Disappearing snow increases risk of collapsing ice shelves in Antarctica

Collapsing ice shelves

Since the 1970s, about 20% of the ice shelves in the Antarctic Peninsula have disappeared. But these floating extensions of glaciers didn't just melt away: instead, they collapsed suddenly and in a matter of weeks. As a result, the mountain glaciers previously feeding the ice shelves sped up by a factor of 3-4, delivering large amounts of ice directly to the ocean. The current ice loss from the Antarctic Peninsula, which we observe with satellites, is largely due to increased ice discharge after the collapse of ice shelves that used to provide stability to the feeding glaciers. Knowing why these ice shelves collapsed, and whether it could happen to other shelves in the future, is crucial to understand the future contribution of Antarctica to the rising seas.

Satellite images of Larsen B collapse (31 Jan - 7 Mar 2002). Animation from the National Snow and Ice Data Center.

Meltwater lakes

Prior to two big and famous collapse events, those of the Larsen A (1995) and Larsen B (2002) ice shelves, satellites recorded extensive ponding of meltwater in lakes on the surface of these ice shelves. It is believed that the water in these lakes can enter into crevasses. The water pressure is enough to widen the crevasses, eventually opening up the crevasse all the way to the ice-shelf bottom. This hydrofracturing mechanism is thought to be the explanation for the sudden ice-shelf collapses we have seen in recent decades.

Detail of the front of Larsen B ice shelf just prior to its collapse in January 2002. Note the extensive melt ponds and rivers at its surface. Photo by NASA.

What has firn to do with all this?

On any glacier or ice shelf, snow falling on its surface first compacts to a substance we call firn, before it finally becomes ice. The layer of firn on an ice shelf can act as a sponge, absorbing the meltwater that is created at its surface. As long as there are enough pores in the firn, meltwater will seep into the firn and refreeze. As a result, no melt lakes can form, and no hydrofracturing occurs. But when the firn layer is gone, any surface melt will form lakes, and hydrofracturing becomes possible (see the illustration below).

An ice shelf with lots of pore space in the firn (left image) holds water away from crevasses and fractures. When the firn is gone (right image), lakes form, and hydrofracturing can occur.

Using a climate model that calculates past and future snowfall and melt, we calculated how the firn layer evolved in the recent past, and how it might evolve in a future climate. We use the amount of pore space to assess which ice shelves could become under threat of collapse in the coming two centuries.

Some lose, some win

First, we demonstrate that there was very little firn left on the Larsen A and B ice shelves when they collapsed. The amount of firn was already low since the start of our simulation in the 1960s. It shows that indeed a depleted firn layer is required before hydrofracturing can occur.

Next, we show that it's mainly the ice shelves in the Antarctic Peninsula that will lose their firn layer in the next 200 years, leaving them vulnerable to collapse. Most of the ice shelves in East Antarctica are simply too cold to develop a significant melt season that could fill the pore space of the firn layer. On the two largest ice shelves (the Ross and Filchner-Ronne shelves), each roughly the size of Spain (or California), we even expect the amount of pore space to increase, just because there will be more snowfall. These ice shelves will thus be safe from hydrofracturing in the next 200 years.

Finally, the total amount of emitted greenhouse gases determines how many ice shelves become under risk of collapse. The amount of future pore space depends mostly on the amount of summer melt, which is in turn determined by temperature. Because the collapse of ice shelves leads to an increase in ice loss, we can limit hydrofracturing-induced sea-level rise from Antarctica by reducing greenhouse gas emissions.

Amount of pore space (in metres) for 12 different ice shelves, during the next two centuries.

Publication

This research appeared in the Journal of Glaciology on 30 January 2014 as an open-access article: Kuipers Munneke, P., S. R. M. Ligtenberg, M. R. van den Broeke and D. G. Vaughan. 2014. Firn air depletion as a precursor of Antarctic ice-shelf collapse. J. Glaciology, 60(220), 205-214. doi:10.3189/2014JoG13J183. This research was financed by the Netherlands Polar Programme, and the European research consortium ice2sea. Both researchers from IMAU (Utrecht University) and the British Antarctic Survey (Cambridge, UK) contributed to this article.