Skip to main content

Half of Antarctic ice shelves could collapse in a flash, thanks to warming

Cracked surface of Larsen b ice shelf
(Image: © Shutterstock)

Antarctic ice shelves can disappear astonishingly fast — sometimes in minutes or hours — as meltwater surges through cracks in their surface. And as the atmosphere warms, this phenomenon may become more commonplace; at least half of the ice shelves on the continent are vulnerable to this process, a new study suggests.

These floating ice sheets ring Antarctica's glaciers and prevent them from sliding into the ocean. Without these icy barriers, glaciers would flow more quickly into the water, causing the continent to shrink and accelerating sea level rise.

The new study, published today (Aug. 26) in the journal Nature, suggests that about 50% to 70% of ice shelves that hold Antarctic glaciers in place could become weak and potentially collapse with surges of meltwater.

Related: 10 signs that Earth's climate is off the rails

"What we find is that the amount of melting is important, but where the melting happens is also important," said lead author Ching-Yao Lai, a postdoctoral researcher in the Department of Marine Geology & Geophysics at Columbia University's Lamont-Doherty Earth Observatory in New York. The research highlights which ice shelves are most likely to crumble, but not when this dissolution is most likely to occur. 

"The time frame over which this process could happen is the biggest question," Christine Dow, the Canada Research Chair in Glacier Hydrology and Ice Dynamics at the University of Waterloo, who was not involved in the study, told Live Science in an email.

Cracks in the ice 

Some ice shelves float on open water and do not slow glaciers' slide into the sea, because there's no land mass for them to brace against. But ice shelves confined to bays and gulfs create a physical barrier that the slow-moving glaciers butt up against. Lai and her colleagues focused on these so-called buttressing ice shelves.

The buttressing shelves accumulate cracks on their surfaces as glaciers push against them from behind, and cracks also appear as the shelf pushes against the curvature of the shoreline. In addition, the ice shelves stretch out as they flow across the ocean water, due to their velocity, Lai told Live Science. "Usually near the front of ice shelves, it gets stretched the fastest and this kind of breaking occurs," she said. 

When warm atmospheric conditions cause the surface of an ice shelf to melt, meltwater can pool in these cracks and cause "hydrofracturing" — a process in which the excess water puts pressure on the ice, deepening the cracks and sometimes triggering a complete collapse of the shelf.  

The melt water "can punch through the ice to the ocean in a matter of minutes to hours, as long as there's enough water available to keep on filling the crevasse and keep up the pressure," Dow said. "The crack in the ice then fills up with ocean water," and the shelf may begin to break apart. Scientists theorize that this is what happened to an ice shelf known as Larsen B, which lost 1,255 square miles (3,250 square kilometers) of ice over the course of a few weeks in 2002, according to The National Snow and Ice Data Center. For context, that area of ice is larger than the state of Rhode Island. 

Related: Photographic proof of climate change: Time-lapse images of retreating glaciers

To determine which buttressing ice shelves are vulnerable to collapse, Lai and her colleagues developed a machine learning model — an algorithm that can be trained to recognize visual features based on past images it has analyzed. The authors trained their model to recognize surface fractures in the ice using satellite images of two ice shelves, named Larsen C and George VI, in the northwest region of Antarctica, and then applied the model to a complete map of Antarctica

With their map of ice cracks in hand, the researchers then determined which fractures in Antarctica's many ice shelves would be prone to hydrofracture, given the pressures exerted on them from surrounding land masses and their movement over the water. Researchers have used this analysis at specific ice shelves before, but "this is the first time it has been applied to the Antarctic as a whole," Dow noted.

The next big question is "how does this hydrofracture process contribute to the sea level rise?" Lai said. To find out, the team will need to pair their model of ice cracks with climate predictions and a model of how ice flows over the bedrock of Antarctica; this ice flows downslope into the ocean "just like rivers flowing downhill," she said. 

In the future, this type of research will help the authors determine how fast and how much sea level rise could change due to cracking ice sheets, given rising atmospheric temperatures, Lai noted. As of now, "estimating time scales on which these ice shelves might break up due to surface melting and hydrofracture is beyond the scope of their study," Alison Banwell, a research scientist at the Cooperative Institute for Research in Environmental Sciences (CIRES) at the University of Colorado Boulder, who was not involved in the study, told Live Science in an email. 

Though the researchers don't have a time frame yet, some scientists predict that climate change may drive massive hydrofracturing events within a matter of decades, according to a 2015 report in the journal Earth and Planetary Science Letters.

"Also, Antarctic ice shelves are currently losing most [of their] mass due to basal melting," or melting on the underside of the ice sheet, "in response to warming ocean temperatures," Banwell added. "It would be interesting to see which ice shelves are most vulnerable due to all three processes," meaning basal melt, surface melt and hydrofracture, combined. Basal melting could make ice shelves more prone to hydrofracture, as the shelves become thin and stretch more easily the more they melt, Lai added.

Originally published on Live Science. 

  • Broadlands
    The climate has warmed in the past. From ~1880 to ~1940. Did these Antarctic ice shelves start to collapse back then? After all the global climate has warmed less than one degree C overall. Or is this just a model of what might happen?
    Reply
  • Chem721
    Broadlands said:
    After all the global climate has warmed less than one degree C overall. Or is this just a model of what might happen?

    It appears that deep water up-welling is driving most of the melting of basal Antarctic ice. The ocean temperature there is much warmer than expected. Here is something that sounds rather alarming from the "Intro" (in the below article) :

    "The Antarctic Peninsula has been one of the most rapidly warming regions of the world during the twentieth century where ~75% of the ice shelves have already retreated over the past 50 years."


    Here is a link to that :

    "Ocean temperature impact on ice shelf extent in the eastern Antarctic Peninsula" (18 January 2019)

    https://www.nature.com/articles/s41467-018-08195-6
    While the temperature variations do not seem extreme, the effects on the ice sure seems apparent.
    Reply
  • Broadlands
    "The Antarctic Peninsula has been one of the most rapidly warming regions of the world during the twentieth century." An interesting quote. Most of the recent climate concern has been about so-called Arctic amplification...the fastest warming part of the globe.

    My quesation was this: The climate has warmed in the past. From ~1880 to ~1940. Did these Antarctic ice shelves start to collapse back then? Or at any time in the past?
    Reply
  • Chem721
    Broadlands said:
    My quesation was this: The climate has warmed in the past. From ~1880 to ~1940. Did these Antarctic ice shelves start to collapse back then? Or at any time in the past?

    Your persistence has paid off. And it has provided me with information not previously known. Below is a quote from my favorite source of information (that Nature article was found on one of their sites. Wiki is an excellent source for references to original papers.)

    The quote is from Wiki on "Retreat of glaciers since 1850*". Just the time period you are asking about, and no doubt for a good reason. You clearly know a lot about this sort of thing.

    "The Little Ice Age was a period from about 1550 to 1850 when the world experienced relatively cooler temperatures compared to the time before and after. Subsequently, until about 1940, glaciers around the world retreated as the climate warmed substantially. Glacial retreat slowed and even reversed temporarily, in many cases, between 1950 and 1980 as global temperatures cooled slightly. Since 1980, a significant global warming has led to glacier retreat becoming increasingly rapid and ubiquitous, so much so that some glaciers have disappeared altogether, and the existence of many of the remaining glaciers is threatened. In locations such as the Andes of South America and Himalayas in Asia, the demise of glaciers in these regions has the potential to affect water supplies in those areas."

    end quote

    Clearly the growth and melting of glaciers has been significant in the last few centuries or so. No doubt you were aware of this. But the trend now appears to be melting beyond the cooling period accumulations, and now we are going into a new phase of melting. Or so the data seem to indicate.

    If you can tolerate all the data, this is very revealing :

    * https://en.wikipedia.org/wiki/Retreat_of_glaciers_since_1850
    Reply
  • mogoso
    90 % of the sea ice shelves breaks off yearly never a boost to sea level. Think on it put some ice in a glass with some water mark the level. When the ice melts does the water in the glass rise, does it over flow. Displacement, it stays the same. Give me something tangible to believe in on man made climate change, other than we need a tax administered by the UN who backs this climate change hocus pocus
    Reply
  • Whitefeather
    Tbe temperature on Earth has consistently but slowly varied from one extreme to the other about every 40,000 years and nature (living things) has accommodated the change by moving in the case of herbivores alternately between the equator and the arctic circle. The big difference now is that our use of fossil fuels has accelerated the change so it is happening much faster than we can cope with. Climate change deniers frequently use the fact that there have always been these swings, to spread misunderstanding and a sense of impotence. Unless we very quickly wake up and adopt a means of mitigating the chaos and disaster awaiting us, it will be too late. We have desploiled a wonderland with our greed and ignorance. Politicians don't help. Read my paper under climate change in live science. It offers an economically sound means to slow the rate of temperature increase down NOW not when it's too late!
    Whitefeather
    Reply
  • daveburton
    The article says, "These floating ice sheets ring Antarctica's glaciers and prevent them from sliding into the ocean."
    That's gibberish:
    Ice sheets don't float.
    Antarctica's glaciers would not "slide into the ocean" if the adjacent floating ice shelves floated away.
    Most readers probably:
    Do not know the difference between "ice shelves" and "ice sheets." In fact, neither does the science journalist who wrote this article, obviously, since she used the terms interchangeably. Ice shelves are floating ice, and ice sheets are the vast masses of grounded ice, which contain 70% of the Earth's fresh water.


    Don't know that ice shelves represent a minuscule fraction of Antarctic ice.


    Don't realize that melting or disintegrating ice shelves don't affect sea-level.


    Don't know that ice shelves are constantly breaking away at the outside edges, in pieces large and small, and they are constantly being replenished with ice flowing downhill from upstream glaciers and ice sheets.


    Don't know that most of the ice in the ice shelves originated as part of the ice sheets and glaciers, which flowed downhill into the ocean. (Most of the rest is snowfall accumulation.)


    Don't know that Antarctica averages more than 40° below zero, so most of it is in no danger of melting.


    Don't realize that the article's claim that "ice shelves confined to bays and gulfs create a physical barrier that the slow-moving glaciers butt up against," and thereby affect ice flow rates, describes an extremely minor and transient effect.Point #7 deserves a bit of elaboration.

    When I say "extremely minor," what do I mean? Well, here's a quote, from a paper about it:
    "Loss of buttressing offsetting half of the tendency for ice‐stream/ice‐shelf spreading for an ice stream similar to Pine Island Glacier, West Antarctica is modeled to contribute at least 1 mm of sea‐level rise over a few decades." https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2004GL022024
    Yes, you read that correctly: "at least 1 mm of sea-level rise over a few decades." Batten the hatches, we're all gonna drown!
    It is also transient, because when ice flow rates from glaciers and ice sheets increase, the influx of ice tends to cause the downstream floating ice sheets to grow again: a negative (stabilizing) feedback mechanism.

    Greenland is similar, but not as cold. Here's an article about Greenland's largest glacier, which feeds a downstream ice shelf that had almost completely disappeared... but is now "growing back."
    https://climate.nasa.gov/news/2852/cold-water-currently-slowing-fastest-greenland-glacier/

    Here's the effect that global warming is having on sea-level:
    https://sealevel.info/1612340_Honolulu_vs_CO2_thru_2020-03_annot1.png
    Reply
  • Broadlands
    A response to Chem721. It is important to distinguish between glacial ice and polar ice shelves as Mogoso pointed out. The back and forth growth and retreat of glaciers is well known. My question was about the polar ice caps, not glaciers.

    A response to Whitefeather is that his acceleration of change since pre-industrial time is just a trivial 0.75°C as of 2019. It was 0.83°C in 2016. There is no meaningful way to mitigate that trivial amount of chaos by moving CO2 around...lowering emissions does not lower the CO2 we have already added.
    Reply
  • Whitefeather
    The point is that the 'trivial' amount of warning you refer to is enormous when applied to the oceans. They have acted like a 'heat sink'. If we had't had them we would have fried to death ages ago.
    The amount of carbon and other trash fuels we have consumed has also meant that we have polluted the atmosphere with gases that hinder the reflection of heat from the Earth. Also, as the ice caps melt they lose their reflective ability so we are now astride an exponential effect. It's accelerating rapidly. Another exponential process is the carbon now escaping from frozen tundra and ice caps as they warm up.
    As long as we cease using hydrocarbons as fuels, we can allow nature to store existing CO2 for us as it's done for millions of years. That's if we stop destroying the Amazon of course.
    I do wonder how many catastrophes need to occur before people realise that global warming is a far greater threat than Covid 19 !!!!!!!!!
    Whitefeather
    Reply
  • Whitefeather
    mogoso said:
    90 % of the sea ice shelves breaks off yearly never a boost to sea level. Think on it put some ice in a glass with some water mark the level. When the ice melts does the water in the glass rise, does it over flow. Displacement, it stays the same. Give me something tangible to believe in on man made climate change, other than we need a tax administered by the UN who backs this climate change hocus pocus
    Reply