Positive feedbacks from the carbon cycle

Two papers appeared in Geophysical Research Letters today claiming that the warming forecast for the coming century may be underestimated, because of positive feedbacks in the carbon cycle. One comes from Torn and Harte, and the other from Scheffer, Brovkin, and Cox. Both papers conclude that warming in the coming century could be increased by carbon cycle feedbacks, by 25-75% or so. Do we think it’s time to push the big red Stop the Press button down at IPCC?

The approaches of both papers are similar. The covariation of temperature versus CO2 (and methane in Torn and Harte) is tabulated for a record in the past. For the Torn and Harte paper, the time frame chosen is the last 360,000 years, while Scheffer et al. focus on the Little Ice Age, from 1500-1600 A.D. In both cases it is assumed that the climate shift is driven by some external thermal driver. As the temperature warms (in the case of the deglaciation) or cools (the LIA), the CO2 concentration of the atmosphere changes in the sense of a positive feedback, rising associated with warming or falling in response to cooling. The changing CO2 drives a further change in temperature.

In general, it is clear that eventually the sense of these articles could be correct. The response of the terrestrial biosphere to rising CO2 could go either way; toward an increase in uptake because of CO2 fertilization or a longer growing season (as we see today) versus an increase in soil carbon respiration in warmer conditions (the reason why tropical soils contain so little carbon). Uncertainties in the response of the terrestrial biosphere to rising CO2 is a major source of uncertainty for the climate change forecast (Cox et al., 2000).

The oceans are presently taking up about 2 Gton C per year, a significant dent in our emissions of 7 Gton C per year. This could slow in the future, as overturning becomes inhibited by stratification, as the buffer loses its capacity due to acidification. Eventually, the fluxes could reverse as with a decrease in CO2 solubility due to ocean warming.

The biggest question, however, before pushing the Stop the Press button at IPCC, is timing. The CO2 transition through the deglaciation took 10,000 years. (Actually this helps to constrain the cause of the CO2 transition, because the air/sea equilibration time scale for CO2 would be considerably shorter than that.) The timescale that seems intrinsic to IPCC is a century or so, during which we should be able to reap only a small fraction of any harvest that takes 10,000 years to grow. The Scheffer et al paper avoids this issue by restricting its attention to a time period of just a century.

Scheffer et al illustrate the potential feedback for the coming century in a figure which looks something like Figure A.

Temperature depends on CO2 concentration via radiative equilibrium in the blue curves, and CO2 concentration in the air is affected by temperature according to the red lines. A rise in CO2 concentration from an external source changes the equilibrium CO2 as a function of T relation toward higher CO2, to the right, labeled “forcing”. The stable final equilibrium is where the two relations cross, with further CO2 degassing from the land or the ocean, so that more CO2 ends up in the atmosphere than would have if there were no feedback (a vertical red line). A climate sensitivity calculated from the coupled system is higher than one that ignores any carbon cycle feedbacks.

The situation today is complicated somewhat by a carbon spike transient. Atmospheric CO2 is rising so quickly that the terrestrial biosphere and the ocean carbon reservoirs find themselves far out of equilibrium. In attempting to keep up, the other reservoirs are taking up massive amounts of CO2. If emissions were to stop today, it would take a few centuries for the atmosphere to equilibrate, and it would contain something like 25% of our emitted CO2.

I would draw our current situation as in Figure B, with CO2 concentration wildly higher than the equilibrium red line, poised to relax toward lower concentrations if emissions stopped. The effect of the carbon cycle feedback is to change the equilibrium atmospheric CO2 that we are relaxing toward. It seems to me that the most important part of the equation for our immediate future is the decay rate of that carbon spike, rather than the equilibrium value that CO2 will relax to in hundreds of years.

122 comments on this post.
  1. John Palkovic:

    This page (http://www.realclimate.org/index.php?p=311) does not render properly, either in Firefox or Safari, on a mac running OS 10.3.9. The “B” graph is about 1/2 covered by the brown list on the right side of the page. I could test from a windows machine, but I’ll bet some of the readers of this site are running macs!

    [Response:thanks! fixed now. David]

  2. Coby:

    Is there any hope for gaining insight into the interaction between a “carbon spike transient” and the kinds of feedbacks these slower processes bring by examining the PETM event?

    [Response: The timing is father fuzzy from this event. There may be an initial 13-c change in planktonic forams (surface ocean) before the benthic ones (deep ocean), but it’s impossible to tell if the invasion took 100 years or 500, which is the information that would relevant to this question. ]

    Re Torn et al. It seems problematic to draw too many parallels between a process that took 10K yrs and one that is playing out over one century, especially if it involves the biosphere which will undoubtably react very differently to such a sudden warming. And what about the multi-century lag between T and CO2 in deglaciation? Why do they expect to see effects in this century?

    Do either paper’s conclusions depend on specific mechanisms or is it more based on observation of past correlations?

    [Response:Not really. I think they’d be stronger if the mechanisms were well-known and quantifiable, more than they are. David]

  3. Hans Erren:

    Why didn’t this huge CO2 feedback not occur in the Eemian?

    [Response:Warmer temperature, but not higher CO2? Good question. David]

    I can only see a response of 10 ppm/K.

  4. Alastair McDonald:

    You are all missing the point. Water vapour is the main greenhouse gas not carbon dioxide. When the Artic ice melts (due to the increase CO2 levels)then the increase in water vapour due to a warmer and wetter ocean surface will have a runaway effect caused by the positive feedback from the saturated vapour pressure which increases exponetially. This will be compounded by the loss of albedo from the sea ice. The time to press the big red Stop button has long passed.

    It is well known that carbon dioxide did not drive the recent deglaciation, nor did it (or methane) drive the rapid warming at the end of the Younger Dryas. It is not the main player in the climate stakes, but by moving the snow line it can alter the planetary albedo and so trigger a rapid warming.

    [Response: This isn’t a correct statement. The amplifying effect of water vapor feedback is already incorporated in the “radiative” temperature vs. CO2 curves in Figs (a) and (b). The detailed shape of those curves is affected by the specific assumptions made about water vapor feedback (specifically the way in which relative humidity in the mid to upper troposphere changes with temperature). –raypierre]

  5. Matt:

    As I read the article, the positive feed back, is either a potosynthesis or respiration.

    This paper, using carbon isotope measurements concludes a net transfer to land of some 600 Gt: Estimate of deglaciation carbon transfer

    Woods Hole puts the amount of upper ocean carbon (first 100 meters?) at less than 800 gigatons while land holds a little less than 2,000 gigatons, and deep ocean has someting like 40,000 gigaton. Prior to fossil fuel use the atmosphere held something less than 600 gigaton. I hope my conversion from petagram to gigaton is correct.

    So, the mind boggles with the various scenarios. It seems that the land response to deglaciation is more complex then a simple scenario favoring respiration early and photosynthesis later, which was my guess. The tropics have less carbon and nothern tundra is suspected to be a net carbon source as it warms, yet the northern tundra gained carbon during deglaciation.

    [Response:You are correct that the 13-C seems to tell us that there was net uptake of organic carbon during the deglaciation, which makes explaining the atmospheric CO2 rise more difficult to explain. Truly the mind boggles. David]

    Given the length of the Holocene, the land carbon, worldwide, must have reached eqilibrium, all things being equal. If I compare inter-glacial time periods and holocene plant growth rates of the northern hemispere I suspect that the net introduction of nothern carbon continued well into the holocene. A simple transfer of tropical carbon north does not explain the additional 600gt total, unless the paper cited is wrong; or the net transfer occured early during deglaciation.

    The deep oceans might have been net producer of carbon until the introduction of fossil carbon and deep upwelling stopped as the fossil carbon atmospheric pressure grew.

    Is there something in the science of ocean carbon that could cause increasing ocean carbon on the surface to push back on the release of deep carbon, a change in circulation, or am I just wrong about upwelling of deep carbon?

    [Response:CO2 invades the ocean in some places, where it’s cold or where CO2 has been stripped out by phytoplankton, and it invades in other places. I don’t think of the deep sea as being a huge source of carbon waiting to get out, I think of it as being in equilibrium with the atmosphere. Of course, if you stirred up the ocean more quickly than biology could keep up, or if you killed the biology, then the deep sea would be a source of carbon to the air. David]

  6. Forrest Curo:

    First of all, when do we start heading for that alleged “equilibrium” CO2 level?” We’re doing a dynamic process here, not a static one, and the warming we get will come from the CO2 we have, not the amount we ought to eventually expect with the current temperature. That leftward arrow is a fiction until people get alarmed in a constructive way, which is not yet a general reaction. For now, diagram A is our actual condition, and we need a serious change in behavior to make B apply (which may not occur until a shortage of fuel-burning humans sets in.)

    [Response:Situation B is where we are, insofar as the CO2 is still invading the oceans and the land. You are right that the equilibration arrow is a fiction because we’re adding CO2 faster than it goes away. If we cut emission by about half, uptake would balance release and atmospheric CO2 would stabilize. David ]

    The immediate effect of these positive feedback processes is not simply an increase in the eventual equilibrium temperature, but an increase in the amount of CO2 entering the atmosphere and increasing the current temperature. Perhaps not a large enough increase to dominate our own contributions, but working to bring us closer to whatever other potentially destabilizing thresholds may lie ahead.

  7. Andre:

    Re #2 The question about the 10ka+ spikes of Carbon dioxide is whether it is cause or effect of something. We have several things to take into account. First, the CO2 contents of the oceans is double digits compared to the atmosphere; next, atmospheric pCO2 is very sensitive to changes in surface curents (eg El Nino), changing sink areas and CO2 venting areas and there is clear evidence for oceanic CO2 exchange . Furthermore, the Thermohaline ocean current underwent strong and quick changes around the end of the ice age (Younger Dryas).

    Therefore, it cannot be excluded that those CO2 spikes were the direct and exclusive result of those oceanic changes (Hodell et al 2001), rather than that it was primarily temperature cause – effect related.

    Moreover, the initial warming is assumed to have started about 19ka ago (Clark et al 2004), whereas the start of the CO2 spike is dated not before 17ka (Monnin et al 2004). Hence maintaining the claim of interaction between the two appears to require some revisiting of the evidence.


    Clark et al (2004), Rapid Rise of Sea Level 19,000 Years Ago and Its Global Implications. Science 21 May 2004: 1141-1144

    Hodell D.A et al (2001) Late Pleistocene evolution of the ocean’s carbonate system, Earth and Planetary Science Letters 192 (2001) 109-124

    Monnin, E., et al 2004. EPICA Dome C Ice Core High Resolution Holocene and Transition CO2 Data. IGBP PAGES/World Data Center for Paleoclimatology Data Contribution Series # 2004-055.

  8. Thomas Lee Elifritz:

    Does that big red button happen to say ‘Do Not Press This Button’?

  9. Edward Greisch:

    Question: Lakes with CO2 at the bottom sometimes turn over and the CO2 kills a whole village. Can the ocean do that too, killing the whole planet? How high above sea level do I have to live to be safe from ocean CO2 overturn?

    [Response: You’re thinking of Lake Nyos, I imagine. The uptake of atmospheric CO2 at the ocean surface could not conceivably create the kind of CO2 profile that lead to the lake-type catastrophic release; CO2 enters through the surface of the ocean, and only gradually is brought to the deep ocean. One of the proposed schemes for CO2 sequestration, however, is to inject liquified CO2 directly to the bottom of the ocean, and here it’s not as immediately obvious that a catastrophic release can be ruled out. This was discussed in the IPCC carbon sequestration report that just came out, and a Lake Nyos type catastrophic release seems to be essentially impossible, because of the greater depths and pressures involved in the ocean, and the greater stratification that is working to inhibit vertical mixing. The ecological effects of ocean floor CO2 disposal provide lots of cause for worry, but catastrophic release is one thing you can probably breathe easy about. –raypierre]

  10. Don Condliffe:

    I was thinking along the lines of comment 6. To continue with the thought it seems to me that identified positive feedbacks to increase the effect of higher CO2 levels operate in the near term. Examples include: albedo change from north pole sea ice melt and northern forest albedo decrease. Other identified positive feedback loops include methane release from warming permafrost and decreased CO2 absorption by warmer seas that also act in the near term. Other positive loops that accelerate glacial flow rates and melt, affect sea level.

    Corrective negative feedbacks to sequester CO2 operate in the longer term.

    It seems we have initiated a global system shift to a different state. To reverse this shift wouldn’t we have to reduce the amount of CO2 in the atmosphere? This implies reducing CO2 emissions to less than the rate of sequestration, not just slowing the rate of increase in emissions. Since current trends are for CO2 emissions to continue to increase is there in fact any real prospect avoiding rapid global warming?

  11. teacher ocean:

    How long would that extra 25% of our emissions remain in the atmosphere if we were to stop emitting CO2 today? Tens of thousands of years? Hundreds of thousands of years?

    [Response:Some of it (10% or so) for hundreds of thousands of years. This was the topic of my first post at realclimate! David]

  12. Stewart Argo:

    The article makes sense to me, as far as the last deglaciation is concerned. However I have a problem with the LIA. If Scheffer et al were able to identify a feedback response, does this imply that the LIA was a global event (or at least more widespread than the IPCC thought it was)?

    [Response: Experiments using coupled climate-carbon cycle models such as Gerber et al (2003) [Gerber, S., Joos, F., Bruegger, P.P., Stocker, T.F., Mann, M.E., Sitch, S., Constraining Temperature Variations over the last Millennium by Comparing Simulated and Observed Atmospheric CO2, Climate Dynamics, 20, 281-299, 2003] suggest that the hemispheric-mean cooling of the LIA indicated by the reconstructions featured in the IPCC TAR are consistent with pre-industrial CO2 variations, and that significant greater hemispheric-mean temperature variations would actually be inconsistent the CO2 record of the past millennium. –mike]

    If there was a feedback within a 100 year period, I guess that it could have been caused by cooling prior to this (the LIA started, I believe, one or two centuries before the period studied). I take your point about a CO2 spike in current times, but do their results imply that the feedbacks respond faster during a cooling period than during a warming period?

  13. Arthur Smith:

    It seems to me there’s a curious quirk of a result like this though, that it actually REDUCES climate sensitivity, as measured by temperature change associated with a given CO2 increase – because the positive feedback is to increase the CO2 for a given temperature.

    That’s doesn’t mean it makes things better – it makes things worse because anthropogenic CO2 is enhanced by naturally released CO2 from the positive feedback. But it’s perhaps a limitation of just looking at a doubling of CO2 in describing the results of climate models.

    [Response:No it doesn’t – because classical ‘climate sensitivity’ is only defined for constant CO2 so these feedbacks just don’t come into it. The confusion between the articifical construct ‘climate sensitivity’ and the real question of how climate is sensitive to increasing emissions is very common though. – gavin]

  14. Ferdinand Engelbeen:

    Over the whole 420,000 years of Vostok data, the linear regression between CO2 levels (lagging with many hundreds to many thousands of years) and temperature (derived from dD), was some 8.1 ppmv/K.

    Taking the 1.5-4.5 K warming of the IPCC range of model projections, this would lead to an extra 12-36 ppmv of CO2 in full equilibrium (after thousands of years…), due to warmer ocean temperatures. Or some 4-13% extra over the 280 ppmv induced by humans to reach a CO2 doubling. Or some 1.56-5.1 K warming (which is lower than the 1.6-6.0 K derived from the model by Sheffer ea.) instead of the original 1.5-4.5 K projection.

    Further I wonder how Sheffer ea. could derive anything from the miniscule changes in CO2 levels in one century of the LIA. The variation in temperature over the past millennium was between 0.2 K (MBH98/99) and 1.0 K (Huang, bore holes), which means a CO2 variation of 1.6-8 ppmv. The effect of this on temperature is unmeasurable… (BTW, the variation in CO2 between MWP and LIA was near 9 ppmv in the Taylor Dome ice core).

    [Response: The numbers you cite for “the variation in temperature over the past millennium” are nonsensical (the range of total variation only differs by about 0.4C for the full range of reconstructions, such as shown in this previous post). For actual quantitative comparisons between observed pre-industrial CO2 variations and what is expected given estimated temperature variations and their uncertainties, see the Gerber et al (2003) paper cited above. The comparisons are consistent with 2xCO2 sensitivities within the conventionally cited range of 1.5 to 4.5, but not much higher. –mike]

    If we will need a panic button or not, depends of what the real effect of a CO2 doubling will be, but the CO2 feedback seems of less importance here…

    [Response: Another mistake you seem to be making is that you’re regressing glacial-interglacial CO2 against Vostok temperature, which is a measure of the Antarctic temperature changes. Then you’re applying that coefficient to the global mean temperature. It’s a fallacy to take the Antarctic temperature swings as representative of the magnitude of the global mean, or even of the Southern Hemisphere mean. –raypierre]

  15. pat neuman:

    Re 14

    The amount of knowledge and evidence already justifies pushing the red button, and has justified pushing it for some time. This weekend, ND and MN has daily max temps in the 90s and humidity, but trees have yet to completely leaf out. Earlier this month, TV meteorologists sounded out below normal May temps … but not anymore. Monthly average temperatures in the Upper Midwest have near or above the historical averages at climate stations since 2001. Nobody wants to get accused of pushing panic buttons and feel a loss of credibility. Too bad it seems to be that way, otherwise we may have acted years ago to reduce emissions.

    [Response:Hang on, my rhetorical fluorish of the “Stop the Press” button at IPCC seems to have confused several people (this post, #14, and #8). I meant to say, should the IPCC, which is soon going to start printing its next Climate Scientific Assessment report, stop the press, to revise its estimate of the climate sensitivity or future warming. I did not mean to imply that we should not be concerned about CO2 emissions (the button being whether to take climate change seriously or not). My apologies for all the confusion. David]

  16. Alexander:

    To the Note 6: “If we cut emission by about half, uptake would balance release and atmospheric CO2 would stabilize. David”
    certainly, this would hold true, if the uptake capacity of atmosphere and for CO2 was constant in the future. But taking in mind increased cumulative CO2 rise (or is this increased rise in CO2 due to increased emissions, not due to decreasing absorbing capacity of ecosystems?) and continuing increasing temperature, the amount of “equilibrium carbon stop” needed will be higher from year to year, or?

    [Response:Yeah, you’re right. Cutting emissions enough to balance natural uptake today would only stabilize CO2 until the ocean and land uptake saturates, at which point CO2 would start to rise unless emission were cut further. David. ]

  17. Alastair McDonald:

    David, re your response about which big Red button should be pressed, it was I who proposed that it should be the Really Big one in #4. There was no need for you to apologise for my corruption of your message :-)

    As Ray pointed out in his reply to Ferdi, the correlation between temperature and CO2 in the Vostock core is irrelevant to climate in the northern hemisphere where virtually all the people of the world live. The real danger from positive feedback is a repeat of the end of the Younger Dryas when Greenland temperatures jumped by 10C in three years and by 20C in thirty years [Alley, R. “The Two Mile Time Machine”, 2000.] That would have led to temperature changes in the bread baskets of the world which would not have been so severe, but the effect to the water cycle stretched as far south as the Amazon Basin [Maslin, M.A., and S.J. Burns, “Reconstruction of the Amazon Basin effective moisture availability over the last 14,000 years” Science, 290, 2285-2287 (2000)]. There is no evidence that positive feedback from CO2 or methane was the driving force for that event.

    Ray argues that the models are already including the effects of water vapour, but I am arguing that they are not doing it correctly. The models cannot reproduce the amount of cooling during the Younger Dryas, and that can only be because the water vapour sensitivity is wrong. In fact, we know that is true from “Anthropogenic greenhouse forcing and strong water vapor feedback increase temperature in Europe” by Rolf Philipona et al. GRL, 2005. Ray’s greenhouse effect of water vapour in the upper troposphere is irrelevant, because the atmosphere is in a state of LTE there! And we know from the MSU and radiosonde measurements that the upper troposphere is hardly warming, and what warming is happening is probably due to the Asian Brown Cloud.

    [Response: What you said I pointed out was not in fact what I pointed out. CO2 is global, and your remarks about regional influence are completely unrelated to my point, which was simply that Antarctic deuterium temperature fluctuations are not the same as global mean temperature fluctuations, which (in the ice age problem) are not even themselves necessarily the temperature that feeds back on CO2. Therefore Engebeen’s correlation likely underestimates the feedback, insofar as one can estimate it at all from data of this sort. The comments in your last paragraph are so completely wrong, and in a way we’ve gone over so many times before, that I won’t bother responding again. –raypierre]

  18. Andre:

    Re #17


    “The real danger from positive feedback is a repeat of the end of the Younger Dryas when Greenland temperatures jumped by 10C in three years and by 20C in thirty years [Alley, R. “The Two Mile Time Machine”, 2000.]”

    This is the interpretation of Alley of isotope variation, it could be something different. But Antarctic isotopes are not representative for global temperature why would Greenland summit isotope-temperatures be representative of NH temperatures? It is not even true for South Greenland:


  19. Bryn Hughes:

    Re 15
    It might have been warmer in the midwest but most of Europe was cooler than average in May.I suppose neither of these facts will bear any relationship to the global temperature after all they are just local events.

    [Response:All of these local effects go into the statistics, so they do influence the mean. But if there are more warm anomalies than cold, the global anomaly will be warm. That’s what averaging means. -gavin]

  20. Gar Lipow:

    I know some global warming deniers who have jumped from “global warming isn’t happening” to “global warming is not human caused” to the latest “it is too late to do anything about it.” The old version of this was simply that global warming is an on/off switch – once any global warming happens, well degree does not matter and that is it, let’s move on. But that I think has proven too obviously wrong – I mean the answer is pretty self-evident; it is like saying there is no difference between a mild sunburn, and catching on fire.

    However I’m seeing feedback seized on as the next version of the ‘too late’ trope. “We have all that methane under the permafrost in siberia, and the hydrates under the iceaps. Feedback has already reached the point where they are all going to be released; this will dwarf human emissions, so there is no point in doing anything about those human emissions.” Early days on this one , but I’m betting you’ll be hearing more about it. The intersting thing about it is it may provide additional evidence of how driven by the loonies the denier side is. This trope is not put foward by the few scientists or even much in evidence among the institutes who claim be scientific. It is strickly among the loonies who don’t even make global warming their main focus – for whom global warming denial is just one belief alongside a number of other “interesting” views. But I’m going to make a prediction. Within the next twelve months you are going to hear well known deniers add the “it’s too late; the climate has tipped; the coming of the methane is inevitable” to their standard list. I’m posting it here, both because it is relevent to the subject you are discussing and because I’m making my prediction publicly, so that the accuracy (or otherwise) may be tested. If a talking point that is not put forward currently by any global warming denier with scientific credentials or even any major denier think tank becomes a position advocated either by such think tanks or those with such credentials, I think it will be a tiny piece of additional evidence that such denial is indeed politics driven and not merely a maverick scientific viewpoint.

  21. pat neuman:

    re 19. I also said that monthly average temperatures in the Upper Midwest were above historical averages since 2001.

    I could have added that annual temperature averages were above historical averages for the last nine years in a row and that Jan-Apr temperature averages in ND and northern MN clearly show warming trends which are based on 100-110 mean temperature averages at NOAA climate stations. However, I already said all that and I provided my link to the plots many times to RC.

    Strong regional trends over a 100-110 year period is much more significant than than one or a couple days or months of above or below the official recent 30 year National Weather Service averages which are used to define normal for temperatures at climate stations.

    re 20. I think you missed that … bwe know what’s happening but our day to day lives are a struggle enough, and besides, what one person does or doesn’t do is insignificant. Get a life.

  22. Fernando Magyar:

    As a non scientist who took statistics a long time ago I struggle to understand the analysis of all the data that is presented here on this site. Though I suspect I am able to get the general gist of the fact that according to the general consensus we may be in for some significant climate change begining now and progressing into the commimg centuries, the consequences of which we probably cannot fully asses at the present time.

    I suspect that Gar Lipow is not far off the mark with regards as he calls them “Loonies” However what I find much more disturbing is those who are crazy like foxes and have enormous vested interests such as the oil and energy industries. They have both the money and the power to sway the general public opinion and are more than able thus to have their way.

    Case in point an article in today’s New York Times business section: http://www.nytimes.com/2006/05/28/business/28coal.html?hp&ex=1148875200&en=2cffad8d8e32d294&ei=5094&partner=homepage
    talks about what is happening with coal. The discussion seems to focus on the application of cleaner tecnnology to reduce soot emissions and reduce things like acid rain. That certainly may sound like a good idea. Though it seems to me to miss the point entirely. I think that any carbon dioxide emitting energy source regardless of how currently inexpensive it may seem at the moment has to have the cost of the long term societal consequences factored in to the bottom line. In other words if the use of cheap coal will in the long term be a big part of accelerating climate change and if that change can be shown to have deleterious effects on the livelihoods of of societies through drought and famine and floods etc… then they should only be allowed to embark on such business ventures if they are held accountable for the consequences from the get go. As I see it now if these scenarios should come to pass they will suffer no consequences whatsoever. So they feel free to pursue their profits at any cost and since they will appear to be selling cheap energy (the real costs of which will be denied) the ignorant public will be a willing participant in this unfolding global fiasco. ‘CO2: they call it pollution, we call it Life!’ is not in my opinion as funny as it seems at first glance.

    [Response:That’s exactly why some kind of regulatory environment for CO2 is needed, and why the market won’t take care of it all by itself. There are regulations in place (despite attempts to weaken and dismantle them) that control soot, sulfur compounts, nitrogen compounds and (soon) mercury, and therefore companies spend money on the technology to reduce emissions of these. However, there are no market signals that represent the environmental costs of CO2 emissions, so a company would have to be crazy — indeed irresponsible in the fiduciary sense — to spend extra money reducing CO2 emissions, for example by building IGCC plants and providing for sequestration. Right now, burning coal in pulverized coal plants makes great economic sense, and there is no reason in the world it won’t expand — as indeed it is doing. The tragedy is that a coal plant will have a 50-60 year capital life, perhaps more. If a company took the long view and figured in the cost of capturing and sequestering CO2, not only would IGCC plants look cheaper than pulverized coal, but even coal with IGCC would start to look less economically attractive than various alternatives. –raypierre]

  23. John L. McCormick:

    Re# 16

    Alexander, you said:

    “But taking in mind increased cumulative CO2 rise (or is this increased rise in CO2 due to increased emissions, not due to decreasing absorbing capacity of ecosystems?)”

    I used available data on global 2003-2005 fossil fuel use and its CO2 emissions and the 2006 Jan-APR Mauna Loa CO2 concentration measurements and the following might address your question:

    Using 2006 JAN-MAY 11 hourly CO2 measurements provided by NOAA Earth Systems Research Laboratoryâ??s Carbon Cycle Greenhouse Gases Group, I derived a January increase of 2.98 ppmv CO2 over 2005 January concentration. Please note, my calculations are based upon preliminary data.

    Is that increase attributable to increased fossil fuel use?

    Global fossil fuel use data from BPâ??s Statistical Review of World Energy, June 2005 provided the following coal, oil and natural gas production for 2003 and 2004. I estimated 2005 data using conservative percent increase.

    2003- 5.185 billion tons
    2004- 5.538 billion tons
    2005- 5.954 billion tons (7.5 percent increase)

    2003- 28.1 billion barrels
    2004- 29.3 billion barrels
    2005- 30.8 billion barrels (5 percent increase)

    Natural Gas
    2003- 92,053 billion cu ft.
    2004- 94,462 billion cu ft.
    2005- 97,298 billion cu ft. (3 percent increase)

    Global Tons of CO2

    2004- 12.671 billion tons
    2005- 13.621 billion tons

    2004- 12.851 billion tons
    2006- 13.494 billion tons

    Natural Gas
    2004- 5.526 billion tons
    2005- 5.691 billion tons

    Total CO2
    2004- 31.048 billion tons
    2005- 32.807 billion tons

    1 GtC corresponds to ~3.67 Gt CO2

    2.12 GtC or ~7.8 Gt CO2 correspond to 1 ppmv CO2 in the

    D. Schimel,et.al,CO2 and the carbon cycle. Pages 76-86
    in [IPCC 95])

    The January 2006 Mauna Loa CO2 increase over 2005 was 2.98 ppmv, or 23 billion tons of CO2.

    So, I conclude that the January 2006 increase was equivalent to 74 percent of total 2004 global fossil fuel CO2 emissions.

    CO2 uptake by ocean and terrestrial sinks is not immediate and uptake is affected by temperature.

    The energy and CO2 data are readily available to compare January 2006 increase against any successive Januaries from 1958 to 2005. If you are interested, I can provide more comparisons.

    I hear many different explanations for the dramatic Jan 2006 over 2005 increase. None were based on analysis – just thow out lines such as fires, El Nino (?), very cold winter in Europe and Russia, China and India energy use. The January CO2 increase is not faulted by the measurement process itself. The Keeling team have a rigid, quality-controlled standard of excellence.

    So, any suggestions to add to the “cause” list?

  24. Alexander:

    I think that, in general, processes in which one component lags another and is the effect of that cause or that they share a common cause, the lagging component will display a diminution in response to changes of a shorter period than the period of the lag.

    A response with a millennium lag will look smoothed on a centennial time scale, one with a century lag will look smoothed on a decadal time scale, etc.

    From the little I know of ice core evidence I am left with the impression that the CO2 response lags considerably => 500 years. Such a CO2 response mechanism, if this is a single mechanism, would be unlikely to be able to track sub centennial variations which is what I believe has been shown for the Little Ice Age record.

    Alternatively there could be different mechanisms or even a reversal from effect to cause between these two examples. It is not clear to me that a single mechanism could be responsible for both.

    I do not know how discriminating the ice core records are on decadal scale nor if anyone has looked for a second response with a short lag in these records. Absence of any short term CO2 signal in the presence of short term termperature signals would speak against a CO2 response to temperature with a short time lag, at least during that epoch.

    Regarding the near future; If the lag is very short (< =10 years) , it might not be possible to extract the CO2 response from the CO2 cause prior to it all being a postmortem exercise unless a significant reduction in emissions occurs. If the lag is long (>200 years) the response will not be significantly excited if the problem is solved in time that remains to do so. I would think that the worry would lie in the lag being of the order of the problem, say 30-60 years as that could give rise to a strong, and so far hidden, response extending for another 30-60 years even if all the combustion stopped tomorrow.

    What, perhaps, is needed is a significant signal in known CO2 emissions that can be looked for in the record to show the lag time in any response mechanism that is in the current regime.

    Sadly I can not think of one, not even the Oil Crisis 1973 or a war, that would be likely to show this up. I do not know if general emissions are sufficiently well recorded to analyse the C02 record for any lag.

    Even if it is, then given that the atmosphere is out of balance due to the rapidity of the rise, one might expect a predominance of an “apparent” CO2 leading signal due to the effects of CO2 sinks, as the CO2 level moves from being a response more to the total emissions to an equillibrium where it reflects more the rate of emissions. Detecting a lagging signal might not therefore be feasible given the uncertainties involved.

    A possibility that such a response exists leaves us with the interesting prospect that until we make a significant dent in the emissions we might never know how bad the problem is.

    I apologise for being long winded and vague on the evidence but if a strong ~50 year C02 reponse to temperature exists then it would be very bad news and if it can not be ruled out then perhaps it should be factored in to the limits of its possible range.

  25. Alexander Harvey:

    My apologies, I did not notice that I am the second Alexander , the long ramble about lagging signals by me.

  26. Wolfgang Flamme:

    Here’s an interesting publication about Vostok ice leads and lags:

    Mudelsee, M., The phase relations among atmospheric CO2 content, temperature and global ice volume over the past 420 ka
    Quaternary Science Reviews 20 (2001) 583-589

  27. Ferdinand Engelbeen:

    [Moderator: Nonsense deleted]

    I know that Vostok dD at most represents SH sea surface temperatures, but – although there were time shifts between SH and NH – the NH temperature variations over the glaciations/deglaciations were not that different.
    – I checked ice sheet growth over the Eemian, which is roughly compatible with Vostok dD (a lag of a few thousand years). This was measured by d18O in N2O, which I suppose represents global ice sheet growth.
    – As ice sheet growth was mainly over land and mainly over the NH, I suppose that NH temperatures were lower in the NH than in the SH during the glacials (and higher during interglacials, as there is more land, which warms faster).
    – If temperatures in the NH were lower during glaciations, then global temperatures were lower than measured as dD in Vostok. And the global temperature variation during the transitions were higher, but result in the same variation in CO2 changes.
    – Consequently, the result of temperature increases even gives less extra CO2 than in my previous calculations…
    – Btw, according to the Gerber model runs, most of the CO2 changes follow temperature changes already within 1-2 decades. Thus it is no wonder that such an effect is not measurable in the current CO2 and temperature trend. Only with fast changes like El Nino, a variation in CO2 increment follows temperature changes after some 6 months. Similar changes are measurable in plant growth.

    [Response: You miss the point about the LGM. The LGM temperature drop pattern is not a very good analog for global warming, since the interhemispheric asymmetry in the former is much greater than the latter, especially as equilibrium is approached. Which temperature do you want to pick to make your analogue? Some estimates of LGM SH midlatitude cooling are as little as 2-3C, putting them in the same ballpark as the temperature change from anthropogenic global warming. That would (naively) imply that if the feedbacks worked the same way you could get maybe 100ppm boost from the CO2 feedback. Maybe you’d even get more, since in the LGM you expose a lot of new forestable land as you deglaciate, tending to offset what the ocean might be doing. It’s really really hard to use the LGM as an analogue for what happens under anthropogenic global warming, but insofar as one is going to do it at all your numbers are way too optimistic. –raypierre]

  28. Gar Lipow:

    >think you missed that … bwe know what’s happening but our day to day lives are a struggle enough, and besides, what one person does or doesn’t do is insignificant. Get a life

    Huh? I’m trying to see how it what you say relates to what I wrote. I’m pretty sure that the only people I talked about were flavors of global warming denialists. I made a testable prediction of the next denialist line. Don’t see how that is such a terrible thing to do. Don’t know what is wrong with people doing what they can.

  29. Grant:

    Re: #23

    First of all, here is a plot of montly CO2 concentration from Mauna Loa since 2000. There are two clear trends. First, there is a steady secular increase, which is signficantly described by a quadratic curve. This fit indicates a rate of increase

    R = 2.11 + 0.0952 (t – 2003.17)

    where R is the increase rate in ppmv/yr., t is the time in years. Second, there is an annual variation.

    We can subtract both the secular increase and the average annual variation, leading to a “CO2 anomaly,” which is plotted here. From this we see that the CO2 anomaly for Jan. 2006 is not at all out of line; in fact it’s very small. We can also see that Jan. 2005 had a rather large *negative* anomaly. Therefore the difference between Jan. 2006 and Jan. 2005 is large, constituting the rate for 2006 (from the above formula, 2.38 ppmv/yr) + anomaly for Jan. 2006 (small positive) – anomaly for Jan. 2005 (sizeable negative).

    However, most of the departure from normal of the Jan.2005 – Jan.2006 difference is due to the large *negative* anomaly from 2005. Hence the exaggerated growth of CO2 from one January to the next really says a lot more about how *low* the level was (compared to long-term trend) in 2005, than about how high it was in 2006.

  30. pat neuman:

    re 27. I was trying to give examples of what people have said to me when I brought up the fact that global warming is out of control and is being driven by our greenhouse gas emissions. When they give up on trying to say global warming isn’t really happening, they’ve said there’s nothing they can do about it anyway which would matter. I agree with you that people need to do what they can to reduce their own emissions, even though it may seem to not matter. It’s a moral issue to do what we know is right, not wrong. It’s wrong to use fuel excessively with no regards for the many very serious consequences from a more rapid rate of global warming.

  31. Hank Roberts:

    Ray, for those who have a single theory they keep reposting, tempting you to either rebut again or again swear off answering — perhaps RC could provide a linkable stock rebuttal page?

    Reminding the posters you’ve rebutted them before won’t stop them; new readers however need to see your early, patient, detailed, footnoted attempts to explain why such ideas aren’t supported in the research.

    Point being to educate new readers who can’t tell which of you to believe in the absence of any footnotes from either of you! (wry grin here). “Use the cites.”

  32. Alex:

    To clarify, are these papers suggesting the warming may be underestimated because of stronger positive feedbacks in the carbon cycle, or were such feedbacks largely excluded from consideration in the IPCC projections?

    Also, I understand there have been studies assessing the effect of CO2 fertilization under real world conditions, but I wonder if there’s been much follow-up on the effect warming will have on long-term sequestration by phytoplankton, and how quickly that might become a factor. This was the last I saw on diatoms.

  33. Hank Roberts:

    Alex, try these (papers citing the article in Science)


  34. Lawrence McLean:

    I would just like to remind all posters that they should always carefully review their posts before submitting them. Real Climate have provided a nice mechanism for this process.

    A submission may be clear in your head, but, sometimes a crucial letter, word or sentence may be missing causing confusion and resulting in to and from posts that would have been unnecessary if the originating post was correctly submitted.

    Pat, I think your post 21 had something missing, I feel you missed a sentence to establish the context for the last paragraph. Also I do not like acronyms, excepting those that are widely understood, for example I am not sure what “bwe” means. :-)

  35. Gar Lipow:

    Re: 30

    >When they give up on trying to say global warming isn’t really happening, they’ve said there’s nothing they can do about it anyway which would matter. I agree with you that people need to do what they can to reduce their own emissions, even though it may seem to not matter. It’s a moral issue to do what we know is right, not wrong. It’s wrong to use fuel excessively with no regards for the many very serious consequences from a more rapid rate of global warming.

    Ah. I’d agree with the provision that it is more collectively wrong than individually wrong. Yes there are things one individual can do. But mostly there are things we as a society can do. I can bicycle when practical but I can’t take a train in my town cause there are none.

    Similarly decarbonizing the grid will be a collective, not an individual choice. A few well-to-individuals can buy hybrids and convert them to plug-in hybrids on their own, or even buy expensive electric cars. But until these are massed produced, available at the lot these won’t be an option for most of us. And so on. So it is not a question of individual guilt – it is a question of social choices.

  36. Pekka Kostamo:

    RE 28: It seems to me that in the anomaly graph there is also some signal (i.e. a period of about 3 years). Of course, the time series is far too short to be convincing, but still it does not appear like just random noise. On even more shaky basis there might be a relation to the Dow Jones index (http://home.xtra.co.nz/hosts/cwwebsite/history_dow_jones_index.htm) which purportedly is somehow dependent on industrial activity level. Some broad trends appear to be coincident.


  37. Grant:

    Re: #36

    I too think the data (the “CO2 anomaly”) is not just random noise. However, the point I was really driving at is that the change from Jan. 2005 to Jan. 2006 is definitely *not* unusual, given the behavior of the CO2 concentration since 2000.

    As for an approximately 3-yr. period in the CO2 anomaly, it’s best to look at the entire time series. I took the whole montly data set from Mauna Loa and generated a new curve in the same manner: subtract from the data a quadratic secular trend and the average annual variation to generate the “CO2 Anomaly.” The result is plotted here. Note that these are *not* the same anomalies plotted before, because the quadratic secular-trend curve which has been subtracted is that based on the entire time series, not just the series since 2000.

    The most notable feature is the precipitous drop in CO2 anomaly in the early 1990s. This was discussed before; posts suggested as possibilities an association with el Nino, as well as with the collapse of the Soviet economy.

    There are visible signs of an ~ 3-yr. period throughout the time series. A Fourier power spectrum is shown here. The most notable feature is the tall peak at very low frequency, which does *not* correspond to any real periodicity. Note there is a significant peak at frequency 0.275 cycle/yr., corresponding to period 3.63 yr. However, this peak could be indicative of a period, or it could indicate only a “characteristic timescale.” It could even represent neither, being only a manifestation of the “red noise” character of the CO2 anomaly curve. In fact the interpretation of peaks in a Fourier power spectrum — especially at low frequencies — is a *very* tricky business (Foster, G. 1996, Astron. J., 111, 541).

  38. jhm:

    I was unaware that temperatures could be derived from ice core data. Could I have a link to an explaination of “dD?” Is this a derivative, or a delta? what is ‘D?’

  39. Hank Roberts:

    jhm, here’s a quick answer, and an example of how you can find such answers.
    Read just the bit I quoted below for the answer to your question; see the whole paper for a good explanation.

    This is from the first hit produced by Google Scholar using this string:

    +temperature +”ice core data” +dd

    Climate and atmospheric history of the past 420,000 years from the Vostok ice core, Antarctica

    JR Petit, J Jouzel, D Raynaud, NI Barkov, JM â�¦ – Nature, 1999 – earth.unh.edu

    Ice cores give access to palaeoclimate series that includes local
    temperature and precipitation rate, moisture source conditions,
    wind strength and aerosol fluxes of marine, volcanic, terrestrial,
    cosmogenic and anthropogenic origin. They are also unique with
    their entrapped air inclusions in providing direct records of past
    changes in atmospheric trace-gas composition….

    The ice record
    The data are shown in Figs 1, 2 and 3 (see Supplementary Infor-
    mation for the numerical data). They include the deuterium content of the ice (dD ice, a proxy of local temperature change), the dust content (desert aerosols), the concentration of sodium …

    That’s been cited by 848 subsequent papers.

    I suggest you read the whole thing, search for it with Google Scholar.

  40. Andre:

    About the ice cores, There is a connection between isotope rates and temperature but especially in (Ant-)Arctic areas also with precipitation. The complexity has been shown in this Ph.D thesis:


    “It is emphasised that isotopic variations in the hydrological cycle are complex, even in a relatively stable area such as Antarctica. Therefore one should be careful when quantifying climate signals over a period of several years on the basis of ice-core data.”

    However since the deep ice cores of Antarctica are very difficult to date with very few hard data points, Jouzel et al (EPICA Dome C) and Petit et al (Vostok) appear to have a used rather simple algoritms to wiggle match precipitation rate with isotope ratio and age. At least this appears when the “layer thickness” is correlated with the isotope ratio. r2>95% That’s highly improbable and it seems to proof direct dependency, or circular reasoning, so to say. Consequently, the variation that Michael Helsen identifies would have definite implications for both chronologies and temperature interpretation.


    For instance, Look at the near identical shape of the last spikes of both deuterium and clearly lagging CO2 (dD). Assuming normal second or higher order response characteristics, this strongly suggest a near simultaneous start.

  41. John L. McCormick:

    RE #36

    Pekka, many thanks for the link to that fascinating chart. It tells much about the industrialization of the US before and after the WWII but it does not have much direct, certain relationship to atmospheric CO2 concentrations. Certainly, the IT, dot.com industry added a lot of growth and gambling to world stock trading. However, a graph of US or world consumption of coal, oil and gas will track with the upward momentumn of Wall Street.

    I will post some data on post-1981 global fossil fuel production in a later submittal. Thanks again for this unique piece of historical record.

  42. wayne davidson:

    What would be the current absolute temperature increase of the atmosphere,
    with respect to the above :

    “Both papers conclude that warming in the coming century could be increased by carbon cycle feedbacks, by 25-75% or so”

    Would like to read IPCC trend compared to the papers suggested increase. in degrees K.

  43. Brian Gordon:

    This page (http://www.realclimate.org/index.php?p=311) does not render properly in Opera, on a PC running XP. Figure A sometimes partially appears, and there’s a Figure B? I usually don’t have problems with this site in Opera.

    As to the feedbacks: Is it just me, or is pretty much every scientific (as opposed to political) estimate of the rate of climate change an underestimate?

  44. pat neuman:

    re 35.

    I still think it is individually wrong to use fuel excessively, and there should be individual quilt for doing that. Two summers ago I visited a place that makes extravagant yachts that use enormous amounts of fuel. Their business was and probably still is booming.

  45. Steve Bloom:

    Re #40: Andre, it appears you saw what you wanted to see in that dissertation. It simply asserts that it is more reliable to average temp data over several years, i.e., temp data plotted for individual years will tend to have larger error bars than a running average. Is this any sort of surprise? Do any scientific conclusions relying on ice core data depend on having precise temps for individual years?

  46. Mark Wiener:

    Methinks arguments and speculation regarding atmospheric Carbon Dioxide will continue for some time. Regardless of possible climatic effects, in my opinion an overage of pollution alone is sufficient justification for burning less fossil fuel. Another justification is to cool off the oil wars.

    A combination of Net Zero appropriate architecture and plug in hybrid vehicles could quickly cut carbon emissions in half or more. Visit calcars.org and read about Net Zero architecture with photovoltaics.

  47. John McCormick:

    Re: #46

    I do not reject your support for Net Zero architecture and plug-in hybrid vehicles. But, I do urge you and others sharing your view to dwell on some of the real facts of life in America and likely all developed nations.

    1) most all of us are in debt up to our ears and are still paying the mortages on our 64 million owner-occupied dwellings. So, we are not about to put a “FREE” sign (who could afford two morgages?) on the lawn and build a Net Zero home.

    2) we drive our cars, SUVs and whatever we own, for about 10 years on average. To abandon our car for a hybrid (even if our line of credit still has some cash available) will flood used car lots with cheaper cars the less wealthy among us will purchase and drive.

    3.)there is no way, no way at all, to “quickly cut carbon emissions in half or more”. That regrettable fact deserves some attention when contributors wing all kinds of carbon-cutting suggestions without trying to imagine and help us see how a credit-strapped America will respond or – even how impractical it might be to plant trees without having, in place, a means to water them.

    I could go on for days with examples of the “lets do this or that” as if the average Joe and Jane are eager to spend big bucks on a “solution” they have no clue will make a difference.

    Personal choices will only go so far but mega-decisions driven by an almost-martial law attitude will start moving the Northern people towards the 60 to 70 to 80 percent carbon reductions needed to help our children survive.

    by John McCormick

    [Response: It’s because of the capital life issues you raise that it’s so important to get moving on putting the right market signals in place right now. Putting strong market incentives in place now will assure that emissions will start to go down in the next 20 years as capital assets are retired and replaced. It would have been better to have started this 20 years ago, but better late than never. –raypierre]

  48. Karen Street:

    Re #47 Raypierre is of course right, that we are currently making big infrastructure changes, in the US and elsewhere. A house or building from 2010 will stand a long time, a car from 2012 will run for a long time, so we want them to be as energy efficient as possible.

    That said, I suspect the typical American could reduce energy use some 10% fairly easily. During the California electric crisis, we cut electricity use 10%, and those interviewed for TV news said, I didn’t really do anything, just turn off the light when I left the room and turned off my computer when I wasn’t using it and … Perhaps carpooling or taking public transit once or twice a week, or taking the train or bus for vacation rather than the plane, and living a little more consciously in the house. Then the policy people and the technology people could work a little less feverishly to find solutions (though reducing carbon emissions 60% or more will still be hard).

  49. John L. McCormick:

    RE #48

    Karen, power control areas are 133 electric grid rings that connect regional electric power to regional demand across the nation. They are connected one to another and are the means to avoid a widescale blackout if a plant or transmission line goes down. Perhaps you know this but not many people have any idea how fragile is the US electric grid now.

    Looking to a possible near-future crash program to reduce US CO2 emissions, power stations will be primary targets. Of the 1,000+ gigawatts of generating capacity, 60 percent are coal, oil,gas-fired and emitted, in 2004, 2.3 billion tons of CO2, or, 32 percent of US 7.12 billion tons.

    Now, big infrastrcuture changes come into play. The top 14 power control areas, in terms of CO2 emissions, emitted about one billion tons of CO2 from their total 212 Gigawatts of capacity. They are all within a 300 mile radius of Louisville. Bottom line: 20 percent of US generators supply electricity to about 100 million customers and deliver 14 percent of US CO2 emissions to the earth’s atmosphere.

    I say this as an avowed environmental activist with 30 years of lobbying for the coal strip mine reclamation law, clean air act and acid rain control and against utility deregulation.

    RealClimate may not be a suitable page to discuss engineering challenges but we do have to pay some attention to those pesky numbers behind all of our hopes that we can somehow make the problem of climate change become more manageable by cutting 60 or 80 percent of our carbon use.

    I too believe we can shave 10 percent off our individual energy use. That is a down payment on the more than $200 billion in infrastructre replacement the fractured and recalcitrant electric utility industry faces.

  50. Ferdinand Engelbeen:

    Re #27 (comment):

    Raypierre, I have purchased (again…) the two articles in GRL to know the details of what was done. For the Torn and Harte paper, which used Vostok data, it is indeed clear that the difference between straight-forward local temperatures (derived from dD) and SH temperatures (derived from corrected dD) over 360,000 years is quite large.

    This gives 9.1 ppmv/K for local Vostok temperatures (over the full 420,000 years, I calculated 8.1 ppmv/K), against 14.6 ppmv/K for the SH. For the LGM-Holocene transition, the (probably SH) slope is 17 ppmv/K.

    But what happened in the NH? That is the important question. For what I have read in the past, the NH has much larger temperature swings between glacials and interglacials, due to more land, which means more ice building during ice ages and higher temperatures in interglacials. But higher NH temperature variations implicate higher global temperature variations and a lower response of CO2, as these are globally known (within the error margins of the ice cores…).

    I didn’t find NH data which corresponds to the full Vostok time record, but the Greenland ice core data span the last glacial, including the LGM-Holocene transition. From the Greenland ice core data, one can derive a local temperature change of ~20 K between the LGM and the Holocene (with an extra peak just before the Younger Dryas), derived from d18O in ice. The corresponding NH (or global?) temperature according to the univ. of Leeds, item 7.5, might be around 10 K between LGM and Holocene. Global temperature shifts then would have been around 8-9 K (with the corrected SH temperature swing and assuming the “Leeds factor” as NH temperature). The change in CO2 levels in the same period is ~80 ppmv, that gives a slope of ~10 ppmv/K. A little more than I derived from the full Vostok (local) temperatures, but less than what is calculated by Torn and Harte.

    If we assume that the 10 ppmv/K is right, then the IPCC model range would be 1.58-5.3 K for 2xCO2. The lower end still seems benign, the middle range (3->3.3 K) still critical and the higher range still a disaster, with or without CO2 feedback…

    I haven’t checked the Scheffer ea. data yet (based on MWP-LIA changes), and rapid changes due to ENSO events may be of interest too.