Melting ‘glue’ sent enormous iceberg to its doom
The thinning of an icy ‘glue’ that holds fractured ice together may drive ice shelf collapse in Antarctica. Ice shelves are massive stretches of ice that build up over many thousands of years. But warming air and rising ocean temperatures have been driving ice shelves to disintegrate. Many of Antarctica’s ice shelves have fractured or collapsed in the past couple of decades, but exactly what’s accelerating the ice loss has been unclear.
To figure this out, a group of glaciologists zoomed in on rifts on Antarctica’s Larsen C Ice Shelf, which calved a city-sized iceberg called A68 in July 2017. The split of A68, an iceberg approximately 2,240 square miles in area, reduced the size of Larsen C by 12 per cent. Larsen C is the third ice shelf on Antarctica’s western peninsula to undergo massive ice loss in the past two decades.
The prevailing theory was that these splits were happening due to a process known as hydrofracturing, in which pools of melted ice on the surface of ice shelves seep through the
cracks and expand once they freeze again.
“But that theory fails to explain how iceberg A68 could break from the Larsen C Ice Shelf in the dead of the Antarctic winter when no melt pools were present,” said Eric Rignot, a professor of Earth system science at the University of California, Irvine.
Rignot and his colleagues analysed hundreds of rifts, or fractures, in the Larsen C Ice Shelf using a model of ice sheets and sea-level changes developed by NASA, as well as data from satellites and research aircraft. They zoomed in on 11 cracks and modelled three melting scenarios. Two of the three scenarios focused on the role of ‘melange’, a mix of windblown snow, frozen seawater and ice shelf fragments that exists inside and around rifts and typically works to seal the fractures.
In the first scenario, the glaciologists modelled what would happen if the ice shelf thinned due to melting. In the second they modelled what would happen if the ice melange thinned, and in the third they modelled what would happen if both the ice shelf and the melange thinned. Their simulations showed that the thinning of the melange controlled the rate at which the rift opened.
If the ice shelf thinned but the melange remained just as thick, the rift widening slowed down with time. In other words, the melange acted as a ‘healing’ glue, fusing parts of the cracks. If both the ice shelf and the melange thinned, rift widening also slowed, but not as much as it did in the first scenario. If the ice shelf remained the same but the melange thinned, as in the second scenario, the average annual rate of rift widening increased from 76 to 112 metres.
Like sea ice, melange is vulnerable to the effects of warming oceans and rising air temperatures. “The melange is thinner than ice to begin with,” said Eric Larour, a NASA Jet Propulsion Laboratory research scientist.
Just 10 to 20 metres of melange thinning is enough to reactivate a rift, or start to unzip it and trigger a major calving event. Reactivating a rift can trigger ice shelves to retreat decades before water ponding would cause hydrofractures on the ice sheet surface.