30 thoughts on “He’s at it again”

  1. Do you disagree with his basic premise that “Climate sensitivity” is based on Charney and includes only fast feedbacks?

    Some years back I was seeing ‘one degree’ argued. Lots of rhetoric, no definitions.

    I learned finally that number came from ‘instant doubling of CO2 with no other changes’ — Lubos explained it to me. Nobody else did.

    Climate writers didn’t LIKE that definition (because it’s physically impossible, duh).

    Me, I just suggested if the string theorists could instantly reduce the total CO2 by half with no other changes, they’d be the heros of the age. Pull the right string, who knows, maybe.

    Now, climate scientists like three degrees.
    Hansen agrees. He says that’s the Charney definition number.

    I think he’s saying that the Charney number — three degrees — is also physically impossible in the long term, that Charney handles fast feedbacks.

    But like the string theorist’s number, events don’t stop where the definition stops.

    I wish I could hire you, William, to do a ‘code review’ on Hansen’s work — there should be a file folder for every footnote in his long list of citations, I’m sure.

    In the earlier thread I said he sounded like a biologist these days to me, and you wrote

    [If Hansen were talking about biology, I’d be happier…..]

    If he weren’t talking about biology, this would be just an interesting theoretical exercise in comparative planetary climate science, wouldn’t it?

    I guess my question to you is —

    What would have to be true for what he’s writing to make sense to you? What assumption or risk does it seem he’s assuming but not supporting, or that’s not supported by his sources?

    [I’m sure we’ve done this before, probably in comments on the earlier Hansen post. To get the 6 oC value you need to include large ice-albedo feedbacks from large ice sheets that simply aren’t there any more. Hansen is the only one who seems to believe this, and he refs pretty well exclusively his own papers in support. There is no good evidence for his less-than-centuries at all. Code reviewing Hansens stuff is near impossible – like I said, we did a journal club on it at work, and the best summary was someone saying that the paper was “like being inside Hansens head”. There are all the ideas floating around, but they are badly connected and poorly supported. Nor do we know what “fast” means. It might mean “< 1kyr", or 5kyr -W]


  2. I believe Hansen replies to your original critique in this paper. In any case, I argue that at worst, “he may be mostly right for a different reason than he thinks”:


    [I agree with quite a bit of what you write. Saying that we can’t go over 350, or at least not for too long, just isn’t helpful. We’re over 350, we will be for ages.

    Hansen does attempt to address the lack of extra slow feedbacks bit, by considering the Cenozoic, but I doubt thats going to convince anyone not already inclined to be convinced. My knowledge of this is fuzzy, but I think both the T and even more the CO2 histories are imprecisely known, so using them to bound sensitivity just isn’t possible. Certainly in political terms, announcing that clim sens is now twice what you said before, for some unspecified value of “slow”, based on such imprecise evidence… is a complete waste of time, and probably counterproductive.

    I would be all for investigating this via calm quiet science papers, but thats not what he is doing -W]


  3. It is simply misleading to say “he’s at it again”. The main point of the new paper is to use Cenozoic data to counter the criticism that the slow forcing mechanisms of the glacial-interglacial cycles are not effective in the modern context. At 35 My there was no effective ice albedo operating because the planet was largely ice free. Climate sensitivity was at ~3C. After that point, triggered by geological drawdown of CO2 via weathering > subduction outgassing, ice albedo starts to kick in, as the Antarctic ice sheet accumulates. Figs 5 and S4 are the key to understanding this. Which is the better fit – 600, 450 or 325 ppm at the 35 Ma trigger point? Hansen et al say somewhere between the 325 and 450 model seems about right, and it is hard to disagree (look also at Fig S9 for proxy alternatives). In addition, as Joe has already said, there are plenty of other non-albedo amplifiers that are likely vulnerable at 1-2C above current temperatures but plausibly “out of reach” at Holocene-type temperatures. It is not implausible to imagine flat spots where one feedback plays out before others kick in. See Lenton et al 2008 PNAS 105: 1786 for a recent discussion of some likely candidates.

    Regarding the implied lag of 700 yrs between the solar trigger and CO2 feedbacks, the shorter (century to decadal scale response) is inferred on the basis that that there must have been some [unknown] period of time between initial forcing and sufficient CO2 feedback to create a net imbalance in airborne CO2 – probably on the order of centuries due to the relatively small size of the net global solar forcing (~0.25W/m2). There is now no need to await a feedback – we are already in substantial energy imbalance due to direct anthropogenic GHG forcing.

    [That was my reaction to his repeat of the idea that a 6 oC clim sens is (a) likely or (b) policy relevant. The Cenozoic evidence I admit to not having done more than skimmed. I don’t think its good enough -W]


  4. so is Hansen saying that If we kept co2 @ 600 ppm for 10,000 years that temperature would rise 6 C??
    Thanks for your time.


  5. > Jacob
    No. Click on Joe Romm’s site link for help, or the Realclimate “Start Here” link.

    William, would you expect the people whose work Hansen has cited to respond in the journals, if they agree or disagree with his reading of their papers? His drafts have seemed to be review papers pulling together a lot of other people’s work, already done quietly as you say.

    Do you know if he’s geting comments from the authors of the work he cites for the possibilities he describes?

    [I have no idea if he’s getting comments. I certainly *haven’t* seen people saying “yeah, he’s right!”. Most of the people he cites are himself, anyway 🙂 -W]


  6. Hank:
    Is he saying that if we (A) magically added (and froze inplace) 285ppm CO2, and let other things run naturally, i.e. other feedbacks including GHG feedbacks for a sufficiently long time we would see 6C? Is we saying that if we (B) added/maintained independently +4w/M**2 by any means (say brightened the sun) we would get that result? Note that (A) and (B) are not equivalent because the additional forcing due to (A) would vary with the (nonmagical) part of the CO2 fraction.

    I (and I expect other readers who don’t have time to laboriously go through RC) would like a simple explanation.


  7. Doesn’t this say that some Antarctic ice is thinning fast (10x) over the 1992-1996 period compared to the rate over the last 14,000 years? Seems relevant to arguments about rates.

    GEOLOGY, March 2008 v. 36 p. 223 -226.

    First exposure ages from the Amundsen Sea Embayment, West Antarctica: The Late Quaternary context for recent thinning of Pine Island, Smith, and Pope Glaciers

    Joanne S. Johnson Michael J. Bentley Karsten Gohl
    Alfred-Wegener-Institut für Polar- und Meeresforschung, Postfach 120161, D-27515 Bremerhaven, Germany
    Dramatic changes (acceleration, thinning, and grounding-line retreat of major ice streams)in the Amundsen Sea sector of the West Antarctic Ice Sheet (WAIS) have been observed during the past two decades, but the millennial-scale context for these changes is not yet known.
    We present the first surface exposure ages recording thinning of Pine Island, Smith, and Pope Glaciers, which all drain into the Amundsen Sea. From these we infer progressive thinning of Pine Island Glacier at an average rate of 3.8 ± 0.3 cm yr-1 for at least the past 4.7 k.y., and of Smith and Pope Glaciers at 2.3 ± 0.2 cm yr-1 over the past 14.5 k.y. These rates are more than an order of magnitude lower than the ~1.6 m yr-1 recorded by satellite altimetry for Pine Island Glacier in the period 1992-1996. Similarly low long-term rates (2.5-9 cm yr-1 since 10 ka) have been reported farther west in the Ford Ranges, Marie Byrd Land, but in that area, the same rates of thinning continue to the present day. Our data provide the first evidence that puts into context recent rates of thinning of the WAIS in the Amundsen Sea Embayment and demonstrates that these are unusually rapid. The data also provide much-needed constraints for ice sheet models, which are the primary tool for predicting the future behavior of the WAIS
    and its likely contribution to sea-level rise.

    [Antarctica isn’t going to disappear in the near or even far future. “dramatic” and “fast” are very subjective terms -W]


  8. mr roberts thankyou for the links.
    my thought is that because we are emitting a pulse of Co2 that
    its decrease over 1000’s of years may make Hansen’s point less
    If I’m reading what David archer has to say over 1/2 of the Co2 pulse would be gone in a thousand years, this may be fast enough to effect some of the slow feedbacks??
    Just a thought.

    Click to access archer.ms.fate_co2.pdf


  9. jacob l — I believe that David Archer’s paper has appeared. If so, you’ll find it on his publications page. Then checking the references to the draft paper by Hansen et al. may well answer your question.


  10. > Hank:
    > Is he saying …

    Don’t ask me, I can’t interpret. I’m just another reader here, trying to ask questions that may be smart enough to be useful.



  11. In the last couple of years there has indeed been a lot of progress on Cenozoic CO2, and co-author Royer is one of the leading lights. Actually that’s a pretty solid-looking set of paleos on the author list. William, I think you’re being a teensy bit unfair when you imply their results aren’t important to the paper’s conclusions.

    [The results are clearly necessary for the papers conclusions. But I don’t think they are good enough. My opinion isn’t that valuable on the Cen Co2 though, so it will be interesting to read informed comment when/if its properly published -W]

    IMHO we can look at the poles right now and see evidence consistent with the sort of effect the paper proposes. The problem is that the trend is too short as yet to prove anything. The IPY may turn out to have been timed really well.

    [Can we? In Antarctica, there is essentially no change in albedo at all. In the Arctic we appear to be losing ice, but its not clear its having any dramaitc effect on the rest of the world. Moreover, all this is occurring on a timescale too short for his “slow” feedbacks (at least I think it is; his timescales are so vague its very hard to know) -W]


  12. William: you said

    [ In the Arctic we appear to be losing ice, but its not clear its having any dramaitc effect on the rest of the world.]

    Given the increasing demand for grains in India and China and Australia’s wheat crop reduced 60 percent in 2007 it is no comfort to hear uncertainty about Arctic meltback effects on the rest of the world.

    A more responsible and reasoned comment would be:


    John McCormick


  13. Dear Hank Roberts,
    you said it better than I ever did.

    When defining climate sensitivity, one must be careful about the time scale of feedbacks that are allowed to operate. The “one degree” sensitivity indeed allows slow feedbacks only.

    Now, one must ask how slow? How do we distinguish fast from slow? Isn’t there a continuum? There is but one can choose an effective separation. For example, one doubles CO2 instantly and sees the temperature rise, looking like it wants to converge to an equilibrium.

    Once the rate of warming drops to 10% of the initial rate right after the doubling (and it drops after a decade or a few decades), we are pretty much finished with the response of the climate system. We multiply the total warming up to the moment by 1.1, to account for the remaining 10% or so, and obtain the sensitivity that I claim is comparable to 1 degree although Richard Lindzen puts it between 0.3 and 0.5 degrees.

    Now, there could be slower feedbacks, including slow warming of much deeper layers of the oceans and rearrangements of the ice and water level in various places of the world that can take time. One can get new effects and a different sensitivity – in both directions. Moreover, this slow experiment is much less clean physically because it depends on the “context” and can have several solutions. There can exist different attractor points where the overall system is headed and they generate different values of sensitivity.

    At any rate, it seems that there exists no good reason to think that the very-long-term sensitivity should be “certainly” higher than the fast-feedback sensitivity. As I wrote, it could also be lower because the climate system can also have a lot of slow negative feedbacks that regulate things, in agreement with La Chatelier’s law.

    The long-term predictions become much more muddy, less reliable, and they can be influenced by bias and irrational pressure much more than the fast-feedback calculations and predictions (the very idea that the slow-feedback sensitivity is higher is just a hypothesis that is convenient for certain people; and some people tend to add slow feedbacks that almost certainly won’t operate in the near-future world because the required ingredients such as large ice sheets no longer exist) which is why I believe that climate science should only discuss them long after it understands the fast-feedback sensitivity properly.

    Climate scientists often tend to think that they can ignore the very-short-term weather and all uncertainties go away. However, most of the derivations and calculations they use are not valid for asymptotically long periods of time. They may be valid for some intermediate time scales, if any. The idea about the “ultimate” long-term trend being easily calculable is silly. One simply has to be careful what effects he neglects – and for most relevant calculations in the climate, one must neglect both very fast processes (weather that hopefully averages out) as well as very slow processes (that are too slow to matter in a foreseeable future).

    But whether each of these groups of “negligible” effects is really negligible depends on numbers and one should try to be careful if he tries to create trustworthy predictions. Some people are not careful and they just assume, as a matter of dogma, that both inconvenient fast and slow processes are irrelevant. They don’t have to be.

    All the best


  14. According to Hansen’s figure 2, man’s very presence managed to prevent the ice-age from coming – shades of Ruddiman. Phew another catastrophe avoided! We should congratulate ourselves.

    That’s a dramatic GHG forcing on the graph. Now if it included all GHG’s (ie H2O too) it would be exactly the same on the left but very much less steep and scary on the right. Overly-dramatic presentation is obviously vital for pushing policy. I think he’s looking for the top science job in the next Democrat government.


  15. JamesG stated According to Hansen’s figure 2, man’s very presence managed to prevent the ice-age from coming – shades of Ruddiman. Nonsense. The next possiblity of a stade (massive ice sheets) is in 20,000 years. W.F. Ruddiman has recently published a defnse of his thesis that humans have stabilized the climate throughout much of the Holocene. This seems to be so.

    JamesG also stated Now if it included all GHG’s (ie H2O too) it would be exactly the same on the left but very much less steep and scary on the right. I doubt this is so and certainly will not take your word for it. Provide evidence.


  16. William: you said

    [ In the Arctic we appear to be losing ice, but its not clear its having any dramaitc effect on the rest of the world.]

    Would we expect to, at this stage?

    Wouldn’t most of the dramatic effects show up after the ice was done melting?


  17. Phil Hays,

    you said: [Wouldn’t most of the dramatic effects show up after the ice was done melting?]

    Would you ask the same of the need to have a driver-side air bag or would you only consider that after they pull the wreck away from the tree?

    My point is simply that a huge amount of world population’s well being is riding on continuing and reasonable assurance that the world’s grain basket can turn out a crop. Is that too much to ask?

    John McCormick


  18. As you note a draft of “Target Atmospheric CO2: Where Should Humanity Aim?” is available at

    Click to access TargetCO2_20080317.pdf

    You say: “To get the 6 oC value you need to include large ice-albedo feedbacks from large ice sheets that simply aren’t there any more.”

    I don’t see how you can get this from his draft. Point A is at a time as the ice sheets are forming, not yet large and he gets 6 deg as follows, (if I followed)

    [But he says himself that a large part of the extra is from ice sheets. This is not in dispute -W]

    To get an explicit calc of the 6 deg C

    1) Add labels for points A and C to Fig 3B.

    T value for A is 4.8 deg C. T value for C is -1 deg C. By my reading of the graph.

    Therefore delta T (deep ocean) between A and C is 5.8 deg C.

    delta T surface is 1.5 times delta T deep ocean therefore = 8.7 deg C is delta T surface.

    2) Add A, C labels to Fig 4C.
    delta F for A to C is 2.6 + 2.7 = 5.3 W/m^2.

    3) 8.7 C / (5.3 W/m^2) gives about 1.5 C/(W/m^2).
    If 4 W/m^2 for doubled CO2, then
    6 deg C for CO2 doubling.

    So if Figs 3 and 4 are good representations, the 1.5 factor is good and the 4 W/m^2 is good, then 6 C for CO2 doubling, in the longer run.
    Don F


  19. > To get the 6 oC value you need to include large
    > ice-albedo feedbacks from large ice sheets that
    > simply aren’t there any more

    Run this by me again with a few extra words added?
    To get the 6 oC value (that much warming)
    you need to include? Or lack?
    large ice-albedo feedbacks (reflection to space?)
    from large ice sheets
    that simply aren’t there ….

    Isn’t this saying that without the large ice sheets there isn’t enough reflection to space, so the planet warms up?

    And right now we don’t have the large ice sheets?

    What’s the sign on this “extra” amount, plus? or minus?

    Sorry, I just find this so terse that I have to ask someone to spell it out for me in excruciating detail.

    As a Unix (VAX 750) sysop told me in the 300 baud modem days, when I was having trouble understanding how to set a file parmission, “That’s trivial.”

    I asked for the definition of trivial he was using.
    He thought a moment, then smiled and said

    “Too simple for me to waste time explaining, and you’ll never figure it out on your own.”

    I humbly come to you with that kind of question, frequently.

    [The 6 oC is the climate sensitivity he is proposing, ie the temperature change corresponding to 2*CO2. At the moment, that value is believed to be around 3 oC, for policy-relevant timescales. Hansen is suggesting that the value is larger. That largeness comes from including feedbacks from longer timescales. But part of the feedback he wants is from albedo feedbacks from icesheets that simply aren’t there any more. Remember its temperature *change* -W]


  20. Hank, I think the idea is that (a large?) part of his temp. change comes from the reduction in the ice cover also reducing the amount of reflection. That is, take away the mirror and you get a lot of extra insolation (as it’s no longer getting reflected). Instead we’re starting without the mirror, so we’re already getting that extra insolation.


  21. Adam, I followed that much, but as I read it, Hansen’s concerned about the loss of reflection (from loss of the summer ice over the area of the Arctic Ocean, during its 24-hour sunlit period).

    Isn’t that comparable to loss of reflection with prior loss of the continental glaciers outside the Arctic Circle, at an earlier time? I don’t have numbers for the surface area, incidence angle, and day length to do the math.


  22. Got me there. I’d always assumed that the ice that was lost covered a larger area (or had the greater effect), but have to admit that I never confirmed that assumption. From what I’ve looked at, Hansen gets his numbers from places like Greenland & Antarctica which won’t (likely) melt at a rate to bring the high sensitivity in within a meaningful time frame. But here I’m rapidly running out of my comfort zone.

    [Hansen is explicit that he is including the large (no longer existent) ice sheets like Laurentide. Why is this confusing? -W]


  23. I thought the idea was to make an analogy between conditions in the the past — start from a baseline with the Laurentide ice sheets, lose those, absorbing more heat, reaching a higher equilibrium over a long time; and then looking at the present, baseline with the Arctic summer ice present, losing that, absorbing more heat, and suggesting the higher equilibrium would be comparable.

    Not pretending I’m reading this as an expert, just trying to see what the resemblance is between past and present.

    You’re saying he’s basing his six degree number using a contemporary baseline that assumes warming starting from the temperature of Earth with the Laurentide ice sheet in existence?

    I admit I assumed he was talking about that as the past comparison era, and was now talking about analogizing that to a baseline starting in, oh, the 1700s or so, beginning with the world as it was at the beginning of human fossil fuel use?

    I guess I misunderstood completely. So you’re saying he’s really written his paper with a completely wrong baseline starting point, for the current warming?

    I better reread it, again.


  24. “Hansen is explicit that he is including the large (no longer existent) ice sheets like Laurentide.”

    Yes, I thought I said that?

    “Why is this confusing?”

    I don’t know how big the Laurentide ice sheet was (I probably did once).

    “Laurentide + Fennoscandinavian + Cordilleran + Patagonian areas > Arctic Sea area”

    Thanks David, that’s what I was assuming but hadn’t checked.


  25. I know there is a way to figure the energy reflected (total sunlit time, and angle of incidence over that time, and the total area of dry ice or snow, vs. wet ice/snow surface). But I didn’t find an answer.


  26. Ok, reading along in the new RC thread on CO2 levels referring to this, I’m still confused about exactly what Hansen’s supposed to have gotten completely wrong here.

    Would someone point to the language in the Hansen draft article that’s got the facts wrong about the existence of ice, on which he bases his argument?

    [If you’re saying you can’t work out what he is going on about (for which I can’t blame you) then thats a fair condemnation of the paper. “Climate forcing in the LGM equilibrium state due to the slow-feedback ice age surface properties, i.e., increased ice area, different vegetation distribution, and continental shelf exposure…” is part of it -W]


  27. From ‘Target Atmospheric CO: Where Should Humanity Aim?‘, given this statement (which is entirely justifiable when you consider current government desires, attitudes, beliefs, intents and actions on both sides of the Atlantic):

    Present policies, with continued construction of coal-fired power plants without CO2 capture, suggest that decision-makers do not appreciate the gravity of the situation.

    what is wrong with the following as a recommendation for a starting point for a new target, to be kept under review:

    We suggest an initial objective of reducing atmospheric CO2 to 350 ppm, with the target to be adjusted as scientific understanding and empirical evidence of climate effects accumulate. … This target must be pursued on timescale of decades, …


    [Two things. Firstly, we ‘ve already missed the target by a long way. There is no hope (short of some unexpected major catastrophe) of any action severe enough to get CO2 this low for ages. Secondly, its not really clear how grave the sitatuation is. Hansens meters of SLR rise in a century is well out on a limb -W]

    William, I can appreciate you may disagree with James Hansen, Makiko Sato, Pushker Kharecha, David Beerling, Valerie Masson-Delmotte, Mark Pagani, Maureen Raymo, Dana L. Royer, and James C. Zachos all deciding to stick their necks out in ‘Target Atmospheric CO: Where Should Humanity Aim?‘ beyond a comfortable climate scientist’s stretch.

    But … how on Earth are we going to bridge the gap between highly knowledgeable scientists who are experts on specific areas, and the rest of us (policymakers and members of the public alike) who need targets to base policy and actions upon?

    At least Hansen himself gets the attention of policymakers and coverage in the mainstream press, such as the Grauniad. Climate communication that actually makes people think more deeply about actions and targets is, after all, exactly what is needed at this stage. Over, and over, and over again. (Do you agree?)

    If you had the chance (with eight peers) to set the world a target for atmospheric CO, what would your recommendation be?

    [Hansen, and his co-workers, are indeed sticking their necks out. You can call that brave if you want, but they do not represent the majority opinion. What is noticeable is the lack of glacios standing up and saying “yes! Hansen is right”. Getting in the paper because you’ve said something wacky isn’t good. Hansen may have “attention” from policymakers but they aren’t listening to what he is saying

    I’m not sure how you set a target CO2. Based on what? Science or economics or precautionary? My own opinion is that we in the West would be happier if we lived on less. How we get to there from here is hard to know. I’m sure people don’t really like commuting across the atlantic, but they do it -W]


  28. My problem with 350 (or 450) as a target, is that it is unreachable for a long time and that means in policy terms that everyone will do nothing for a long time, wake up and do the Pielke: Impossible, we can’t do that in the time that is left, adapt. My targets would be a schedule of emission rates broken down by years, so that there would be annual goals that are a) reachable and b)testable. The targets should be by country, and within countries by regions.


  29. it seems that there exists no good reason to think that the very-long-term sensitivity should be “certainly” higher than the fast-feedback sensitivity. As I wrote, it could also be lower because the climate system can also have a lot of slow negative feedbacks that regulate things, in agreement with La Chatelier’s law.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s