They call it the Doomsday Glacier. A chunk of Antarctic ice as big as Florida and two thirds of a mile thick, the Thwaites Glacier disgorges into the ocean in a remote region of West Antarctica. Glaciologists say it may be on the verge of total collapse, which could swamp huge areas of low-lying coastal land around the world within a few decades. Now, ambitious plans to save it are set to become an early test of whether the world is prepared to enact massive geoengineering efforts to ward off the worst effects of climate change.

Recent monitoring by uncrewed submarines and satellites, along with ice-sheet modeling, suggest that the Thwaites Glacier and its adjacent smaller twin, the Pine Island Glacier, may already be in a death spiral — eaten up by the intensifying speed and warmth of the powerful Antarctic Circumpolar Current. If they are past a point of no return, say researchers involved in the studies, then only massive human intervention can save them.

Meanwhile, some of his collaborators, fearing the logistical complications of such a task, are pondering an even more mind-bending idea. They want to substitute the physical curtain with a giant “bubble curtain,” created by a constant injection of bubbles of air or cold surface water.

Opponents of the plans, including many glaciologists, say such outlandish proposals are a dangerous diversion from the real task of mitigating climate change by curbing carbon emissions. But advocates say the two glaciers can’t wait. “We can’t mitigate our way out of this,” says Moore. “We need other tools.”

West Antarctic ice sheet and CO2 greenhouse effect: a threat of disaster
Douglas I. Benn & David E. Sugden
Pages 13-23 | Published online: 14 Feb 2021

ABSTRACT
Over 40 years ago, the glaciologist John Mercer warned that parts of the West Antarctic Ice Sheet were at risk of collapse due to the CO2 greenhouse effect. Mercer recognised the unique vulnerability of ice sheets resting on beds far below sea level (marine-based ice sheets), where an initial warming signal can initiate irreversible retreat. In this paper, we review recent work on evidence for ice sheet collapse in warmer periods of the recent geological past, the current behaviour of the ice sheet, and computer models used to predict future ice-sheet response to global warming. Much of this work points in the same direction: warming climates can indeed trigger collapse of marine-based portions of the West Antarctic Ice Sheet, and retreat in response to recent warming has brought parts of the ice sheet to the threshold of instability. Further retreat appears to be inevitable, but the rate of collapse depends critically on future emissions.

Introduction
The title of this paper is identical to that used by John Mercer in Nature over 40 years ago, in which he predicted that anthropogenic climate change could threaten the stability of the West Antarctic Ice Sheet (Mercer, 1978). The introduction to Mercer’s paper has an impressively modern touch and highlights the existence of a marine basin beneath the West Antarctic Ice Sheet. He argues that the stability of the ice sheet covering the basin depends on the presence of shallow topographic thresholds and ice shelves around its periphery. With continued rise in CO2 this stability is threatened, bringing a danger of ice-sheet collapse and eventual global sea-level rise of several metres. Mercer warned that we should keep an eye on ice shelves in the Antarctic Peninsula as an early warning sign. Since then, several ice shelves in the Peninsula have been lost, such as the Larsen B Ice Shelf in 2002 (Cook & Vaughan, 2010; Scambos et al., 2000) and rapid changes have occurred around the margins of the Amundsen Sea (Christie et al., 2016; MacGregor et al., 2012).

The word ‘collapse’ is often used in the context of ice sheets and glaciers, but is seldom defined. Here, we use it to mean an irreversible process of mass loss initiated when some trigger causes the system to cross a threshold into instability. Most land-based glaciers do not exhibit this kind of behaviour. Although warming may cause rapid melting of land-based glaciers, their responses to climatic signals are typically linear and reversible. That is, if the climate signal changes, melting will slow down and the glacier may stabilise or even grow again. Marine-based glaciers, on the other hand, can undergo irreversible retreat in response to warming of the atmosphere and oceans, and ice loss may continue until all is gone even if the initial signal is removed. Another important point about the idea of ‘collapse’ is that it may happen over short or long timescales. In the case of a floating ice shelf such as Larsen B, collapse may occur over a few days once the critical stability threshold is crossed (MacAyeal et al., 2003). On the other hand, collapse of a large marine-based ice sheet may play out over hundreds or a thousand years. The key issue is not the rate of ice loss, but the fact that the system has no stable state after the initial ‘push’.

So, how much of the Antarctic Ice Sheet is at risk of collapse, and how fast could ice be lost to the ocean? There are three main approaches to answering these questions. First, study of the long-term history of an ice sheet reveals what has happened in the past. It is particularly useful to examine what happened during climates warmer than present, for example in the Pliocene 5.3–2.6 million years Ma ago and the last interglacial period some 120 thousand years ka ago. Second, we can study the current behaviour of the ice sheet alongside observations of the oceans and atmosphere, to identify areas of rapid change and understand the key processes at work. Third, computer modelling techniques allow us to perform experiments with ‘virtual glaciers’, and to study system response to changing conditions. Computer models can help us understand what has happened in the past, analyse the controls on ice-sheet behaviour in the present, and address the all-important question of how rapidly the ice sheet may change in response to alternative greenhouse gas futures.

…