This post continues my post from May, which was written to lend some historical background to the recently released news that the large marine glaciers emptying into the Amundsen Sea seem to have passed a point of no return, and will continue to collapse until they are gone, whereupon the rest of the West Antarctic Ice Sheeet (WAIS) may well follow. Total sea level rises should be 2–3m within a few centuries, though the exact timescales could be faster or slower. The above video from the NASA Jet Propulsion Laboratory features Eric Rignot, one of the leaders of the two teams who have reached this conclusion. (The other team is led by the University of Washington’s Ian Joughin, on whom more below.)
The history in Pt. 1 concluded circa 1980, when geologist John Mercer (1922-1987) connected the prospect of WAIS collapsing to global warming. Shortly thereafter the University of Maine’s Terry Hughes—who had previously linked future WAIS collapse to an ongoing global retreat from the Last Glacial Maximum (18,000 years ago)—identified (pdf) the Amundsen Sea glaciers as WAIS’s “weak underbelly,” which would be the “mechanism for disintegration of [WAIS] during a proposed Super Interglaciation triggered by CO2-induced climatic warming.” This post addresses what occurred in the intervening decades to convince the glaciological community of the assertion.
The first thing was to find out whether marine ice sheets actually behaved in the way that Hans Weertman’s 1974 theoretical model (pdf) said they, in principle, could. To do this, research teams undertook an extensive program of field research throughout the 1980s and 1990s on the ice streams of the Siple Coast, where WAIS meets the Ross Ice Shelf. This sector drains a larger amount of ice from WAIS than any other, and in this period estimates of how quickly WAIS could collapse were based primarily upon how quickly these ice streams could transport ice from WAIS’s interior. This research involved answering questions such as how easily ice streams stagnated and revived, and how fast they could potentially move. Topics of investigation included the movement of the streams from year to year, and the friction at their beds and sides.
The movement of the Siple Coast ice streams is, in any case, constrained by the “buttressing” effect of the Ross Ice Shelf, which is almost the size of France. It is sufficiently far south that, while it could disintegrate (like Larsen B on the Antarctic Peninsula dramatically did in 2002), it was not expected to do so in the next century or so. Although it was known that there were no major ice shelves to buttress the glaciers draining into the Amundsen Sea, the region was not given substantial attention in this period, mainly because it is logistically difficult to access, but also because the ice flow through them is much less than through the Ross Ice Shelf and was assumed to be slow. In an interview with me, NASA’s Bob Bindschadler, one of the key leaders of work on the Siple Coast, emphasized that the University of Alaska’s Keith Echelmeyer (1954-2010) was instrumental in continually reminding researchers of the importance of investigating these glaciers.
Meanwhile, however, theoreticians began to regard the underlying principle that marine ice sheets are inherently unstable to be grounded in a presupposition that they were. Kees van der Veen wrote in a 1985 paper that Weertman’s model, expanded upon by glaciologist Bob Thomas, contained a “a natural built-in instability.” In 1993 the theoretician Richard Hindmarsh further reinforced this diagnosis. According to my conversations with glaciologists, it was not until Christian Schoof’s 2007 paper, “Ice sheet grounding line dynamics: steady states, stability, and hysteresis,” (pdf) that a case for the instability of marine ice sheets was convincingly grounded in basic physical principles.
Modelers likewise had trouble replicating WAIS disintegration. For many years, the main modeling tradition was based in Europe, and led by Hans Oerlemans, van der Veen (who is now in Kansas), and Philippe Huybrechts, whose three-dimensional model of Antarctic ice set the modeling standard for the 1990s. These models developed more or less independently of empirical research on WAIS, not least because models and computing power were not yet capable of handling ice sheet behavior at the scale of ice streams. It has only been since 2000 that new models have been developed by researchers such as Frank Pattyn (but also others) that have proven capable of coupling climate change to ice stream behavior and ice sheet behavior. It is this tradition of modeling that led to Ian Joughin’s model, which now clearly predicts WAIS disintegration through the Amundsen Sea embayment.
Then there are advances on the empirical side of the coin. In our interview, Bob Thomas (who conducted the first glaciological field investigations of the Ross Ice Shelf in the 1970s) remarked how he felt that attention was focused for too long on the Siple Coast at the expense of studying the ice sheet as a whole. But it was not until 2002 that he was able, with a Chilean team, to do an extensive survey of the Amundsen Sea sector from the air. Before that, though, the European Remote Sensing satellites (ERS-1 and 2) had also begun to survey Antarctica extensively, beginning in 1992. By the end of the 1990s, teams led by Duncan Wingham in the UK, and Eric Rignot at the Jet Propulsion Laboratory, used multiyear data to investigate the flow of the glaciers into the Amundsen Sea, discovering the rapidity of the drawdown of ice in that sector. It is the continuation of this work, of course, that has led to Rignot’s group’s current conclusions.
A couple years ago I asked both Wingham and Rignot via email whether these investigations were motivated by the question of possible WAIS collapse. Both, interestingly enough, answered that they were not; they were motivated by their desire to use data gathered from the specialized instrumentation on ERS-1. In any event, the evidence they gathered for a major drawdown captured their (and particularly Rignot’s) interest in the subject, and, within a few years, the interest of the researchers already working on the WAIS question. The results were certainly instrumental in building up support for a field research program on Pine Island Glacier, which is currently underway.
The final piece of our puzzle is the solidification of the links between climate change and WAIS collapse, which, in the years following Mercer’s paper had still not been intensively investigated until recently. (The deeply frigid environment of WAIS means that atmospheric warming is not expected to impact it directly as it does Greenland, e.g., via melting.) I admit, I am most fuzzy on this part of the history, but I gather that the critical advance was an increased oceanographic understanding of the role of deep-water currents reaching near the surface to melt the grounding lines of the Amundsen Sea glaciers. Stanley Jacobs of Columbia University seems to be a key figure here. While the instability of marine ice sheets is innate, and responsible for the irreversibility of the collapse process that now seems to be underway, anthropogenic warming has pretty clearly played a key role in bringing us to this point with respect to the Amundsen Sea embayment, and will continue to play a role in speeding the collapse process.
Note: this post is a very rough sketch of a set of research traditions that have expanded markedly in recent years. It will no doubt have glossed over the contributions of key figures, and will have done violence to the finer contours of the history. My article in Climatic Change on this subject is only slightly more satisfactory. There is a great deal of work that could be done to expand and deepen understanding of this history, and to broaden it beyond a narrow focus on the WAIS collapse question. Unfortunately, I have not managed to construct an opportunity to pursue this work.
The most recent (2011) review of the scientific literature is by Joughin and Richard Alley. As always, the comments section is open for anyone who would like to challenge my history, or to expand on its admittedly skeletal account.