- Education and Science»
- Geology & Atmospheric Science
ICE Core CO2 Records - Ancient Atmospheres Or Geophysical Artifacts ?
Ice cores from Greenland and Antarctica are sacred pillars of evidence, establishing benchmarks of carbon dioxide (CO2) in Earth’s prehistoric atmospheres. But can ice actually preserve atmospheric air (free of contamination or chemical alteration) for thousands, tens of thousands, or hundreds of thousands of years?
Fossil vs. Artifact
A fossil is a remnant or impression of something that existed in past geological ages – an ancient form preserved in its original appearance.
An artifact is a product with an artificial character due to extraneous agency – an object that is left over from a sequence of creative events.
The question is: Do ice cores contain the fossils of Earth’s prehistoric atmospheres, or do ice cores contain artifacts of geophysical processes that continuously remix gases in glacial ice?
In 1992, a group of three otherwise respectable scientists answered this question with a 57-page article that went against the mainstream view . These three dissenters (led by Zbigniew Jaworowski) subsequently endured professional contempt , career endangerment , and harsh criticism . Between 1992 and 2007, Jaworowski produced three more notorious articles condemning ice core science , , .
Mainstream scientists argue convincingly against Jaworowski, and they continue to reject him today. A notable exception is Emeritus Professor of Organic Chemistry, Joel M. Kauffman, who uses Jaworowski’s case to help reexamine the highly controversial issue of human-caused climate change .
Jaworowski and his followers insist that ice cores are invalid media for determining CO2 concentrations in Earth’s prehistoric atmospheres, because:
- Ice sheets are NOT closed systems that trap gases mechanically and preserve them indefinitely. Instead, liquid saline water can exist in ice at temperatures below –70° C, within a permeable ice sheet where a capillary liquid network acts as a sieve to redistribute elements, isotopes and micro-particles.
- Air recovered from old ice is contaminated during field and laboratory operations.
- The widely accepted pre-industrial atmospheric CO2 level of 290 ppm rests on biased rejections of high CO2 readings in ice cores. Early studies on ice cores consistently showed a range of CO2 readings that were higher than later studies – in one case, a study by the same researcher on the same ice core showed different numbers in different years.
D. Raynaud and coauthors  admit that “several processes could cause the gas record measured in ice samples to be different from the original atmospheric composition.” They list the following processes:
- absorption of gases on the surface of snow and ice crystals,
- separation by gravity and molecular diffusion of the gases in the ice column,
- alteration of gas composition by formation of air hydrates at great depths in the ice sheet or by presence of drilling-induced fractures or thermal cracks in ice samples,
- alteration by chemical interaction between gases and ice on long time scales.
Raynaud assures us that researchers can test and calculate the effects of these interfering physical processes, citing papers by Ethridge and Neftel – two of the authors whom Jaworowski contests.
Critique Of Raynaud
In key sentences, Raynaud’s language is obscure and lacking in convincing detail – he appears to elevate strong statements of confidence above decisive descriptions of procedures that eliminate uncertainties in ice-core gas measurements. Raynaud easily dismisses ice-core CO2 testing methods before the 1980’s, claiming that these methods were inappropriate and that reliable results came only later. This argument (if true) overcomes Jaworoski’s objection of suspiciously higher CO2 measurements in ice cores prior to the 1980’s. A similar argument (used by Harro A. J. Meijer against Ernst-Georg Beck rules out much of the 19th century air measurements of CO2. Lack of knowledge, lack of skill, and lack of proper methodology all are mainstream arguments that reject CO2 measurements higher than 290 ppm before the year 1958.
Regarding the soundness of ice core science, Raynaud and coauthors conclude:
“It is now firmly established that after selecting appropriate sampling sites, the ice core record of greenhouse gases provides the most direct evidence for past atmospheric change.”
Expanding on Raynaud’s confidence, Christopher Readinger  writes:
“All of the modern analytical techniques used to extract these proxy records have been developed and honed over time, and with the assistance of better technology and new ideas, more accurate methods of ice core analyses are being developed.”
Looking at Readinger’s end list of limitations, however, a person could easily embrace the doubt that Jaworowski advises.
Greenland and Antarctica Ice Cores
Digging through the literature, I stumbled upon a paper from year 2000 by Bernhard Stauffer and Jurg Tschumi  laying out problems that can vex ice core analysis. These authors acknowledge the reality of artifacts, even though they do not state outright that these artifacts disable reliable ice core analysis. Their presentation is terse and non-opinionated, yet their sentences speak clearly.
For example, in one sentence, they report, “Despite the relatively good agreement between the GRIP, the Dye 3 and the Camp Century record for the glacial part, it became obvious that the results do not represent a reliable record of the atmospheric CO2 concentrations.”
Later, they clarify:
“We performed very detailed measurements along short sections of the GRIP core and observed large variations over distances of a few centimeters in sections which show high CO2 concentrations. … As mentioned above, such short-term variations cannot reflect variations of the atmospheric CO2 concentration, it has to be an artifact. Delmas  suggested that the surplus CO2 is produced by an acid carbonate reaction in the ice. Another possibility to produce CO2 would be the oxidation of organic material in the ice.”
These authors do not convey the same confidence as Raynaud and Readinger, but by their association with the colloquium on ice core science in which their paper appeared, I can only assume that they support ice core science. To give a fuller flavor of their tone, I present Stauffer’s and Ischumi’s conclusions, exactly as written:
- CO2 can be produced and probably also depleted by chemical reactions occurring between impurities in glacier ice. For the production of CO2 oxidation reactions are as important as acid carbonate reactions.
- Antarctic ice core records compared to those from Greenland are generally less affected by such chemical reactions, due to the lower impurity concentrations in Antarctic ice.
- The most reliable results concerning reconstruction of ancient atmospheric CO2 concentrations are obtained with ice samples containing a low concentration of carbonates and H20 2. Measurements can be considered reliable only if a detailed high resolution record along a few annual layers shows that the scattering of the results is of the order as the analytical uncertainty.
Measurements on ice where air is enclosed in clathrates show additional difficulties due to a fractionation of air components. This holds especially for the zone in which both air bubbles and clathrates coexist.
It is very important to develop dry extraction methods, which allow for the extraction of gases with an efficiency close to 100 %. The sublimation technique is at present the one which IS most promising.”
The Real Picture
From this review so far, a more realistic picture of ice core science begins to emerge – a picture that might be more tolerant of Jaworowski, whose jobs have placed him in the trust of others as a contamination expert.
Jaworowski argues that mainstream researchers are dreaming if they think that they can extract reliable information from the conundrum of possible contamination problems in ice cores. Mainstream researchers, on the other hand, claim that they know about these problems and have them under control. Brooks Hurd , an expert in purity of gases in the semiconductor and other industries, echoes Jaworowski’s doubt, when he writes, “concerns about loss and contamination of atmospheric components from sampling handling prior to analysis should be a major issue in deep ice core CO2 analyses.”
A Young Science
Analyzing ice cores is a relatively young science, compared to other established sciences  , but even in 1993, a researcher named Robert J. Delmas  cautiously discussed the presence of excess CO2 in Greenland ice samples. In a seemingly reluctant tone, he pointed out interactions between acid and alkaline impurities that could lead to this excess CO2. He admitted that the excess CO2 might form either in the ice sheet itself or during ice-core storage. Although he avoided questioning low published CO2 values (180 – 200ppm), he concluded:
“The consequence of this carbonate decomposition is that rapid CO2 fluctuations associated with climate change could well NOT be an atmospheric reality, at least not as they are now reconstructed from Greenland ice core measurements.”
He was careful to emphasize, however, that his findings cast little doubt on Antarctic ice cores or on data about ice ages in both Polar Regions.
Gases In Ices
Ices of the Greenland and Antarctic ice sheets, thus, remain the premier records of Earth’s ancient atmospheres. Gases trapped in these gigantic glaciers continue to stand as trustworthy fossils of prehistoric air.
Air fossils are certainly an intriguing idea, but how can air remain undisturbed for millennia, in a medium that melts and refreezes easily, flows and reforms plastically, and appears and disappears cyclically? To help assess this question, let us look briefly, at how gases interact with glacial ices.
In ice sheets, ice transforms through different stages, from snow on the surface to dense ice at great depths. New snow compresses steadily under the accumulating weight of more falling snow. Old snow compresses into granular ice called “firn”. Grains that make up firn squeeze together even tighter to form successively denser layers of ice. This process continues until it produces the densest glacial ice. Throughout the transformation from snow to firn to glacial ice, atmospheric air intermingles with the ice sheet in a progression of its own :
- Air mixes with snow.
- Snow compresses under its own weight into granular ice (firn), between whose grains air continues to circulate.
- As firn compresses further, air becomes more stagnant and more closed off from the atmosphere.
- At deeper layers in the ice sheet, air becomes trapped as air bubbles in dense glacial ice.
- At the greatest depths, air bubbles get squeezed out altogether, and individual air molecules become imprisoned in cages of hydrogen bonds, to form a class of crystal-like solids called “air hydrates” or “clathrates”.
According to ice core scientists, trapped air provides a record of the original atmosphere that became sealed in the ice sheet. While complexities in selecting and analyzing air samples most certainly exist, ice core scientists again claim confidence in methods that untangle the measurements of such samples.
For example, Michael Bender  says, “Measured concentrations of gasses in ice cores and firn air need to be corrected for effects of gravitational fractionizaton and thermal fractionization.” He also points out, “there are substantial uncertainties associated with age [of ice and air] limiting our ability to interpret some records.” His discussion makes clear that researchers have to deconvolute air measurements according to ice flow models limited by basic unknowns and basic assumptions whose sanctity few people question.
Werner F. Kuhs  admits, “in several respects, a deeper understanding of the chemico-physical behavior of air in contact with ice is not yet obtained.” Kuhs makes several important points:
- Air hydrates (clathrates) most likely behave as rigid units when embedded in an ice sheet subject to visco-plastic deformation processes.
- Even so, there is evidence that diffusion of water molecules in the microscopic channels between ice grains occurs faster than the visco-plastic deformation process.
- At high pressures deep in the ice sheet, the interactions between air and ice is more complex, and how this interaction takes place on a microscopic level is still an open question.
In a very technical paper, A. W. Rempel and J. S. Wetttlaufer  provide insights into the latest understanding of glacial ice:
- Careful studies reveal ice core segments where appreciable changes in the original deposits have occurred.
- Microscopic boundaries between ice grains can connect as a fluid network of veins where liquid transports and mixes impurities.
- Downward diffusion through this network of veins eventually outpaces the flow of the ice, thereby separating substances from the ice with which they were deposited.
Scientists require a model that includes such a capillary process to better interpret ice records for reconstructing prehistoric climates.
More On Diffusion
In the JOURNAL OF GLACIOLOGY (2008), Jinho Ahn  and coauthors, in the article, CO2 Diffusion In Polar Ice, write, “The processes of gas diffusion related to variable physical properties of ice are still not well known.” While never questioning the fundamental premise of constructing ancient atmospheres from ice core CO2 measurements, they nonetheless point to undeniable facts:
- CO2 diffusion in ice after the air is trapped in bubbles is poorly understood, because the extremely small rate of CO2 diffusion has not been accurately determined.
- Other types of diffusion (for example, via liquid in ice grain boundaries or veins) may also be important, but their influence has not been quantified.
- Substantial CO2 diffusion may occur in ice on timescales of thousands of years.
Expanding on the second fact above, Ahn and coauthors write:
“Processes other than volume diffusion may be important but are difficult to quantify. For example, there is evidence of the existence of melt at triple junctions of grain boundaries in polar ice (Mulvaney and others, 1988). Thus, CO2 may dissolve and migrate in the liquid vein, while noble-gas species, with lower solubility, may mostly stay at the original sites. If this is the case, the diffusion via the liquid vein or ice grain boundaries may be governed directly by the grain growth rate, as suggested from an ion chemistry study (Barnes and others, 2003).”
Additional important points by these authors are:
- Diffusion of CO2 may significantly increase with greater depth in the ice due to geothermal warming.
- The relationship between the solubility of CO2 in ice and temperature is unknown.
Ice Sheet Creep, Deformation, Recrystallization, And Plastic Flow
Besides chemical instabilities in ice sheets, there are also physical instabilities in ice sheets.
The enormous ice sheets of Greenland and Antarctica contain ice that flows towards the ocean under the force of its own mass, creeping like a viscous fluid along thousands of years. Flow occurs mainly through deformation. Deformation occurs mainly at the base, as immense weight causes ice crystals to crush over one another and recrystallize. Such deformation and recrystallization are what govern the slow motion and changes of ice sheets.  The huge time-scales and relatively high temperatures of the creeping process ceaselessly changes the microscopic structure of ice.
With this understanding of ice sheets, Sergio H. Faria  makes a bold claim:
“In spite of the fact that recrystallization phenomena have essentially a thermodynamic character, most polycrystalline ice-sheet models proposed so far have been based on ad hoc theories, without corroboration of a rigorous thermodynamic analysis.”
Paul D. Bons  and coauthors echo Faria’s assessment:
“The main complicating factor is that several coupled competing processes affect the micro structuring (e.g., diffusional creep, grain boundary sliding and recrystallization) … This means that the true dynamics of ice flow, including the possibility of shear localization, are probably not fully captured by the usual flow laws for ice.”
Both Faria and Bons emphasize how ice flow continually reworks the microscopic structure of ice, yet neither author questions whether ice-core CO2 records are reliable. The most Bons concedes is this:
“… the interpretation of the complex deformation history of polar ice is challenging and continues to be a matter of intense research.”
Ice Flow Assumptions And Ice Core Drilling
According to G. Durand  and coauthors, scientists drill ice cores exactly at the tops of ice domes, where three main assumptions apply:
- Ice flows symmetrically around the vertical axis.
- Ice thins because of vertical compression only, without horizontal sheer.
- Ice thins increasingly with increasing depth, in a smooth, monotonous manner.
These assumptions guide scientists in dating ice and its contents. These assumptions, however, ignore any possible local deformation, which Durand shows to be incorrect. What he shows is that flow disturbances are detectable from almost the top of the ice sheet to the depths of the ice sheet, increasing in number and intensity with depth. Durand, thus, suggests that ice core scientists reconsider the current standard dating charts.
The findings of Durand, Faria, and Bons support the idea that ice cores provide only a static snapshot of the cumulative deformational history of an ice sheet. If such slow, continuous, fluid deformation occurs over millennia, a person might wonder whether a slow, continuous, fluid remixing of contents also occurs. Suddenly, Jaworowski’s objections come back to haunt the discussion again.
It is hard to conceive that the physics of ice is separate from the chemistry of ice – that the contents of ice are separate from the continuity of its millennial flow. How can something as fragile as air retain any semblance of its original identity, and how can researchers find any point of lasting stability from which to determine a true signature of any ancient atmosphere? No outstanding piece of literature so far seems to answer this question convincingly.
Living Breathing Ice Sheets
In the temporal frame of human experience, the monolithic ice sheets of Greenland and Antarctica appear static and pristine. In a greater temporal frame, however, and often on smaller scales of measure, a more dynamic picture unfolds. Very slow, very small changes occur constantly in ice sheets, for hundreds, thousands, tens of thousands, and hundreds of thousands of years. In addition to the dynamic features already mentioned above, there are still others:
Greenland SupraGlacial Lakes And Antarctica SubGlacial Lakes
- NASA  reports that the surface of Antarctica at any one location, in any one moment, belies its true, complex character, which includes numerous, active pools of water underneath the massive ice sheet.
- Benjamin E. Smith  and coauthors write that, as of year 2008, the number of subglacial lakes in Antarctica is 280, the majority of which lie under the plateau of the East Antarctic ice sheet.
- Sarah B. Das  and coauthors document an efficient subglacial drainage system in the Greenland ice sheet that disperses surface lake water. The authors highlight a particularly rapid lake-draining event in 2006 whose flow rate exceeded the average flow rate of Niagra Falls.
Hugh Corr and David Vaughan  report a volcanic eruption under the West Antarctic Ice Sheet as recent as 2000 years ago. Episodes of water production and release from the volcano probably affected ice flow. Ongoing volcanic heat production might affect contemporary ice dynamics of this glacial system.
Microbial Growth, Chemical Activity And Cosmic Radiation In Glacial Ice
- The Ice Core Working Group , in the year 2003, discussed new discoveries of microbial life in glacial ice and subglacial lakes. One interesting question these scientists raised was whether there are metabolically active microbes in liquid water veins of solid glacial ice. They also stated:
“It is important that future ice-coring projects include biological studies in the planning stages so that appropriate measures can be taken to avoid microbial contamination. Much of the past biological research on ice cores has had to battle contamination problems and has often yielded equivocal results.”
- P. Buford Price and Todd Sowers  support the view that, far below the freezing point, liquid water inside ice and permafrost is available for metabolism. They explain in detail:
“The thermodynamic stability of ion-rich liquid veins at triple grain-boundaries in polycrystalline ice and of thin films of unfrozen water on microbial surfaces in permafrost enables transport of nutrients to and waste from them. Certain impurities such as mineral acids or salts can reduce the freezing point of water in narrow intergranular veins to as low as -90°C. Acidophilic psychrophiles in a Greenland mine survive the cold winters at -30°C, probably by taking advantage of such a habitat.”
- A. J. Colussi and M. R. Hoffman  challenge the premise that chemical activity is permanently arrested in deep ice cores, stating that such activity poses a genuine problem in assessing Earth’s ancient atmospheres. They point to photochemical reactions driven by cosmic radiation penetrating deep into glacial ice. M. I. Guzman  and coauthors offer a related study on persistent chemical activity in glacial ice.
In this article, I have pointed out a number of active processes that occur in the ice sheets of Greenland and Antarctica:
- deformation and recrystallization
- air diffusion
- capillary liquid transport
- subglacial and supraglacial water flow
- subglacial volcanic activity and geothermal heat production
- microbial respiration
- chemical reactions
Not only might these processes take place currently at any one moment in any one location of an ice sheet, but also these processes have taken place continuously, in combination, cumulatively, at different levels and locations in a growing and flowing ice sheet, over many thousands of years or more.
Whether or not current ice flow models adequately take into account the collective effect of all these processes remains an open question. In my judgment, whether any model at all can predict or reconstruct the complex effect of these processes on a fragile air inclusion remains doubtful. Consequently, my main question remains unsettled – Are ice core CO2 air samples actually atmospheric fossils or geophysical artifacts?
1. Z. Jaworowski, T. V. Segalstad and N. Ono (1992)), Do Glaciers Tell A True Atmospheric CO2 Story?, THE SCIENCE OF THE TOTAL ENVIRONMENT, vol 114, p 227-284
2. Jim Easter (April 2005), Hans Oeschger’s Letter To ESPR, SOME ARE BOOJUMS [web log]
3. (May 2007) The Ice Core Man [published online], NATIONAL POST, CanWest MediaWorks Publications Inc.
4. Jim Easter (May 2005), The Golden Horseshoe Award: Jaworowski And The Vast CO2 Conspiracy, SOME ARE BOOJUMS [web log]
5. Zbigniew Jaworowski (Spring 1997), Ice Core Data Show No Carbon Dioxide Increase, 21st CENTURY, p 42-52
6. Zbigniew Jaworowski (March 2004), Climate Change: Incorrect Information On Pre-Industrial CO2, [statement intended for congressional hearing]
7. Zbigniew Jaworowski (March 2007), CO2: The Greatest Scientific Scandal of Our Time, EIR SCIENCE, p 38-53
8. Joel M. Kauffman (2007), Climate Change Reexamined, JOURNAL OF SCIENTIFIC EXPLORATION, vol 21, no 4, p 723-749
9. D. Raynaud, J. Jouzel, J.M. Barnola, J. Chappellaz, R.J. Delmas, C. Lorius (February 1993), The Ice Record Of Greenhouse Gases, SCIENCE New Series, Vol 259, no 5097, p 926-934
10. Christopher Readinger (February 2006), Ice Core Proxy Methods for Tracking Climate Change, PROQUEST [website]
11. Bernhard Stauffer and Jiirg Tschumi (2000), Reconstruction Of Past Atmospheric CO2 Concentrations By Ice Core Analyses, in PHYSICS OF ICE CORE RECORDS, Edited by T. Hondoh, Hokkaido University Press, Sapporo, p 217-241
12. Brooks Hurd (November 2006), Analyses of CO2 and Other Atmospheric Gases, in AIG NEWS, p 10-11
13. University of Copenhagen, Center for Ice and Climate, Neils Bohr Institute, The History of Danish Ice Core Science [webpage]
14. Chester C. Langway, Jr. (2008) THE HISTORY OF EARLY POLAR ICE CORES, U. S. Army Corps of Engineers [ebooklet]
15. Robert J. Delmas (May 1993), A Natural Artefact In Greenland Ice-Core CO2 Measurements, TELLUS B, Vol 45 Issue 4, p 391-396
16. Werner F. Kuhs, Gas Hydrates (Clathrate Hydrates) [webpage]
17. Michael Bender, Todd Sowers and Edward Brook (August 1997), Gases In Ice Cores, PROCEEDINGS NATIONAL ACADEMY OF SCIENCES, vol 94, p 8343–8349, Colloquium Paper [This paper was presented at a colloquium entitled CARBON DIOXIDE AND CLIMATE CHANGE, organized by Charles D. Keeling, held November 13–15, 1995 at the National Academy of Sciences, Irvine, CA)
18. Werner F.Kuhs, Alice Klapproth and Bertrand Chazallon (2000), Chemical Physics Of Air Clathrate Hydrates, in PHYSICS OF ICE CORE RECORDS, Edited by T. Hondoh, Hokkaido University Press, Sapporo, p 373-392
19. A.W. Rempel and J.S. Wettlaufer (2003), Segregation, Transport, And Interaction Of Climate Proxies In Polycrystalline Ice, CANADIAN JOURNAL OF PHYSICS, vol 81 issue 1/2, p 89-97
20. Jinho Ahn, Melissa Headly, Martin Wahlen, Edward J. Brook, Paul A. Mayewske, Kendrick C. Taylor (2008), CO2 Diffusion In Polar Ice: Observations From Naturally Formed CO2 Spikes In The Siple Dome (Antarctica) Ice Core, JOURNAL OF GLACIOLOGY, Vol. 54, No. 187, p 685-695
21. Maurine Montagnat and Paul Duval (2004), The Viscoplastic Behaviour Of Ice In Polar Ice Sheets: Experimental Results And Modelling, COMPTES RENDUS PHYSIQUE, 5, p 699-708
22. Sergio H. Faria (September 8, 2006), Creep And Recrystallization Of Large Polycrystalline Masses. III. Continuum Theory Of Ice Sheets, PROCEEDINGS OF THE ROYAL SOCIETY A, vol 462, no 2073, p 2797-2816
23. Paul D. Bons, Sérgio H. Faria, and Sepp Kipfstuhl (2008), Modelling Of Deformation And Recrystallisation Microstructures In Polar Ice: Numerical Studies [joint project], Institute for Geosciences, Tubingen University
24. G. Durand, F. Graner and J. Weiss (July 2004), Deformation Of Grain Boundaries In Polar Ice, EUROPHYSICS LETTERS, vol 67, no 6, p 1038-1044
25. NASA (April 2007), Sub-glacial Lakes – Antarctica, EARTH OBSERVATORY [website]
26. Benjamin E. Smith, Helen A. Fricker, Ian R. Joughin, Slawek Tulaczyk (2009), An Inventory Of Active Subglacial Lakes In Antarctica Detected By ICESat (2003–2008), JOURNAL OF GLACIOLOGY, Vol. 55, No. 192, p 573-595
27. Sarah B. Das, Ian Joughin, Mark D. Behn, Ian M. Howat, Matt A. King, Dan Lizarralde, Maya P. Bhatia (February 2008), Fracture Propagation To The Base Of The Greenland Ice Sheet During Supraglacial Lake Drainage, revised version uploaded to SCIENCE
28. Hugh F. J. Corr and David G. Vaughan (January 2008), A Recent Volcanic Eruption Beneath The West Antarctic Ice Sheet, NATURE GEOSCIENCE, vol 1, p 122-125
29. Ice Core Working Group (June 2003), U. S. ICE CORE SCINECE: RECOMMENDATIONS FOR THE FUTURE [a report based on the workshop, The Future Of U.S. Ice Coring Science, March 20-21, Arlington, Virginia], p 25-28
30. P. Buford Price and Todd Sowers (March 2004), Temperature Dependence Of Metabolic Rates For Microbial Growth, Maintenance, And Survival, PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES, vol 101, no 13, p 4631-4636
31. A. J. Colussi and M. R. Hoffmann (2003), In Situ Photolysis Of Deep Ice Core Contaminants By Cerenkov Radiation Of Cosmic Origin, GEOPHYSICAL RESEARCH LETTERS, vol 30, no 4, p 1195-1198
32. M. I. Guzman, M. R. Hoffmann and A. J. Colussi (May 2007), Photolysis Of Pyruvic Acid In Ice: Possible Relevance To CO and CO2 Ice Core Record Anomalies, JOURNAL OF GEOPHYSICAL RESEARCH, vol 112