How low will it go? Scientists to use improved forecast system to predict Arctic ice melt

How low will it go? Scientists to use improved forecast system to predict Arctic ice melt

CPOM scientists at Reading University have been closely involved in developing a new, more accurate forecast system to predict whether Arctic sea ice will fall to a record low level this summer, described in a study published in the journal Earth’s Future.

Arctic ice extent is already at its lowest winter level ever recorded, ahead of the big melt that takes place every spring and summer.

It has always been difficult to estimate exactly how much ice will disappear by the end of summer, but this new forecasting model based, on satellite measurements, will allow improved predictions to be made much earlier.

CPOM researcher David Schroeder, one of the study’s authors, said: “Our work will allow us to accurately predict the summer minimum levels of Arctic ice in April, rather than June as is currently the case. The new forecast model takes into account far more data, such as ice concentration, and focuses on key locations that offer clues to the bigger picture.

“But the proof is in the pudding. We have already put our theory to the test using historic data, but this will be the first time it has been used to predict future melting so far ahead. The timing is apt, too, as our forecast should give a strong indication of whether we are about to see a record low summer ice extent this September.”

Arctic sea ice. Credit: ESA

Accurate forecasts of summer Arctic sea ice are valuable for shipping companies and native people that depend on sea ice for hunting, while many animal and plant species are impacted directly by changing conditions.  However, uncertain weather through spring and summer make the forecasting of Arctic sea ice for a given year extremely challenging.

The new model analyses the physical characteristics of the sea ice cover as the melt season develops, to more precisely estimate if the amount of sea ice come September will be more or less than expected from the long-term trend.

The study found that the forecasts based on melt onset at the time at which sea ice starts to melt and open water appears in the Arctic Ocean, were most reliable in early spring, while sea ice coverage-based predictions were more reliable from June onwards.

The forecasts focus specifically on regions that historically correspond with how much sea ice remains come the September minimum extent. The predictions become more accurate with each passing month, as the model integrates more near-real-time information about sea ice melt and the distribution of open water areas across the Arctic Ocean and surrounding seas.

To test the model, the team went back in time and made predictions for each year of the satellite record using historical data of the Arctic sea ice conditions. They then evaluated the results against both the actual minimum extent for that year and what the long-term trend would have predicted.

Lead author Dr Alek Petty, a sea ice researcher at NASA’s Goddard Space Flight Center, said: “We found that our forecast model does much better than the linear trend at capturing what actually happened to the sea ice in any specific year. Our model is very good at catching the highs and the lows.

“The absolute values? Not exactly, but it tends to do very well at seeing when the sea ice extent is going to go up and when it’s going to go down compared to what we might be expecting for that year.”

The research also showed that models can produce reliable forecasts of sea ice not only for the whole Arctic, but for discrete regions; specifically, the Beaufort and Chukchi seas north of Alaska.

“The state of sea ice has a large impact on the Alaskan hunting communities,” Dr Petty said. “If they know ahead of time what the sea ice cover is going to be like that year, they might be able to infer the availability of the species they hunt.”

Future research will explore synthesising different sea ice measurements into the same model to improve the reliability of the forecasts.

Full reference:

A. Petty, D. Schröder, J. Stroeve, T. Markus, J. Miller, N. Kurtz, D. Feltham, D. Flocco (2017). ‘Skillful spring forecasts of September Arctic sea ice extent using passive microwave sea ice observations. Earth’s Future. DOI: 10.1002/2016EF000495

Satellites track variations in retreat of West Antarctic glaciers

An almost 25-year long record of elevation change of some of Antarctica’s fastest flowing and receding glaciers reveals important differences in the timing of retreat between single glaciers and the pace by which the ice loss spreads to Antarctica’s interior, according to a study now published in Geophysical Research Letters.

Combining data from five separate satellite missions, research carried out by CPOM shows how these glaciers have been thinning in patterns spreading upwards from the ice edge.

Focusing on Pine Island Glacier – which contributes more to sea level rise than any other ice stream on the planet – alongside the neighbouring Thwaites and smaller Pope, Smith and Kohler (PSK) Glaciers, the study shows how glacier heights have been falling, most likely as the glaciers are thinning due to the warmer sea temperatures recorded around Antarctica in recent decades. What is particularly interesting, however, is that they have all been responding differently over the last two decades.

Lead author Dr Hannes Konrad explained: “Scientists generally agree that it is warm ocean water that melts the floating part of the glacier, which then allows the glacier to flow more easily because it’s no longer held back by the floating ice shelf. As the glacier flows faster, it starts to become thinner.”

“If there’s not enough snow and ice accumulating higher up to compensate, the glaciers lose more and more of their mass as they flow towards the sea, and that’s exactly what we are seeing here, but the detail varies considerably between the three systems, and even within each glacier.”

Bringing together data from 1992 (when satellite altimetry records began) to the present day, the team reconstructed surface heights along a series of flowlines to see how thinning at the grounding lines had affected the glacier further inland.

By 1992, all three were already experiencing height loss at or near the grounding line, with Pine Island Glacier losing height by around 1m every year, although the interior’s surface was stable. Thinning at Pine Island Glacier then spread steadily, firstly up the glacier’s main trunk, and then further inland.  While the pace at which it spread across the glacier surface varied, rates of thinning reached up to 13km/yr in places.

Changes at Thwaites Glacier were more erratic: the surface at the grounding line was already falling by up to 3 m/yr at the start of the satellite record, but thinning ceased around 2000. Since restarting in 2004, thinning has spread at similar rates to those seen at Pine Island Glacier, but the offset of about ten years means that it has not spread so far inland.

Finally, although PSK experienced the largest falls in surface height (up to 7m/yr), most likely beginning around 1980, the thinning spread much more slowly than at Pine Island Glacier or Thwaites Glacier.

Possible reasons include differences in glacier catchment size, bedrock, topography and hydrology, but what remains clear is that, over the past 25 years, all three systems have seen thinning which has spread from the grounding line across the glacier surface.

With climate models indicating that ocean warming is set to continue and the glaciers expected to retreat further, even without the influence of warm water due to their resting on downwards-sloping bedrock, further ice melt at the glacier margins and subsequent surface lowering seems likely.

Dr Konrad summarised: “Combining these altimetry records has given us a unique insight into how some of Antarctica’s most important glaciers are changing, including the fact that they do not all respond in the same way.  We are even seeing interesting differences between the tributaries of individual glaciers.”

Stressing the importance of continued monitoring, CPOM Director Professor Andy Shepherd added: “As well as being able to routinely monitor the polar ice sheets as a whole, these results show the ability of satellites to pinpoint how individual glaciers are responding to environmental change.”

“The next steps are to refine our calculations of ice loss and sea level rise from the Antarctic ice sheet as a whole, and, in turn, improve our models of what might happen in the future”.

Konrad, H. et al. (2016) Uneven onset and pace of ice-dynamical imbalance in the Amundsen Sea Embayment, Geophysical Research Letters, doi:10.1002/2016GL070733

Lowest early-winter sea ice growth on record

There is set to be only 10,500 cubic kilometres of sea ice in the Arctic this month, the joint lowest of any November on record, according to satellite observations processed by CPOM. Although sea ice volume did not hit a record low at the end of summer, early winter growth has been 9 % slower than usual, leading to today’s conditions.

Trends in early-winter Arctic sea ice volume recorded by CryoSat. Sea ice growth this month has been 9 % lower than usual, and November 2016 is tied as a record low.

The measurements were made using the European Space Agency’s CryoSat satellite, which is dedicated to monitoring the polar regions and is uniquely able to detect the thickness of sea ice floes. Not only are these observations vital for tracking climate change, they are also an essential resource for maritime operators who increasingly navigate ice infested waters.

This year, the US National Snow and Ice Data Centre reported that the extent of Arctic sea ice fell to 4.1 million km2 on 7th September tied with 2007 as the second lowest on record. However, CryoSat shows that the ice was thicker at the end of summer than in most other years – 116 cm, on average – and so there was substantially more ice than in two other years (2011 and 2012). Thicker ice can occur if melting is lower, or if snowfall or floe-compaction is higher.

CPOM Research Fellow Rachel Tilling explained, “Sea ice is highly mobile and susceptible to deformation, so we need to measure its thickness as well as its extent to be sure how much is there.”

However, although the Arctic sea ice pack usually gains 161 cubic kilometres per day in November, this year’s growth has been 10 % lower at 139 cubic kilometres per day, and it is estimated that the final amount will rise to only 10,500 cubic kilometres by the month end.

This would essentially tie with conditions in the Novembers of 2011 and 2012, when levels of Arctic sea ice were at their lowest on record for this time of year.

Although sea ice in the central Arctic is currently thicker than it was in 2011 and 2012, there is far less ice in more southerly regions such as the Beaufort, East Siberian and Kara Seas.

Maps of November Arctic sea ice thickness recorded by CryoSat-2. Although ice is thicker than usual north of Canada, there is less ice overall in southerly regions such as the Beaufort, East Siberian and Kara Seas.

Rachel Tilling added: “Because CryoSat can measure Arctic sea ice thickness in autumn, it gives us a much clearer picture of how it has fared during summer. Although sea ice usually grows rapidly after the minimum extent each September, this years’ growth has been far slower than we’d expect probably because this winter has been warmer than usual in the Arctic.”

As demand for information on Arctic conditions increases, CryoSat has become an essential source of information for polar stakeholders ranging from ice forecasting services to scientists studying the impacts of climate change.

CPOM Director and Principal Scientific Advisor to the CryoSat mission, Professor Andrew Shepherd, summarised: “In its short, six years of life, we have learnt more about Arctic sea ice from CryoSat than from any other satellite mission. But to really understand the role that sea ice plays in the climate system, and the restrictions it places on maritime operations, we must ensure that its measurements are continued into the future.”

A complete assessment of 2016 sea ice conditions will be available from CPOM in the coming weeks.

CryoSat reveals recent Greenland ice loss

The most recent, detailed picture of ice loss from Greenland is presented in a new study, published online today in Geophysical Research Letters.

A team of scientists, led by researchers from the UK Centre for Polar Observation and Modelling at the University of Leeds, combined data from the European Space Agency’s CryoSat mission with a regional climate model to map changes in Greenland ice sheet mass.

They found that between 2011 and 2014 Greenland lost around one trillion tonnes of ice.  This corresponds to a 0.75 mm contribution to global sea level rise each year, around twice the average of the preceding two decades.

Importantly, CryoSat’s sophisticated radar altimeter has also allowed the researchers to map the complex regional pattern of imbalance that lies behind this headline figure.

“CryoSat’s radar really brings into focus our view of the ice sheet, revealing which glaciers are exhibiting the greatest signs of change,” explained lead author Dr Mal McMillan. “This helps us to study Greenland’s individual outlet glaciers, which in turn allows us to better understand the contribution they have made to global sea level rise.”

The team also mapped annual changes in the ice sheet mass over the survey period. They found large variations in the amount of ice lost from year to year, with the highest losses by far occurring in the warm summer of 2012 and smaller losses in other years. These measurements highlight how sensitive Greenland’s ice is to sudden changes in its surrounding environment.

CryoSat’s measurement of Greenland ice losses are in close agreement with those computed from NASA’s GRACE gravimetry mission, which carries sensors that are specially designed to weigh changes at the scale of the entire ice sheet.

CPOM Director and principal scientific advisor to the CryoSat mission Professor Andy Shepherd summarised: “Having already delivered assessments of Antarctica and Arctic sea ice, these findings from Greenland are a major step towards completing CryoSat’s picture of how Earth’s ice is changing. As well as being able to routinely monitor polar ice sheet losses as a whole, CryoSat is a tool for pinpointing glaciers that are responding to environmental change.”

Further information:

The full reference is: McMillan, M. et al. (2016) A high resolution record of Greenland mass balance, Geophysical Research Letters 43, doi:10.1002/2016GL069666.

Arctic sea ice shows resilience

CPOM scientists have found that Arctic sea ice volume increased markedly in the summer of 2013 compared to the previous three years, as  less of the ice melted due to lower air temperatures.

The research, published in Nature Geoscience, suggests that the ice is more sensitive to changes in summer melting than it is to winter cooling, an important finding which will influence predictions of its future change.

Thick, heavily deformed multi-year sea ice in the Lincoln Sea, off the north coast of Greenland, 29th March 2014 (Credit: Tom Armitage)

The team used measurements from the European Space Agency’s CryoSat-2 mission, which was designed especially to detect changes in sea ice volume. Although there was a 14% reduction in summertime Arctic sea ice volume between 2010 and 2012, this was followed in 2013 by a 33% increase compared to previous levels.

Lead author and CPOM researcher at University College London (UCL) Rachel Tilling explained: “Until 2013, the volume of Arctic sea ice was going down, but in 2013 there was a sharp increase, we think because that summer was much cooler than recent years.“

The lower summer temperature in 2013, conditions more typical of the late 1990s, allowed thick sea ice to persist northwest of Greenland, because there were fewer days when it could melt.

Rachel added: “Although models have suggested that Arctic sea ice volume is in long term decline, we know now that it can recover by a significant amount if the melting season is cut short.“

Arctic sea ice extent has been steadily falling since the late 1970s. However, until now it has been difficult to assess how the volume has changed, because the ice pack drifts, and because there have historically been very few sea ice thickness measurements from across the region.

To detect changes in Arctic sea ice volume, the team combined 88 million measurements of sea ice thickness recorded by CryoSat-2 since it was launched in April 2010 with maps of its changing extent.

Rachel explained:

“Maps of sea ice extent have been extremely useful, but our measurements of its thickness now complete the picture, because they reveal what’s going on below the water – where most of the action happens.” 

Although the first five years of CryoSat-2 measurements have revealed important information on the state of Arctic sea ice, the record is still too short to establish a long-term trend.  Nevertheless, understanding what controls the amount of sea ice puts us one step closer to making reliable predictions of how long it will last.

CPOM Director and co-author Andy Shepherd explained:

“What’s clear, from the CryoSat-2 measurements, is that the volume of Arctic sea ice is more sensitive to summer melting than it is to winter cooling. Since we expect temperatures to rise in the future, I suspect that the events of 2013 have simply wound the clock back a few years on the long-term pattern of decline.“

Professor Shepherd added:

“Arctic sea ice is a key component of Earth’s climate system, and so we must now use CryoSat-2’s measurements to improve predictions of future climate change and, at the same time, make sure we do not lose this unique capability to monitor Arctic sea ice when the mission ends.“

The full paper is: Tilling, R.L., Ridout, A., Shepherd, A., Wingham, D.J. Increased Arctic sea ice volume after anomalously low melting in 2013, Nature Geoscience, Vol. 8, August 2015 doi:10.1038/NGEO248.  Archive data of spring (March/April average) and autumn (October/November average) sea ice thickness from autumn 2010 onwards will be made available on the CPOM sea ice page.

Sea ice: a story of cracks

CPOM’s Harry Heorton blogs about sea ice formation and dynamics and why it’s important for global climate on the Reading Meteorology Department blog here.

Coulombic shear faulting (cracking) in a sea ice sample. Credit: Illescu and Schulson (2004)

Arctic ice cap slides into the ocean

Satellite images have revealed that a remote Arctic ice cap has thinned by more than 50 metres since 2012 – about one sixth of its original thickness – and that it is now flowing 25 times faster.

Heavily crevassed terminus of Kapp Mohn outlet glacier, Austfonna, in May 2013, after 25-fold increase in flow speed (Dunse et al., The Cryosphere Discussions, 2014). Credit: Thorben Dunse, University of Oslo

CPOM scientists combined observations from eight satellite missions, including Sentinel-1A and CryoSat, with results from regional climate models, to unravel the story of ice decline.

The findings show that over the last two decades, ice loss from the south-east region of Austfonna, located in the Svalbard archipelago, has increased significantly. In this time, ice flow has accelerated to speeds of several kilometres per year, and ice thinning has spread more than 50km inland – to within 10km of the summit.

Rate of ice cap elevation change between 2010 and 2014 observed CryoSat, overlaid onto an image acquired by the Sentinel-1A satellite. Red regions show where the ice surface has lowered due to ice loss. Credit: CPOM/GRL

“These results provide a clear example of just how quickly ice caps can evolve, and highlight the challenges associated with making projections of their future contribution to sea level rise,” said the study’s lead author Dr Mal McMillan.

The study, published in Geophysical Research Letters and reported online today by the European Space Agency (ESA), is the first to make use of measurements from ESA’s latest Earth observation satellite, Sentinel-1A.

Sentinel-1A, the first satellite developed for Europe’s Copernicus programme, was launched in April last year, while CryoSat has been in orbit since 2010.

Dr McMillan said: “New satellites, such as the Sentinel-1A and CryoSat missions, are essential for enabling us to systematically monitor ice caps and ice sheets, and to better understand these remote polar environments.”

Melting ice caps and glaciers are responsible for about a third of recent global sea level rise. Although scientists predict that they will continue to lose ice in the future, determining the exact amount is difficult, due both to a lack of observations and the complex nature of their interaction with the surrounding climate.

“Glacier surges, similar to what we have observed, are a well-known phenomenon,” said CPOM Director Andy Shepherd.

“However, what we see here is unusual because it has developed over such a long period of time, and appears to have started when ice began to thin and accelerate at the coast.”

There is evidence that the surrounding ocean temperature has increased in recent years, which may have been the original trigger for the ice cap thinning.

“Whether or not the warmer ocean water and ice cap behaviour are directly linked remains an unanswered question. Feeding the results into existing ice flow models may help us to shed light on the cause, and also improve predictions of global ice loss and sea level rise in the future,” said Professor Shepherd.

Long-term observations by satellites are the key to monitoring such climate-related phenomena in the years and decades to come.

The paper, Rapid dynamic activation of a marine-based Arctic ice cap, was published online by Geophysical Research Letters as an Early Access article on 23 December 2014.

Migrating supraglacial lakes could trigger future Greenland ice loss

Supraglacial lakes on the Greenland ice sheet can be seen as dark blue specks in the centre and to the right of this satellite image. Credit: USGS/NASA Landsat

The finding follows a new study, published in Nature Climate Change, in which the future distribution of lakes that form on the ice sheet surface from melted snow and ice ,called supraglacial lakes, have been simulated for the first time.

Previously, the impact of supraglacial lakes on Greenland ice loss had been assumed to be small, but the new research has shown that they will migrate farther inland over the next half century, potentially altering the ice sheet flow in dramatic ways.

Dr Amber Leeson, who led the study, said: “Supraglacial lakes can increase the speed at which the ice sheet melts and flows, and our research shows that by 2060 the area of Greenland covered by them will double.”

Supraglacial lakes are darker than ice, so they absorb more of the Sun’s heat, which leads to increased melting. When the lakes reach a critical size, they drain through ice fractures, allowing water to reach the ice sheet base which causes it to slide more quickly into the oceans. These changes can also trigger further melting.

Dr Leeson explained: “When you pour pancake batter into a pan, if it rushes quickly to the edges of the pan, you end up with a thin pancake. It’s similar to what happens with ice sheets: the faster it flows, the thinner it will be.

“When the ice sheet is thinner, it is at a slightly lower elevation and at the mercy of warmer air temperatures than it would have been if it were thicker, increasing the size of the melt zone around the edge of the ice sheet.”

Until now, supraglacial lakes have formed at low elevations around the coastline of Greenland, in a band that is roughly 100km wide. At higher elevations, today’s climate is just too cold for lakes to form.

In the study, the scientists used observations of the ice sheet from the Environmental Remote Sensing satellites operated by the European Space Agency and estimates of future ice melting drawn from a climate model to drive simulations of how meltwater will flow and pool on the ice surface to form supraglacial lakes.

Since the 1970s, the band in which supraglacial lakes can form on Greenland has crept 56km further inland. From the results of the new study, the researchers predict that, as Arctic temperatures rise, supraglacial lakes will spread much farther inland – up to 110 km by 2060 – doubling the area of Greenland that they cover today.

Dr Leeson said: “The location of these new lakes is important; they will be far enough inland so that water leaking from them will not drain into the oceans as effectively as it does from today’s lakes that are near to the coastline and connected to a network of drainage channels.”

“In contrast, water draining from lakes farther inland could lubricate the ice more effectively, causing it to speed up.”

Ice losses from Greenland had been expected to contribute 22cm to global sea-level rise by 2100.  However, the models used to make this projection did not account for changes in the distribution of supraglacial lakes, which Dr Leeson’s study reveals will be considerable.

If new lakes trigger further increases in ice melting and flow, then Greenland’s future ice losses and its contribution to global sea-level rise have been underestimated.

CPOM Director Andy Shepherd, who co-authored the study, said: “Because ice losses from Greenland are a key signal of global climate change, it’s important that we consider all factors that could affect the rate at which it will lose ice as climate warms.

“Our findings will help to improve the next generation of ice sheet models, so that we can have greater confidence in projections of future sea-level rise. In the meantime, we will continue to monitor changes in the ice sheet losses using satellite measurements.”

The paper, Supraglacial lakes on the Greenland ice sheet advance inland under warming climate, was published in Nature Climate Change on 15 December 2014.

CryoSat extends its reach on the Arctic

Five years’ ice thickness change. Credit: ESA/CPOM

CryoSat has delivered this year’s map of autumn sea-ice thickness in the Arctic, revealing a small decrease in ice volume. In a new phase for ESA’s ice mission, the measurements can now also be used to help vessels navigate through the north coastal waters of Alaska, for example.

Measurements made during October and November show that the volume of Arctic sea ice now stands at about 10,200 cubic km – a small drop compared to last year’s 10,900 cubic km.

The volume is the second-highest since measurements began in 2010, and the five-year average is relatively stable. This, however, does not necessarily indicate a turn in the long-term downward trend.

“We must to take care when computing long-term trends as this CryoSat assessment is short when compared to other climate records,” said CPOM Director Andy Shepherd.

“For reliable predictions, we should try other approaches, like considering what is forcing the changes, incorporating the CryoSat data into predictive models based on solid physics, or simply waiting until more measurements have been collected.”

CryoSat was designed to measure sea-ice thickness across the entire Arctic Ocean, enabling scientists to monitor accurately the overall change in volume.

While the amount of ice normally fluctuates depending on the season, longer-term satellite records show a constant downward trend in ice extent during all seasons, in particular in summer, with a minimum occurring in the autumn of 2012.

Establishing whether the ice volume is following a similar trend is one of CryoSat’s key mission objectives.

“October is interesting because it is the first month we get data directly following the sea-ice minimum in September, so that’s where we see the largest interannual variability in our volume estimates,” said Rachel Tilling, who is working on the CryoSat measurements as part of her PhD studies.

Launched in 2010, CryoSat has long surpassed its planned three-year life. At the mission’s recent mid-term review, it was further extended until February 2017.

Ice thickness for operational applications. Credit: ESA/CPOM

Tommaso Parrinello, ESA’s CryoSat Mission Manager, said CryoSat has already achieved outstanding results, both within its original mission objectives and for unexpected applications.

“Looking ahead, we are working hard to prototype new operational capabilities so that the measurements can be used for routine assessments in climate science and for services affected by Arctic sea ice.”

To test this, scientists have produced an assessment of sea-ice thickness north of Alaska and eastern Russia with data acquired over the last month. Products like this could prove useful for maritime services, such as shipping and exploration.

Arctic sea ice: a cool start to spring

CPOM’s Danny Feltham describes the current state of the Arctic sea ice and explains the uncertainties on Reading University’s Weather and Climate blog.

Arctic sea ice. Credit:Tom Armitage