Short-term stability and long-term collapse: exploring the complex behavior of the Antarctic ice sheet

Yes, Antarctica is gaining ice. And yes, the Antarctic ice sheet is vulnerable to a major collapse. These simple, seemingly contradictory statements highlight the complexities of understanding the behavior of the Earth's vast southern continent. But, after a recent paper discovered that ice mass is actually increasing in Eastern Antarctica, headlines have abounded with simple declarations: "Antarctica gaining mass" or "Antarctica gaining ice despite climate change".

While these statements are true, they miss the dynamic processes that are occurring in different regions of Antarctica, all of which contribute to the short-term and long-term future of Antarctic ice sheets that impact climate change. So I thought I would explore two recent studies - one showing the increased ice mass in the east¹, the other portending western ice sheet collapse² - to provide some details to an Antarctic ice story that can be easily oversimplified.

The Eastern Rise

The major factors determining net Antarctic ice mass are snowfall and sublimation or drift loss, as seen in the figure below. Snow falls over grounded ice and freezes, creating more ice. This ice then flows toward the glacial edges. If the edge is near solid ground, melting and runoff reduces ice mass, a common occurrence in some ice sheets like Greenland. However, in Antarctica, ice loss primarily occurs through the interaction of the ocean with the edge of the sheet (left side of figure). Here, the circulating ocean heats the underside of the ice, which eventually breaks off and melts into the water.

Scientists have known about this competition between snowfall and drift melting, but most studies could only predict a net gain or loss of ice mass with a high degree of uncertainty.³ This has changed with a recent paper that provides a more definitive picture of how snowfall is increasing ice mass in eastern Antarctica.

In the study, Zwally and colleagues measured mass changes from 1992-2008 in two ways: (1) using satellite data to track changes in elevation, and (2) identifying 27 drainage systems along the ice sheet perimeter to track flow rates via melting or drift loss. By taking into account the density of ice, changes in mass could b determined from the elevation data. Using these methods, gains or losses in ice were summed over a 50 km x 50 km grid covering the entire ice sheet to determine whether or not Antarctica is losing or gaining ice.

The map of total mass changes (dM/dt) is shown below - the top and lower map visualize changes between 1992-2001 and 2003-2008, respectively. As you can see, mass is increasing in both time periods across Eastern Antarctica (orange/red/purple colors). Even regions in the northern part of West Antarctica show increasing mass due to snowfall. Ice loss generally occurs along the West Antarctic edge (green/blue colors) in an area known as the Amundsen basin. We'll return to this region in the next section as loss in this area is quite important when predicting the future impact of Antarctic ice loss on sea level rise.

In the end, mass increases from snowfall outpace ice loss and lead to a net gain in ice mass during both time periods. In particular, Eastern Antarctica (EA) shows a 136 Gigaton/acre (Gta^-1) mass increase from 1992-2001, more than enough to compensate the 16 Gta^-1and 29 Gta^-1losses along the Western Antarctica (WA) and Antarctic Peninsula (AP) to result in a net mass increase of 112 Gta^-1. Similar trends from 2003-2008 demonstrate a net increase of 82 Gta^-1.

To put these numbers in perspective, one Gigaton of water corresponds to a volume of one cubic kilometer (km). The ocean spreads across about 360 millions square kilometers, so one Gt of water added to the ocean would raise its sea level only about 2.8 microns, less than the width of a human hair! Based on the data above, the 82 Gta^-1net increase in ice mass from 2003-2008 corresponds to 0.23 mm/acre of sea level depletion.

So it is this increase in net ice mass that has caught the attention of the mainstream media, claiming that climate change has not had the dire effects predicted by the scientific community. It is true that these findings are important: they indicate that Antarctic ice loss has not yet contributed to sea level rise since any ice loss is compensated through evaporation and subsequent precipitation over the ice sheet. But what about this troublesome spot along the Amundsen basin of Western Antarctica showing dramatic ice loss? Do we have reason to worry about what could happen in this region?

The Western Collapse

It turns out that we do. A second study² published in PNAS has implemented a computational model to predict just what will happen decades from now if ice around the Amundsen Basin continues to disappear. The sobering results predict over three meters of sea level rise due to a collapse of the Western Antarctic sheet, highlighting the extraordinary impact of local changes in the sheet even if the continent as a whole is gaining ice.

To understand how such a collapse can occur, we must first take a look underwater. Most of the western ice sheet is grounded to rock below sea level, (see the grounding line in the figure below), with ice shelves that float out into the ocean. As warm water circulates beneath the ice shelves, ice can melt and push the grounding line farther inland (as well as melt or break off large chunks of the ice shelves). However, unlike this figure, the bedrock in Western Antarctica slopes downward as it moves inland, which leads to a highly unstable grounding line.⁴ This means that, as more ice melts and the grounding line pushes farther inland, the system could destabilize and lead to an unstoppable retreat of the ice sheet towards its center.

With this possibility in mind, Feldmann and Levermann sought to model the behavior of the Amundsen Glacier and western Antarctic ice sheet as a whole over thousands of years to see how this destabilization would affect long-term ice loss. The model is quite complex (feel free to take a look at Reference 2 for details) but it involves coupling the ice sheet and the ocean thermally and mechanically, taking into account temperature effects and physical stresses that dictate flow and melting rates. Beginning with conditions that simulate the current state of the ice sheet, the researchers 'push' the behavior of the Amundsen Glacier slightly in the direction expected from current ice loss rates (this 'push' is known as a perturbation in these models). The model then evolves the ice sheet over space and time, measuring ice gains and losses across the Western ice sheet for hundreds and thousands of years.

The results provide a clear, stark message: if we assume that the Amundsen Glacier is indeed destabilized, which current evidence suggests, then the entire Western ice sheet will discharge into the ocean within 3000 years, leading to a 3 meter rise in sea level within 10,000 years. This extreme loss is due to a 'point-of-no-return' scenario: once the grounding line begins retreating, ice will continue to melt inland, allowing water to carve a path inward until it breaks off the whole western sheet.

It is true that this process unfolds over a very long time period, giving us plenty of time to respond if this scenario does occur. But the world would look quite different. In the US alone, 12.3 million people would lose their homes.

Thinking Globally and Locally

Both of the studies described above provide important new information about how climatic changes affect the future of the Antarctic ice sheet. But a simple message of ice mass increase or decrease does not give a clear or helpful picture about what we can expect global warming to do to the massive continent. Snowfall may be increasing ice across Antarctica as a whole and preventing short-term sea level rise, but local changes in the Amundsen Basin can result in catastrophic changes over long timescales. This is a perfect example of chaotic behavior, in which small perturbations can push a system into an entirely new direction that can be difficult to see initially or stop. Both perspectives, the short and long views, must be balanced when considering climatic changes, their impact on our global society, and what policies we develop in response to them.

References