State of the Planet

News from the Columbia Climate School


Deep Ocean Heat Is Melting Antarctic Ice

Like dirt swept under the carpet, some of the human-made heat produced over the last century has been getting soaked up by the world’s oceans, and sinking into deep waters. Now, it is coming back to haunt the surface, in a very sensitive place: western Antarctica, where vast ice sheets meet the ocean. The result appears to be that ice is rapidly being eaten from the bottom, says Douglas Martinson, a polar scientist at Lamont-Doherty Earth Observatory, who presented the findings Monday at the fall meeting of the American Geophysical Union.

Warm waters carried by the Antarctic Circumpolar Current are brushing the ice front in the western part of the continent, in the area of the Bellingshausen Sea.
Warm waters carried by the Antarctic Circumpolar Current are brushing the ice front in the western part of the continent, in the area of the Bellingshausen Sea. (Click to enlarge. Courtesy Douglas Martinson, Lamont-Doherty Earth Observatory)

Martinson said that heat stored in deep waters far from Antarctica is being pushed southward and becoming entrained in the Antarctic Circumpolar Current, a vast, wind-driven water mass that constantly circles the frozen continent. The evidence comes from 18 years of Antarctic voyages Martinson has made to measure water temperature, salinity and other qualities at different depths. He called the increases in ocean heat in the past few decades “jaw dropping.” Temperatures have risen only a few degrees above the melting point–but that is all it takes to cut at the ice front. “This is like a huge freight of hot coals–fresh, hot water being delivered right to the the front door,” he said.

This raises the specter of sea-level rise driven by melting in this region–but there is a larger implication, said Martinson. Even if all sources of human-produced carbon dioxide in the air were cut off, the built-up heat will remain in the ocean for many years to come. “Pretend your brains out that the politicians did something to stop global warming tomorrow.  Even if they did, we will still have decades and decades of upwelling of that warmed water eating ice,” he said.

Read a Discovery News article about Martinson’s talk.

Columbia campus skyline with text Columbia Climate School Class Day 2024 - Congratulations Graduates

Congratulations to our Columbia Climate School MA in Climate & Society Class of 2024! Learn about our May 10 Class Day celebration. #ColumbiaClimate2024

Notify of

Inline Feedbacks
View all comments
13 years ago

This may be a naive question, but could you explain how warm water sinks in the first place?

Doug Martinson
Doug Martinson
13 years ago

Jason — Your comment is not naive, it is very common. When we say “warm” water in polar regions, we are talking relative to the freezing point. So in the above report, warm refers to waters that are 3.5 to 4 degrees warmer than freezing. That is blazing hot for ice (water will melt over 4300 times more ice than air given air and water of the same volume and temperature). But, this hot water is still so cold I wear special scuba gloves when sampling it. Anyway, in most of the oceans where the water temperature is actually warm as you and I would think when touching it, the density of water is controlled by temperature, but in cold regions it is controlled by salinity. So, my “warm” water that is really cold, sinks when it gets saltier (it gets saltier when sea ice forms, since that process only freezes the pure water trapping the ocean salts in a dense salty brine that drains into the ocean raising its salinity and increasing its density so that it sinks). Hope this clarifies this reasonable question for you…

David Lewis
David Lewis
13 years ago

I’m trying to understand what scientists know about what the changes in the Southern Westerlies mean for the Antarctic region and for global climate change.

I’ve heard Joellen Russell vividly describe that the intensification of these winds and the movement of their average location towards the South Pole means that the ACC will also strengthen which will increase the upwelling of deeper warmer water. She says that the melting you report on the Antarctic Peninsula as well as growth of sea ice where its coldest in the Ross Sea are exactly what she expects to see if the ACC is causing more upwelling.

The global effect would be, according to her, that the increased ocean mixing done by the ACC will tend to counter the effects of ocean stratification that will be caused by an increased hydrological cycle elsewhere and cause more heat to be stored in the oceans than otherwise, perhaps enough to put a bit of a brake on the rising trend in the global average surface temperature chart. She also says modelling shows more CO2 ends up in the ocean over decades and centuries than would otherwise be the case.

This could cause a political effect – a lot of attention and debate centers around ups and downs in that global average surface temperature chart and a new drag on it may add to the forces of doubt in some minds about whether action to limit the accumulation of greenhouse gases is necessary. As you say, the heat will come back to haunt us, whether more of it is ending up in the ocean as Russell says, or not.

There is the paper by Boning, who in “The response of the Antarctic Circumpolar Current to recent climate change”, which says his group examined Argo float data and they feel “the transport in the ACC and meridional overturning in the Southern Ocean are insensitive to decadal changes in wind stress”.

I see your report describes heat “stored in deep waters far from Antarctica” is “pushed” there, presumably by some other currents, where it is “entrained” by the ACC and ends up melting your ice. You say nothing about whether you think the ACC has strengthened, or if the upwelling water is getting further towards the continent because the force driving it has more power.

Can you clarify or add more detail? Thank you.

13 years ago

Doug: I assume this is primarily effecting shelf ice, which is small compared to the sheet itself. Not to minimize the situation in the least, but will greater shelf melt effect the rate of sheet advance or melt? What kind of volumes are we talking here and what kind of ocean level (MSL) change could we see in 10-20 years as result?

Doug Martinson
Doug Martinson
13 years ago

David and Barry —

First I want to thank you (and Jason previously) for your informed and thoughtful comments/questions. I am going to answer your questions/comments in context by giving a more detailed overview of my AGU presentation, which is briefly summarized in the report by science reporter Kevin Krajick. Obviously his summary left out unnecessary details, some of which directly speak to your questions, hence my approach of a more detailed summary.

My talk was motivated by glaciological observations that the accelerated glacial melt in the Western Antarctic Ice Sheet (WAIS) is primarily a consequence of warmer ocean water coming in contact with the base of the major ice streams that drain the WAIS as they flow into the ocean. As they enter the ocean, these streams form shelves that spread out and eventually encounter geographical features that buttress the flow, slowing or even stopping the drainage of the ice sheet; melting of the shelves keeps the ice stream flowing, and increasing that melt increases the flow – perhaps many times faster. As the physical oceanographer working on the western Antarctic Peninsula in the Palmer Long Term Ecological Research (LTER) project, downstream of the WAIS, I have an unprecedented gridded ocean record of 18 years for which I have been documenting the ocean heat. Historical data in the same location shows that the ocean heat content has increased approximately exponentially since the 1960s (see Martinson et al., 2008). The warm water carrying the heat is Upper Circumpolar Deep Water (UCDW); it is circulated around the Antarctic by the ACC, and because of the proximity of the ACC adjacent to the continental margin in the western Antarctic (and only in this region), this is the water accounting for the ocean heat in both my area along the peninsula and the WAIS melt region. These observations provide support for increased ocean heat observed in my region is the same as that driving accelerated WAIS glacial melt. The point of my presentation was to discuss the source of this increased heat content in an effort to better assess the future of accelerated glacial melt (and sea level rise). There are 3 (not independent) potential mechanisms for the increased ocean heat: (1) stronger westerlies driving increased tilt of ACC isopycnals (bringing the warmer deeper waters to the level they can be upwelled onto the continental shelves) or enhanced eddy production fighting the increased tilt (eddies are the primary mechanism for modern-day delivery onto the continental shelf) or variants of this, (2) advection of warmed global deep waters to the ACC, and eventually to the western Antarctic, (3) increased coastal upwelling.

The first mechanism related to the stronger westerlies seems to be an obvious conclusion (accepted almost universally, and it was certainly my solution of least astonishment = Occam’s razor). But, while accepted at the gut level by most, many of us worried that the tilt of the isopycnals of the ACC, the dynamic response to increased wind forcing driving increased transport, may have already reached its limit, since the increased tilt generates increased instability in the form of eddies that tend to flatten the tilt. At some point the tilt would likely achieve a maximum, for which any increased wind stress will simply go into more eddy generation and no more tilt or increased transport. The Boning et al. study addressed this explicitly through models and data. They find no evidence of increased tilt or transport over the past decades, but there is more enhanced eddy production. So, based on their findings, I dropped this as a primary mechanism, and left it that maybe the increased eddy production may lead to more eddy heat flux onto the continental shelf (something we are now evaluating). Mechanism #2 seemed to be a longshot, but I was motivated by the important studies of Barnett et al. (2001 and 2005) showing that if one adds a global ocean model to the base of the global climate models that a huge fraction of the heat entering the atmosphere is absorbed by the ocean, eventually making it to the longer term deep ocean mobile reservoir (recent estimates suggest that 90% of the heat added to the atmosphere during the decades of global warming has ended up in the ocean). Barnett’s work also showed why the global climate models were overestimating the amount of atmospheric warming. If the deep oceans did warm, we know that there are southward currents that would advect some of this heat to the ACC, and from there to the western Antarctic. Finally, #3, local coastal upwelling is clearly an active mechanism today as we document in my 2008 paper where it accounts for approximately 84% of the heat content rise in the 1990s and early 2000.

Preliminary conclusions: (1) based on the Boning et al. work, increased tilt of ACC by stronger westerlies plays reduced, if any, role. (2) The recent NOAA State of the Planet 2009 report shows the result of 8 independent studies analyzing the heat content of the global oceans — these all show more-or-less the same approximate exponential warming as we see in my western peninsula observations. So, this mechanism must surely contribute to the west Antarctic increased ocean heat content since there is no question that some fraction over some time scale will be advected to the ACC. The question now is: how much of the deep ocean heat will be vented to the south, and over what timescales and via what paths. From that we can estimate how long we might expect accelerated glacial melt. And, the scary implication is that even if somehow global warming was stopped soon, we are likely to have decades of this heat moving south and continue the accelerated melt and sea level rise. (3) Local coastal upwelling is also important. So these conclusions suggest that we must now determine what fraction of the increased heat content can be attributed each of the contributing mechanisms in order to apply our warming to that of the WAIS primary melt region (e.g., coastal upwelling on the open shelf of the western Antarctic Peninsula is not likely to account for any warming in the WAIS melt region since that warm water must move under the ice shelves via mechanisms not related to coastal upwelling). And of course, the inevitable final caveat (this is what drives politicians crazy): the climate is undergoing a change in forcing and as a consequence, the Earth’s climate is trying to find a new equilibrium state. This will come through feedback mechanisms kicking in, as well as perhaps changes in local interactions or relative roles of currently active mechanisms.

Here are some points I did not address above:

Barry — I’m not sure what kind of sea level rise we are talking about because the glaciologists are now trying very hard to relate glacial melt rates to ocean temperature before it enters the ice shelf cavity. At this time, I am simply supplying them evidence that their mechanism of accelerated glacial melt is from enhanced ocean heat content is consistent with observations, and the extent and potential source of this increased heat content may even be such as to be alarming.

David – If you want more specific comments to each paragraph in your comments, contact me directly and we can discuss that (perhaps via phone).

Hope this clarifies the science for you, if not, don’t hesitate to contact me directly…

12 years ago

This is one of the most worrying studies I’ve read so far. How quickly does it change? It is accelerating as I presume? But in what way, smoothly, or do you find it able to ‘jump’?

Combine it with Hansen’s thoughts about the West Antarctica, and its geology, will it lead to the ice sheets able to ‘break of’?

And what about the under ice flows lubricating the inner of Antarctica? Will they too speed up? Or they won’t be affected?