We know what global temperatures are like now, from direct measurement around the globe. And we know quite a lot about what temperatures were like over the past few hundred years thanks to written records. But what about further back than that? How do we know what temperatures were like a thousand years ago, or even hundreds of thousands of years ago? There is, of course, no written record that far back in history – but there is a chemical record, hidden in the ice of Antartica and Greenland.
Towards the end of July I was in Brookings, South Dakota, for the ChemEd 2017 conference at South Dakota State University. While I was there, I had the opportunity to visit their ice core lab, where they analyse sections of ice cores brought back from Antarctica. Things get described as mind-blowing far too easily these days, but the sheer amount of information about our planet’s climate and atmosphere that can be obtained from ice is genuinely worthy of the description. From these unassuming columns of ice scientists can determine past temperatures and climates, and can also give a humbling perspective on how human activities can have serious impacts on our atmosphere.
The science of analysing ice cores is relatively new; the idea that they could give us information about the atmosphere and conditions of the past didn’t gain serious traction until the 1950s, when the first ice cores were drilled. These early cores were only drilled to a depth of around 100 metres, and the low quality of the cores recovered prevented any significant analytical work once they were recovered. The first ever continuous ice core all the way down to the bedrock in Greenland was drilled in 1966, and was 1388 metres long. The longest ice core ever obtained, at Vostok in Antarctica in 1999, reached a depth of 3623 metres.
At this point you might be wondering how an ice core 3623 metres long could possibly be brought up to the surface in one piece. This is indeed as impractical as it sounds; the core isn’t brought up in one section, but in a large number of much shorter sections, typically around 4 metres in length. The drill itself is a self-contained unit, lowered down in to the ice and connected to the surface by a thick steel cable. The teeth on the drill head rotate and cut into the ice, while the inside of the drill is hollow to collect the ice core being drilled. When drilling of a core is complete, the cutting head breaks the core off at the base so it can be pulled back up the shaft.
During drilling, the ice cuttings from the drill have to be collected in order to keep the bore hole open. In addition, as the bore hole gets deeper, there’s a further problem to deal with when ice cores are removed: at depths greater than 350 metres, the pressure of the surrounding ice can force the borehole closed. For this reason, a drilling fluid is used. The drilling fluid has to have a high enough viscosity to balance out the pressure of the surrounding ice, while also having a freezing point low enough to remain liquid in the icy temperatures. Ester-based fluids are often used (commonly palm oil derivatives), with thousands of litres being needed to drill deep ice cores.
Once the ice core is recovered, it needs to be preserved for analysis. Of course, there is no danger of the ice melting in the subzero temperatures of Antarctica or Greenland, but the air bubbles within the ice can be valuable for analysis, and even below zero degrees these can be affected. For this reason, a freezer is necessary even in Antarctica to keep the ice cores at the correct temperature, around –23˚C. In the USA, the cores will eventually end up at the National Ice Core Laboratory in Colorado – actually more of a repository for ice cores rather than a laboratory. Sections of the ice cores are then removed and sent to laboratories around the country for a range of different analyses.
The analysis of ice cores is where they get really interesting. They consists of thousands of years of snowfall, with each year’s snowfall compacting the previous years’, until they form layers of ice. Around thirty centimetres of ice accumulates each year; two hundred years after it fell as snow, the ice will be buried 60 metres below the surface, and any remaining air will be trapped in the ice as air bubbles. At this point, you can visibly see the layers in the ice produced year-on-year, as the bigger air bubbles in the summer layers due to less dense snowfall makes them appear lighter than darker winter layers.
The oldest ice core yet recovered was drilled at Dome Concordia in Antarctica in 2004, giving us an ice record extending back 800,000 years ago. Scientists can use a number of methods to ascertain the age of ice cores; one method is simply counting the yearly layers, though this gets harder as you go further back in time in the ice record as the layers stretch and thin over time. Electrical conductivity can also be used to date the cores, as the acidity of snowfall varies between the summer and winter of each year, changing the conductivity of the ice. Another method which can be used is the determination of isotope ratios in the ice. Isotopes are atoms of the same element that are made up of the same number of protons and electrons, but a varying number of neutrons. They can undergo the same chemical reactions as each other, but can behave differently in physical processes, which can be used to our advantage when analysing ice cores.
Oxygen isotopes are of particular interest when it comes to ice core analysis. Oxygen-16 (8 protons, 8 electrons, 8 neutrons) is by far the most abundant isotope of oxygen, accounting for almost 99.8% of oxygen atoms, but oxygen-18 (8 protons, 8 electrons, 10 neutrons) has an abundance of 0.2%. Because these isotopes behave identically chemically, a small number of water molecules in any body of water will contain two hydrogen atoms and an oxygen-18 atom, as opposed to an oxygen-16 atom. However, water molecules containing the lighter oxygen isotope are slightly more likely to evaporate due to their lower molecular mass, so as water from the oceans evaporates it becomes depleted in oxygen-18. At higher temperatures, slightly more oxygen-18-containing water molecules will evaporate than at lower temperatures, and when this water eventually falls as precipitation, the discrepancy is preserved. Consequently, for warmer years, scientists expect to see more oxygen-18 in ice than they do in colder years. Since isotopic composition of snowfall lines up pretty well with the temperature when it fell, this allows past temperatures to be calculated. A similar method can also be used with hydrogen isotopes, 1H and 2H (deuterium).
Analysis of ice cores can also tell us about the amount of carbon dioxide in the atmosphere over the past 800,000 years. This is accomplished by running meltwater from the cores through a gas chromatograph, which gives the carbon dioxide concentration in the trapped air. The ice core records from several locations show that in the past 800,000 years the highest pre-industrial carbon dioxide levels in the atmosphere reached around 300 parts per million. This compares to a peak of 412 parts per million this year. When isotopic temperature records and carbon dioxide concentration records from ice core analyses are put side-by-side, the correlation between the two is hard to miss – the two follow each other almost identically.
Ice cores can give us other insights into human effects on the atmosphere, some positive and some negative. For example, levels of sulfate in the atmosphere in the past can also be determined; sulfate levels show peaks for volcanic eruptions, but also show an increasing trend since industrialisation and the increased combustion of coal. However, since the early 1970s its levels in the atmosphere have decreased according to the ice record, likely as a direct consequence of the US Clean Air Act and similar air quality standards being developing in other industrialised countries.
Another atmospheric pollutant linked to fuel combustion is lead. Lead was added to petrol in the form of tetraethyl lead from the 1920s until the late 1970s before its eventual phaseout after its potential health consequences became apparent. Recent ice core studies have shown that, even before this, smelting of ores had increased the concentration of lead above natural levels, and that only during the Black Death did lead levels fall back to a natural minimum. Ice core data also shows starkly the effect that leaded petrol had on atmospheric lead levels, as well as clearly showing the positive effect of the leaded petrol ban.
Thomas Midgley Jr was the American chemist and mechanical engineer who helped develop tetraethyl lead, and in the 1930s he was also partly responsible for the development of another environmental pollutant that can be tracked in the ice record: Chlorofluorocarbons, or CFCs. These were gases used in refrigerators, air conditioning units, and aerosols, until their ability to react with and deplete ozone in the ozone layer became apparent. Since then, their phaseout has been timetabled by the international agreement known as the Montreal protocol, and in most industrialised countries their production and use has now ceased. Their usage, and that of the closely related hydrochlorofluorocarbons (HCFCs), is thought to have contributed to the increase in perchlorate levels seen in ice core records over the past 30 years. Surprisingly, considering the environmental impacts, the National Ice Core Laboratory in Colorado still uses a HCFC refrigerant for its ice core freezers.
As far as ice core science goes, this article has really just covered the tip of the iceberg. It’s a fascinating area of science which gives us a unique insight into what our planet was like hundreds of thousands of years ago, and shows us how recent human activities have affected our atmosphere. Sadly, it’s a discipline which is threatened on multiple fronts; not only are increasing global temperatures causing increased melting of the very subject of the research, but cutbacks in science funding are leaving the resources available for further research rather thin.
This is a shame since there’s still lots left for ice cores to reveal: just last year plans were being put in place to drill an ice core that could give atmospheric information up to 1.5 million years back. This could provide vital information on why the cycle by which ice ages come and go seems to have lengthened in the past million years, based on information obtained from ocean sediments. It could also give us more confidence in future climate predictions. Let’s hope that the potential value of this information is not overlooked in the future.
Enjoyed this post & graphic? Consider supporting Compound Interest on Patreon, and get previews of upcoming posts & more!
The graphic in this article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. See the site’s content usage guidelines.
References & Further Reading
- The two-mile time machine (£) – R B Alley
- A brief history of ice core science over the last 50 years – J Jouzel
- A short primer on ice core science – H Fischer
- Drilling through ice and into the past – M R Albert
- Changes in environment over the past 800,000 years from chemical analysis of the EPICA Dome C ice core – E W Wolff and others
- Drilling an ice core – North Greenland Eemian Ice Drilling
- Ice core drilling – P Neff
2 replies on “The science of ice cores: Atmospheric time machines”
Your enthusiasm for this subject is palpable. Thanks for the long and interesting post.
Thanks, glad the enthusiasm came through!