Sea level rise: New iceberg theory points to areas at risk of rapid disintegration

In events that could exacerbate sea level rise over the coming decades, stretches of ice on the coasts of Antarctica and Greenland are at risk of rapidly cracking apart and falling into the ocean, according to new iceberg calving simulations from the University of Michigan.

“If this starts to happen and we’re right, we might be closer to the higher end of sea level rise estimates for the next 100 years,” said Jeremy Bassis, assistant professor of atmospheric, oceanic and space sciences at the U-M College of Engineering, and first author of a paper on the new model published in the current issue of Nature Geoscience.

Iceberg calving, or the formation of icebergs, occurs when ice chunks break off larger shelves or glaciers and float away, eventually melting in warmer waters. Although iceberg calving accounts for roughly half of the mass lost from ice sheets, it isn’t reflected in any models of how climate change affects the ice sheets and could lead to additional sea level rise, Bassis said.

“Fifty percent of the total mass loss from the ice sheets, we just don’t understand. We essentially haven’t been able to predict that, so events such as rapid disintegration aren’t included in those estimates,” Bassis said. “Our new model helps us understand the different parameters, and that gives us hope that we can better predict how things will change in the future.”

The researchers have found the physics at the heart of iceberg calving, and their model is the first that can simulate the different processes that occur on both ends of the Earth. It can show why in northern latitudes-where glaciers rest on solid ground-icebergs tend to form in relatively small, vertical slivers that rotate onto their sides as they dislodge. It can also illustrate why in the southernmost places-where vast ice shelves float in the Antarctic Ocean-icebergs form in larger, more horizontal plank shapes.

The model treats ice sheets-both floating shelves and grounded glaciers-like loosely cemented collections of boulders. Such a description reflects how scientists in the field have described what iceberg calving actually looks like. The model allows those loose bonds to break when the boulders are pulled apart or rub against one another.

The simulations showed that calving is a two-step process driven primarily by the thickness of the ice.

“Essentially, everything is driven by gravity,” Bassis said. “We identified a critical threshold of one kilometer where it seems like everything should break up. You can think of it in terms of a kid building a tower. The taller the tower is, the more unstable it gets.”

Icebergs do have a tendency to form before that threshold though, Bassis suspects, due to cracks that are already there-either formed when capsizing bergs crash into the water and send shockwaves through the surrounding ice, or when melted water on the surface cuts through. The former is believed to have led to the Helheim Glacier collapse in 2003. The glacier had begun to retreat slowly in 2002, but suddenly gave way the following year when the thinner ice had broken away, exposing a thicker ice coast.

The latter-melted water pools-are occurring more frequently due to climate change, and they’re believed to have played a role in the rapid disintegration of the Antarctica’s Larsen B ice shelf, which crumbled over about six weeks in 2002.

When the researchers added random cracks to their model, it could mirror both Helheim and Larsen B.

A third feature is also required for the most dramatic ice collapses to occur. Icebergs can’t float away and make room for more icebergs to break off the main sheet unless the system has access to open water. So areas that border deep, unobstructed ocean rather than fjords or other waterways are at greater risk of rapid ice loss. The researchers point to the Thwaites and Pine Island glaciers in Antarctica and the Jakobshavn Glacier in Greenland, which is already retreating rapidly, as places vulnerable to “catastrophic disintegration” because they have all three components.

“The ice in those places gets thicker as you go back. If our threshold is right, then if these places start to retreat as you expose the thicker calving font, they’re susceptible to catastrophic breakup,” Bassis said.

Retreat of the current ice coasts in these places areas could occur via melting or iceberg calving.

Sea level rise: New iceberg theory points to areas at risk of rapid disintegration

In events that could exacerbate sea level rise over the coming decades, stretches of ice on the coasts of Antarctica and Greenland are at risk of rapidly cracking apart and falling into the ocean, according to new iceberg calving simulations from the University of Michigan.

“If this starts to happen and we’re right, we might be closer to the higher end of sea level rise estimates for the next 100 years,” said Jeremy Bassis, assistant professor of atmospheric, oceanic and space sciences at the U-M College of Engineering, and first author of a paper on the new model published in the current issue of Nature Geoscience.

Iceberg calving, or the formation of icebergs, occurs when ice chunks break off larger shelves or glaciers and float away, eventually melting in warmer waters. Although iceberg calving accounts for roughly half of the mass lost from ice sheets, it isn’t reflected in any models of how climate change affects the ice sheets and could lead to additional sea level rise, Bassis said.

“Fifty percent of the total mass loss from the ice sheets, we just don’t understand. We essentially haven’t been able to predict that, so events such as rapid disintegration aren’t included in those estimates,” Bassis said. “Our new model helps us understand the different parameters, and that gives us hope that we can better predict how things will change in the future.”

The researchers have found the physics at the heart of iceberg calving, and their model is the first that can simulate the different processes that occur on both ends of the Earth. It can show why in northern latitudes-where glaciers rest on solid ground-icebergs tend to form in relatively small, vertical slivers that rotate onto their sides as they dislodge. It can also illustrate why in the southernmost places-where vast ice shelves float in the Antarctic Ocean-icebergs form in larger, more horizontal plank shapes.

The model treats ice sheets-both floating shelves and grounded glaciers-like loosely cemented collections of boulders. Such a description reflects how scientists in the field have described what iceberg calving actually looks like. The model allows those loose bonds to break when the boulders are pulled apart or rub against one another.

The simulations showed that calving is a two-step process driven primarily by the thickness of the ice.

“Essentially, everything is driven by gravity,” Bassis said. “We identified a critical threshold of one kilometer where it seems like everything should break up. You can think of it in terms of a kid building a tower. The taller the tower is, the more unstable it gets.”

Icebergs do have a tendency to form before that threshold though, Bassis suspects, due to cracks that are already there-either formed when capsizing bergs crash into the water and send shockwaves through the surrounding ice, or when melted water on the surface cuts through. The former is believed to have led to the Helheim Glacier collapse in 2003. The glacier had begun to retreat slowly in 2002, but suddenly gave way the following year when the thinner ice had broken away, exposing a thicker ice coast.

The latter-melted water pools-are occurring more frequently due to climate change, and they’re believed to have played a role in the rapid disintegration of the Antarctica’s Larsen B ice shelf, which crumbled over about six weeks in 2002.

When the researchers added random cracks to their model, it could mirror both Helheim and Larsen B.

A third feature is also required for the most dramatic ice collapses to occur. Icebergs can’t float away and make room for more icebergs to break off the main sheet unless the system has access to open water. So areas that border deep, unobstructed ocean rather than fjords or other waterways are at greater risk of rapid ice loss. The researchers point to the Thwaites and Pine Island glaciers in Antarctica and the Jakobshavn Glacier in Greenland, which is already retreating rapidly, as places vulnerable to “catastrophic disintegration” because they have all three components.

“The ice in those places gets thicker as you go back. If our threshold is right, then if these places start to retreat as you expose the thicker calving font, they’re susceptible to catastrophic breakup,” Bassis said.

Retreat of the current ice coasts in these places areas could occur via melting or iceberg calving.

Researchers shed new light on supraglacial lake drainage

Supraglacial lakes – bodies of water that collect on the surface of the Greenland ice sheet – lubricate the bottom of the sheet when they drain, causing it to flow faster. Differences in how the lakes drain can impact glacial movement’s speed and direction, researchers from The City College of New York (CCNY), University of Cambridge and Los Alamos National Laboratory report in “Environmental Research Letters.”

“Knowledge of the draining mechanisms allows us to improve our understanding of how surface melting can impact sea-level rise, not only through the direct contribution of meltwater from the surface, but also through the indirect contribution on the mass loss through ice dynamics,” says Dr. Marco Tedesco, the principal investigator and lead author.

Dr. Tedesco is an associate professor in CCNY’s Department of Earth and Atmospheric Sciences at CCNY and is currently serving as temporary program director for the National Science Foundation’s Polar Cyberinfrastructure Program. The research described in the paper was funded before Dr. Tedesco accepted the position at NSF.

NSF supported the research along with NASA’s cryosphere program, the Natural Environment Research Council, the U.S. Department of Energy’s earth systems modeling program, St. Catherine’s College (Cambridge), the Scandinavian Studies Fund and the B.B. Roberts Fund.

Over the past decade, surface melting in Greenland has increased considerably.

Previous research already suggested that the water injected from the rapid draining of the supraglacial lakes controlled sliding of ice over the bed beneath it. However, there was no evidence of the impact of the slow draining mechanism, which the paper identified.

Professor Tedesco and colleagues documented that supraglacial lakes have two different drainage mechanisms that cause them to empty rapidly or slowly. The findings are based on analysis of data collected in 2011 from five GPS stations the team installed around two supraglacial lakes in the Paakistoq region of West Greenland.

The smaller of the two lakes, Lake Half Moon, overflowed its banks and drained from the side to reach a moulin. It took approximately 45 hours to empty. The larger lake, Lake Ponting, drained through a crack in the ice beneath it and was voided in around two hours.

“At first, a crack in the ice beneath the lake may be small, but it deepens as water enters it because the pressure of the water overcomes the compressive action of the ice, which is trying to close the crack,” Professor Tedesco explains. “When the crack reaches the bed beneath the glacier, which could be 1,000 meters or more below the surface, the lake empties rapidly, like a bathtub after its plug is pulled.”

Drainage from both lakes accelerated glacial movement. However, water from Lake Ponting caused the glacier to move faster and further. While the slower drainage from Lake Half Moon caused the glacial pace to increase from baseline values of 90 – 100 meters per year to a maximum of around 420 meters a year, glacial movement in the area affected by Lake Ponting reached maximum velocities of 1,500 – 1,600 meters per year, nearly four times greater.

The drainage of the two lakes impacted the glacier’s trajectory differently, as well. The emptying of Lake Half Moon via the moulin did not change the direction of glacial movement. However, when Lake Ponting drained a slight southerly shift in the glacier’s direction was detected.

“Because the different draining mechanisms affect ice velocity, they could also affect the amount of ice lost through calving of glaciers, which results in icebergs,” Professor Tedesco points out. “Because what happens on a glacier’s surface impacts what is going on below, researchers are trying to look at glaciers as a system instead of independent components,” he adds.

“The surface is like the skin of a tissue and the subglacial and englacial channels that develop because of the surface water act like arteries or veins that redistribute this water internally.”

Researchers shed new light on supraglacial lake drainage

Supraglacial lakes – bodies of water that collect on the surface of the Greenland ice sheet – lubricate the bottom of the sheet when they drain, causing it to flow faster. Differences in how the lakes drain can impact glacial movement’s speed and direction, researchers from The City College of New York (CCNY), University of Cambridge and Los Alamos National Laboratory report in “Environmental Research Letters.”

“Knowledge of the draining mechanisms allows us to improve our understanding of how surface melting can impact sea-level rise, not only through the direct contribution of meltwater from the surface, but also through the indirect contribution on the mass loss through ice dynamics,” says Dr. Marco Tedesco, the principal investigator and lead author.

Dr. Tedesco is an associate professor in CCNY’s Department of Earth and Atmospheric Sciences at CCNY and is currently serving as temporary program director for the National Science Foundation’s Polar Cyberinfrastructure Program. The research described in the paper was funded before Dr. Tedesco accepted the position at NSF.

NSF supported the research along with NASA’s cryosphere program, the Natural Environment Research Council, the U.S. Department of Energy’s earth systems modeling program, St. Catherine’s College (Cambridge), the Scandinavian Studies Fund and the B.B. Roberts Fund.

Over the past decade, surface melting in Greenland has increased considerably.

Previous research already suggested that the water injected from the rapid draining of the supraglacial lakes controlled sliding of ice over the bed beneath it. However, there was no evidence of the impact of the slow draining mechanism, which the paper identified.

Professor Tedesco and colleagues documented that supraglacial lakes have two different drainage mechanisms that cause them to empty rapidly or slowly. The findings are based on analysis of data collected in 2011 from five GPS stations the team installed around two supraglacial lakes in the Paakistoq region of West Greenland.

The smaller of the two lakes, Lake Half Moon, overflowed its banks and drained from the side to reach a moulin. It took approximately 45 hours to empty. The larger lake, Lake Ponting, drained through a crack in the ice beneath it and was voided in around two hours.

“At first, a crack in the ice beneath the lake may be small, but it deepens as water enters it because the pressure of the water overcomes the compressive action of the ice, which is trying to close the crack,” Professor Tedesco explains. “When the crack reaches the bed beneath the glacier, which could be 1,000 meters or more below the surface, the lake empties rapidly, like a bathtub after its plug is pulled.”

Drainage from both lakes accelerated glacial movement. However, water from Lake Ponting caused the glacier to move faster and further. While the slower drainage from Lake Half Moon caused the glacial pace to increase from baseline values of 90 – 100 meters per year to a maximum of around 420 meters a year, glacial movement in the area affected by Lake Ponting reached maximum velocities of 1,500 – 1,600 meters per year, nearly four times greater.

The drainage of the two lakes impacted the glacier’s trajectory differently, as well. The emptying of Lake Half Moon via the moulin did not change the direction of glacial movement. However, when Lake Ponting drained a slight southerly shift in the glacier’s direction was detected.

“Because the different draining mechanisms affect ice velocity, they could also affect the amount of ice lost through calving of glaciers, which results in icebergs,” Professor Tedesco points out. “Because what happens on a glacier’s surface impacts what is going on below, researchers are trying to look at glaciers as a system instead of independent components,” he adds.

“The surface is like the skin of a tissue and the subglacial and englacial channels that develop because of the surface water act like arteries or veins that redistribute this water internally.”

The contribution of the Greenland ice sheet to sea-level rise will continue to increase

New research has shown surface ice melt will be the dominant process controlling ice-loss from Greenland. As outlet glaciers retreat inland the other process, iceberg production, remains important but will not grow as rapidly.

The Greenland ice sheet is often considered an important potential contributor to future global sea-level rise over the next century or longer. In total, it contains an amount of ice that would lead to a rise of global sea level by more than seven metres, if completely melted.

Changes in its total mass are governed by two main processes – fluctuations in melting and snowfall on its surface, and changes to the number of icebergs released from a large number of outlet glaciers into the ocean.

The ice loss from the ice sheet has been increasing over the last decade, with half of it attributed to changes in surface conditions with the remainder due to increased iceberg calving – the process by which ice detaches from the glacier to become an iceberg.

Researchers from the Vrije Universiteit Brussel, funded by ice2sea, a European Union project, tackled the question of how both processes will evolve and interact in the future. This was done with a computer model, which projects the future ice sheet evolution with high accuracy using the latest available techniques and input data.

They devised a method to generalize projections made in earlier research which concerned just four of Greenland’s outlet glaciers. By doing so they could apply the earlier findings to all calving glaciers around the Greenland ice sheet. Their results indicate a total sea-level contribution from the Greenland ice sheet for an average warming scenario after 100 and 200 years of 7 and 21 cm, respectively.

The balance between the two processes by which ice is lost is, however, changing considerably in the future so that iceberg calving may only account for between 6 % and 18 % of the sea-level contribution after 200 years. This is important, because variations in outlet glacier dynamics have often been suspected to have the potential for very large sea-level contributions.

Lead author Dr Heiko Goelzer, of the Vrije Universiteit Brussel, says,

“Our research has shown that the balance between the two most important mass loss processes will change considerably in the future so that changes in iceberg calving only account for a small percentage of the sea-level contribution after 200 years with the large remainder due to changes in surface conditions.”

The limited importance of outlet glacier dynamics in the future is the result of their retreat back onto land and of strongly increasing surface melting under global warming, which removes ice before it can reach the marine margin.

Ice2sea coordinator Professor David Vaughan, of the British Antarctic Survey says,

“This scenario is no reason to be complacent. The reason the significance of calving glaciers reduces compared to surface melting is, so much ice will be lost in coming decades that many glaciers currently sitting in fjords will retreat inland to where they are no longer affected by warming seas around Greenland.”

The contribution of the Greenland ice sheet to sea-level rise will continue to increase

New research has shown surface ice melt will be the dominant process controlling ice-loss from Greenland. As outlet glaciers retreat inland the other process, iceberg production, remains important but will not grow as rapidly.

The Greenland ice sheet is often considered an important potential contributor to future global sea-level rise over the next century or longer. In total, it contains an amount of ice that would lead to a rise of global sea level by more than seven metres, if completely melted.

Changes in its total mass are governed by two main processes – fluctuations in melting and snowfall on its surface, and changes to the number of icebergs released from a large number of outlet glaciers into the ocean.

The ice loss from the ice sheet has been increasing over the last decade, with half of it attributed to changes in surface conditions with the remainder due to increased iceberg calving – the process by which ice detaches from the glacier to become an iceberg.

Researchers from the Vrije Universiteit Brussel, funded by ice2sea, a European Union project, tackled the question of how both processes will evolve and interact in the future. This was done with a computer model, which projects the future ice sheet evolution with high accuracy using the latest available techniques and input data.

They devised a method to generalize projections made in earlier research which concerned just four of Greenland’s outlet glaciers. By doing so they could apply the earlier findings to all calving glaciers around the Greenland ice sheet. Their results indicate a total sea-level contribution from the Greenland ice sheet for an average warming scenario after 100 and 200 years of 7 and 21 cm, respectively.

The balance between the two processes by which ice is lost is, however, changing considerably in the future so that iceberg calving may only account for between 6 % and 18 % of the sea-level contribution after 200 years. This is important, because variations in outlet glacier dynamics have often been suspected to have the potential for very large sea-level contributions.

Lead author Dr Heiko Goelzer, of the Vrije Universiteit Brussel, says,

“Our research has shown that the balance between the two most important mass loss processes will change considerably in the future so that changes in iceberg calving only account for a small percentage of the sea-level contribution after 200 years with the large remainder due to changes in surface conditions.”

The limited importance of outlet glacier dynamics in the future is the result of their retreat back onto land and of strongly increasing surface melting under global warming, which removes ice before it can reach the marine margin.

Ice2sea coordinator Professor David Vaughan, of the British Antarctic Survey says,

“This scenario is no reason to be complacent. The reason the significance of calving glaciers reduces compared to surface melting is, so much ice will be lost in coming decades that many glaciers currently sitting in fjords will retreat inland to where they are no longer affected by warming seas around Greenland.”

Data show Antarctic ice stream radiating seismically





Douglas Wiens (left) and a colleague ready equipment to emplace seismographs in Antarctica during a 2001 expedition. Data gathered for this project, called TAMSEIS, provided evidence that an Antarctic ice stream radiates seismic waves twice daily that are equivalent to a magnitude seven earthquake. - Image courtesy of Doug Wiens
Douglas Wiens (left) and a colleague ready equipment to emplace seismographs in Antarctica during a 2001 expedition. Data gathered for this project, called TAMSEIS, provided evidence that an Antarctic ice stream radiates seismic waves twice daily that are equivalent to a magnitude seven earthquake. – Image courtesy of Doug Wiens

A seismologist at Washington University in St. Louis along with colleagues at Pennsylvania State University and Newcastle University in the United Kingdom have found seismic signals from a giant river of ice in Antarctica that make California’s earthquake problem seem trivial.



Douglas A. Wiens, Ph.D., professor of earth and planetary sciences in Arts & Sciences, and colleagues combined seismological and global positioning system (GPS) analyses to reveal two bursts of seismic waves from an ice stream in Antarctica every day, each one equivalent to a magnitude seven earthquake. The GPS analyses were performed by Pennsylvania State and Newcastle University researchers.



The ice stream is essentially a giant glacier 60 miles wide and one-half mile thick. The data show that the river of ice moves about 18 inches within ten minutes, remains still for 12 hours, then moves another eighteen inches. Each time it moves, it gives off seismic waves that are recorded at seismographs all around Antarctica, and even as far away as Australia.



Seismic waves from what are loosely called “glacial earthquakes,” mainly near Greenland, were originally reported in 2003, and the numbers have been increasing in recent years. Some scientists think the waves come from a phenomenon of calving, during which a big chunk of ice breaks off of a glacier and floats away in the ocean. This is a very violent activity that could generate strong seismic signals. The new results show, however, that at least some of the glacial earthquakes are produced by sudden sliding off large ice sheets.



The Antarctic signals were first detected by seismographs deployed by Wiens and his colleagues in 2001-2003 at a location about 500 miles away from the ice stream.



“At first we didn’t know where the waves were coming from, but eventually we were able to narrow down the source to the ice stream.” Wiens said.

Slower than a real earthquake



Prior to this discovery, researchers were not aware that ice streams radiated seismic waves.



“By some measures, the seismic impact is equivalent to a very large earthquake, but it doesn’t feel like it because the movement is much slower than a real earthquake,” Wiens said. “The data look an awful lot like an earthquake, but the slip lasts for 10 minutes, while on the other hand an earthquake of this size would last for just ten seconds. I guess you could call it an earthquake at glacial speed. This is very strange behavior, and we need to understand more about it.”



GPS instruments placed directly on the ice stream can detect where the slipping motion begins and where it stops. Scientists describe the motion as “stick-slip,” which is the classic motion of earthquakes. Stick-slip occurrs when the area around a fault moves slowly but the fault is stuck, remaining stationary until the stress builds up and the fault finally slips.



“The GPS shows us directly how the ice stream moves,” Wiens said. “The slip starts in a certain part of the ice stream and then it moves out, rather like a landslide might start at a certain point and then move out to envelop an entire mountainside. The GPS tells us which part moved first and what other parts moved next and so forth.”



The data show that the slip always starts from the same spot on the bed of the ice stream – what glaciologists call a “sticky” spot – which has more friction than the surrounding part of the bed.


A slip, not a creep



“Glaciologists had thought that they understood how glaciers move, and they thought the ice moves move slowly and continuously by creep, but now this indicates that they move with a fast slip, almost like an earthquake,” Wiens said.



Wiens said that it is important to understand the physics behind what is controlling this kind of slip.



“This stick-slip phenomenon may provide a clue about what makes these ice streams move faster or slower,” Wiens said. “This particular ice stream has been slowing down over the last few decades, and no one knows why. ”



Wiens plans to study seismic records of stick-slip events going back several decades to see if there are changes, and also to search for similar signals from other ice streams.



“We need to understand what controls the speed of the ice streams, because that factor will affect how fast the ice in Antarctica will go away and sea level will rise as global warming melts the West Antarctic Ice Sheet.”



The study was published in the June 5 issue of Nature Online and was funded by the National Science Foundation.