Last week, British scientists announced a disturbing finding - a crack in the Larsen C ice shelf in the Antarctic Peninsula had dramatically accelerated its spread, increasing 18km in length in the space of a month.
This means the floating ice shelf, which is nearly as big as Scotland and the fourth largest of its kind in Antarctica, is poised to break off a piece nearly 5180sq km in size, or over 10 per cent of its total area. An ice island the size of a small US state would then be afloat in the Southern Ocean.
That's dramatic enough, but there is uncertainty in the science world about what would happen next.
On the one hand, the researchers with Project Midas, who announced the growth of the rift, have published research suggesting that, in their words, it "presents a considerable risk to the stability of the Larsen C Ice Shelf".
If they're right, it's hard to understate how big a deal it is - Antarctica has lost ice shelves before, but not one so enormous. Not only would a loss of Larsen C change the map of the Earth itself; the shelf holds back glaciers capable of contributing about 10cm of global sea level rise over time.
The rift is now largely following the second, and worse scenario from their research, said glaciologist Daniela Jansen of the Alfred Wegner Institute in Germany - if it is correct. "This calving will be a test for our calving front stability criterion," she said by email.
However, other analyses have suggested that most of the ice that would be lost is so-called "passive ice" that does not play a key role in holding the glaciers behind the shelf in place. And some scientists have expressed scepticism about whether what's happening at Larsen C is "cause for alarm".
Time, ultimately, will show who is right. But in the meantime we can sketch, a little, the kinds of things that scientists are thinking about as they watch all of this unfold.
First, they are very mindful that what's happening to Larsen C could be part of a broader pattern for ice shelves both globally and on the Antarctic Peninsula - including Larsen A, which collapsed in 1995, and most of all Larsen B.
Much of Larsen B, located just north of C on the other side of a peninsula, collapsed dramatically in 2002 - and scientists have shown that this is probably not something that has happened in the last 12,000 years or even, perhaps, in more than 100,000 years, or since the last interglacial event.
What's more, the collapse followed closely upon a large break in 1995, a possible analogue for what is now happening to its southern cousin.
Larsen C has also apparently been in place for a very long time. And whether or not the pending loss of its enormous iceberg would ignite a fast retreat, researchers agree the event will bring it down to a size smaller than anything humans have observed before.
Indeed, whether the retreat is slow or fast, a number of scientists have also documented that Larsen C is changing in a way suggestive of a weakening due to a warming climate. In all ice shelves, ice flows slowly from inland glaciers out through the shelf towards the sea. But recently, the flow rate at the northern sector of Larsen C has sped up. "This is a sign of instability, akin to the 20 per cent increase in the flow speed of Larsen B between 1996 and 2000 prior to its collapse," noted a 2011 scientific paper by University of California at Irvine geoscientist Eric Rignot and his colleagues.
Larsen C is also very slowly lowering in the water. This is happening because the ice that comprises the shelf is getting thinner, whether due to melting from above or from below. The thinning process is slow enough that it would not result in any sudden loss of the ice shelf, but it's still another sign of weakening.
What's striking is that all of this is occurring even though, because it is farther south along the Antarctic Peninsula, Larsen C is in a colder place than the other ice shelves that have already collapsed. Scientists have suggested that there is a kind of threshold in annual temperatures for where ice shelves cease to be feasible: a temperature above minus-9C, averaged annually. Larsen C turns out to be right at the cutoff for this to happen.
"Larsen C is at the limit of break up, its northern side (which is not far from Larsen B) is close to the limit of viability of ice shelves," said Rignot. "The southern side is more protected."
So with all of that in mind, what happens next?
Let's start with the argument from the researchers affiliated or working with Project Midas, who do think this one break could destabilise the shelf, and liken it to what happened with the Larsen B ice shelf nearby.
Their model of the ice shelf suggests that its front, facing the ocean, could remain in a partly unstable state after the break and lose considerably more ice afterward. The reason, they say, is that that break will remove regions of ice that were more stable because in these areas, the ice was subject to offsetting stresses and counter-stresses that had the effect of hemming it in and keeping it in place. But if some of these ice regions are gone, the argument goes, that would leave behind ice whose main tendency is to flow outward, toward the sea.
In addition to these balancing stresses near its front, the ice shelf also contains what scientists call a "compressive arch," which you might think of as a line across the shelf where the stresses on the ice switch from causing it to be compressed, to causing it to stretch outward. This is further inland, but if that area is breached, the ice shelf is expected to collapse. At present, it looks like the break will cut through some of the arch, but not all of it.
If the calving continues and goes past the compressive arch . . . then the ice shelf will break up
SHARE THIS QUOTE:
But despite such debates, other researchers remain unsure just how badly the breakup will damage the larger shelf.
"We studied the current rift in the past few years, it has been progressing rather 'normally,' the recent acceleration in the rift progression is 'expected' in my opinion," said Rignot by email. "The consequences on the rest of the ice shelf are not clear at this point. If the calving continues and goes past the compressive arch . . . then the ice shelf will break up."
A recent study in Nature Climate Change, however, found that the calving event would largely remove "passive ice" from Larsen C, ice that is not doing a great deal of work to hold glaciers back, and so will be "unlikely to instantly produce much dynamic change." Granted, that study agreed that the remaining ice shelf will have a concave shape, just as Larsen A and B did before they collapsed.
"These authors have results that indicate a low level of concern, but not zero," said Richard Alley, a glaciologist at Penn State University, referring to the Nature Climate Change study.
If scientists are hedging a bit about what happens next at Larsen C, perhaps that shouldn't be surprising in light of the fact that we are about to see a rare event. Moreover, it's one whose precise effect on the shelf ultimately depends on incompletely understood and highly complex processes of ice fracture - one that the models may or may not be able to capture.