Ice Ages #2

132 Comments

  1. MattN
    Posted May 1, 2008 at 10:06 AM | Permalink

    Someone asked over on RC some time ago “what would need to happen to discredit AGW theory?” Gavin’s response was (IIRC):”At least 10 years of non or negative temperature trend.”

    Given that NASA has announced the PDO phase shift and report after report is now surfacing claiming there will be no warming for the foreseeable future because of it, what will Gavin and his lackeys say in 2018 when is has been 20 years without eclipsing the 1998 record?

  2. yorick
    Posted May 1, 2008 at 10:22 AM | Permalink

    What exactly does that have to do with ice ages? Beyond the idea that synchronized PDO minima in conjunction with a solar minimum at the proer place in te Milanchovich forcing tables might lead to a couple years without appreciable summer at high lattitudes, snow doesn’t melt, and albedo, yada yada yada. How do you know the scenario is false? Easy, you can’t model it with a GCM, therefore it can’t happen.

  3. Posted May 1, 2008 at 11:02 AM | Permalink

    [snip – no need to vent]

  4. mbabbitt
    Posted May 1, 2008 at 11:35 AM | Permalink

    An outsider’s question:
    Was this PDO shift predicted or taken into account by anyone in the AGW community?
    Coming out now and saying that GW has taken an unexpected vacation seems a bit silly.
    If the global community of scientists with all of their sophisticated models and computations could not predict this dramatic shift to a cooling sycle, how can anyone in the general public have any faith in any of their predictive statements? This lack of forewarning just doesn’t pass the smell test, especially when they are now stating that in 2015 warming will begin again? Reminds me of Paul Erlich and his population disaster predictions — just keep moving the goal post forward.

  5. Posted May 1, 2008 at 11:41 AM | Permalink

    yeah, sorry Steve.

  6. Posted May 1, 2008 at 12:15 PM | Permalink

    4 (babett): It has long been appreciated that there might be a roughly 60-year ‘wave’ in PDO, e.g.:

    and that therefore a negative phase is ‘due’ or, at least, not surprising.

  7. MarkW
    Posted May 1, 2008 at 12:18 PM | Permalink

    It has long been appreciated that there might be a roughly 60-year ‘wave’ in PDO, and that therefore a negative phase is ‘due’ or, at least, not surprising.

    Apparently the cold phase of the PDO can halt global warming, but the warm phase has no impact on it whatsoever.

  8. Posted May 1, 2008 at 12:23 PM | Permalink

    7 (MarkW):

    Apparently the cold phase of the PDO can halt global warming, but the warm phase has no impact on it whatsoever.

    I think that you are wrong on this. Clearly the warm phase must make a significant contribution to the run of the temperature and be conducive to warming. You should be able to see that.

  9. MarkW
    Posted May 1, 2008 at 12:27 PM | Permalink

    I can see it, but the IPCC apparently can’t.

    If they do, could you point me to the paper in which they specify the percentage of the warming over the last 30 years that is due to the just ended PDO warm phase.

  10. tetris
    Posted May 1, 2008 at 12:38 PM | Permalink

    Leif,
    Re:7 and 9. Is it possible that Mark is being a tad sarcastic? Maybe he’s making the point that “natural variation” is readily counted/discounted when it impacts one way, but not the other way around?

  11. Posted May 1, 2008 at 12:44 PM | Permalink

    7-15 should be snipped, of course.

  12. Sam Urbinto
    Posted May 1, 2008 at 12:48 PM | Permalink

    It’s a wry observation and an indictment of those that claim things one way are a disaster and the other way are not important. And then perhaps switch viewpoints when the evidence changes.

    Inconsistency.

  13. Posted May 1, 2008 at 12:54 PM | Permalink

    Leif, I believe MarkW was engaging in sarcasm. At least that’s how I read it.

    MBabbitt (#4) I have to date only seen one paper referenced, the Tsonis et al paper, that addressed the combination of issues A secondary link is here: http://motls.blogspot.com/2007/08/mechanism-for-major-climate-shifts.html

    … By studying the last 100 years of these cycles’ patterns, they find that the systems synchronized several times.

    Further, in cases where the synchronous state was followed by an increase in the coupling strength among the cycles, the synchronous state was destroyed. Then. a new climate state emerged, associated with global temperature changes and El Nino/Southern Oscillation variability.

    The authors show that this mechanism explains all global temperature tendency changes and El Nino variability in the 20th century.

    Major climate shifts have occurred or will occur around 1913, 1942, 1978, 2033, and 2072 according to the authors of this recent paper, who also predict a 0.2 Celsius cooling between 2005 and 2020 which should be followed by a 0.3 Celsius warming until 2045 or so – then cooling for the rest of the 21st century.

    Tsonis, Anastasios A.; Swanson, Kyle; Kravtsov, Sergey:
    ‘A new dynamical mechanism for major climate shifts’
    Geophys. Res. Lett., Vol. 34, No. 13, 12 July 2007

  14. Sam Urbinto
    Posted May 1, 2008 at 12:56 PM | Permalink

    Yep. Maybe stated a bit better this way: Unthreaded #34 #12 MarkW said:

    The thing I find interesting is this view that while the cool phase of the PDO can completely over power global warming, the warm phase of the PDO has no affect whatsoever on it.

  15. Posted May 1, 2008 at 1:17 PM | Permalink

    18 (Sam): in general, sarcasm is not useful. What Mark could have said simply and directly, is that the IPCC [and their ilk] seems to be biased in that particular way, but then, of course, we all knew that, so what’s the utility?

  16. Sam Urbinto
    Posted May 1, 2008 at 1:19 PM | Permalink

    Leif #22 “the IPCC [and their ilk] seem to be biased in that particular way”

    {witty observation} Never hurts to mention that from time to time.

    🙂

  17. Posted May 1, 2008 at 1:34 PM | Permalink

    23 (Sam): never hurts, but it can be said plainly and directly. The problem with sarcasm is that it has a tendency to spill over into ad-hominem comments, like #19.

  18. Patrick M.
    Posted May 1, 2008 at 1:41 PM | Permalink

    re 4 (mbabbitt):

    From what I’ve seen on the PDO, nobody seems to be able to predict it. That’s one reason I don’t have much faith in the GCM’s. If they can’t predict the PDO how can they hope to predict the global temp?

  19. Jon-Anders Grannes
    Posted May 1, 2008 at 1:55 PM | Permalink

    Re 1 and 6

    How could UNEP/IPCC claim that “we” where responsible for climatic change the last 50 years?

    When a simple PDO change(nature) according NASA have the potensial too totally undermine that claim?

  20. Sam Urbinto
    Posted May 1, 2008 at 2:34 PM | Permalink

    Leif #23: I agree. It can detract.

    Jon-Anders #25: There is no doubt population, industrialization and urbanization have an effect upon the climate, the problem is quantifying and qualifying it, and then proving there’s a net increase.

    What the climate change aggregators are doing is trying to give a certain impression. When they should be saying “We don’t know, but we think…” Trouble is, there’s usually no good idea for thinking that way any more than for thinking that coin will be tails next flip. Or more that die will be a 6 on the next roll.

    Hey, that’s good. Somebody’s asking us to bet money the next roll will be a 2 and then the thing will hang in mid-roll and balance on an edge, never falling for a milllllyyyuuuuun years.

  21. Willem de Lange
    Posted May 1, 2008 at 3:07 PM | Permalink

    The PDO shift to a cool phase was predicted more than a decade ago. I’ve even published on the consequences for New Zealand back then. There were a flurry of papers prompted by work at the University of Washington (Mantua and colleagues). However, one of the papers I have cited most often to address the PDO contribution to global warming is

    Karl, T.R., Knight, R.W. and Baker, B., 2000. The record breaking global temperatures of 1997 and 1998: evidence for an increase in the rate of global warming? Geophysical Research Letters, 27(5): 719-722.

    One of the reasons I got annoyed with the IPCC when I was involved, was an almost universal dismissal of the contribution of natural forcing to the observed warming – particularly the role of decadal-scale climate variability such as the PDO. In discussions with one of our Nobel Peace Prize winning meteorologists back in the 1990s, their position was that the greenhouse effect had overwhelmed natural variability and it just would not be possible for the PDO to switch.

  22. lgl
    Posted May 1, 2008 at 3:44 PM | Permalink

    #6 Leif

    It’s the Triple Saros cycle
    Wiki: “However, if one waits three Saros cycles, the local time of day of an eclipse will be nearly the same. This period of three Saros cycles (54 years 1 month, or almost 19756 full days), is known as a Triple Saros”

    And the sum of the Triple Saros and the 74,4 years AMO cycle will peak every 200 years, also a known climate cycle.
    http://www.virakkraft.com/AMO-NPI-PDO.ppt

  23. yorick
    Posted May 1, 2008 at 3:47 PM | Permalink

    I have heard that too, that any cooling was impossible. Sic semper doomsday cults.

  24. Posted May 1, 2008 at 3:51 PM | Permalink

    28 (Lgl): or perhaps it is the The Damhsa (Gaelic for Dance) Theory formulated by analyzing climate data and applying the General Relativity Theory and orbital forcing to the time series and proposing a solution to the variable data?

  25. lgl
    Posted May 1, 2008 at 4:01 PM | Permalink

    #30
    and that theory shows that the winter NPI cycle is 54 years?

  26. CWells
    Posted May 1, 2008 at 4:03 PM | Permalink

    MarkW#9-
    Regarding your request for info on PDO; Fawcett & Jones have a new paper, “Waiting for Global Cooling” available at http://www.aussmc.org/documents/waiting-for-global-cooling.pdf where they attempt to reduce the global temp signal by, not the PDO, but by the ENCO. Close to what you were asking about, but not quite, and their math is not fully disclosed. However, look at their graphs………w/o ENSO influence, the effect of PDO is clearly apparent; ’05 to ’46, strong warming during the warm PDO, ’46-’77 falling and flat temps during the cold, and a new rise after ’77 during the last warm PDO. They clearly have reduced the noise, but I don’t think they see yet what they may have uncovered. I.E.- direct correlation of PDO to temps!!
    CWELLS

  27. Posted May 1, 2008 at 4:08 PM | Permalink

    31 (lgl): hard to say, but have a look.

  28. srp
    Posted May 1, 2008 at 4:28 PM | Permalink

    I have a toy called a Randomly Oscillating Magnetic Pendulum (ROMP). It has a magnetic pendulum bob and comes with a set of moveable magnets that can be positioned anywhere on the base to attract or repel the pendulum. When you release the pendulum over the base and its arrangement of magnets, it describes a fairly complex, only-partly predictable twisting trajectory with non-monotonic amplitude. Friction (at the top of the pendulum hinge, and presumably from the air) eventually causes it to run down and stabilize. As a minor collector of desk toys I highly recommend this one.

    Are the climate oscillations like the non-monotonic oscillations of my ROMP? I mean, are there endogeonous states, independent of any shocks or forcings, where the energy of the fluids is “trapped” by constraints and so oscillates rapidly but then “breaks out” into a longer-amplitude, lower velocity track before getting trapped again? And if so, are they repeatedly “pumped” and rejuvenated by external energy or should we expect them to run down just like my pendulum?

  29. lgl
    Posted May 1, 2008 at 4:30 PM | Permalink

    33, yeah, I knew your were joking. The rule of ice age transitions is easy:
    Cold phases end close to every n*41 kyr where n=1,2 or 3 at the 22 kyr precession cycle minimum closest to the obliquity maximum closest to the peak of the 95 kyr precession cycle envelope. A new cold phase starts before or at the following obliquity minimum. (I.e. The next ice age begins within 10 kyr from now)

  30. Posted May 1, 2008 at 4:34 PM | Permalink

    35 (lgl): yeah, I couldn’t resist. I thought you were joking too.

  31. lgl
    Posted May 1, 2008 at 4:47 PM | Permalink

    36, of course not. Don’t you agree there’s a 18yr and a 3*18yr=54yr cycle in the NPI?

  32. Posted May 1, 2008 at 5:07 PM | Permalink

    37 (lgl): There does seem to be a ~60 year quasi-period some of the time, but I do certainly not agree that the period is 19756 days.

  33. lgl
    Posted May 1, 2008 at 5:23 PM | Permalink

    38, and I didn’t say 19756 days. Average, you know we are not dealing with Swiss clockwork here. There’s 1,5 cycle between 1908 and 1990, that’s 1 year off.

  34. Posted May 1, 2008 at 6:02 PM | Permalink

    39 (lgl): that’s right you only said “almost 19756 full days”. I do not consider it credible that the Saros cycle has anything to do with PDO or ENSO.

  35. Dennis Wingo
    Posted May 1, 2008 at 6:21 PM | Permalink

    (32) CWells an Mark

    Here is a paper from 2001 that has some very interesting data on the PDO, and how the theory of it came about.

    http://www.iphc.washington.edu/Staff/hare/html/papers/pcworkshop/pcworkshop.pdf

  36. Basil
    Posted May 1, 2008 at 7:53 PM | Permalink

    #39 (Leif): “I do not consider it credible that the Saros cycle has anything to do with PDO or ENSO.”

    What about the lunar nodal cycle, at least for the PDO? There are papers claiming an 18.6 year cycle for certain climate dynamics in the Arctic, if I’m not mistaken. I think for “bidecadal” periodicities, it is sometimes difficult to distinguish between the impact of the lunar nodal cycle, and the Hale cycle, and that both may be influencing such periodicities.

    Basil

  37. Posted May 1, 2008 at 8:03 PM | Permalink

    42 (Basil): There are many possible periods and it can be hard to keep them apart. What I mean by ‘no credence’ is that there is no general agreement on any of them, and I’ll tend to be ‘mainstream’ on this since Climate is on the edge of my expertise. I’m a ‘solar’ guy, remember. Just because I do not consider something credible does not mean that there are not multitudes of supporters and enthusiasts with a different opinion. [How is that for triple negative…]

  38. Steve McIntyre
    Posted May 1, 2008 at 8:15 PM | Permalink

    #42. Basil, this thread is about Ice Ages.

  39. bender
    Posted May 1, 2008 at 9:11 PM | Permalink

    Good answers exist for #24 and #25. But PDO and GCMs are OT.
    (Andrew, take note of these arguments.)

  40. Posted May 1, 2008 at 9:34 PM | Permalink

    47 (bender): take them over to Svalgaard#6, where they are welcome, as we are discussing if the Sun is a factor in these.

  41. Dr.M.Pharaoh
    Posted May 2, 2008 at 3:43 AM | Permalink

    Could someone clear up a question for me. Is there a causative link between the PDO and the indo-pacific warm pool extent and temperature. Such a large area can dump large volumes of water vapour into the atmosphere.

    Mark

  42. CWells
    Posted May 2, 2008 at 10:16 AM | Permalink

    DennisW:
    Thanks for the article- didn’t have that one.
    But to my earlier point- going to Fawcett/Jones; after smoothing/refining out ENSO, doesn’t the remaining temp curve follow PDO?? And NOT the continuous CO2 increasing curve/line from Mauna Loa?? (Sorry, maybe someone better at graphics than I can get the plots and move them here. SteveMc??) And do not give me the particulate arguement for the flat period ’46-’77; there was particulate before and after, and there wasn’t continuous volcanic eruptions during the period either.
    I’m excited that I think I see something and want someone to explain it if possible. Thanks.
    C

  43. CWells
    Posted May 2, 2008 at 10:20 AM | Permalink

    I hope #48 isn’t too off topic as it relates to the possible cooling PDO and if it will contribute to a global/regional temp cooling.
    Thanks,
    C

  44. Posted May 2, 2008 at 10:37 AM | Permalink

    47-49: take those over to Svalgaard #6, please as they are OT here.

  45. Tim
    Posted May 2, 2008 at 12:31 PM | Permalink

    And for all these possible connections, let me just say because my stat teacher drilled it into us over and over. Correlation is not causation. Let’s not make the same mistakes the IPCC true believers do.

  46. Jon T
    Posted May 2, 2008 at 2:38 PM | Permalink

    If, as seems to be concensus by ommission or questions, we don’t know how much warming the PDO has been adding, then we dont know how much cooling it will now add. If as some imply a significant proportion of the warming was from the PDO then it should be the same for cooling and we are now in for a massive change in temps. From my limited reading if various people are also correct that lower solar levels are also probable thru cycles 25/25 and that they result in lower temps too (Q is PDO solar related) then the downturn could be ice age inducing.

    Sorry if i am just repeating others views?

    PS – i thought the thread was about possible ice ages

  47. CWells
    Posted May 2, 2008 at 3:31 PM | Permalink

    To directly address this idea of Ice Age- based on the ice core data I’ve seen, the peek interglacial periods do not look to be very long compared to the complete cycle (glacial through to peak warm through next cooling phase). Cooling appears to take twice to three times (even longer) as long on average as the warming up side.
    Where are we in the currect cycle?
    Looking at the Vostok and other core data it seems we’ve been at peak temp longer than the prior six cycles. Are we past the flip point versus the prior cycles timewise(ie, perhaps we’ve ameloriated the cooling with our CO2 contribution) or is the actual max timeframe still in front of us?
    C

  48. Pofarmer
    Posted May 2, 2008 at 5:07 PM | Permalink

    A substantial contribution to the
    record warmth of 1998 came from the very strong El Niño of 1997/98 and, when the annual
    data are adjusted for this short-term effect (to take out El Niño’s warming influence), the
    warming trend is even more obvious.

    I thought this quote from the Fawcett/Jones paper kind of stood ou. Yes, if we ignore the effect of the earth on Climate, then man’s influenc eis much more apparent.

    They get paid for this?

  49. Evan Jones
    Posted May 2, 2008 at 7:12 PM | Permalink

    Was this PDO shift predicted or taken into account by anyone in the AGW community?

    Not really. PDO was not supposed to be a significant factor. the post-1910 rise was volcanoes. The post-1950 drop was aerosols. The post-1978 rise was CO2.

    Remember how the D’Aleo correlation got the raz over in Tamino Hall?

    The rest of us, of course, have been banging on PDO for over a year and pointing out that it was around 30 years into a 30-year phase.

    And predicting how when it happened it would be put down as a mere masking of the One True AGW. (It would seem our climate-political model was more accurate than the IPCC’s.)

  50. Evan Jones
    Posted May 2, 2008 at 7:30 PM | Permalink

    I’m a ‘solar’ guy, remember.

    Once we we do in the AGW hotfoot crowd, youse guys and us “Axe” Moerner’s sea witches duke it out!

    The only ones we hate more than the hotfoots are the blasted sun worshipers! SPLITTERS!

    #B^1

  51. Evan Jones
    Posted May 2, 2008 at 7:37 PM | Permalink

    And for all these possible connections, let me just say because my stat teacher drilled it into us over and over. Correlation is not causation. Let’s not make the same mistakes the IPCC true believers do.

    Mmmm. Yes. But it’s a starting point. And if it’s an up-down-up-down-up correlation like the PDO a presumption of causation carries a certain amount of bucckshot. Something to be going on with.

    And lack of correlation may be held to demonstrate lack of causation.

  52. Evan Jones
    Posted May 2, 2008 at 7:37 PM | Permalink

    And for all these possible connections, let me just say because my stat teacher drilled it into us over and over. Correlation is not causation. Let’s not make the same mistakes the IPCC true believers do.

    Mmmm. Yes. But it’s a starting point. And if it’s an up-down-up-down-up correlation like the PDO a presumption of causation carries a certain amount of bucckshot. Something to be going on with.

    And lack of correlation may be held to demonstrate lack of causation.

  53. lgl
    Posted May 3, 2008 at 3:51 AM | Permalink

    42# Basil

    Yes you’re right about the arctic, http://ansatte.hials.no/hy/climate/preEcoSysTromsø_070314.pdf
    (and n-pacific, http://www.pices.int/publications/presentations/PICES_15/Ann15_S3/S3_Tadokoro.pdf

    And it’s no wonder that the 54 yr tide will be more powerful than the 18,6 yr tide because every 54 yr an eclipse returns not only to roughly the same latitude but also to the same longitude

    http://eclipse.gsfc.nasa.gov/SEsaros/SEsaros.html
    “For solar eclipses, this results in the shifting of each successive eclipse path by ~120º westward. Thus, a saros series returns to about the same geographic region every 3 saroses (54 years and 34 days).”

  54. dearieme
    Posted May 3, 2008 at 1:29 PM | Permalink

    Leif, I am not being in the least sarcastic when I point out that many months ago I introduced the term “Global Tepiding” and accordingly look forward keenly to receiving my Nobel Prize.

  55. Posted May 3, 2008 at 2:19 PM | Permalink

    60 (derieme): ‘predicting’ a long-awaited cooling ain’t going to get you the Prize. Proving conclusively what people have failed [and not for want of trying] for 400 years, that there is a day-by-day relationship between lower troposphere temperatures and the sunspot number just might because of the enormous economic impact of such a tight relationship. And I’m not being sarcastic either [although I think you were a little bit less than serious in claiming you were not].

  56. Geoff Sherrington
    Posted May 21, 2008 at 5:05 AM | Permalink

    A paper, maybe a draft, fell out of the Net today and I am puzzled by it. Author is J J Drake, who was active at McMaster Uni in the 1970-80 period, if the same person. If it is correct, it is fundamentally important.

    http://homepage.ntlworld.com/jdrake/Questioning_Climate/userfiles/Ice-core_corrections_report_1.pdf

    The methodology plots CO2 concentration for Vostok versus the difference between ice age and CO2 age there (as derived from isotopes in ice and gas, I presume) and by simple maths derives a new interpretation for CO2 variation in Vostok, which is about level at 325 ppm +/- 25 ppm for 400,000 years. This “adjusted” CO2 fails to correlate with deuterium temperatures deduced by Jouzel at al.(1987, 1993, 1996) on the Jouzel time scale.

    Dated 17 May 2008.

  57. Pat Keating
    Posted May 21, 2008 at 6:49 AM | Permalink

    56 Geoff

    Yes, an interesting paper.
    However, don’t fully understand the IGD quantity. What are the exact definitions of “Ice Age” and “Gas Age”?

  58. Posted May 21, 2008 at 7:12 AM | Permalink

    57 (Pat): Jonathan Drake sent me this
    http://www.ess.washington.edu/Surface/Glaciology/courses/ess431/LECTURES/2006/Lect_15_Ice_Cores_2_2006.pdf

  59. Posted May 21, 2008 at 7:30 AM | Permalink

    57 (Pat):
    The central Figure is this one:

    I asked Drake this question:
    “If I were to integrate the CO2 concentration over the entire ice core from the bottom to the top and calculate the number of molecules of CO2, I would find two very different values using the red curve or using the green curve. The red curve number would be about double the green curve number. Where did all these molecules [the difference] go [or come from – with the difference calculated the other way]? The number of molecules in the entire column should be constant [apart from a tiny loss at the surface] no matter how the molecules diffuse inside the column. And if it is not constant, why?”
    ———
    I did not understand his answer:
    “I don’t know the origin of the trend in IGD, I can only assume it comes from the methods used in its calculation. It is the relationship between CO2 and IGD that is responsible for the integral difference you refer to. There is no effective change in the number of particles.

    The average IGD is about 4000 years which corresponds to an original CO2 level of approximately 235 ppmv (green curve), and so the average CO2 will be close to this value.

    If I had chosen to, I could have picked IGD = 4000 as my reference point and the trend in the red curve (corrected CO2) would have appeared at a mean value of about 235 ppmv. However, there was no justification to do this. It makes far more sense to set IGD = 0. That is when the ice and gas are the same age. This gives the additional benefit of synchronising the absolute ages by definition.

    Note the choice of reference IGD does not affect the shape of the corrected CO2 curve, merely its average level.

    The process effectively pegs the value of IGD rather than letting it modulate the data. So instead of having the tail wagging the dog, we have the dog wagging its tail!”

  60. Phil.
    Posted May 21, 2008 at 8:05 AM | Permalink

    Re #56 et seq

    I don’t see the logic for this correction, no where is it spelled out what the mechanism is. The age difference seems to correlate strongly with temperature. If the concentration of CO2 was always ~325ppm what is the source of the 180ppm samples? This would suggest that the CO2 (the heavier gas) gets diluted preferentially when it’s cold (similar to Leif’s point), again what’s the mechanism? If the air is trapped at ~325ppm why does the heavier gas diffuse out preferentially?

  61. Posted May 21, 2008 at 9:44 AM | Permalink

    60 (Phil.):

    why does the heavier gas diffuse out preferentially?

    And where does it go? Unless lost at the surface [which makes the whole thing meaningless] the CO2 molecules will have to stay in the ice column and so the total [or average – if you prefer] amount would be the same no matter what the difference IGD is, me reckon.

  62. Pat Keating
    Posted May 21, 2008 at 9:51 AM | Permalink

    58 Leif
    Thanks for that URL. There’s a lot of material there.
    I think the answer to my question is there but I’m not sure whether Drake’s ice age is determined by stratigraphy or carbon dating. I assume the gas age is determined by carbon dating. Hopefully, the ice age is also, so that he is subtracting apples from apples.

  63. Pat Keating
    Posted May 21, 2008 at 9:58 AM | Permalink

    60, 61 Leif Phil

    why does the heavier gas diffuse out preferentially?

    As I understand it, because they move to colder ice. So they leave when the air is consistently above the ice temperature.

  64. MBK
    Posted May 21, 2008 at 9:59 AM | Permalink

    It seems to be just a statistical point with no mechanisms given.

    To 59 (Leif), the correction uses a factor from extrapolation to zero IGD, so it is not “redistributing” the original integral but assumes that depending on IGD some of the CO2 has effectively vanished from the enclosed bubbles. This is why the corrected integral is higher. The missing CO2 by this logic would have to have gone into the surrounding ice.

  65. SteveSadlov
    Posted May 21, 2008 at 10:07 AM | Permalink

    Are we there yet?

    (Written as an unnnnnnnnnnnnnnprecedented near-summer-solstice height lowering event unfolds along the US West Coast ….)

  66. MBK
    Posted May 21, 2008 at 10:08 AM | Permalink

    to 61: if only the gas in the bubbles is used for ppm determination, then there can be a difference if it goes in the ice.

  67. Posted May 21, 2008 at 10:16 AM | Permalink

    64 (MBK):

    The missing CO2 by this logic would have to have gone into the surrounding ice.

    And what keeps it from diffusing back into the bubbles? after all it has several hundred thousand years to do so.

  68. Posted May 21, 2008 at 10:27 AM | Permalink

    Conventional wisdom seems to say that CO2 does not go into the ice:

    http://www-das.uwyo.edu/~geerts/cwx/notes/chap01/icecore.html
    It appears that the air bubbles trapped in the ice represent the atmospheric composition at the time of snow deposition, in other words gas diffusion through the ice is negligible.

  69. MBK
    Posted May 21, 2008 at 10:33 AM | Permalink

    67, 68 (Leif) good question… I am completely out of my depth in terms of this methodology of course! But, it doesn’t sound too impossible to me that there should be a factor acting monotonically as to remove the CO2 from a bubble if the bubble moves, without putting it back. Say if such a thing as IGD exists in the first place it means that the bubble as a physical feature must have moved within the ice. And from the IGD – CO2 concentration correlation we infer with a high degree of plausibility that the more the bubble is different from the surrounding ice, i.e., the more it moved, the less CO2 it contains. How? I don’t know. But a connection does not seem too implausible.

  70. MBK
    Posted May 21, 2008 at 10:36 AM | Permalink

    Actually rethinking this, CO2 is well capable of sublimating for instance, I wouldn’t see why it should not also diffuse into ice. Maybe not significantly enough for the surface CO2 to influence deeper ice layers significantly in meters or kms depth, but maybe enough to alter a mm wide bubble’s CO2 concentration significantly.

  71. Phil.
    Posted May 21, 2008 at 10:42 AM | Permalink

    Re #56 et seq

    So why do these mechanisms apparently correlate with the air temperature at the time the ice was laid down? As Leif points out there’s been plenty of time for diffusion since!

  72. bender
    Posted May 21, 2008 at 11:28 AM | Permalink

    #4 mbabbitt

    Was this PDO shift predicted or taken into account by anyone in the AGW community?

    A PDO shift to positive in 1976 was not predicted because EOF analysis is a relatively new technique (~20 years old) that came after the shift to positive. A shift back to negative will not be known until 20+ years after the fact. This assumes, of course, that the PDO spatial anomaly does not redefine itself under GHG forcing. Which it may. PDO is a post-hoc algebraic invention, not a physical dynamical feature.

    Coming out now and saying that GW has taken an unexpected vacation seems a bit silly.

    Gavin Schmidt has warned that noisy departures from an overall warming trend are entirely possible. He has suggested IIRC that that a 10-year cool anomaly would be a major concern, but that anything short of that would represent nothing more than a bit of noise. If you are an even-handed skeptic, we’ll see you back here in 2-4 years to revisit your position.

  73. bender
    Posted May 21, 2008 at 11:45 AM | Permalink

    #7 MarkW

    Apparently the cold phase of the PDO can halt global warming, but the warm phase has no impact on it whatsoever.

    More twisted logic on PDO. Who says cool phase of PDO “halts” warming? Who says that GHGs do *not* influence PDO (e.g. causing it to run warmer more frequently)? Your supposed double-standard is a straw man that will not be going into the database.

    General comment. PDO seems to have become the skeptic’s new poster child. You had better read that literature if you don’t want to look foolish. And I warn you: it’s a minefield.

    Andrew, count.

  74. Phil.
    Posted May 21, 2008 at 12:01 PM | Permalink

    On going through the presentation that was sent to Leif I’ve come up with the following:

    The concentration of ‘heavy’ N2 in firn is explained by a Soret effect whereby in periods of steady state, gas with heavy isotopes is
    a)Drawn downward by gravity
    b)Drawn upward to cold surface
    These competing effects minimize differences between free atmosphere and closing bubbles at depth.

    If this mechanism is applied to CO2 then during the long steady cold periods the CO2 concentration should be accurate, i.e. ~200ppm.

    After rapid climate warming, gas with heavy isotopes is
    a)Drawn downward by gravity
    b)Drawn downward away from warm surface
    These complementary effects maximize differences between free atmosphere and closing bubbles at depth.

    If this mechanism applies to CO2 then the concentration in the bubble will be higher than the surface and younger.

    It seems to me that this explains the observations and indicates that the ‘correction’ is inappropriate.
    What do you think Leif?

  75. Richard Sharpe
    Posted May 21, 2008 at 12:47 PM | Permalink

    bender says:

    Gavin Schmidt has warned that noisy departures from an overall warming trend are entirely possible. He has suggested IIRC that that a 10-year cool anomaly would be a major concern, but that anything short of that would represent nothing more than a bit of noise. If you are an even-handed skeptic, we’ll see you back here in 2-4 years to revisit your position.

    I always find bender’s statements useful. The appropriate approach should be: We shall see.

    Of course, investigating those periods where we think there were increases in GHGs to see what their impact was might also give us extra info, although we might not be able to relate them to such things as PDO etc.

  76. bender
    Posted May 21, 2008 at 1:21 PM | Permalink

    #75
    1. But I AM a nag. And for good reason. You and many others give skeptics a bad name with your spurious arguments. You keep up with the literature and I’ll keep up with the conversation. No deal, hunh?
    2. So you refute some fringe straw man alarmist making an illogical argument, so what. Your problem is the non-alarmist consensus middle grounders, who are not so easily marginalized. Their arguments are not as trivial as you make them sound. If circulatory modes are not only unstable, but amenable to perturbation by GHGs, then you have a devil of a problem. Don’t you think?

  77. Chris Knight
    Posted May 21, 2008 at 1:39 PM | Permalink

    Iceages seem to be strongly associated with changes in the orbital eccentricity, rather than the absolute eccentricity of the orbit.

    At present, our orbit is becoming less eccentric, reducing the differential summer heating of the SH at perihelion cf. the NH, due to solar irradiation.

    I guess it would be possible to calculate the differences between hemispheric aphelion/perihelion TSI at the Earth’s surface over the course of the last 40000 years, having precession and obliquity data to hand, and see how the ice core temperature record ties in – or has this already been done?

    Presumably in a century, or so, we will be able to measure the IGD of ice laid down in the late 20th century, and test the corrections of Drake, mentioned above.

    The above plot used uncorrected CO2 from the Vostok data series, Barnola et al, Nature, 329, 408-414 (1987)

    Eccentricity data from: http://www.jgiesen.de/kepler/eccentricity1.html

  78. bender
    Posted May 21, 2008 at 1:55 PM | Permalink

    #76

    I always find bender’s statements useful. The appropriate approach should be: We shall see.

    Thank you. The problem, however, is that you have a bolus of lobbysits claiming that there is no time to wait and see. Your call for patience is not going to rouse as much response as a call for precautionary action. That’s precisely why the GCMs were invented. So that we would be empowered to act in *advance* of crisis. You can dispute their equations, but you can not dismiss them on principle. So, sorry, wait and see is not a viable option. Yes, we shall see. No, we can not wait.

  79. Richard Sharpe
    Posted May 21, 2008 at 2:17 PM | Permalink

    Bender says:

    Yes, we shall see. No, we can not wait.

    Act in haste, repent at leisure.

  80. bender
    Posted May 21, 2008 at 2:22 PM | Permalink

    #80 Yes, there is a cost to urgency. This is life.

  81. Sam Urbinto
    Posted May 21, 2008 at 5:00 PM | Permalink

    77 bender

    Noticing that some in the debate say there will be no warming because of cool modes that have in the past denied the warm modes have an impact on warming is hardly making the argument that cool modes mean no warming and warm modes mean no cooling or vice versa or some mix. It’s simply pointing out the inconsistency (the straw-manishness if you prefer) in somebody else saying there’s a causal linkage.

    “If circulatory modes are not only unstable, but amenable to perturbation by GHGs, then you have a devil of a problem.”

    Are the modes stable or unstable? Are they amenable to peturbations by GHGs or not? Although one would assume that if “GHGs” includes water vapor, the answer is probably the GHG are part of the modes themselves, and not a separate thing to preturb them or not. The answer could be anything, in the absense of “the literature” to the contrary….

    Come on bender, give us the reference you’re alluding to! Pretty please.

  82. Steve McIntyre
    Posted May 21, 2008 at 6:50 PM | Permalink

    #74. I’ve asked Jonathan Drake for an exact data citation He said:

    In the document I used the following:

    http://cdiac.esd.ornl.gov/ftp/trends/co2/vostok.icecore.co2 (Vostok, used as main example)

    http://cdiac.ornl.gov/ftp/trends/co2/lawdome.combined.dat (Law Dome). I took mean data values from all available points that corresponded to differing (ice – gas) ages, IGD. This amounted to five points.

    http://www1.ncdc.noaa.gov/pub/data/paleo/icecore/antarctica/taylor/taylor_co2-latequat.txt (Taylor)

    ftp://ftp.ncdc.noaa.gov/pub/data/paleo/icecore/antarctica/vostok/vostok_deld.txt (Temperature series)

    I replicated his first graphic from the first data set. I don’t understand the idea behind “Compensated” CO2 but I confirmed his observation of the very strong inverse correlation between CO2 levels and Gas Age Differential. In warm periods (high CO2), the differential is about 2000 years, while in cold periods (low CO2) the differential increases to 7000 years.

    It is a a surprisingly large difference. Has this been explained or even discussed in the journal literature?

  83. Phil.
    Posted May 21, 2008 at 8:51 PM | Permalink

    Re #82

    In warm periods (high CO2), the differential is about 2000 years, while in cold periods (low CO2) the differential increases to 7000 years.

    As I said in #74 it makes more sense that the temperature is the driver, using a mechanism similar to that outlined here

  84. Steve McIntyre
    Posted May 21, 2008 at 9:20 PM | Permalink

    I looked at the reference in #83 and did not see any explicit report of a changing differential between ice age and gas age. PErhaps one of the mechanisms in this PPT mcan be shown to explain it, but there’s no effort in the PPT linking the quantitative difference between 2000 years and 7000 years to this effect or any attempt to link that quantitative difference to the effect; indeed the changing differential is nowhere explicitly reported in this PPT. An explanation based on one of the effects in this PPT might very well “make sense”; I don’t know about the phenomenon to have any opinion one way or the other; but this PPT is not an explicit discussion of this effect in the peer reviewed literature and your #83 is not an answer to my question.

  85. Geoff Sherrington
    Posted May 21, 2008 at 10:21 PM | Permalink

    Re J J Drake referenced at # 56 et seq

    Because JJ Drake does not give mechanisms, we have to start thinking from first principles. The ice age-gas age gap (the IGD of Drake) has long been known, see

    Barnola, J.-M., D. Raynaud, C. Lorius, and N.I. Barkov. 2003. Historical CO2 record from the Vostok ice core. In Trends: A Compendium of Data on Global Change. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge, Tenn., U.S.A.

    “Because air bubbles do not close at the surface of the ice sheet but only near the firn-ice transition (that is, at ~90 m below the surface at Vostok), the air extracted from the ice is younger than the surrounding ice (Barnola et al. 1991). Using semiempirical models of densification applied to past Vostok climate conditions, Barnola et al. (1991) reported that the age difference between air and ice may be ~6000 years during the coldest periods instead of ~4000 years, as previously assumed.”

    Drake is challenging this explanation. Barnola et al treat IGD as a constant of uncertain magnitude, linked to temperature. Drake plots it as a variable with a reasonable linear fit, not stated as related to temperature, but related to CO2 in ice core.

    Drake makes the assumption that the age of the ice should be the same as the age of the gas trapped in it. This seems reasonable when one asks how the (air + CO2) of today reaches down 90 m to be frozen in old ice at the firn-ice transition at Vostok. This is hardly a quantitative process by the standards of analytical chemistry.

    Historical reconstruction from isotopes involves some empiricism. For example, it is often stated that warmer water, on evaporation, is depleted in O-18 and that when transported to a cold place and deposited as ice, there is a corresponding heavy oxygen isotope depletion. The same argument is given for the hydrogen isotope deuterium. I have long regarded this as a qualitative process, yet it is used to derive quite precise quantitative relationships now frequently quoted. The calibration of these relationships involves an element of circular logic. Part of the logic requires an explanation such as the firn-ice interface.

    For example, the derivation of temperature from oxygen isotopes is no longer the same if the Drake explanation is accepted, see his second figure showing the deuterium temperature from Jouzel et al. This is because his recalculated CO2 levels disrupt the circular logic. He is essentially inferring that the seas have been at more constant temperature than is now assumed, for 400,000 years. Or that the qualitative explanation for fractionation needs re-examination. I would not jump to the conclusion that his IGD varies from cold eras to warm eras. He simply argues that CO2 hardly varies.

    I shall be looking for publications where carbon isotopes can be used to help interpret what happens as ice forms over 40,000 years and what the IGD of such shallow ice is. Then I shall ask why Vostok is different. It would also be interesting to view the IGD of the top 90m of the Vostok holes.

  86. Geoff Sherrington
    Posted May 21, 2008 at 10:39 PM | Permalink

    #85 continued after more work

    Fow Law Dome, Antarctic, near Casey Station:
    http://www-das.uwyo.edu/~geerts/cwx/notes/chap01/icecore.html

    and confirmed by drilling at Taylor Dome

    It appears that the air bubbles trapped in the ice represent the atmospheric composition at the time of snow deposition, in other words gas diffusion through the ice is negligible. For instance, the CO2 concentration in air bubbles dated to be from 1958 or later agrees very well with direct free-atmospheric CO2 measurements, which have been made since then.

    (My bold) Why does Vostok need the firn-ice transition explanation?

  87. Phil.
    Posted May 21, 2008 at 11:13 PM | Permalink

    Re #84

    As I outlined in #74 the ppt explains how N15 is concentrated as a result of the Soret effect in conditions of rapid warming (showing spikes of N15 during periods of rapid warming), a similar effect for CO2 would increase the concentration of CO2 at such times and decrease its age relative to the surrounding ice. Drake’s concept of correcting for an unknown mechanism in such a way that renders the concentration of CO2 ~constant and thereby creating an unknown sink for CO2 requires some theoretical justification.

  88. Geoff Sherrington
    Posted May 22, 2008 at 12:35 AM | Permalink

    Re # 87 Phil

    Might you please expand upon your “unknown sink”

    thereby creating an unknown sink for CO2 requires some theoretical justification.

    Thanks Geoff.

  89. Posted May 22, 2008 at 12:41 AM | Permalink

    The main question is why would there be a correlation between ice age – gas age difference and CO2 levels at all? The difference in ice age – gas age is mainly a question of snow accumulation. The fastest accumulating ice cores have the smallest difference. This is a pure physical process. If that influences the CO2 content (or CO2/O2/N2 ratio’s) during the ice formation is a quite different question.

    The Law Dome ice cores have the most detailed atmosphere to firn to ice air composition data available. See Etheridge e.a.

    It is theoretically possible that wind shear gives some changes in the composition of the air in the snow vs. the atmospheric composition. From a physical point of view, this should give a lower CO2 value just below the surface, but no gradient further down, except if there was a gradient in wind speed over the years (which need to be highly correlated to temperature changes, in the case of Law Dome very unlikely for the last 100 to 1000 years). There is no physical process that can explain the reduction in CO2 levels downward, if there was a fractionation at the surface due to (relative) constant wind shear. To the contrary, the higher levels in the firn just below the surface should (and do) migrate downward, but as the sealing/closing compete with the migration speed, one sees a smoothed CO2 gradient between sub-surface and deeper layers of firn. In the Law Dome firn record, there is no difference between atmospheric and sub-surface CO2 levels, which means that there is little to no fractionation between atmosphere and near surface firn air. The same is true for firn air vs. fully closed ice air bubbles at the same depth. Thus there is no fractionation due to the closing process itself either.

    Further, Law Dome had three drilling cores, two at very high accumulation sites (1.2 m ice equivalent/year) and one at a somewhat slower accumulation site (0.6 m/year). The latter was used for the longer-term (1,000 years) measurements. Despite the difference in accumulation rate and the ice age – gas age difference (10 years vs. 40 years), the CO2 levels of the same gas age are similar.

    Last but not least, all ice cores show similar CO2 variations for the same gas age in overlapping periods, despite the very huge differences in accumulation rate and ice age – gas age difference, from the coastal cores towards the near South Pole inland cores.

    All together, it seems to me a classical example of A (temperature) driving B (ice age – gas age difference, due to accumulation rate changes) and C (CO2 levels) independent of each other, where the “correction” of C using B has no physical meaning at all…

  90. Phil.
    Posted May 22, 2008 at 1:20 AM | Permalink

    Re #89

    All together, it seems to me a classical example of A (temperature) driving B (ice age – gas age difference, due to accumulation rate changes) and C (CO2 levels) independent of each other, where the “correction” of C using B has no physical meaning at all…

    It certainly seems that way to me.

    Re #88

    Might you please expand upon your “unknown sink”

    Just that if you take Drake’s ideas at face value he comes up with a constant CO2 of about 325ppm which implies there has to be a mechanism to substantially reduce the CO2 concentration in the bubbles compared with the atmosphere during low temperature periods i.e. the ‘unknown sink’.
    This is something Drake appears to ignore.

  91. Geoff Sherrington
    Posted May 22, 2008 at 3:36 AM | Permalink

    Re # 90 Phil

    I have a neutral position while delving into this paper By JJ Drake.

    Might you please go to his figure comparing “corrected” CO2 with that of Barnola. Both show an approximately level CO2 concentration with excursions, one higher than the other. So what is the basic difference between the two plots, Drake averaging say 325 ppm and Barnola averaging say 225 ppm?

    Sure, the higher one is constructed and the lower one is measured. But you are accepting that the measurement is correct, when Drake is giving you evidence that it might not be.

    I do not easily buy the business of the 90m of ?? at the top of the Vostok core. Why should the Vostok core not preserve a permanent record until frozen at 90 m depth, when at Law Dome and Taylor Dome this effect is 10-40 years? Note that at Law Dome the pre-industrial level suggested by Etheridge at al is some 280 ppm CO2 in atmosphere, closer to Drake than to Barnola.

    Re Ferdinand Engelbeen # 89

    The main question is why would there be a correlation between ice age – gas age difference and CO2 levels at all?

    The main answer is that it is shown to exist by the Drake data. The most obvious explanation is that the measured CO2 ppm are affected by the burial process. Next most obvious is that they are affected in the recovery process before lab analysis.

    What is your explanation for an ice age-gas age difference of thousands of years at Vostok, and none at shallow depth at Law and Taylor Domes, where recent rises in atmospheric CO2 allow a better range of examination of both CO2 and time scales?

  92. MarkW
    Posted May 22, 2008 at 5:29 AM | Permalink

    How is my argument untenable, I’m just using their own words to demonstrate the indefensibility of their position.

    They are the ones claiming that PDO has stopped global warming until 2015.
    They are the ones who previously claimed that PDO had no affect on global temperature.

    Why is pointing this out a bad thing?

    I know that you have your knickers in a knot about PDO, that’s your problem.

  93. Sam Urbinto
    Posted May 22, 2008 at 11:43 AM | Permalink

    As far as ice bubbles etc, there is a close match between Law Dome and Mauna Loa in the period when they overlap, they don’t diverge very much. However, nothing says that you can compare levels at one of the poles in ice with air in another hemisphere above the water. But there is that match, so it’s likely you can. I’d say.

    MarkW, I’m still not sure why the idea that some long-term weather pattern that for some reason changes direction is such a point of contention with anyone. Or coming to the conclusion (or not) that that pattern changing is a likely or the the most likely answer (or not) is evil wrong bad skeptical. Or why “using their own words to demonstrate the indefensibility of their position” means holding that same position. I’m ambivalent neutral agnostic on this all myself, but it seems bringing up a point to discuss it or as a thought equates to slavishly believing in that point.

    I’d still be curious to see what literature claims the rather vague “GHG cause PDO shifts” or whatever it’s supposed to be claiming. I don’t see the point here.

  94. Posted May 22, 2008 at 12:43 PM | Permalink

    Re #91 Geoff,

    It is a matter of accumulation speed. Vostok is far inland Antarctica at 3,500 m height and has very little snowfall over a year. The amount of snow is directly correlated with (semi) global humidity and this again is directly correlated with temperature. From AGU (relative old story):

    Because the accumulation rate is governed by saturation water vapor pressure, past accumulation may be estimated from the temperature record. Accumulation rates inferred in this way are supported by measurements of beryllium 10 (10Be), an isotope produced by the interaction of cosmic rays and the upper atmosphere, can be used to determine past snow accumulation in Vostok ice. Deposition of this cosmogenic isotope is assumed to be constant.

    Not completely correct, as 10Be production is variable with solar variations…

    Thus during ice ages, there is less snowfall and during interglacials (like the current) there is more snowfall.

    What does that mean for the ice age – gas age difference?

    Ice is formed from snow that was deposited at a certain moment. If one can count the layers (which is the case for high-accumulation cores), one can know the exact year of deposit of some snow/firn/ice. For Vostok that is more difficult, as the layers are very thin at the top and absent at depth, and one need models and other proxies to estimate the age of the ice. See e.g. here and here.

    While the ice layers build up, at a certain mass, the snow is compressed into firn and gradually air trapped in the snow crystals is completely enclosed in ice. This happens somewhere between 70-90 m snow/firn depth. As long as there are open pores, there is an exchange of air between the surface and the firn possible. But the speed with which this happens diminishes with depth, as the pores are more and more closing. Thus the exchange of air between surface and depth is highly dependent of the speed with which the 70-90 m snow/firn/ice is formed. For the fast Law Dome cores (1,2 m ice equivalent/year) the exchange goes on for the upper 20 m and ultimately the difference in age between the age of the snow deposit (turned into ice) and the average air composition is about 10 years. For the slower third core of Law Dome (0.6 m ice equivalent/year), air exchange lasts longer, as there is more time before the bubbles are sufficiently sealed to have a slower exchange than the ice formation. This leads to a difference of 40 years between ice age and average gas age (which is a smoothed average of air composition over decades).

    Vostok, with its very low accumulation, needs hundreds to thousands of years to reach the 70-90 m firn/ice layer thickness, which is influenced by temperature/humidity during the ice ages and interglacials…

    Snow reflects the composition of the water at the moment of deposition, which reflects the composition of the (ocean) water at the origin of water vapour which formed snow. There is some fractionation in isotopes during evaporation and (partly) deposit during transport to the place of ultimate deposit, but this is accounted for (with some discussion…). The heavy isotopes of oxygen (18O) and hydrogen (D – deuterium) increase in ratio with increasing (ocean) temperature at the moment of evaporation. This can be used to reconstruct the (ocean) temperature at the time of evaporation, which is practically the same as the time of deposit.

    Thus we know with reasonable certainty that the ice age – gas age difference is a matter of snow deposit, which varies with humidity, which is caused by temperature changes. At the other side, temperature changes influence CO2 levels: higher temperatures give higher CO2 levels. Thus both IGD and CO2 changes are caused by temperature changes and both show a high correlation with temperature, therefore we see a remarkable correlation between IGD and CO2 levels, but there is no physical connection between IGD and CO2 levels at all…

    Hope this made things a little more clear…

  95. Posted May 22, 2008 at 12:56 PM | Permalink

    Re #93 Sam,

    The Law Dome ice core air bubbles have an exact match with South Pole atmospheric CO2 data during the overlap period of 20 years (1959-1978) within the accuracy of the measurements (+/- 1.2 ppmv – 1 sigma). While the Mauna Loa data (as good as all NH CO2 data) show far larger seasonal variations and the SH data show a one year lag vs. the NH data, all (ten) CO2 base stations show near exact the same yearly averages and same trends in the past (up to five) decades…

  96. Sam Urbinto
    Posted May 22, 2008 at 2:07 PM | Permalink

    #95 Ferdinand Thanks. So you’re saying these are adjusted to match time-wise after the fact? Okay.

    This is the data I have for Law Dome v Mauna Loa. Yes, the trend 1959-1978 is off around 1.5 ppmv or so (which I’d previously done and why I said they matched!). 🙂

    315.6 316.00
    316.3 316.91
    317.0 317.63
    317.7 318.46
    318.4 319.02
    319.2 319.52
    320.0 320.09
    320.8 321.34
    321.8 322.13
    322.8 323.11
    323.8 324.60
    324.8 325.65
    325.8 326.32
    326.9 327.52
    328.0 329.61
    329.2 330.29
    330.3 331.16
    331.5 332.18
    332.6 333.88
    333.7 335.52

    That data is 2004ish though, and depending on what set, the mauna loa is different slightly. (For example, I have for 2004 377.43 and their website has 377.55)

    I’d say that it certainly looks like it matches too. 🙂

    But the interesting thing is Mauna Loa versus global marine in the present:

    1998 366.50 365.48
    1999 368.14 367.57
    2000 369.41 368.77
    2001 371.07 370.36
    2002 373.16 372.39
    2003 375.80 374.94
    2004 377.55 376.76
    2005 379.75 378.78
    2006 381.85 380.90
    2007 383.72 382.65

    As far as growth per year.

    1980-2007 for Mauna Loa is a mean of 1.68 ppmv per year growth with a median of 1.46 and .59 σ The range is .49 to 3 per year. Last year was 2.14 The processesing is:

    The estimated uncertainty in the annual mean is the standard deviation of the differences of annual mean values determined independently by NOAA/ESRL and the Scripps Institution of Oceanography.

    1980-2007 for marine global is a mean of 1.65 ppmv per year growth with a median of 1.65 and .58 σ The range is .68 to 2.94 per year. Last year was 2.23 The processing is:

    The uncertainty in the global annual mean is estimated using a monte carlo technique that computes 100 global annual averages, each time using a slightly different set of measurement records from the NOAA ESRL cooperative air sampling network. The reported uncertainty is the mean of the standard deviations for each annual average using this technique. Please see Conway et al., 1994, JGR, vol. 99, no. D11. for a complete discussion. Units are ppm.

  97. Posted May 22, 2008 at 3:42 PM | Permalink

    Re #96 Sam,

    I suppose that they followed the gas data for what they found top down in the firn and matched that for the start or end of the overlap with the atmospheric data. See the graph from Etheridge e.a. (1996) for firn/ice CO2 levels with depth:

    This matches quite good with the slope of the CO2 levels in the atmosphere at the South Pole:

    The South Pole data match a little better than the Mauna Loa data, as the lag of CO2 between the NH and the SH slowly grows, which gives a slight difference in slope (this points to a NH source of the extra CO2, where 90% of the emissions are). There is little difference in the slope and yearly averages of the CO2 measurements from near the North Pole (Barrow) to the South Pole, except for the NH-SH lag:

  98. Sam Urbinto
    Posted May 22, 2008 at 5:14 PM | Permalink

    Ferdinand #97 Very nice. The variations in the last graph look around 1% so I’d just say they’re all the same. Trends look like around 10% over the last 25 years.

    So what does a 4% per decade rise in carbon dioxide level as a trend do when it’s (on its own) about ~50% of the radiative forcing part of the atmosphere, ignoring all but radiative forcing? Not much is what I’d say taking everything into consideration.

    So is there some consensus on a correlation between carbon dioxide and ice ages, in what direction, over what time period, or just a bunch of conflicting guesses?

  99. Phil.
    Posted May 22, 2008 at 5:48 PM | Permalink

    Re #91 Geoff

    Might you please go to his figure comparing “corrected” CO2 with that of Barnola. Both show an approximately level CO2 concentration with excursions, one higher than the other. So what is the basic difference between the two plots, Drake averaging say 325 ppm and Barnola averaging say 225 ppm?

    Sure, the higher one is constructed and the lower one is measured. But you are accepting that the measurement is correct, when Drake is giving you evidence that it might not be.

    I wouldn’t say that they both show a ‘level CO2 profile with excursions’
    The original data show a strong correlation with temperature whereas Drake’s ‘correction’ has removed this.
    Drake found a correlation between [CO2] and IGD and decided to correct for to 0 IGD without any mechanism to justify this approach.
    Since the IGD appears to correlate as well with Temperature one could equally correct for constant surface temperature, I don’t think anyone would think that to be justified?
    If we use reverse logic and accept that Drake’s procedure is correct and that there really is a constant [CO2] then what is the explanation for the rise in the [CO2] measurement during periods of high temperature. Drake provides no theoretical background for this.
    I think Ferdinand’s summary makes sense and certainly agrees with my interpretation of the data:
    “Thus we know with reasonable certainty that the ice age – gas age difference is a matter of snow deposit, which varies with humidity, which is caused by temperature changes. At the other side, temperature changes influence CO2 levels: higher temperatures give higher CO2 levels. Thus both IGD and CO2 changes are caused by temperature changes and both show a high correlation with temperature, therefore we see a remarkable correlation between IGD and CO2 levels, but there is no physical connection between IGD and CO2 levels at all…”

  100. Posted May 23, 2008 at 12:49 AM | Permalink

    Re #98 Sam,

    There is a reasonable constant ratio of about 8 ppmv/°C over the ice ages. There is a lag of CO2 after temperature of 800 +/- 600 years during the warming up towards an interglacial and of several thousands of years during a cooling towards a new ice age. For Vostok (420 kyr) here is the graph:

    The lag proves that temperature changes cause CO2 changes, but that doesn’t exclude that CO2 may have a feedback on temperature. Current GCM’s imply that 50% of the ice age – interglacial temperature transition is caused by GHGs, as feedback. In reality there is little proof for that. The end of the Eemian (see here neither the LGM-Holocene transition (see here show much feedback from decreasing/increasing CO2 levels…

    The 8 ppmv/°C ratio is quite constant over (very) long-term transitions, which include vegetation and icefield border changes, ocean current changes,… For recent short term temperature variations, the ratio is about 3 ppmv/°C, superposed on the about 1.7 ppmv/year increase we see as result of human emissions. This is mainly a result of fast response in partial pressure of CO2 in the ocean’s surface to temperature changes.

  101. Posted May 23, 2008 at 12:51 AM | Permalink

    The correlation graph didn’t show up, here a second try:

  102. Posted May 23, 2008 at 1:00 AM | Permalink

    Too early in the morning here. I did send the wrong graph, which was made by André van den Berg from data of the Dome C ice core in Antarctica. It shows very little, if any feedback from CO2 on temperature in the last warming up to the current Holocene. The correlation graph is here (if I paste the right URL…):

  103. Geoff Sherrington
    Posted May 23, 2008 at 5:24 AM | Permalink

    Re # 99 Phil & 3 100 Ferdinand Englebeen

    Why is it so vital in your explanations to invoke temperature? To my knowledge, temperature is an estimate derived from mechanisms that could plausibly be among those Drake questions (though he states that he accepts the isotopic analysis and does not propose mechanisms). So we are on wavelength – what is the method of calibration of temperature estimates in the pre-thermometer era, say for the last 400,000 years?

    Let’s leave temperature out of the loop for a moment and deal with water, air and carbon dioxide. As I understand, it is possible to estimate the concentration of CO2 in air bubbles, but not possible to isolate the CO2 and to determine the deltaO-18 of the CO2. The deltaO-18 of the air can be measured, as it can be from the ice surrounding the gas bubbles.

    By what mechanism should the deltaO-18 of the ice differ significantly from that of the air in enclosed bubbles? What is the measured difference?

    The snow-firn-ice transition proposed at Vostok is intuitively implausible. For it to work, it more or less requires that the composition of all air in the top 90m or so is the same as the present atmosphere, plus or minus a few years of change. So what has happened to the air that was in this (supposedly porous) layer 1,000 years ago, 2,000 years ago? Has it been flushed out by new air? Do no trapped bubbles survive this top 90m of burial?

    What is wrong with an ab initio assumption that the pre-industrial CO2 concentration in air was essentially constant for 400,000 years as Drake depicts? That would be an intuitive starting assumption. It’s only when you make a cause-effect with temperature that you have to start looking for CO2 variation. Sure, you find variation in measurements, but are these artefacts or climate effect?

    Temperature complicates matters. We all know that there has been some debate about whether T lags or leads CO2 events in the Vostok core. What do you say to this abstract?

    Ice-core measurements of carbon dioxide and the deuterium palaeothermometer reveal significant covariation of temperature and atmospheric CO2 concentrations throughout the climate cycles of the past ice ages. This covariation provides compelling evidence that CO2 is an important forcing factor for climate. But this interpretation is challenged by some substantial mismatches of the CO2 and deuterium records, especially during the onset of the last glaciation, about 120 kyr ago. Here we incorporate measurements of deuterium excess from Vostok in the temperature reconstruction and show that much of the mismatch is an artefact caused by variations of climate in the water vapour source regions. Using a model that corrects for this effect, we derive a new estimate for the covariation of CO2 and temperature, of r2 = 0.89 for the past 150 kyr and r2 = 0.84 for the period 350–150 kyr ago. Given the complexity of the biogeochemical systems involved, this close relationship strongly supports the importance of carbon dioxide as a forcing factor of climate.

    Kurt M. Cuffe & Françoise Vimeux 2 August 2001
    http://www.nature.com/nature/journal/v412/n6846/full/412523a0.html;jsessionid=14959F0C2C521F4C393466AC501213A5

    Is it valid to generally assume that a deuterium palaeothermometer exists for widespread use? In a closed system, if there is enrichment of deuterium in one place (let’s call it the source area), there has to be depletion in another. Since we might not know the source area, or its properties, are we not assuming too much too glibly? I do not know if the ice at Vostok was formed from material taken from elsewhere on the Antarctic ice plateau, by erosion or sublimation, or from oceans beyond, or which oceans. I do not know the source(s) of the CO2 in the air bubbles.

    There seems to be too much circular logic going on. One even feels that wishful thinking enters the interpretation at times. What is the most firm peg to which the circle can be pinned?

  104. John Drake
    Posted May 23, 2008 at 7:53 AM | Permalink

    re #56 et seq
    I assure you that the JJ Drake in question is NOT the one who from McMaster University, who is me!

    In general, papers without proper attribution and the author’s contact information should be regarded very cautiously. Perhaps those who have been in contact with Jonathan Drake, another supposed author, could provide these details?

    John Drake
    Professor, Geography & Earth Sciences
    McMaster University
    Hamilton, ON Canada

    Steve:
    Nice to have a visitor from McMaster. One of my nieces recently graduated from McMaster.

  105. Posted May 23, 2008 at 11:40 AM | Permalink

    Re #102 Geoff,

    There is no vital reason to invoke temperature, it is only quite realistic: there is a reasonable correlation between temperature and CO2 levels and there is a reasonable correlation between temperature and snow layer thickness, where it is proven that temperature is the driving force, at least in current times. The opposite, that layer thickness or CO2 levels are driving temperature is more discutable… And that layer thickness is the common determinator of CO2 levels and temperature needs some more detailed physical explanation, before one can accept that.

    As far as I know, there is no/little connection between d18O of ice and air in the ice cores. d18O in ice is from water and its level is influenced by evaporation and deposit of water/water vapour/rain and snow/ice. d18O in air bubbles is from surrounding air and is influenced by plant growth in general (which prefers the lighter isotopes, as good for carbon as for oxygen). The d18O is mainly measured in N2O (which has the advantage of containing only one oxygen molecule) and the value is inversely correlated with ice sheet formation. as far as I understand the mechanism is more ice sheets – less vegetation – more light isotopes and v.v. Something similar can be seen in the d13C levels of CO2 in the ice core gas bubbles.

  106. Posted May 23, 2008 at 1:12 PM | Permalink

    Re #102 Geoff (continued),

    Sorry, hit the submit button by accident…

    The reduction of both heavy isotopes in the atmosphere shows that ice sheets growed at the cost of vegetation and thus that – in general – global temperatures decreased.

    Other methods were used too: the temperature of the bore hole was measured and more or less conserved the (local) temperature in the period of deposit. This confirmed the temperature excursions, and could be used to calibrate the isotope changes. See Jouzel e.a.: http://www.ipsl.jussieu.fr/GLACIO/hoffmann/Texts/jouzelJGR2003.pdf

    I don’t see any reason why the Vostok snow-firn-ice transition should be different than the Law Dome transition, except that it takes a much longer period for the bubbles to fully closure. That means that short time (less than 600 years) relative small variations in CO2 levels would be smoothed out. But an increase or decrease to a different level that is sustained for thousands of years remains measurable. The current upper 20 m at Vostok indeed may show near the same CO2 levels as the current atmosphere, with some gradient (but I have no figures for that). Below 20 m (in the Law Dome ice core) there is little exchange of air anymore. That only means that enclosed bubbles reflect more or less the current atmosphere, while the ice is already 600 years old. Nothing unusual, see the first graph of Law Dome in #97.

    What is wrong with the assumption that CO2 levels were near constant over the ice ages is that we see today that the levels are not constant and covary with temperature: over the seasons and year-by-year. Temperature has a fast (3 ppmv/°C) effect on CO2 levels, as well as over the seasons as for year-by-year temperature changes. Thus why should that be different for longer term? Again, there is no known physical link between IGD and CO2 levels, but both have a physical link with temperature. Moreover, ice cores with completely different accumulation rates (and thus IGD) show the same CO2 excursions, something impossible if IGD governed the CO2 levels. See: http://www.sciencemag.org/cgi/reprint/283/5408/1712.pdf

    The Cuffe and Vimeux abstract is quite misleading: at all time spans in the Vostok ice core, CO2 lags temperature changes. That is the only point that can’t be denied. There is a nice correlation between temperature and CO2 levels, that is proven for one direction, with temperature as leading factor, but unproven the other way out. The fact that a corrected temperature reconstruction increased the correlation doesn’t prove that CO2 has any effect on temperature (neither disproves that), the more that at the end of the Eemian, CO2 levels only started to drop after temperature was already at minimum and had no measurable effect on temperature. The temperature correction doesn’t change that. See the graph of the end of the Eemian (with corrected and uncorrected temperature trends) here. Thus there is no evidence for CO2 as important forcing factor in the ice cores…

    About the ice cores CO2 and water: this is mainly based on what currently is observed. For CO2, it is quite clear: over 95% of the earth’s atmosphere has the same yearly averages and trends (within 5 ppmv), with only the lag between NH and SH as main difference. For water, that depends of the ice core distance to the oceans and the bore hole surface heighth. Near the coast, most precipitation comes from the nearby oceans. More inland, at higher altitudes, the precipitation comes more generally from the wider SH oceans. If evaporation happens at higher temperatures, the ratio of heavy atoms increases in the vapour. This changes the remaining isotopic composition of the oceans. Jouzel and others used corrections for these changes… See Jouzel e.a. in the forementioned article.

    For CO2, the Law Dome ice core is interesting enough, as Etheridge e.a. described in detail all the measures that were taken to prevent contamination, cracks and migration and other possible pitfalls in ice core measurements. Moreover, he compared different drilling methods and compared gas composition of fully closed and not closed bubbles in the firn and he compared ice core / firn air composition with atmospheric measurements, including an overlap period. All data together confirm that the gas in the bubbles is really a smoothed average of the composition of the atmosphere at the time of closing, where the smoothing time period depends of snow accumulation speed (thus bubble closing time)…

  107. Sam Urbinto
    Posted May 23, 2008 at 6:24 PM | Permalink

    Ah, Ferdinand #100, I am aware of most of that (in ice cores temperature leads carbon dioxde by ~700 years, recent years have about a 2 ppmvish per year rise) (and we only have 200 or so years of air measurement after all), but it’s not the data over it, it’s the consensus over it! It’s obvious that all the atmosphere and hydrosphere have to do is change the water vapor content or lapse rate to adjust for more or less of the so-called long-lived IR absorbing/emitting gases.

    Saying carbon dioxide causes temperature is like saying the anomaly causes population.

    🙂

  108. Phil.
    Posted May 23, 2008 at 10:12 PM | Permalink

    Re #103

    It isn’t ‘vital’ to invoke temperature except that it does change so you can’t ignore it! I basically agree with Ferdinand.

    What is wrong with an ab initio assumption that the pre-industrial CO2 concentration in air was essentially constant for 400,000 years as Drake depicts? That would be an intuitive starting assumption. It’s only when you make a cause-effect with temperature that you have to start looking for CO2 variation. Sure, you find variation in measurements, but are these artefacts or climate effect?

    Nothing wrong with that initial assumption, although that’s not Drake’s assumption it’s his result.
    But as I’ve said before such a high constant value isn’t feasible.
    Look at the result in the powerpoint for [N15] which surely is constant in the atmosphere, in the bubbles it is constant except for periods of rapid temperature change when it spikes. The expectation would be the CO2 would do the same which is incompatible with Drake’s result, you’d also expect [C14O2] to increase by the same mechanism which would make it appear ‘younger’.

  109. Posted May 23, 2008 at 10:54 PM | Permalink

    104 (Drake):
    This is the information I have from an email to me:
    “Leif,
    I am a Physicist in the UK and have been researching some of the aspects regarding AGW. I have seen many of your posts on forums and thought you would be interested in a paper that I have just finished.
    It is titled “A Simple Method to Correct Carbon Dioxide Concentrations in Ice Core Data for Ice / Gas Age Difference Perturbations.”. However, it ultimately suggests that CO2 ice-core data used as evidence of AGW is compromised.
    Jonathan Drake

  110. Posted May 23, 2008 at 10:58 PM | Permalink

    109 (Leif): and the email address was: jdrake@ntlworld.com

  111. Geoff Sherrington
    Posted May 24, 2008 at 2:58 AM | Permalink

    Re # 104 John Drake,
    Please accept my apologies for confusing you with Jonathan Drake of the UK. I did insert a caveat (“if they are indeed the same person”) hoping that I would be corrected if you were not. Some of your citations were in a similar research area so it was not a hard mistake to make, given the information from which I was working. As host Steve McIntyre says to you, I also say welcome to CA. Regards Geoff.

  112. Richard Sharpe
    Posted May 24, 2008 at 9:55 AM | Permalink

    I note that in relation to an earlier claim by Phil that David Smith (I think) had ignored inflows from the Bering Strait, the average depth of the Bering Strait seems to be 30-50m and it is some 92km wide (it seems), while the average depth of the Greenland sea is around 1450m and is several hundred km while the Barents sea is something like 400m deep and also a lot wider than the Bering Strait.

  113. Phil.
    Posted May 24, 2008 at 7:22 PM | Permalink

    Re #112

    And then as now I gave the supporting evidence for it, e.g. here
    In particular slides 7, 8, 9 & 10.

    “Model and observational uncertainties of importance to global climate
    1. Northward heat transport from the N. Atlantic/Pacific to Arctic Ocean
    2. Arctic sea ice thickness and volume
    3. Freshwater export from the Arctic to North Atlantic ”

    “• Pacific Water enters via narrow (~60mi) Bering Strait
    • outflow through Fram Strait prevents Atlantic Water inflow
    • Atlantic Water entering through Barents Sea losses ~98% of heat to atmosphere”

    “ATTRIBUTION OF CHANGE:
    Pacific Water forcing of sea ice in the western Arctic Ocean”

    ” Up to 60% of recent decrease of sea ice in the Western Arctic can be due to oceanic forcing:
    •northward inflow of Pacific Water – increased inflow of warmer water
    •Less ice allows more solar absorption, which leads to warmer ocean, which in turn will melt more sea ice (the so called ice­ ­albedo feedback)”

  114. Richard Sharpe
    Posted May 24, 2008 at 7:53 PM | Permalink

    Same old misinformation, eh Phil, since there is more recent information about that:

    NASA Examines Arctic Sea Ice Changes Leading to Record Low in 2007

    Nghiem said the rapid decline in winter perennial ice the past two years was caused by unusual winds. “Unusual atmospheric conditions set up wind patterns that compressed the sea ice, loaded it into the Transpolar Drift Stream and then sped its flow out of the Arctic,” he said. When that sea ice reached lower latitudes, it rapidly melted in the warmer waters.

    You seem unable to find more recent and relevant information about these things.

    How about you tell us who you are so we can check whether or not you really have more experience than Steve M with journal submissions, eh?

  115. Phil.
    Posted May 24, 2008 at 9:03 PM | Permalink

    Re #114

    Hardly misinformation, unless you think that the Navy has no idea about what’s happening in the Arctic ocean?
    Rather ironic since I regularly post Quikscat images here.
    Perhaps you’d like to check in which direction the Transpolar Drift flows!

  116. Geoff Sherrington
    Posted May 24, 2008 at 9:30 PM | Permalink

    Re # 108 Phil The Drake article (continued)

    By now I have corresponded with the author Jonathan Drake who sent me material including the following abstract:

    CO2 in Natural Ice
    Stauffer, B | Berner, W
    Symposium on the Physics and Chemistry of Ice; Proceedings of the Third International Symposium, Cambridge (England) September 12-16, 1977. Journal of Glaciology, Vol. 21, No. 85, p 291-300, 1978. 3 fig, 5 tab, 18 ref.

    Natural ice contains approximately 100 ppm (by weight) of enclosed air. This air is mainly located in bubbles. Carbon dioxide is an exception. The fraction of CO2 present in bubbles was estimated to be only about 20%. The remaining part is dissolved in the ice. Measurements of the CO2 content of ice samples from temperate and cold glacier ice as well as of freshly fallen snow and of a laboratory-grown single crystal were presented. It is probable that a local equilibrium is reached between the CO2 dissolved in the ice and the CO2 of the surroundings and of the air bubbles. The CO2 content of ancient air is directly preserved neither in the total CO2 concentration nor in the CO2 concentration in the bubbles. Possibly the CO2 content of ancient air may at least be estimated if the solubility and the diffusion constant of CO2 in ice are known as a function of temperature.

    (My bold). Has this been disproven or overtaken by more recent research? Do you know if residual ice is analysed for carbon after the laboratory extraction of CO2 from ice bubbles?

    You have to be careful comparing nitrogen to carbon isotopes in natural processes. In the present discussion, C as CO2 is quite soluble in ice and water, whereas N as N2 is not nearly so. Of course, nitrogen oxides are more soluble. When you mention N-15 as an indicator of CO2 behaviour, can you reference the chemical form of the N-15 you are discussing?

  117. Posted May 25, 2008 at 3:26 AM | Permalink

    Re %116 Geoff,

    The cited article is from… 1977. Since then, more is known about what happens with CO2 in ice. Part of it can be dissolved in liquid water, which is present in all ice, even at -40°C (the average Vostok ice core temperature) together with impurities. The latter is what causes the liquid phase. At the other side, more impurities (salts) reduce the solubility of CO2 in water. Vostok has very low temperatures and only a small amount of salts, thus little water and little CO2 in the liquids.

    I remember an article (but didn’t recover it), where crushing ice at low temperature was compared with total melting/high vacuum/refreezing for CO2 measurements. The two methods differed with a few % in CO2 levels.

    Current techniques have little problems with CO2 eventually dissolved in liquid water, see the Dome C core measurements for the LGM-Holocene transition:

    For CO2 measurements, each sample (of size 2.5 cm by 2.5 cm by 1.5 cm) is cracked in an evacuated and cooled (-30°C) needle cracker. The air from opened bubbles expands into a laser absorption cell. The CO2 concentration of the gas in the absorption cell is measured by tuning an infrared laser six times over the absorption line of a vibration-rotation transition of the CO2 molecule. Reference gas from the Scripps Institution of Oceanography at 251.7 ppmv is used for the calibration of each measurement. Control measurements are routinely (after about five measurements) performed by using other reference gases from the Scripps Institution of Oceanography at 321.06 and 342.03 ppmv to check the linearity of the system. The linearity of the measurement technique for CO2 concentrations lower than the calibration reference gas can be checked indirectly by generating control measurements at lower pressure, generating similar absorption line depths than for measurements with lower CO2 concentrations. Measurements on bubble-free single-crystal ice samples, to which reference gas is added and which are cracked afterward, yield a reproducibility of ±1.5 ppmv.

    The latter shows that little CO2 (if any) is dissolved in ice itself.

    About the whole discussion: while samples for CO2 measurements needs to be carefully threated to prevent losses and artefacts, the same samples are used for CH4 measurements, which causes far less problems for clathrate formation and/or fractionation and/or solubility in water veins. But these show roughly the same (even better) correlation with temperature as CO2.

    Thus all together, it is quite realistic that temperature has a direct influence on CO2 and CH4 levels, something that can be seen up to current times. The alternative hypothesis, that IGD (thus accumulation rates) has a direct influence on CO2 (and CH4) levels has no physical base at all…

  118. Geoff Sherrington
    Posted May 25, 2008 at 5:21 AM | Permalink

    Re # 117 Ferdinand Engelbeen

    I am not promoting the counter argument that IGD has a direct influence on CO2 in a primary sense. My neutral stance has been stated. It is more that a correlation coefficient of 0.9 or so (here, between ice bubble CO2 concentration and IGD) is rare in Earth Science systems and on examination commonly yields information. I am trying to find what that information is. The explanation of accumulation rates goes towards an explanation but I think might not be the whole story.

    The last part of your quotation is worded quite loosely and means little to me as an analyst. The test is susceptible to specific events being present or absent. I shall attempt to read the whole paper. It leaves me wondering why Stauffer and Berner, quoted above, were so wrong, if they were.

    Measurements on bubble-free single-crystal ice samples, to which reference gas is added and which are cracked afterward, yield a reproducibility of ±1.5 ppmv.

    Of course it is relatively easy to recover added gas from a pure single ice crystal free of bubbles. But Stauffer and Berner were working with natural materials as found. The date of publication should not affect the observations. Besides, it is accuracy and not reproducibility which is under test here.

    There is still no way for me to accept the classic descriptions for the cause of the IGD. They are so tied in with temperature, and temperature is in turn tied up with them. The circularity of argument remains a problem for me.

    Thank you for persevering with me. This is not my primary field, though I am quite familiar with overlap work like isotopic and chemical analysis and interpretation, plus preparation of material for analysis.

    Are you not puzzled also by the 0.9 correlation coefficient? Are you not concerned that adjacent materials seem to have ages up to 6,000 years apart? Do you know if residual ice is analysed for carbon after the gas extraction process from the bubbles?

  119. Posted May 25, 2008 at 3:59 PM | Permalink

    Geoff,

    To begin with your last paragraph: I have no problems with the 0.9 correlation coefficient between IGD and CO2 levels. Similar high correlations are found between IGD, CO2, CH4, annual snowfall,… and temperature. The basic problem is to find out what cause and what effect is in each case. We know from current times that there is a direct cause and effect from temperature on average annual snowfall for near all ice cores. Annual snowfall and IGD are inversely correlated, as bubble closing depth is near constant, only somewhat influenced by snow/firn/ice temperature. And as CO2 follows temperature too, both show similar correlations with temperature, and thus with each other, while they (probably) have not the slightest relation with each other.

    Further, you definitely should read the Etheridge e.a. Law Dome ice core article, it gives in very detailed form what and how was drilled and measured (including cores with different IGD: 10 and 40 years). That means that a factor 4 in IGD gives no difference in measured CO2 level (or a factor 4 in CO2 level with zero IGD “correction”).
    The only exception could be that IGD has a direct relationship with water veins in the ice. But then one should expect: higher temperatures = more water = more loss of CO2 from the bubbles, but we see the reverse… Again, what is observed is that ice cores with extremely different IGD profiles (from less than ten years to many thousands of years) show similar (within 5 ppmv) CO2 levels in bubbles of the same gas age. And CH4 levels with far less possible problems go up and down with temperature in the same way as CO2.

    Current methods use dry crushing for ice and only measure (part) of the bubbles composition, sometimes a lot of measurements at once: CO2, CH4, N2O and different isotopes. I haven’t found measurements of residual CO2/air in the rest of the ice, I only remembered the full sample melting comparison, which showed only a small difference with the dry crushing method.

    Further, I was searching for the full text of Stauffer and Berner, as I have the impression that the figures given are for snow/ice at near melting point, not for the very low temperatures and very low dust levels, which are found in Vostok or other ice cores. Both Stauffer and Berner use Antarctic ice core CO2 data without hesitation in later work, which seems quite strange if their remarks were true for all ice cores… If you have the full article, please forward it to me (see my email address via my family web site).

  120. Geoff Sherrington
    Posted May 26, 2008 at 6:20 AM | Permalink

    Re # 117 Ferdinand Engelbeen

    The technique for separation of CO2 from ice that you mention in 117 is undated in you post but is similar to others of the 2005+ era. Drilling at Vostok was in two stages, the first to about 200,000 years and there were papers from the early-mid 1980s about oxygen isotopes.

    So the explanation of Drake’s figure might simply begin and end with the premature adoption of imprecise CO2 recovery methods. Many authors, even now, are unhappy with the age-depth relations in Vostok and there is a trend with later papers to contort the derived temperature levels ever closer to the derived CO2 values, so that one cannot be said to cause the other, as was the early view – CO2 changes followed temperature changes. How inconvenient!

    For those wondering what we are discussing, this graph from J J Drake of England is central:

    There is an age for ice derived mainly from isotopes and there is an age for the CO2 in ice bubbles, but the two ages are reported as thousands of years apart. One wonders which age should be used for correlation with other cores and other parameters.

    All explanations are welcomed.

  121. Andrey Levin
    Posted May 27, 2008 at 1:03 AM | Permalink

    Re#119, Ferdinand:

    …0.9 correlation coefficient between IGD and CO2 levels. Similar high correlations are found between IGD, CO2, CH4, annual snowfall,… and temperature.

    The problem I have is that “consensus science” jumps from 0.9 correlation to claiming that their reconstructions of ABSOLUTE numbers of CO2, CH4, temperature, etc. are exact.

  122. Posted May 27, 2008 at 12:00 PM | Permalink

    Re #116 Geoff,

    The original article of Berner and Stauffer seems not available on line, but I found a few remarks about its content, which strengthens my suspicion that the whole article is about snow/ice at near melting point. See the chapter “Gas segregation through the liquid vein system” at page 617 of the investigation of the basal ice of Taylor glacier.

    Models of gas segregation through melting–refreezing processes already exist in the literature (e.g. Weiss and others, 1972; Berner and others, 1977; Souchez and others, 1993)

    and

    Building on the fact that gases entrapped in air bubbles can dissolve within liquid water (e.g. Weiss and others, 1972; Berner and others, 1977; Dash and others, 1995), it is reasonable to assume that, like other types of impurities, gases can segregate into the pre-melt phase.

    and

    If so, since internal meltwater is initially almost free of any impurities (Berner and others, 1977), the relative concentration of highly soluble gas species will increase incrementally in the liquid phase, until a steady state is achieved.

    It looks like that this has no bearing for most of the Antarctic ice cores, which are formed at (very) low temperatures, where no summer melting and refreezing occurs…

  123. Posted May 27, 2008 at 12:34 PM | Permalink

    Re #121 Andrey,

    With current analytical techniques, it seems that CH4 levels in ice bubbles are quite accurate and resemble past temperatures, CO2 levels accurate within a few ppm for adjacent cores (Law Dome) and within 5 ppmv for far distant ice cores. Ice and gas age determination is more problematic with lower layer thickness (and deeper parts of cores) and absolute temperature values are discutable, as these depend of isotope interpretation. But that is mainly a discussion about the amplitude of the ice ages – interglacial temperature difference, less a matter of timing (lags) or correlation with CO2 levels.

  124. Posted May 27, 2008 at 12:55 PM | Permalink

    Sorry made a mistake, CH4 levels in ice bubbles resemble past air composition, not (directly) temperature, be it that CH4 levels closely follow temperature variations…

  125. Geoff Sherrington
    Posted May 27, 2008 at 10:55 PM | Permalink

    Re Ferdinand Engelbeen 124

    Or maybe it is that CH4 closely follows bubble air composition which includes CO2 which is adjacent to ice that has a delta O-18 that does not closely follow the delta O-18 of the CO2 whose origins we do not know so we cannot determine the corrrelation between temperature and delta O-18 so we do not really know if CO2 correlates with temperature or not or if ice does or not because we can get get tritium ages on it as well because we do not know how to calibrate the tritium geothermometer or the O-18 one in absolute degrees C?

    Three questions.

    1. Opinion – if you sampled the surface of the Antarctic, would you find a constant set of surface snow isotope signatures from site to site?

    2. If the Vostok ice core nearly reached the postulated lake above bedrock, and is at a guesstimate of 420,000 years old at the bottom, was there a time not long before then when there was no ice there? (if you reply that the weight of ice is causing basal plastic flow outwards to the sea, have we any measurements showing older ice flowing to the sea?)

    3. Given the difficulty in estimating time with depth at Vostok, how can we be sure that there were not lengthy periods of denudation, erosion or sublimation leading to healed unconformities which are being interpreted as sudden climate changes?

  126. Posted May 28, 2008 at 4:46 PM | Permalink

    Re #125 Geoff,

    Sometimes one need to have a look at the literature to know what is going on in the scientific world… Past temperatures in ice cores are deduced from deuterium and 18O changes, both depend of the surface watertemperature at the moment of evaporation and the fractionation between evaporation and precipitation. This gives a quite good qualitative indication of large changes like an ice age – interglacial, but there still is discussion about the real magnitude. See: http://www.ipsl.jussieu.fr/GLACIO/hoffmann/Texts/jouzelJGR2003.pdf
    Besides ice cores, we have ocean sediments which give some similar qualitative and to a certain extent quantitative ideas about what happened in the past times (see a few examples in http://www.pnas.org/cgi/content/full/94/16/8343 ). Further land based observations like lake sediments give us some idea about the regional climate of the past, like a much warmer Eemian in Alaska. Thus all together, we have a pretty good idea what happened in the past, but quantification still has challenges.

    As already said, d18O from the ice originates from the oceans. This has no direct connection with d18O from the inclused atmosphere, which is mainly measured in N2O. This is influenced by plant life (as good as d13C is) and thus by expanding/receding ice sheets. This is visible in the Vostok (and other) data as a lag of ice sheet formation after temperature (see e.g. the Eemian and later at http://www.ferdinand-engelbeen.be/klimaat/eemian.html ). Thus while there is no direct correlation, there is an indirect one and a lag, as ice sheets need a lot of time to be build and melt…

    Abiout your questions:

    1. No. The snow/ice near the coast reflects the ocean temperatures near Antarctica, more inland snow/ice refects in general SH ocean temperatures. Recently there was reconstruction of (near) Antarctic temperatures over the past 200 years based on coastal ice cores: http://www.agu.org/pubs/crossref/2006/2006GL027057.shtml but see also the comment on that article: http://ff.org/centers/csspp/library/co2weekly/20061013/20061013_02.html
    Here as well as in the past decades, temperature variations of most coastal stations and these of the Peninsula go in opposite direction.

    2. Near the base of all central/thick ice sheets, (earth) temperature and ice pressure do melt the bottom layers. The lower parts of the Greenland (near) summit cores are a mix of ice, water and debris (including organic stuff of pre-glacials and bacteria). One can’t see anything anymore from the oldest layers. Lake Vostok may have collected ice meltwater of millions of years…

    3. One can compare ice core events from high resolution (but smaller time frames) to low resolution (longer time frames) in incremental steps for the overlapping periods. That is done for e.g. Law Dome to Taylor Dome and Taylor Dome to Vostok (see http://www.ncdc.noaa.gov/paleo/taylor/indermuehle00grl.pdf ) and Vostok with Epica. Even Vostok was compared to Greenland and the latter layers were counted (yes 120,000 years or so). Further, specific historical events (Younger Dryas) can be used to synchronise different ice cores and ice cores with marine sediments and land based data. This doesn’t exclude that some parts of the ice core(s) may be of insufficient quality, but that can be compensated by other cores (sometimes from the same drilling site).

  127. Richard Sharpe
    Posted May 28, 2008 at 7:05 PM | Permalink

    Oh dear, I guess I missed the memo:

    Polar robots to explore the Arctic

    It’s now almost certain that the world’s ice shelves are melting. And while satellites provide lots of data about their evolution, ground-based weather stations could be even more useful. But if scientists can no longer stay on fragile and volatile ice sheets, what can they do? They can use specially designed robots called SnoMotes developed by U.S. researchers. ‘The SnoMotes work as a team, autonomously collaborating among themselves to cover all the necessary ground to gather assigned scientific measurements.’ More importantly, a SnoMote is an ‘expendable rover that wouldn’t break a research team’s bank if it were lost during an experiment,’ according to the lead researcher. But read more…

  128. Andrey Levin
    Posted May 29, 2008 at 3:51 AM | Permalink

    Re#124, Ferdinand:

    I do not have any reasonable doubt that RELATIVE values of CO2, CH4, temperature, etc. distinguishing between Ice Ages and Interglacials extracted from deep ice cores are valid.

    However, as anyone having experience with indirect analytical measurements would know, calibration of such relative values to get absolute scale numbers is whole lot different, and highly subjective task.

  129. Geoff Sherrington
    Posted May 29, 2008 at 5:28 AM | Permalink

    Re # 127 Richard Sharpe

    Can we please have a Starbucks SnoMote fitted for highly expensive and time consuming dendrochronology sampling too?

  130. Geoff Sherrington
    Posted May 29, 2008 at 6:02 AM | Permalink

    Problems with ice cores (continued)

    After further emails with Jonathan Drake, I mention this paper and this author again. Jarworowshi, Z, EIR March 2007.
    http://www.warwickhughes.com/icecore/zjmar07.pdf

    Despite the strong words and political comments, Jaworowski (who is very highly qualified) goes into great detail about why ice core methods could have problems. Personally, I think he cuts some corners here and there and needs to add recent data. He has been mentioned before on CA and I do not want to cover that ground again. But does he have some valid points to make, here and in his earlier similar paper of 1994?

    Re Ferdinand # 126

    In your response to my question 2, you mention that ice lakes might have existed at Vostok for millions of years. But my point was that reported measurements with ice stop just above the lake, so where has all the previous, older than 420,000 years ice gone to? My expectation would be that the Vostok locality would be covered by some large thickness of ice before 420,000 years because the proxy temperatures do not rise for a long time as that time gets longer. Or, to phrase it another way, on what surface did the 420,000+ year old ice deposit originally? Was it on rock or on even older ice? If on the latter, where is it now?

    Re my question 3, is it enough to say that some parameters move similarly in drill holes spaced globally? What is to stop a past global unconformity affecting all of the hole sites simultaneously? Might that be a possible explanation of the high and asymmetric gradients adjacent to the “temperature” peaks? If there was 100,000 years of no deposition (to make a point), just erosion, would we be able to detect it? I think not, using present assumptions.

    Steve:
    We’ve got enough issues going. I don’t want the board to get caught up in Jarworoski issues right now.

  131. Posted May 29, 2008 at 2:29 PM | Permalink

    Re #130 Geoff,

    As Steve asked, I am not going into a discussion about what Jaworowski says, as that is not appropriate here. One remark only, that he mainly references his own (outdated) work and that many of his objections based on (possible) ice core problems were answered by the 1996 work of Etheridge e.a..

    About the Vostok ice core: Vostok ice comes from a summit 300 km away of Vostok. Its basal 200 m ice is mixed with ice from an under ice lake, somewhat 50 km upstreams of Vostok. This causes that the layers back to about 420,000 years ago can be used, but not deeper (be it with some expansion, to 440,000 years). If the mixed in upstream lake ice is older or not, that is a different question. Anyway, Vostok is on a glacier which slowly fades away towards the oceans. See: http://www.cosis.net/abstracts/EGU05/03522/EGU05-J-03522-2.pdf
    Lake Vostok is much older (from http://www.globalchange.umich.edu/globalchange1/current/labs/Lab9/Vostok.htm ):

    under the 4km ice sheet at Vostok lies a lake the size of Lake Ontario that has been cut off from Earth surface conditions for perhaps 15 million years.

    As the story tells, basal ice not only is disturbed and mixed with debris and older ice, but it is also warmed by the earth’s inner furnace, be it slowly, due to the high melt energy needed for ice and the slow heat transfer through rock.

    A similar story is somewhere available for the Greenland ice core.

    Older ice is found in the Epica C core, which goes back to over 700,000 years.

  132. Posted Mar 5, 2009 at 3:10 AM | Permalink

    Sorry to step in and wake up an old thread but I thought I would have a go at graphing the Vostok data myself. I took a bit of a different tack and resampled onto a common time axis then plotted as a state space analysis.

    http://lnx-bsp.net/agw/

    Comments welcome. In light of the controversy regarding age measurement and variation between age of ice and age of gas, in the light of Drake’s plot above it might be worth trying a bunch of similar state space plots with alternative age estimation algorithms. I’ve got the resampling code available too so people can try it at home.

%d bloggers like this: