Arguments
East Antarctica Ice-Sheet more vulnerable to melting than we thought: new research
Posted on 31 July 2013 by John Mason
Missing contributor to the 22 +/- 10m Pliocene sea-level rise identified
We know from satellite measurements that the Greenland and West Antarctic ice sheets (GIS and WAIS respectively) are losing mass in response to global warming, and that, in the case of the partly sea-based West Antarctica ice-sheet, basal melting of the ice by warmer ocean-water is likely to be a key mechanism. In the case of the East Antarctica Ice-Sheet (EAIS), the situation has been less clear: thinning of ice shelves and acceleration of glaciers have been described in some areas but it has to date given an impression of relative stability. New research, however, has found that it might not be as resilient to warming as we thought, especially in areas where the bedrock is low-lying.
The research, published in July 2013 in the journal Nature, concerns data collected from marine sediments comprising much (5.3-3.3 million years ago) of the Pliocene Series (spanning 5.3-2.588 million years ago) off the coast of East Antarctica. Its key finding is that during the Pliocene there occurred a series of long, warm intervals during which parts of the East Antarctic Ice-Sheet margin retreated hundreds of kilometres inland. This finding is of importance to our understanding of future global warming and its effects, because the climate during the Pliocene was similar to that predicted for the latter part of the current century and atmospheric greenhouse gas concentrations were similar to those of the present day.
Modelling has already suggested that low-lying areas of the EAIS are candidate zones for Pliocene melt. In fact, some extra ice-melt is likely required to explain global sea-level changes during the Pliocene. That there were periods of significant sea-level rise is understood, but estimates of the amount vary, leading to the figure of 22 +/- 10 metres. What seems likely, however, is that sea-level rise due to the collapse of the GIS and WAIS would, at around 12m, appear to be insufficient to accomplish such an inundation. That these ice-sheets periodically collapsed during Pliocene times has been covered this year at Skeptical Science, here.
Looking for indicators
What kind of evidence would be an indicator of ice-sheet margin retreat? One line of enquiry would be to look for sediments that have come from an area now buried by ice. Retreat involves melting and meltwaters erode the bedrock and transport it as sediments - sands, gravels, silts and muds - down to the sea. So if you want to check whether erosion and sediment transport were taking place during the Pliocene, the best bet is to sample the sediments on the sea-floor dating from that time, by drilling through them and taking cores. This is what the research team did. A borehole was drilled at Integrated Ocean Drilling Program Site U1361 (fig. 1), in 3,465 m of water, situated 310 km offshore of the Adélie Land margin of East Antarctica. A continuous section through the Pliocene, comprising some 75m of drillcore, was recovered.
Fig. 1: map of Antarctica, showing the approximate location of the submarine borehole drilled into the Pliocene sediments. The shaded area is discussed in some detail further into this article.
The Pliocene sedimentary sequence in the drillcore consisted of sixteen distinct layers, alternating between diatom-rich silty clay and diatom-poor clay layers with silt laminations. Diatom-rich sediments point to multiple extended periods of increased biological productivity related to less sea ice and warmer spring and summer sea surface temperatures. This sequence is in good agreement with data from marine and land-based records from the Antarctic Peninsula margin, the Ross Sea and other areas. Reconstructions indicate that during the Pliocene there were prolonged (~200,000 years) warm intervals, when spring and summer sea surface temperatures were between 2 and 6oC above modern levels, with a general background theme of warmer-than-present temperatures.
Provenancing - finding where sediments were eroded from
In order to provenance sediment sources, it is necessary to look at both the composition of the sediment and the geological characteristics of the adjacent land. The part of East Antarctica that is relevant in this case consists of a number of distinct geological terranes with rocks dating from deep in the Precambrian to the Cenozoic occurring in different areas. Areas now covered in ice have been mapped in terms of their magnetism by aeromagnetic surveys.
When sediments are fine-grained, as these clays and silts are, geochemistry is a useful method. The team examined the neodymium and strontium isotope ratios of the sediments and compared them to the known values for the various geological terranes on land. They found that the diatom-rich sediments, deposited during the warm intervals, were predominantly composed of material from one terrane - the Jurassic to Cretaceous volcanic rocks and associated sedimentary rocks of the Ferrar Large Igneous Province (FLIP). This FLIP 'fingerprint' was found to be restricted to the Pliocene warm intervals and was absent from the overlying younger sediments.
Only one problem - from where, now under the ice inland, was the FLIP material eroded? Turning to the aeromagnetic data, they identified a district whose aeromagnetic characteristics strongly resembled those from an area of exposed FLIP bedrock in South Victoria Land. This was the Wilkes Subglacial Basin (fig.2) - or rather basins - for it contains two deep (~2km) trough-like structures known as grabens: recent subglacial topographic data compilations demonstrate that these troughs are directly connected to the Southern Ocean below sea level. Furthermore, recent geophysical data from aerial surveying suggest that one of the two troughs - known as the Central Basin - contains unconsolidated sediments thought to have been derived from the FLIP.
Fig. 2: topographical map of Antarctica showing bedrock heights beneath the ice (highest ground is deep red, lowest deep blue): the Wilkes Basin, highlighted, is below sea-level. From the Bedmap 2 Project.
The authors propose that the enhanced erosion of FLIP material in the Central Basin required multiple retreats of the ice margin inland by several hundred kilometres. Both modelling and observations suggest that ice-sheet retreat, in areas where the ice-base lies beneath sea-level, is driven by sub-surface melting at the ice edge, in response to warm ocean temperatures. Warm Pliocene ocean waters would have facilitated such a retreat of this part of the EAIS into the Central Basin, occurring at the same time as ice shelf collapse and ice margin retreat in other circum-Antarctic locations, such as the Ross Sea.
Sea level rise implications in the Pliocene and in the coming centuries
What contribution to Pliocene sea-level rise would this EAIS retreat have had? Existing ice sheet models imply that the figure is somewhere between 3 and 16 metres. The new findings will aid the refinement of such modelling: more importantly, the data from the Pliocene show that parts of the EAIS are more capable of dynamic and radical responses to warmer climate conditions than was thought to be the case. As with the research into the Pliocene Arctic, the results point out a stark scenario:
* the Pliocene was a time when greenhouse gases were at similar levels to today;
* the Pliocene was generally warmer than today and sometimes considerably warmer for prolonged periods;
* the Pliocene saw the repeated collapses of the Greenland and West Antarctic Ice-sheets during the prolonged warmer periods;
* the Pliocene also saw ice-sheet margins retreating significantly in East Antarctica during the prolonged warmer periods;
* the Pliocene saw global ice retreat leading to global sea level rise of 22 +/- 10 metres during the prolonged warmer periods.
The implications of the points above are clear: if we want to avoid transitioning to a warm Pliocene climate, including substantial sea-level rise, we need to make the transition back from the Pliocene atmosphere that we have created in the past decades, to one that favours relative climate stability.
John, thanks for another excellent post. The USGS, in one estimate, has stated that a total collapse of both the Greenland and Antarctic ice sheets could potentially raise eustatic sea-level 260 feet. Of course this will not happen over night, but there is mounting evidence that the process(es) have started in both the Northern (Greenland) and Southern (Antarctica) Hemispheres . And of course, the vast majority of mountain (valley, piedmont, alpine) glaciers are also retreating rapidly. The situation needs to be brought to the public and politicians and through posts such as yours and others at SkS (and RealClimate.org, Tamino, and others) are helping. Tom
Thanks John, for the summary of this very interesting, I would say a breakthrough article: we are gathering more and more evidence from paleo to our understanding of climate change.
I guess, the they are quoting eustatic SLR, right? I don't have access to the full text so cannot find out, myself. I'm also interested in the detail where the relatively large uncertainties (+-10m) come from. If we are take into account the SLR adjustment to the changed gravitational pull of the melting IS, then the SLR number should go higher although I'm not sure by how much.
Hi Chris - yes the uncertainties are large,much of which is down to the fact that isostatic adjustment may have affected raised shorelines in either direction. One good example of that is in E England where there are sea-cliffs composed of Pliocene/early Pleistocene fossiliferous marine sediments and the deposits extend a fair way inland, although the area in question is slowly sinking at present.
It is certainly possible and may be likely for the polar ice sheets to disappear, causing sea level rise (SLR) of 22 +/- 10 metres over coming millennia. Of more immediate concern is what can be expected to occur over the course of this century. The Letter from C.P. Cook et al (2013) implies that what occurred in the Pliocene is a reasonable indicator to what may happen in the immediate future. That seems questionable, as is the suggestion that some 50% of future SLR could come from ice mass loss from the East Antarctic Ice Sheet.
SLR by 2100 is more likely to come from ice mass loss from West Antarctica (WAIS) where warm ocean currents are already melting ice at glacier mouths and attacking areas of the WAIS resting on the seabed. Atmospheric warming does not appear to contribute to ice mass loss from either the EAIS or WAIS, other than the “Peninsula”.
This is not the case in the Arctic where loss of ice from the Greenland Ice Sheet (GIS) and Canadian Islands is caused by rising atmospheric temperature and a warming Arctic ocean. The latter is caused by penetration of warmer sea currents and loss of albedo causing increased exposure to sunlight. Further, loss of land based ice is more likely to accelerate due to Arctic amplification contributed to by methane emissions and evidenced by temperature rise at over twice the global average.
By contrast atmospheric temperature amplification is not evident in the Antarctic which is insulated by relatively stable circumpolar winds, persistent sea-ice coverage and the loss of tropospheric ozone. All have the effect of maintaining the coldest atmospheric temperatures in the world. Warmer bottom currents from the tropics do reach the EA coast and there is evidence that these enable increased ice loss from some EAIS glaciers. However, the EAIS is entirely land based and, unlike the WAIS which is a marine ice sheet, relatively impervious to warm ocean currents.
Both the WAIS and EAIS are loosing ice mass but the latter is doing so at a much slower rate. For these reasons it is argued that SLR to 2100 is most likely to come from the GIS with exposure to Arctic amplification and WAIS which is vulnerable and exposed to warm ocean currents. EAIS seems unnlikely to be a major contributor this century.
Finally, is it legitimate to compare conditions during the Pliocene, which took hundreds of millennia to evolve, with present conditions which have taken just a few decades to evolve thanks to human intervention. Do present EAIS conditions equate to those which prevailed in the Pliocene?
Agnostic,
I agree with you that nothing drastic is likely to happen to the EAIS for the time being, with sea-level rise coming from the other sources you cite.
However, I think we need to take a good hard look at the Pliocene, because we have driven one parameter straight into that era in a matter of a few centuries. How our current climate evolves in response to having a Pliocene atmosphere imposed upon it remains to be seen, but we need to be aware of what is possible....