ice sheet

Geosciences Column: The best spots to hunt for ancient ice cores

Geosciences Column: The best spots to hunt for ancient ice cores

Where in the world can you find some of Earth’s oldest ice? That is the question a team of French and US scientists aimed to answer. They recently identified spots in East Antarctica that likely have the right conditions to harbor ice that formed 1.5 million years ago. Scientists hope that obtaining and analysing an undisturbed sample of ice this old will give them clues about Earth’s ancient climate.

The team published their findings in The Cryosphere, an open access journal of the European Geosciences Union (EGU).

Why study ancient ice?

When snow falls and covers an ice sheet, it forms a fluffy airy layer of frozen mass. Over time, this snowy layer is compacted into solid ice under the weight of new snowfall, trapping pockets of air, like amber trapping prehistoric insects. For today’s scientists, these air bubbles, some sealed off thousands to millions of years ago, are snapshots of what the Earth’s atmosphere looked like at the time these pockets were locked in ice. Researchers can tap into these bubbles to understand how the proportion of greenhouse gases in our atmosphere have changed throughout time.

As of now, the oldest ice archive available to scientists only goes back 800,000 years, according to the authors of the study. While pretty ancient, this ice record missed out on some major climate events in Earth’s recent history. Scientists are particularly interested in studying the time between 1.2 million years ago and 900,000 years ago, a period scientifically referred to as the mid-Pleistocene transition.

In the last few million years leading up to this transition, the Earth’s climate would experience a period of variation, from cold glacial periods to warmer periods, every 40,000 years. However, after the mid-Pleistocene transition, Earth’s climate cycle lengthened in time, with each period of variation occurring every 100,000 years.  

Currently, there isn’t a scientific consensus on the origin of this transition or what factors were involved. By examining old ice samples and studying the composition of the atmospheric gases present throughout this transition, scientist hope to paint a clearer picture of this influential time. “Locating a future 1.5 [million-year]-old ice drill site was identified as one of the main goals of the ice-core community,” wrote the authors of the study.  

The quest for old ice

Finding ice older than 800,000 years is difficult since the Earth’s deepest, oldest ice are the most at risk of melting due to the planet’s internal heat. Places where an ice sheet’s layers are very thick have an even greater risk of melting.

Mesh, bedrock dataset (Fretwell et al., 2013; Young et al., 2017) and basal melt rate (Passalacqua et al., 2017) used for the simulation. Credit: O. Passalacqua et al. 2018.

“If the ice thickness is too high the old ice at the bottom is getting so warm by geothermal heating that it is melted away,” said Hubertus Fischer, a climate physics researcher from the University of Bern in Switzerland not involved in the study, in an earlier EGU press release.

Last summer, a team of researchers from Princeton University announced that they had unearthed an ice core that dates back 2.7 million years, but the sample’s layers of ice aren’t in chronological order, with ice less than 800,000 years old intermingling with the older frozen strata. Rather than presenting a seamless record of Earth’s climate history, the core can only offer ‘climate snapshots.’

Finding the best of the rest

The authors of the recent The Cryosphere study used a series of criteria to guide their search for sites that likely could produce ice cores that are both old and undisturbed. They established that potential sites should of course contain ice as old as 1.5 million years, but also have a high enough resolution for scientists to study frequent changes in Earth’s climate.

Additionally, the researchers established that sites should not be prone to folding or wrinkling, as these kinds of disturbances can interfere with the order of ice layers.

Lastly, they noted that the bedrock on which the ice sheet sits should be higher than any nearby subglacial lakes, since the lake water could increase the risk of ice melt.

Magenta boxes A, B and C correspond to areas that could be considered as our best oldest-ice targets. Colored points locate possible drill sites. Credit: O. Passalacqua et al. 2018.


Using these criteria, the researchers evaluated one region of East Antarctica, the Dome C summit, which scientists in the past have considered a good candidate site for finding old ice. They ran three-dimensional ice flow simulations to locate parts of the region that are the most likely to contain ancient ice, based on their established parameters.

By narrowing down the list of eligible sites, the researchers were able to pinpoint regions just a few square kilometres in size where intact 1.5 million-year-old ice are very likely to be found, according to their models. Their results revealed that some promising areas are situated a little less than 40 kilometres southwest of the Dome C summit.

The researchers hope their new findings will bring scientists one step closer towards finding Earth’s ancient ice.

By Olivia Trani, EGU Communications Officer

Geosciences Column: Fire in ice – the history of boreal forest fires told by Greenland ice cores.

Burning of biomass contributes a significant amount of greenhouses gases to the atmosphere, which in turn influences regional air quality and global climate. Since the advent of humans, there has been a significant increase in the amount of biomass burning, particularly after the industrial revolution. What might not be immediately obvious is that, (naturally occurring) fires also play a part in emitting particulates and greenhouse gases which can absorb solar radiation and contribute to changing Earth’s climate. Producing a reliable record of pre-industrial fire history, as a benchmark to better understand the role of fires in the carbon cycle and climate system, is the focus of research recently published in the open access journal, Climate of the Past.

Forest fires.  Credit: Sandro Makowski (distributed via

Forest fires. Credit: Sandro Makowski (distributed via

Did you know the combustion of biomass can emit up to 50% as much CO2 as the burning of fossil fuels? The incomplete burning of biomass during fires also produces significant amounts of a fine particle known as black carbon (BC). Compare BC to more familiar greenhouse gases such as methane, ozone and nitrous oxide and you’ll find it absorbs more incoming radiation than the usual suspects. In fact, it is the second largest contributor to climate change.

NEEM camp position and representation of boreal vegetation and land cover between 50 and 90 N. Modified from the European Commission Global Land Cover 2000 database and based on the work of cartographer Hugo Alhenius UNEP/GRIP-Arendal (Alhenius, 2003). From Zennaro et al., (2014).

NEEM camp position and representation of boreal vegetation and land cover between 50 and 90 N. Modified from the European Commission Global Land Cover 2000 database and based on the work of cartographer Hugo Alhenius UNEP/GRIP-Arendal (Alhenius, 2003). From Zennaro et al., (2014). Click to enlarge.

The boreal zone contains 30% of the world’s forests, including needle-leaved and scale-leaved evergreen trees, such as conifers. They are common in North America, Europe and Siberia, but fires styles in these regions are diverse owing to differences in weather and local tree types. For instance, fires in Russia are known to be more intense than those in North America, despite which they burn less fuel and so produce fewer emissions. All boreal forest fires are important sources of pollutants in the Arctic. Models suggest that in the summertime, the fires in Siberian forests are the main source of BC in the Artic and shockingly, exceed all contributions from man-made sources!

To build a history of forest fires over a 2000 year period the researchers used ice cores from the Greenland ice sheet. Compounds, such as ammonium, nitrate, BC and charcoal (amongst others), are the product of biomass burning, and can be measured in ice cores acting as indicators of a distant forest fires. Measure a single compound and your results can’t guarantee the signature is that of a forest fire, as these compounds can often be released during the burning of other natural sources and fossil fuels. To overcome this, a combined approach is best. In this new study, researchers measured the concentrations of levoglucosan, charcoal and ammonium to detect the signature of forest fires in the ice. Levoglucosan is a particularly good indicator because it is released during the burning of cellulose – a building block of trees – and is efficiently injected into the atmosphere via smoke plumes and deposited on the surface of glaciers.

The findings indicate that spikes in levoglucosan concentrations measured in the ice from the Greenland ice sheet correlate with known fire activity in the Northern Hemisphere, as well as peaks in charcoal concentrations. Indeed, a proportion of the peaks indicate very large fire events in the last 2000 years. The links don’t end there! Spikes in concentrations of all three measured compounds record a strong fire in 1973 AD. Taking into account errors in the age model, this event can be correlated with a heat wave and severe drought during 1972 CE in Russia which was reported in The New York Times and The Palm Beach Post, at the time.

Ice core. Credit: Tour of the drilling facility by Eli Duke, Flickr.

Ice core. Credit: Tour of the drilling facility by Eli Duke, Flickr.

The results show that a strong link exists between temperature, precipitation and the onset of fires. Increased atmospheric CO2 leads to higher temperatures which results in greater plant productivity, creating more fuel for future fires. In periods of draught the risk of fire is increased. This is confirmed in the ice core studied, as a period of heightened fire activity from 1500-1700 CE coincides with an extensive period of draught in Asia at a time when the monsoons failed. More importantly, the concentrations of levoglucosan measured during this time exceed those of the past 150 years, when land-clearing by burning, for agricultural and other purposes, became common place. And so it seems that the occurrence of boreal forest fires has, until now, been influenced by variability in climate more than by anthropogenic activity. What remains unclear is what the effects of continued climate change might have on the number and intensity of boreal forest fires in the future.

By Laura Roberts Artal, EGU Communications Officer



Zennaro, P., et al.: Fire in ice: two millennia of boreal forest fire history from the Greenland NEEM ice core, Clim. Past, 10, 1905-1924, doi:10.5194/cp-10-1905-2014, 2014.

Imaggeo on Mondays: Carving polar canyons

This week Ian Joughin, a research scientist from the Polar Science Center at the University of Washington, takes us on the polar express to put glacial processes into perspective and find out what makes a moulin…

“Water filled canyon” by Ian Joughin, distributed by the EGU under a Creative Commons licence.

“Water filled canyon” by Ian Joughin, distributed by the EGU under a Creative Commons licence.

This canyon formed when a melt lake on the surface of the Greenland Ice Sheet overflowed and created a stream that extended out toward a crevasse field. This outflow stream filled a crevasse, causing it to fracture under the pressure of the liquid, creating a hydrofracture that ran through the full thickness of the ice sheet. This fracture created a conduit to the base of the ice sheet, known as a moulin, through which the surface water drained to the bed.

Surface water entering a moulin on Athabasca Glacier (a much smaller Moulin than the one what would have drained the Greenland lake). (Credit: Wikimedia Commons user China Crisis)

Surface water entering a moulin on Athabasca Glacier. (Credit: Wikimedia Commons user China Crisis)

Over the course of several years, the turbulent overflow stream melted the ice down to create this canyon. By the time this photo was taken, snow had dammed canyon near the lake outlet, meaning it no longer actively drained the lake.

Most of the water in the photo is from melt at the sides of the canyon. The ice is flowing at approximately 100 m/yr, slowly moving the stream outlet toward higher ground so it is unlikely that the lake will overflow at this location again. And instead, we have found a new canyon forming in a lower part of the lake basin.

By Ian Joughin, University of Washington

The EGU’s open access geoscience image repository has a new and improved home at! We’ve redesigned the website to give the database a more modern, image-based layout and have implemented a fully responsive page design. This means the new website adapts to the visitor’s screen size and looks good whether you’re using a smartphone, tablet or laptop.

Photos uploaded to Imaggeo are licensed under Creative Commons, meaning they can be used by scientists, the public, and even the press, provided the original author is credited. Further, you can now choose how you would like to licence your work. Users can also connect to Imaggeo through their social media accounts too! Find out more about the relaunch on the EGU website. 

Imaggeo on Mondays: Getting a handle on Greenland’s glaciers

The picture below shows several small glaciers surrounding the Greenland ice sheet, in Tassilaq, near Kulusuk, East Greenland. The dark lines are glacial moraines, responsible for the transport of rock material from mountains towards sea.

The photographer, Romain Schläppy, highlights that “an important scientific topic consists to place the recent and ongoing Greenland warming in the broader context of past changes in south Greenland land climate, vegetation, sedimentation and ice history”. Indeed, with the recent report produced by the Ice2Sea programme, there is a lot of work being done to investigate glacial mass balance, with one particularly cool model looking at the how the edges of the Greenland ice sheet are changing in the greatest detail.

“The power of ice” by Romain Schläppy, distributed by the EGU under a creative commons licence.

“The power of ice” by Romain Schläppy, distributed by the EGU under a Creative Commons licence.

Most models separate large regions into squares, for surface modelling, or cubes, for something a little more 3D. This makes all the data that goes into a model easier to handle as you simplify the variation in, say, runoff rate, over a large area into a single value for runoff. While this makes information easier to handle, you also lose a lot of resolution, not something you want when big changes are happening on small scales.

This is the case in the Greenland ice sheet. The edges are advancing and retreating year in and year out, as they are influenced by the climate and conditions of the ocean around them, but the centre of the ice sheet remains relatively stable. This means that parameters such as meltwater runoff will be changing lots at the glacier front and relatively little in the middle.

To combat this, climate modellers have produced a new model using triangular blocks rather than square ones, so instead of having many equally large simplifications, you can have large, simple triangles where there’s not much going on and tiny ones to capture all the detail where the excitement is happening!


Vaughan, D.G., Aðalgeirsdóttir, G., Agosta, C. et al. From Ice to High Seas, The ice2sea Consortium: 2013.

Imaggeo is the EGU’s online open access geosciences image repository. All geoscientists (and others) can submit their images to this repository and since it is open access, these photos can be used by scientists for their presentations or publications as well as by the press and public for educational purposes and otherwise. If you submit your images to Imaggeo, you retain full rights of use, since they are licensed and distributed by the EGU under a Creative Commons licence.