The Intergovernmental Panel on Cllimate Change (IPCC) estimates pre-industrial levels of CO2 based largely on information derived from ice cores. These are kilometres long cylinders of ice drilled (in short sections) from the Greenland and Antarctic icecaps.
The theory is that air bubbles trapped in the ice are samples of the ancient atmosphere, and thus give an accurate reading of CO2 levels in those ancient times. The ice cores tell us that the pre-industrial level of
CO2 was about 290 parts per million (ppm).
Of course it’s not that simple. There are many factors to take into account, because ice from deep down in the ice cap is under enormous pressure, and that pressure is released when the cores are brought to the surface.
A brief summary of some of the processes going on in ancient, deep ice can be found here:
Zbigniew Jaworowski, a multidisciplinary scientist, has long been a critic of the ice core CO2 record:
He makes a number of claims:
1. Ice core results reported before about 1985 showed higher CO2 levels than those reported since 1985, which he details in his 1997 paper.
2. Low pre-industrial concentration of CO2 in ice cores is an artefact caused by more than 20 physical-chemical processes. For example, CO2 is more than seventy times as soluble in water as is nitrogen, and liquid water is present in deep ice, even at -73C.
3. Drilling and retrieval of ice cores, including their decompression as they are brought to the surface, drastically disturbs the sample.
4. Subsurface melting of snow due to absorption of solar radiation is common in very cold sites, and may occur several times a year. This implies that the practice of dating ice cores using isotopes of oxygen, which assumes the temperature peaks are yearly events, is wrong.
The net result of these effects is to lower the CO2 levels in ice compared to that found in the atmosphere, by 30 – 50%, and to minimise fluctuations in CO2 in ice compared to the atmosphere.
Jaworowski illustrates this by comparing ice core data to other CO2 proxy data, such as stomata indices, which gives higher atmospheric CO2, with more variation than the ice core data over the last 8,000 years.
There is no doubt that ice core analysis is a highly technical subject, and scientists must make many assumptions when interpreting the data. One assumption that I personally found very interesting was that air inclusions in the snow and firn, or compacted granular snow, are well mixed with the atmosphere. In other words, the air in the hard packed snow is basically the same as the atmospheric air, down to the level of the ice. This could be 50 – 120 metres below the surface. Imagine hard packed snow that far down having a connection (and circulation with) the surface air. To me this is very counter-intuitive, but apparently Argon dating confirms it.
Jaworowski claims this is wrong, and that impenetrable layers, caused by melting and refreezing, cut off air inclusions from the atmosphere much earlier. This is the basis of his criticism that ice core data from the Siple core showed similar CO2 levels as the atmosphere did 83 years later, and thus the air in the ice core was assumed to be 83 years younger than the ice it was in.
Jaworowski has been criticised here:
The late Hans Oeschger, a prominent ice core scientist, was also critical of Jaworowski, and wrote a letter in 1995 expressing his views :
Following are some other links.
Jaworowski interview in May this year:
Diffusion of CO2 in atmospheric gas analysis (on pp10 – 11)
From Paul Williams who lives in the Adelaide Hills, Australia
Other contributions to this blog from Paul Williams include:
Hockey sticks and ancient pine tree, at http://www.jennifermarohasy.com/blog/archives/001546.html
Measuring Atmosheric C02, at http://www.jennifermarohasy.com/blog/archives/001983.html
Build Dykes to Beat Global Warming, at http://www.jennifermarohasy.com/blog/archives/001983.html