A Critique of Ice Core Dating and Interpretation
(this is of the 2007 report)
The interpretation and dating of the Greenland and Antarctica ice cores’ has played a major role in supporting global warming arguments. While extremely interesting and important for the study of our world, I believe the ability of scientists to interpret and date the cores has been largely over estimated. This paper will be an attempt to critique the methods used to interpret and date the cores, and the theories and assumptions that lay behind the methods. In doing so I will show that the methods used and theories behind ice core dating are both subjective and theory dependent and that past weather interpretation, while valuable, is far more problematic than has been presented to the average layman through the media.
Anyone reading the media treatment of ice core interpretation would almost certainly get the idea that the dating of the ice cores is as simple as counting tree rings and that past weather interpretation from ice cores was something akin to looking at last nights animal tracks down on the sandbar. Nothing could be further from the truth.
The methods for identifying an annual layer in an ice core sample are based on recognizing the patterns of seasonal freezes and melts, normal patterns of dust levels, and of oxygen isotope ratios which are normally associated with temperature change. In the upper layers, this is fairly straightforward, though even there seasonal storms can mimic and be mistaken for annual layers. However, as the depth of the core increases the difficulty in accurately identifying yearly markers magnifies. At 300 meters it becomes impossible to rely on oxygen isotopes because of diffusion, so other methods must be used (Meese 26412).
Despite the media hoopla over volcanic referencing, it really isn’t a very valuable tool past 1000 B.P. (Schwander et al. 4243). In fact only a handful of volcanoes in the last 1000 years have actually been correlated absolutely by comparing the glass and tephra composition (Meese 26412). Smaller local eruptions can disguise larger eruptions and what appears to be volcanic spikes (an increase in acidity associated with volcanic activity) can be caused be other processes. Since the use of oxygen isotope measurements (the measurement of oxygen 18 and 16 ratios in the ice) and volcanic referencing become unviable fairly quickly, scientists must turn to other methods.
From the literature I read, researchers begin to depend heavily upon the LLS method (laser light scattering) for counting dust layers, and ECM, a method used to electronically measure changes in acidity. Both these methods have difficulties and increasingly the scientists must fall back on background assumptions and a greater degree of subjectivity begins to come into play. Both methods allow non seasonal inputs that can mimic annual events and can artificially up the overall year count (Meese, 26414).
A classic example of the type of subjectivity that can come into play here involves the use of the LLS method. Apparently the number of dust layers that can be visualized and counted depends on the size of laser used to do the counting. This became important when it was discovered that there was a 25,000 year discrepancy between the GRIP core and the Vostok ice core (a discrepancy between a reference point such as a volcanic spike which was believed to correlate both cores to the same date). The chief scientist went back and recounted the “visual stratigraphy” using the 8mm LLS beam that had been used to count the previous 1800 meters of the core. No additional annual layers could be discovered. However a 25,000 year gap is a huge discrepancy and couldn’t be allowed to remain. By switching to a 1mm LLS beam she was able to recount that section of the core and discover another 25,000 annual layers that had been “invisible” previously (Meese 26416). Now the two cores could be said to be in agreement. It must be noted however, that if the apparent discrepancy between the two cores hadn’t been noticed, the missing 25,000 years wouldn’t have been required and both cores would still have been said to be in agreement. One wonders how many more thousands of years could be found if the 1 mm LSS laser had been used on the rest of the core. This is what is known in the scientific community as “data massage”.
At this point it becomes obvious that the ages given for the bottom half of the ice cores aren’t derived by actually counting annual layers. Instead they are based on mathematical models which in turn are based on uniformitarian assumptions about the age of the earth and past precipitation rates, acceptance of the orbital theory of multiple ice ages, and other factors (Dansgaard 377). Before we look into these assumptions and how they work, I want to look at another example of the subjective manner in which data can be manipulated and then used to correlate various ice core interpretations.
In 1993 the European scientists studying the GRIP core discovered what they believed were abrupt climate swings in the period known as the Eemian. They claimed these finding were supported by pollen counts, sea core sediments and from isotope records from the edge of the Greenland Ice Sheet (GRIP 204). By 1997 however, it was decided by the Americans that what had been interpreted as weather changes was actually caused by other processes in the ice sheet that had nothing to do with the weather. Lehman tells us in his Nature article that,” these results later turned out to be artefacts (sic) of ice flow disturbance” (Lehman 117). All the correlating data would be reexamined and revised to fit the new scenario and assumed to have been misread or to have been in agreement due to chance. Almost every data point used to date or interpret weather from the cores can be caused or imitated by other chance processes and so interpretation is never final.
Tens of thousands of years can come and go and whole periods of past weather interpretations can be dismissed, and the correlating factors revised. Certainly, the process is far more tentative and the correlating data far more plastic than the public has been led to believe. As Lorius stated in an article from the journal, Global and Planetary Change, “ … beyond the Holocene period, absolute dating of the ice is not possible and ice core chronologies are then established by using glacial models or by comparison with other paleorecords with inherent limitations in accuracy” (Lorius et al. 98).
So we’re not really counting tree rings here. What kind of figure you come up with for the age of the bottom end of the cores depends on what kind of model you adopt and feed into your computer. In the model used for both the Greenland and Antarctica cores, certain main assumptions are taken for granted and these greatly control what the end result will be. First they believe that the trigger causing ice ages is small variations in the Earths orbit around the sun (the Milankovitch Theory). Because these well documented variations change the way that sunlight is distributed to the Earth (but not the amount of radiation) this has been thought to be the possible climate trigger for dramatic cooling. Because they also assume an age for the Earth in the hundreds of millions of years and knowing that orbital variations are cyclic, they believe there must have been at least 30 ice ages, maybe more. The assumption of the age of the earth is built into what is known as the uniformitarian theory of geology. This is the idea popularized by Charles Lyell that all the geological formations we observe were caused not by catastrophic events, but only by the normal processes we observe today. They also assume that precipitation rates have remained static and that the present ice sheet has been in place for at least 100,000 to 300,000 years. This dictates the width of the annual layers they feed into the model and thus the number of years they can fit into the bottom part of the ice cores (Dansgaard 377). They also use this orbital theory to date both sea cores and ice cores and it is on the Milankovitch theory that the whole scheme rises or falls. Since they assume multiple ice ages, and they believe that all the oxygen isotope changes in the sea cores are associated with these changes they attempt to “tune” cores in accordance with the known orbital variations. A critique of the dating methods of sea cores, the age of the Earth based on the geological column and the assumption of uniform precipitation rates for past ages is beyond the scope of this paper and probably unnecessary. The idea of multiple ice ages caused by orbital variations has sufficient difficulties to make it at least far more doubtful than is generally admitted.
First, let me make it clear that the orbital theory is just one in a long line of Ice age theories. It wasn’t accepted until the late sixties and not because it had answered all the contradictions and challenges it faced. However scientists abhor a vacuum and the orbital theory has now become fact in the minds of many scientists, similar to the Big Bang theory in astronomy. Let’s focus on a few of the problems with this theory and the assumptions that support it.
The orbital variations in the Milankovitch theory, though real, allow for only minimal change in radiation forcing, and it’s generally admitted that this is a major problem for the theory (John 42.). Another problem is that the times of some of the greatest climatic transitions as perceived by scientists, do not coincide with the times of orbital variation (Paillard 325). In other words, even with the ability to massage the data, the shoe just doesn’t fit.
Another problem, which I found quite interesting, is that there really is very little physical evidence for more than one ice age. For a long time, the majority of scientists believed that there had been only one ice age, but the idea of orbital forcing and one other interpretation of data helped to change their minds. This was the evidence of multiple layers of till at the edge of major glaciations. This was then interpreted as evidence of multiple ice ages. However since that time it has been discovered that these layers are actually caused by the movement of one glacier, and do not need to imply multiple glaciations (Derbyshire 77).
Recent studies have reinterpreted areas previously thought to have been glaciated multiple times, and have discovered most evidence points to just one glaciation.
Lack of erosion, insufficient till and the fact that the majority of till hasn’t traveled far from its source rock, are all strong evidence for one ice age, not many (Beany and Shaw, p.53, Feininger, ch.11). The fact that all of the existing till can all be accounted for by one ice age (Sugden and John 133), and the common knowledge that large parts of the northern hemisphere were never glaciated at all, greatly undercuts the assumption of multiple ice ages.
Finally, if there really have been multiple ice ages you would expect to find evidence for some of them in the ice cores themselves. After all if there have been over thirty ice ages in the past, evidence for at least two or three of them should be found in the cores. However, even though researchers were quite confident more than one ice age would be drilled through, the evidence never materialized (Dansgaard 379, GRIP members, p.203). Because of this lack of evidence, some scientists have postulated that almost all of the ice caps melted away some time during the last interglacial (Koerner et al. 19).
To summarize, the ability of orbital variations to force climate change appears negligible, and the timing of orbital changes doesn’t match with some of the main climate changes believed to have been discovered in the various data. The physical evidence for more than one ice age is arguably weak on land and non-existent in the ice cores themselves. If orbital variations didn’t cause the perceived climate changes, they can no longer be used as a tool to date the bottom part of the ice cores. Because of that, and the missing evidence for multiple ice ages, the need or evidence for hundreds of thousands of years disappears. Therefore the time frames adopted for the various glacial models are probably in error.
In conclusion, while I’m all in favor of pioneering new technology and energy sources, I don’t believe that orbital forcing together with increased CO2 were the causes of major climate changes of the past. I believe the top half of the ice cores have been dated reliably within a margin of error, and are valuable for weather interpretation. Despite the hoopla and many so called correlations, I don’t believe the bottom half of the cores have been dated accurately, and though some of the oxygen isotope changes may be related to climate change, the time frame is undefined. While I applaud the hardiness and efforts of the scientists and their crews who braved severe arctic conditions to collect ice cores, we need to remember that they are just as human as we are, and just as fallible.
Therefore, caution and restraint should be exercised before using ice core data as a basis for promoting far reaching legislation. I also believe the ice sheets may have achieved their present state more rapidly and recently than is generally allowed and that scientists should be looking for a different and possibly more catastrophic climate trigger as a cause for an ice age.
Dansgaard, W., et al
One Thousand centuries of climatic record on the Greenland Ice Sheet.
.Journal of Science, 166, pp 377-381, 1969
Beany, C. and Shaw,
The Sub Glacial Morphology of Southern Alberta: evidence for erosion.
Canadian Journal of Earth Sciences, Jan. 2000, pp.51-61
Glaciers and Environment, The Winters of the World, John, B. Ed.
John Wiley and Sons, 1979.
USGS paper cited in Frozen in Time, Oard, M. ICR monograph, ch.11 2004
Climate Instability during the last interglacial period recorded in the GRIP core.
Nature, 364, July 15, 1993, pp. 204-207.
Ice Ages: solving the mystery, Enslow press, New Jersey, 1979.
Koerner, et al,
Ice Core Evidence for Widespread Arctic Glacier Retreat in the Last Interglacial,
Annals of Glaciology. 35, June 2002, pp.19-24.
Sudden End of an interglacial, Nature, 390, 1997, pp.117-118.
Lorius, et al,
cited in The Frozen Record, Oard, M., ICR monograph, 2005
Meese, D.A., et al.
The Greenland Ice Sheet Project 2 Depth Scales: methods and results.
Geophysical Research, 102, (c12) pp. 26411-26423, 1991.
Glacial cycles: Towards a new paradigm, Reviews of Geophysics,
pp 325-246, 2001.
Sugden, J. and John, B.
Glaciers and Landscapes, a Geomorphologic approach, Edward Arnold Pub.
Schwander, J. et al
EPICA Dome Concordia Core. Geophysical Review Letters, 28,