The Naked Scientists

The Naked Scientists Forum

Author Topic: "Another Hockey Stick Illusion"?  (Read 48042 times)

Offline Bored chemist

  • Neilep Level Member
  • ******
  • Posts: 8669
  • Thanked: 42 times
    • View Profile
"Another Hockey Stick Illusion"?
« Reply #75 on: 01/05/2011 12:48:13 »
And, once again, for those who still don't get it.
"why do paleo-climatologists use collision diameter in preference to kinetic diameter when considering the migration of air molecules through firn and ice"
because it gives the right answer.

Offline Pete Ridley

  • Full Member
  • ***
  • Posts: 60
    • View Profile
"Another Hockey Stick Illusion"?
« Reply #76 on: 01/05/2011 21:44:01 »
In response to my original question Professor Wolff has provide two excellent overviews of the fundamental issue of how valid are the attempts to reconstruct past atmospheric composition from air recovered from ice sheets. I thank him for making the time to put together those submissions and hope eventually to cover all questions about the validity of those reconstructions. It is sensible to tackle in a structured manner all of the different complex processes that distort the composition of the original atmospheric air, To this end I have asked the founder of this blog/forum, Chris Smith, if it is possible to expand the thread question to read “Another Hockey Stick Illusion?” – “Do glaciers tell a true atmospheric CO2 story?”. That then allows us to follow the relevant processes from the top of the ice sheet down to the depths of the ice beneath.

I first would like to focus on the matter covered in my original question, which is
.. why do paleo-climatologists use collision diameter in preference to kinetic diameter when considering the migration of air molecules through firn and ice? ..
I shouldn’t have to keep repeating this but there are several people who have commented here who seem not to have read it properly. Professor Wolff went to a great deal of trouble to explain his understanding of it in his very first comment here (27th April @ 14:53:30) and I can’t improve on that.

Once a convincing answer has been given to my original question I would like to return to those other important points made by Professor Wolff that do not directly address my original question but are relevant to “Do glaciers tell a true atmospheric CO2 story?”.

In that comment Professor Wolff said
.. The underlying issue is whether we can believe that the air in bubbles in ice cores is an un-fractionated representation of the atmosphere .. when we crack open a bubble of air, does it contain exactly the relative proportions of different molecules as the air in the atmosphere ..
Note that reference to “air”, , not N2, not O2, not Ar, not CO2, etc. etc. etc. but “air” – all of it.

He then went on to say
.. you are more concerned about whether there is a further fractionation at the final stages of enclosure, related to the size of the molecule – do some smaller molecules more easily escape enclosure leading to a fractionation? ..
Note this time that Professor Wolff focuses on “the final stages of enclosure”, i.e. the deep firn, which is precisely where my original question is targeted at.

Professor Wolff then said
.. Certainly such a fractionation can exist for some atoms and molecules: it is very strong for neon (Ne), noticeable for O2 and Ar (at the permil to 1 percent level) but the literature says that there is no fractionation for CO2 (or CH4) compared to N2 ..

The issue in question should by that stage have been reasonably clear to anyone following this thread.

Let’s look now at what “the literature says” about size-dependent fractionation in deep firn as it approaches close-off, using three papers that Professor Wolff made reference to. First let’s see what is said in the 1996 paper “Natural and anthropogenic changes in atmospheric CO2 over the last 1000 years from air in Antarctic ice and firn” by Etheridge et el. As I pointed out almost a year ago elsewhere, that paper (and as far as I could find then, no other that Etheridge was involved in) made any reference to size-dependent fractionation of air in the deep firn, the only references to fractionation being to the gravitational type. I propose to return to this paper at the appropriate point in the wider discussion about “Do glaciers tell a true atmospheric CO2 story?”.

The papers that are much more relevant to my original question are the two from 2006 that Professor Wolff makes reference to. First let’s look at what “Evidence for molecular size dependent gas fractionation in firn air derived from noble gases, oxygen, and nitrogen measurements” by Huber, Severinghaus, et al. ( says about this. I make the assumption that all eight co-authors can reasonably be called paleoclimatologists. The next point to clear up is whether or not their paper is about the migration of air molecules through firn and ice. The title of the paper certainly suggests so, the paper discusses the molecules of all of the different atmospheric air components and their diffusion within the ice sheet and most importantly
.. In the present study, however, we are mainly interested at the bottom of the firn air column ..
This is precisely the area on which I am focussing, as Professor Wolff acknowledged when saying in his first comment here
.. you are more concerned about whether there is a further fractionation at the final stages of enclosure, related to the size of the molecule – do some smaller molecules more easily escape enclosure leading to a fractionation? ..

At the bottom of the firn air column, as close-off is approached, pores on the surface of the closing air bubbles and the channels that link air pockets must at some stage approach molecular size, i.e. the firn becomes a nanoporous medium. Gas purification experts use nanoporous media (e.g. zeolites) in industrial applications and numerous papers and patents can be found that describe the process. All of these refer to molecular kinetic diameter, not collision diameter, e.g. see the 2002 paper “Application of natural zeolites in the purification and separation of gases” by Ackley et al. ( and  “Sol-Gel Processing of Inorganic Membranes for Natural Gas Purification” ( which says
.. To efficiently separate CO2 (kinetic diameter = 0.33 nm) from CH4 (kinetic diameter = 0.38 nm) or (harder still) N2 (kinetic diameter = 0.364 nm) from methane, it is necessary to accurately control the average pore size between about 0.3 and 0.4 nm and achieve a narrow pore size distribution ..

So, I repeat my original question
.. why do paleo-climatologists use collision diameter in preference to kinetic diameter when considering the migration of air molecules through firn and ice? ..

As I discussed over a month ago on the Climate Conversation Group’s “Fallen snow” thread (which I provided a link to in my first post on this thread)
.. Huber et al. .. say “ .. a critical size of 3.6 Å implies diffusion through channels of about the same dimension .. an outlet from a closing bubble .. changes its dimension steadily from open to closed. .. Diffusion coefficients of gases in ice show a size dependence similar to our findings beside for Ar and O2 .. ”. They should have included CO2 along with those two but didn’t because (I believe) they were focussing on the close-off fractionation of other gases and were misled into ignoring CO2 through using collision not kinetic diameter ..

I won’t say much here at this stage about the paper “Fractionation of gases in polar ice during bubble close-off: New constraints from firn air Ne, Kr and Xe observations” by Severinghaus & Battle ( In Section 1.2 and 3.1 of the paper they provide a good description of the process that I am focussing on and they discuss the results of their
.. simple model of the bubble close-off fractionation …
They go on to say
.. The model presumes that fractionation is caused by selective permeation of gas through the ice lattice from slightly overpressured bubbles. ..
I’m puzzled by that reference to “ice lattice” because there is another process that is thought to take place after close-off. This is a process related to the one that I am focussing on at present and I would like to consider this one later, however, it is noticeable that they again refer to the 0.36nm molecular size. They say
.. The large atoms Kr and Xe do not appear to be fractionated by this process, despite the large size difference between the two gases, suggesting a threshold atomic diameter of 3.6Å above which the probability becomes very small that the gas will escape from the bubble. These findings have implications for ice core and firn air studies that use gas ratios to infer paleotemperature, chronology and past atmospheric composition ..
Once again, they ignore kinetic diameter in favour of collision diameter (see Table 1).

If anyone is interested, more comments of mine on this paper can be found on the Climate Conversation blog threads:
- “It’s not warming, you nitwit — it’s cooling” ( in my comment of 21st March @ 11:26 and
- “Fallen snow” ( in my comment of 23rd March @ 11:17.

Professor Wolff, are you at liberty to attach a copy of your 2011 paper “Greenhouse gases in the Earth system: a palaeoclimate perspective” so that subsequently we can look at the evidence provided therein supporting your conviction that
.. the ice-core record provides a faithful record of changing atmospheric composition. ..
( That paper may provide the answer to the question used by Professor Jaworowski et al. as the title to their 1992 paper “Do glaciers tell a true atmospheric CO2 story?” (

I referred to that paper in my comment of 13th April @ 21:58:19) and as I said then
.. When I discussed this last June with Professor Zbiniew Jaworowski, whose 1992 paper first drew my attention to this issue, he expressed the opinion that “This is a highly specialized field of science. My impression is that it is a terra incognita for glaciologists”. Subsequently I have asked the same question of Professors Richard Alley, Jeffey Severinghaus and Michael Bender without receiving any worthwhile justification for their use of collision diameter. ..

Best regards, Pete Ridley

Offline ericwolff

  • First timers
  • *
  • Posts: 3
    • View Profile
"Another Hockey Stick Illusion"?
« Reply #77 on: 02/05/2011 12:12:24 »
Pete and other readers,

My first post (353863) was intended to be a very clear and direct answer to your original question; additionally I have since contacted Professors Alley and Severinghaus: although I have not seen their correspondence with you, they both tell me that they felt they had already answered you, in similar terms, in private e-mails in the past.  Given this, I am surprised that you are still claiming your question has not been answered.  I will therefore make one closing attempt to be absolutely clear.  Your specific question was:

"why do paleo-climatologists use collision diameter in preference to kinetic diameter when considering the migration of air molecules through firn and ice?".
The problem that we are all having is that this is a false or loaded question: the implication elsewhere in your posts is that the idea that CO2 is unfractionated compared to N2 comes from authors making this assumption.  In fact it is quite the other way round: the empirical evidence that CO2 is not fractionated on enclosure (as well as the observation that Ar is less fractionated than O2) is what led these authors to hypothesise that collision diameter was the controlling variable.  So the specific answer is that they use collision diameter because this is what allows them to rationalise the data they observe.

I think that none of us has a definite molecular-level understanding of the physical process occurring at closeoff, and it would be great if someone can do the experiments in the lab to understand that better.  But it won't alter the empirical facts.

Incidentally it may be worth mentioning that, in their expts, Severinghaus and Huber used trace gases (such as Ar and O2) that are understood to have been invariant in concentration over recent decades, precisely so they could look at the diffusion and enclosure processes free from any assumptions about temporal change in concentration. 

You raised several other points in different posts, and I simply don't have time to answer them all.  Your questions about kinetic diameter imply you have got quite deep into the literature, so I am surprised that you are confused about the kink in firn air profiles, as this is clearly explained in several of the papers.  It is the result of moving from the diffusive zone (in which the air is ageing according to diffusive mixing with the air above) into the non-diffusive zone, where the air is ageing with the ice, ie much faster ageing. Just to give a schematic example:
Imagine a site at which the snow accumulation rate is 1 m ice equivalent per year, and the non-diffusive zone starts at 70 m snow depth (which because the firn is much less dense near the surface, might be 50 m in ice equivalent).  This is similar to the situation at Law Dome, but I have chosen to use a hypothetical site with round numbers for simplicity.  At 70 m, the ice is 50 years old, but the air is about 10 years old.  At 80 m, the ice is 60 years old, and the air is 20 years old.  Thus the CO2, would change from about 385 ppmv at the surface (2010 concentration, South Pole) to 365 ppmv at 70 m, then 350 ppmv at 80 m.  This would of course show a sharp kink if plotted against depth.

Finally, I am surprised you are citing Jaworowski as an expert.  His monographs about ice core CO2 have been comprehensively refuted by those who actually work on the topic, most notably some 15 years ago by Professor Hans Oeschger.  I myself looked into his work a decade ago, but found it so full of misunderstandings (if I am generous) that I did not bother to formulate any specific refutation, but such material is easily available on the web.  The saddest part is that Jaworowski cites as issues aspects of ice core analysis that many of the pioneers of the field were very aware of, and spent decades carefully studying and overcoming.  It is therefore galling to see their efforts ignored.

I will be happy to answer other well-formulated questions about ice cores, but I don't think I have any more to add to this thread.

Offline yor_on

  • Naked Science Forum GOD!
  • *******
  • Posts: 12001
  • Thanked: 4 times
  • (Ah, yes:) *a table is always good to hide under*
    • View Profile
"Another Hockey Stick Illusion"?
« Reply #78 on: 03/05/2011 01:58:36 »
Thanks Eric. And, you're most welcome back whenever you find the time and inclination. It's been a, ah, interesting journey, with a lot of twists if I may say so, but myself I think your posts have been oasis's of clearness.

I will close this thread now, as I think your post says it all.

The Naked Scientists Forum

"Another Hockey Stick Illusion"?
« Reply #78 on: 03/05/2011 01:58:36 »


SMF 2.0.10 | SMF © 2015, Simple Machines
SMFAds for Free Forums