West Antarctic Ice Sheet may be on “the brink of change”

A project to map the Weddell Sea area of the West Antarctic Ice Sheet from the air has revealed that this largely unexplored region is potentially on the threshold of change.

Radar mapping of the ice-covered landscape has uncovered a deep sub-glacial basin close to the edge of the ice sheet.

The study, published in Nature Geoscience, found that the basin measures 100 by 200 km and is well below sea level, nearly 2km deep in places. The ice sheet, currently grounded above the deep basin, may be more unstable than previously thought and could quickly undergo ice loss.

In a related paper published in Nature, models reveal that the Weddell Sea region may experience warmer ocean conditions at the end of the 21st century, which could provide the trigger for ice sheet change.

Professor Martin Siegert of the University of Edinburgh, who led the project, said: “This is a significant discovery in a region of Antarctica that at present we know little about. The area is on the brink of change, but it is impossible to predict what the impact of this change might be without further work enabling better understanding of how the West Antarctic Ice Sheet behaves.”

Our colleagues at the Canadian SMC collected the following expert commentary. Feel free to use these quotes in your reporting. If you would like to contact a New Zealand expert, please contact the SMC (04 499 5476; smc@sciencemediacentre.co.nz).

Dr. Jeremy Fyke, Postdoctoral Research Associate, Glacier  & ice sheet modeller, Los Alamos National Laboratory, Santa Fe, NM, USA. (Formerly with the Climate Modelling Group, Victoria University of Wellington, NZ).

“Antarctic ice shelves are massive floating glaciers that ring the Antarctic Ice Sheet.  These ice shelves receive ice from the world’s largest glaciers, which flow off the Antarctic continent and lose ice primarily through melting on their undersides when the come into contact with warm Southern Ocean water (little melting occurs on the surface of the ice shelves, because the Antarctic air is generally below freezing year-round).  Critically, if the ocean water under an ice shelf warms, even by a few degrees Celsius, the ice shelf will melt much faster.  If it melts faster, it will thin and cause an acceleration in the flow of grounded glacial ice into the ocean.  This acceleration of glacial flow, when applied to the massive glaciers that drain the Antarctic ice sheet, can easily result in multiple meters of global sea level rise over timescales of hundreds to thousands of years.  Recent geologic expeditions to the Antarctic coast have confirmed that this behavior last occurred millions years ago when climate conditions were very similar to those of the present day (but much cooler than temperatures that are predicted in the future under human-driven climate change).

“Thus, ocean-driven melting under Antarctic ice shelves is a critical process in the planetary climate system, because significant changes to Antarctic ice shelf melt rates can ultimately result in large shifts in global sea level.   However, until now a lack of advanced Southern Ocean models and a poor understanding of the sub-shelf marine environment has largely prevented the climate science community from predicting the impact of future human-induced climate change on Antarctic ice shelves.

“Now, Hellmer et al. (Nature, 2012) provide new scientific findings that for the first time quantify the impact of future human-induced climate change on Antarctic ice shelf behavior.  The authors use a high-resolution ocean model to accurately simulate Southern Ocean currents, especially those that flow past and under the large Antarctic ice shelves.  In particular they focus their attention on ocean patterns near the Ronne-Filchener Ice Shelf, which at 449,000 km2 (the size of France) is the 2nd-largest floating ice shelf on the planet.  When they run their model with predicted future human greenhouse gas emissions, a reduction of offshore marine sea ice drives large changes in Southern Ocean circulation patterns.  These circulation changes cause deep, warm ocean currents to rise and invade the marine cavity under the Ronne-Filchener Ice Shelf.  The resulting increase in sub-shelf water temperatures drives a whopping (1800%) increase in melt rate during the mid-21st century, which is capable of
rapidly thinning the massive ice shelf and potentially drastically accelerating ice loss from the Antarctic continent.

“The finding, of increased melt rates due to ocean circulation changes, is reproduced consistently in several simulations with independent ocean models and different emission scenarios. This reproducibility lends confidence to the study’s conclusions.  Furthermore, the increase in melting they project under the Ronne-Filchener Ice Shelf
is probably conservative given that ice shelves in their ocean model do not change shape with time and their computationally expensive simulations end at year 2200 or earlier.

“Ultimately, this study implies that regional Southern Ocean circulation changes due to human-caused planetary warming will cause dramatic increases in melt rates under the massive Ronne-Filchener Ice Shelf (and probably other Antarctic ice shelves as well). This melting has the potential to drive Antarctic ice loss and sea level to heights that have not been seen for millions of years.  Ultimately, it is a profoundly disturbing statement from the climate science community about the consequences of unabated greenhouse gas emissions.”

Dr. Martin Sharp, Arctic & Alpine Research Group, Department of Earth & Atmospheric Sciences, University of Alberta,  comment:

1. The important result is that the basal melting rate of the Ronne-Filchner ice shelf has the potential to change dramatically as a result of changes in the delivery of warm ocean water into the cavity beneath the ice shelf. The projected changes are large enough (roughly 20x increase in melt rate) to significantly alter the mass balance of
the entire Antarctic ice sheet. The projected higher annual mass loss rate from beneath the ice shelf is equivalent to roughly 80% of estimates of the current total annual net mass loss from the entire ice sheet.

2. This is a model based study. An ice-ocean model is driven by output from 2 versions of a single climate model, forced with 2 different IPCC atmospheric greenhouse gas concentration scenarios that lead to climate warming. It is not based on observations. In my view, this means that the study demonstrates something that potentially could happen under specific circumstances – it should not be viewed as making a definitive projection, or as showing something that already is happening. Different climate models forced in different ways might lead the ice-ocean model to produce quite different results – either more or less dramatic.

3. Having said that, the model results do demonstrate something that could be watched for in observations – and the changes suggested are sufficiently large that it might be a good idea to try to do this.

4. The mechanism involved is as follows: as the atmosphere warms, the difference in temperature between the atmosphere and the ocean is decreased, and the ocean losses less heat to the atmosphere as a result – so the ocean warms up. At the same time, downward longwave radiation from the warmer atmosphere increases and this reduces the thickness and concentration of sea ice in the Weddell Sea, allowing the ice pack to become more mobile.This increases stress transfer to the surface layer of the ocean and causes the surface current that flows around Antarctica to veer to the left towards the front of the the Ronne-Filchner ice shelf (due to the effect of the Earth’s rotation). As the flux of warm water into the cavity below the ice shelf increases, the mean temperature of water in the cavity increases, driving up the rate of melting of the base of the ice shelf.

5. The paper is clearly written and the study has been carefully conducted. The authors are clear (and realistic)  about both the implications and limitations of the results. We should take them for what they are and not  over-react.

Dr. Christian Schoof, Associate Professor, Department of Earth and Ocean Sciences, University of British Columbia, comments:

“This paper shows that we may see a drastic increase in melting of floating ice tongues off the coast of Antarctica during this century. If this does occur, it is likely that ice flow from the interior of Antarctica towards the coast will accelerate, pushing more ice into the ocean. Just like adding ice cubes to a glass of water raises the water level in the glass, pushing more ice into the oceans will increase sea levels.

“The potential of ice stored inland in Antarctica to raise sea levels is large: West Antarctica alone contains enough ice to raise sea levels on average by 3.5 metres. Sea level rise on that scale will not happen instantly, and is likely to take several hundred years. But sea level even on a smaller scale of tens of centimetres is enough to put coastal infrastructure and settlements into jeopardy.

“The paper by Hellmer and colleagues shows that warm water that naturally circulates around Antarctica could be forced closer to shore. This would occur when sea ice around Antarctica melts due to warming temperatures. Sea ice acts as a lid on the ocean. It stops winds from moving ocean water effectively. With the lid gone, winds are much more efficient at moving ocean water – think of how winds make the ocean surface rough, which doesn’t happen with a layer of ice between wind and water.

“If this water flows in close enough to reach the floating ice tongues, it will dramatically increase melting. We have seen this elsewhere in Antarctica, where the result has been dramatic ice loss from the continent.
What is new here is that this may happen in the big bays off West Antarctica, where the potential for ice loss is even larger.

“The methodology used is sound by today’s standards. It uses a computer model of ocean and atmosphere movements to calculate how much warm water will reach the bottom of the floating ice tongues. Confidence in the model comes from its ability to reproduce ocean behaviour during the 20th century [i.e. in the past]. As with all projections into the future, we will only be able to see exactly how well the model works for a warmer climate after the fact – so in 50 to a 100 years’ time.”

Dr. J. Graham Cogley, Professor of Geography, Department of Geography, Trent University, comments:

“We have known for at least ten years, and with growing confidence, that very rapid rates of mass loss from the frontage of the West Antarctic Ice Sheet on the Amundsen Sea [below the “AP” in Hellmer et al.’s Figure 1] are attributable to flows of warm water from the continental shelf offshore. Everything is relative: these “warm” waters are sometimes below the freezing point of fresh water. Nevertheless they are transferring large amounts of heat to the undersides of the ice shelves in that region.

“The result is intense thinning of the ice shelves – but they don’t get significantly thinner. Instead they exert less  buttressing of the grounded ice that feeds them. The grounded ice speeds up dramatically and, to cut a long story short, ice gets “sucked” out of the ice sheet and into the ice shelf at rates much faster than before the warm water arrived.

“This is why we should all be concerned. The ice in the ice shelf makes its contribution to sea-level rise when it flows across the grounding line – the line at which the grounded ice begins to float. The Amundsen coast of the West Antarctic Ice Sheet is a major player in present-day sea-level rise for this reason.

“The Hellmer study is a projection into future decades, not a description of what is happening now. But it is valuable because it dramatizes what is very likely to happen to a part of the margin of Antarctica that is relatively sheltered today. There is no reason to expect that continued global warming will not have consequences like those described by Hellmer and his co-authors, and on the scale that they suggest. The best time to start trying to avert these consequences was a couple of decades ago.

Dr. Shawn Marshall, Canada Research Chair in Climate Change, Cryosphere Climate Research Group, Department of Geography, University of Calgary, comments:

“This is solid new work which increases the emphasis on what glaciologists are just starting to understand: that in Antarctica, ice sheet collapse and sea level rise are all about the ocean.  The ice shelf in the Weddell Sea sector is one of the two ‘linchpin’ ice shelves in West Antarctica, and it would be a frightening prospect for global sea level rise if this unhinged. It has long been known that the West Antarctic Ice Sheet is intrinsically unstable, and it is currently losing mass, but I am not sure how many glaciologists expect the ice sheet to unravel this century. Their prediction that circumpolar deep water could get into this sector of the ice sheet this century is the first time anyone has put a timetable on large-scale ice sheet changes.  Even if there is only a 10% chance that this proves true, this has huge implications for forecasts of sea level rise.  It could easily double the current IPCC [International Panel on Climate Change] projections for 2100.

“The mechanisms of ocean-ice shelf interactions are only just being understood in the scientific community, but everything we have learned in the last decade (e.g., in the Amundsen Sea and in Greenland) tells us that this is a real threat to the ice sheet.  Certainly waters that are 2°C in contact with the ice shelf would lead to its quick demise. It is just a question of whether the Antarctic sea ice changes really occur as forecast in the model, as this leads to the increased mixing and shifts in ocean circulation that deliver the warm water to the ice sheet. So far, Antarctic sea ice has proven to be very resilient to climate change, so we have not yet seen the initial stages of what is projected in this climate model.”

Dr. Jeffrey Kavanaugh, Associate Professor, Department of Earth and Atmospheric Sciences, University of Alberta, comments:

“This paper investigates the impact that a warming climate will have on the circulation of waters in the Ronne-Filchner Ice Shelf region of the Weddell Sea basin.  Circulation patterns in this basin are important, as they control the extent to which the ice shelf is subjected to melt by (relatively) warm ocean currents. They also play
an important role in global ocean circulation patterns by generating the coldest, deepest current in the world’s oceans.  These currents are the result of a few details particular to the region, including: (1) the sea floor inland of the continental shelf slopes downward to the south – that is, towards the continent, rather than away from it;
(2) the presence of the ice shelf; and (3) that present-day wind and water circulation patterns here encourage the formation of sea ice. Because salt is rejected during sea ice formation, the freezing of surface waters increases the salinity (and hence density) of waters below, which then descend down-slope to contact the ice shelf where it
meets the sea floor.  Contact between this “shelf water” and the ice shelf results in modern-day basal melting rates averaging ~0.2 meters per year.

“Model investigations by the authors show that continued warming in the region will likely result in a reorganization of circulation patterns in the Weddell Sea basin.  The primary cause of these changes will be a decrease in local sea ice production, which will allow greater coupling between surface winds and water currents (even in the absence of increased wind speeds).  Model results indicate that as early as 2070, the relatively warm coastal ocean current will be consistently directed southwards, towards the Ronne-Filchner Ice Shelf via the Filchner Trough.  This will warm the waters contacting the underside of the ice shelf by ~2°C, resulting in a 20-fold increase in the rate of melting (to an average of ~4 meters per year).  This result appears robust, with the onset of these flow changes varying by a couple of decades depending on the climatic scenarios tested.  Two different coupled ice-ocean models of different architecture are also employed. Results for both are roughly similar, with the finite-element FESOM model (able to resolve finer features and additional dynamics) showing earlier increases in basal ice melting rates.

“The Antarctic Peninsula, which borders the Weddell Sea on its western edge, is a region of particularly rapid warming.  In the past two decades, several of the peninsula’s ice shelves have collapsed, including the Larsen B (2002) and Wilkins (2008) Ice Shelves.  Because the Ronne-Filchner Ice Shelf is situated further south, air temperatures are significantly colder than for the peninsular ice shelves, and surface melting is unlikely to soon become significant. The ice shelf is, however, subject to increased basal melt the temperature of waters contacting its underside increase.  This is precisely what the modeling work presented here suggests will happen.
This would likely result in a number of changes not explicitly modeled here, but that might further accelerate change of the ice sheet – including thinning of the ice shelf, acceleration of the ice streams feeding the ice shelf, and a retreat of the junction between floating and grounded portions of the ice shelf.  Changes along these lines are currently being observed in the Amundsen Sea region of West Antarctica, increasing concerns about the stability of the marine-based West Antarctic Ice Sheet (whose collapse could increase global sea level by several meters).”

Dr. Robert Bindschadler, Emeritus Scientist, (and former Chief scientist), Cryospheric Sciences Laboratory, Hydrospheric and Biospheric Sciences, NASA Goddard Space Flight Center, comments:

“Ice shelves are thinning in multiple locations around Antarctica and the grounded glaciers are responding to this by accelerating and thinning.  These connections have been emerging in various scientific papers over the past 5 years.  The limitation has been being able to see into the future.  Hellmer’s paper shows how this troubling process could rapidly spread to one of the really large ice shelves.   This is a BIG DEAL because the ocean’s impact on the ice shelves/ice sheets is THE way ice sheets can lose the largest amount of ice in the shortest amount of time.  We understand the climate links, but we know the details well enough.  This work is a large step in the right direction to be able to get credible numbers in the hands of those who will have to decide how to adapt to these changes in the next few decades.

“Once we do, we can model how the ocean and gravity fields adjust. (The redistribution of the water from these large masses of ice melting actually changes the Earth’s shape and its gravity field, as meltwater flows from poles into the ocean and across the globe).

“Once we can model the adjustment of ocean and gravity fields, it is at this point that coastal dwellers at any place on the planet will know whether they will see more than the average, less than the average or the average increase to mean global sea level.  The average rise is likely to be daunting-1 meter by the end of this century is certainly possible, although the first half-century will see “just” ? of this rise with the next 2/3 coming in the latter 50 years. Sea level rise is not constant — the rate of rise accelerates over time due to various positive feedback cycles. And this is just the beginning.  Ice sheets have always shrunk (and sea level has always risen) when the planet warms.  Reversing that trend will take centuries, so we had better prepare for adapting to rising sea level.  We have hardwired it into our long-term future.”