RealClimate logo

Wilkins ice shelf collapse

Filed under: — gavin @ 6 April 2009

Since people are wanting to talk about the latest events on the Antarctic Peninsula, this is a post for that discussion.

The imagery from ESA (animation here) tells the recent story quite clearly – the last sliver of ice between the main Wilkins ice shelf and Charcot Island is currently collapsing in a very interesting way (from a materials science point of view). For some of the history of the collapse, see our previous post. This is the tenth major ice shelf to collapse in recent times.

Maybe we can get some updates and discussion of potential implications from the people working on this in the comments….?

613 Responses to “Wilkins ice shelf collapse”

  1. 601
    Nylo says:

    Today The Independent (UK) informs that a piece of ice the size of New York has collapsed. However it also informs that (I copy literally) “The loss of ice shelves does not raise sea levels significantly because the ice is floating and already mostly submerged by the ocean”. Is this true? I never heard of it before. It also warns that there is a risk of sea-level rise anyway because “their loss will allow ice sheets on land to move faster, adding extra water to the seas”. However the Wilkins is not a glacier-fed ice shelf. How much and how fast would that effect be felt?

    [Response: The almost negligible impact of floating ice shelf collapse on sea level is well known (the only reason it is not zero is because of the fact that the ice is fresh and the sea water it displaces is salty). However, the acceleration of the upstream glaciers after a collapse has only recently been observed and happens relatively quickly. After Larsen-B collapsed, the feeding glaciers sped up by about 400% (see here). But why do you think the Wilkins ice shelf doesn’t come from a land based glacier? It definitely does – you just don’t get that thickness of ice from in-situ formation. Here is an old (pre-collapse) image showing the upstream glaciers on Alexander Is. – gavin]

  2. 602
    Nylo says:

    Gavin, about what leads me to think that it is not glacier-fed, look at this paragraph from (it’s a NASA site, I separated parts of the link on purpose for it to pass posible spam filters, you only need to join them):

    “The Larsen Ice Shelf is “typical” in that it is primarily fed by a land-based glacier. The Wilkins Ice Shelf is somewhat unusual in that only the southern end of the shelf appears to be fed by land-based ice; the rest of the shelf may have formed from accumulation of sea ice that held fast to the coastline through many seasons, as well as snow cover. Glaciologists estimate that the part of the Wilkins Ice Shelf that formed from sea ice may be 300 to 400 years old, and the part that is fed by glacier flow is older, perhaps up to 1,500 years old. Because the Wilkins Ice Shelf is only marginally fed by glacier flow, however, its collapse was not expected to have the same impact on sea level rise as the collapse of the Larsen B potentially could”.

    As you see, it does claim that you can get “that thickness of ice from in-situ formation”. Actually, it says that you can get it in only 400 years (this means, I suppose, that it would have started to happen more or less at the Little Ice Age). And appart from that, do the parts of the Wilkins Ice Shelf that are disintegrating belong to the Southern, glacier-fed part of the shelf? Because if they happen to belong to the Northern part, it seems to me like there would be nothing to worry about at all, with no glacier ice accelerating anywhere.

    Best regards.

    [Response: Ok, in the absence of any more detailed info, this seems to be correct. I am nonetheless surprised that you can get 300m thick landfast sea ice, maybe someone who knows more about the details can comment. – gavin]

  3. 603
    Hank Roberts says:

    Just speculating:
    “… may have formed from accumulation of sea ice that held fast to the coastline through many seasons, as well as snow cover” sounds like no one has reported on a core through that ice. I started looking, but can only see abstracts without heading out to a library. Anyone know for sure?

    I found mention of increasing snow accumulation generally, not specific to Wilkins:
    A doubling in snow accumulation in the western Antarctic Peninsula since 1850
    “… medium depth (136 metres) ice core drilled in a high accumulation site (73.59°S, 70.36°W) on the south-western Antarctic Peninsula during 2007. The Gomez record …”

    and investigation of the ice using remote sensing:

    “… Examination of synthetic aperture radar data collected over the southeastern Antarctic Peninsula
    shows that features sometimes mapped as ice shelves are more likely composed of numerous ice
    tongues interspersed within a matrix of fast ice and icebergs. The tongues are formed by the
    seaward extension of numerous small mountain glaciers that drain from the Antarctic Peninsula.
    Once afloat, the tongues intermingle with a matrix of fast ice and brash. Examination of 1997
    Radarsat-1 image mosaics shows that southeastern Antarctic Peninsula composite-ice-shelves
    covered an area of about 3500 km2. Similar to ice tongues around the rest of Antarctica, these
    features are highly fragmented and likely to be susceptible to mechanical failure….”

    which includes mention of a prediction:

    “Mercer (1978) postulated that Antarctic ice shelves would be the component of the Antarctic
    glacier system most responsive to “greenhouse” warming. He predicted a southerly retreat of ice
    shelf margins starting with the ice shelves in the Antarctic Peninsula ….”

    Structure of Eastern Antarctic Peninsula Ice Shelves and Ice Tongues from Synthetic Aperture Radar Imagery

    But didn’t turn up any specific reference to ice core drilling on the Wilkins. Someone who actually knows the science might be able to Survey article:

  4. 604
    Hank Roberts says:

    Sorry about the bad line breaks.

    Closest drill site I could find (would at least give snowfall info) is mentioned here:
    Those folks would know what else is available; no mention of actual cores drilled through the Wilkins.

    There’s a map — that makes clear the Wilkins doesn’t have a vast area of glacier feeding it; it’s tucked in among islands.

    Gavin earlier remarked on the odd fracture pattern; if the ice accumulates but is broken (by tides?) and infilling with snow to recreate a flat surface, year after year, that would fit the radar result mentioned earlier

  5. 605
    Ike Solem says:

    Hank Roberts, Why would a core drilled through the Wilkins Ice Shelf tell much about the forces responsible for the thinning of the ice bridge and the eventual break, followed by increased instability and breakup?

    ScienceDaily (Apr. 29, 2009) — Satellite images show that icebergs have begun to calve from the northern front of the Wilkins Ice Shelf – indicating that the huge shelf has become unstable. This follows the collapse three weeks ago of the ice bridge that had previously linked the Antarctic mainland to Charcot Island.

    Prior to the breakup, many other fracture zones were created throughout the ice shelf, and the icebergs are breaking off at those zones.

    There are two possible reasons for the breakup, one better understood than the other:

    “The Antarctic Peninsula has experienced extraordinary warming in the past 50 years of 2.5°C, Braun and Humbert explained.”

    That’s well documented – but the ocean looks like it played a role as well:

    We demonstrate here, for the first time, that the adjacent ocean [to the WAP] showed profound coincident changes, with surface summer temperatures rising more than 1°C and a strong upper-layer salinification. Initially driven by atmospheric warming and reduced rates of sea ice production, these changes constitute positive feedbacks that will contribute significantly to the continued climate change.

    Thus, it seems pretty clear that atmospheric and oceanic temperature increases were primarily forced by CO2 from fossil fuel combustion, and that’s what lead to the breakup of the Wilkins Ice Shelf.

    I don’t think anyone will be able to pin that on ‘natural variability’.

  6. 606
    Nylo says:

    I guess that they think about in-situ formation because of the on-land ice sheet near the Wilkins not moving at all, not pushing the shelf. But I wonder if a drill hole could give more information as to how the ice initially formed. After all, even the ice carried over by a glacier was snow in the beginning, so it mustn’t be very different from the ice we would have under in-situ formation. The drill hole would need to show proof of some kind of stress suffered by the ice while being pushed by the glacier, and I doubt that such a thing can be learnt from a drill hole. But I know very little about this procedure anyway…

  7. 607
    John Finn says:

    Attributing the causes of various warming/cooling events is becoming an increasingly complex science.
    While arctic ice melt and the wilkins collapse are caused by human emissions of greenhouse gases, it seems the 100,000 sq km per decade increase in antarctic sea ice is linked to the ozone hole ( and, presumably, human CFC emissions.

    I’d almost forgotten about CFCs and ozone depletion. I wonder if aerosols will make a comeback soon.

    [Response: Oh yes, we’d forgotten about those. Please try to pay attention. – gavin]

  8. 608
    mauri pelto says:

    Wilkins Ice Shelf is recognized as being dominantly fed by surface accumulation with basal melting the principal source of mass balance loss, except for the rapid calving events. Radio echo sounding of WIS from the 1970’s indicates much thicker ice feeding the southern portion of the ice shelf from Haydn Inlet and Schubert Inlet. Further ice velocities from SAR interfertometry indicates strong flow from Schubert Inlet and from the Lewis Snowfield, which is just west of Schubert Inlet. There is a marked change in ice thickness north of a line running roughly from Charcot Island-Burgess ice Rise to Petrie Ice Rise to Dorsey Island. North of this line both thickness and velocity are much reduced. This suggest that the southern sector which is clearly fed by the aforementioned is not feeding the northern section of the ice shelf significantly. The aforementioned rise and islands direct flow more westerly. The flow into the northern sector of the ice shelf from Gilbert Glacier and Haydn Inlet is hard to discern. Whether the ice shelf to the south began as landfast sea ice or not the key process of basal melt and surface accumulation that dominate the mass balance would remove the evidence of the origin. An ice core would penetrate the brine saturated melting base of the ice shelf. How would the original ice have persisted? Because we have two substantial feeder glaciers from Haydn and Gilbert it is not unlikely that greater flow would have fed WIS more significantly in the past.

  9. 609
    John Finn says:

    Re: #

    I’d almost forgotten about CFCs and ozone depletion. I wonder if aerosols will make a comeback soon.

    [Response: Oh yes, we’d forgotten about those. Please try to pay attention. – gavin]

    Yes, Gavin, thanks. It was, in fact, a joke(or obviously not). I’m well aware that several of your most recent posts have been on aerosols.

  10. 610
    Phil. Felton says:

    Re 608: Thanks Mauri for that information, do you have any references?

  11. 611
    Hank Roberts says:

    Another question for Mauri — this abstract:

    Can you (or anyone with expertise in the area) interpret the last sentence? Are they saying that according to their calculations, 98% of the lakes on the Greenland ice sheet have enough water in them to force cracks down through the ice? Or are they saying that a single lake large enough to force cracking of the ice would contain 98% of the total water in all the lakes? Sorry, all I can see is the abstract and I can’t figure it out.

  12. 612
    David B. Benson says:

    Hank Roberts (611) — Here is the conclusion from the paper:
    “Our calculations show that lakes that are only ~250–800 m across and 2–5 m deep contain a sufficient volume of water to drive a water-filled crack to the base of a 1 km-thick ice sheet. Lakes that are smaller may also be drained, however it requires fractures that are fed by multiple basins. This range in lake sizes represents the majority of supraglacial lakes in the ablation zone along the western margin of the Greenland Ice Sheet. Thus we propose that a large fraction of the melt water produced in the summer (on the order of several cubic kilometers) could rapidly reach the base of the ice sheet via this mechanism.”

  13. 613
    Hank Roberts says:

    Thank you David.