surge of grant applications to search for annual changes in any characteristics of known currents. I’d expect most of them to be checking currents in tropical areas near popular resorts. I’d also expect no surge in reported observations. I hope I’m wrong on both predictions (oops, make that projections.)

Comments are closed.