Unforced variations: Nov 2011

Once more unto the open thread…

341 comments on this post.
  1. Edward Greisch:

    Andy Revkin at
    has a new sensitivity “we estimate a lower median (2.3 K) and reduced uncertainty (1.7 to 2.6 K 66% probability).”
    that he got from:
    So Andy Revkin says that there is no problem.

    I replied that there is still the same problem.

  2. sidd:

    Mr. Per Wikman-Svahn writes on the 25th of Nov 2011 at 9:25 AM:

    Re:Supralinear or exponential acceleration of sea level rise

    Sea level rise comes from thermal expansion (arguably linear), export from land based aquifer to ocean (almost zero), and ice melt. Of these, the third seems the most plausible candidate for nonlinear increase. It is hard to see how the ocean heat could penetrate GIS other than through giant rainstorms, but WAIS shows us how fast breakup can go on retrograde seabed. In this regard the Pollard and DeConti results are both illuminating and scary.


  3. prokaryotes:

    CO2 climate sensitivity ‘overestimated’?

    Lead author Andreas Schmittner from Oregon State University, US, explained that by looking at surface temperatures during the most recent ice age – 21,000 years ago – when humans were having no impact on global temperatures, he, and his colleagues show that this period was not as cold as previous estimates suggest.
    “This implies that the effect of CO2 on climate is less than previously thought,” he explained.
    Climatologist Andrey Ganopolski, from Potsdam Institute for Climate Impact Research, Germany, went further and said that he would not make such a strong conclusion based on this data.
    “The results of this paper are the result of the analysis of [a] cold climate during the glacial maximum (the most recent ice age),” he told BBC News.
    “There is evidence the relationship between CO2 and surface temperatures is likely to be different [during] very cold periods than warmer.”
    Scientists, he said, would therefore prefer to analyse periods of the Earth’s history that are much warmer than now when making their projections about future temperatures.

    Duh…. the last cold snap to conclude about CS is more than a stretch! Although current CS estimates (3C) are maybe even an underestimation!
    And great timing for this “study” to appear and a bit lame of the BBC to not make it more clear that this study is rather weak!

  4. Hank Roberts:


    November 25, 2011, 12:29 p.m.
    Renewable energy is surpassing fossil fuels for the first time in new power-plant investments, shaking off setbacks from the financial crisis and an impasse at the United Nations global warming talks.

    Electricity from the wind, sun, waves and biomass drew $187 billion last year compared with $157 billion for natural gas, oil and coal, according to calculations by Bloomberg New Energy Finance using the latest data. Accelerating installations of solar- and wind-power plants led to lower equipment prices, making clean energy more competitive with coal.

    Oil near $97 a barrel Oil near $97 a barrel
    Biofuels, wind power show gains, but hurdles remain Biofuels, wind power show gains, but hurdles remain
    Energy secretary refuses to apologize over Solyndra loan Energy secretary refuses to apologize over Solyndra loan

    “The progress of renewables has been nothing short of remarkable,” United Nations Environment Program Executive Secretary Achim Steiner said in an interview. “You have record investment in the midst of an economic and financial crisis.”

    The findings indicate the world is shifting toward consuming more renewable energy even without a global agreement on limiting greenhouse gases. …”

  5. Edward Greisch:

    Go to americanselect.org/
    Americans Elect is a place to nominate 3rd party candidates and campaign for questions to ask candidates. Please go there and campaign for GW questions and candidates. Let’s nominate Mike Mann for president and Gavin Schmidt and Ray Pierrehumbert for the US senate.

    303 prokaryotes: I hope you replied that to dotearth.

  6. Jon:

    400: My evidence for the claim that “Most of the people who ‘believe’ in AGW don’t know what it is that they believe in” is purely anecdotal, but it’s the only empirical evidence I have the authority to offer. It’s based on 20+ years of regularly reading journalists and columnists failing to explain ‘the’ science clearly, and creating space for madness; 20+ years of listening to highly educated non-scientists of many shapes and sizes at 1 leading UK independent school, 4 different universities in 3 different countries, and various other theatres along the way, in four different continents, opining, or shrugging they’re shoulders and going with the flow; but above all it’s based on being raised in an academic family, surrounded by climatologists and evironmentalists who taught me to be cautious, and by working in a major university where one can talk to four different scientists at lunch, none of whom might be deniers but /all/ of whom have some very different views on what it is that one should be most cautious about, and what it is that it is safe to say. I would not expect most people to be able to explain how their TV set works. Is it all that outlandish to suggest that the foundations of faith among the non-scientific (you know: /most/ people) may not actually be all that flattering for scientists? Don’t you think this might — just a wee bit — help explain why we are where we are?

    401 (Frank): You say “if you keep using the ‘um, I’m not a scientist, this stuff’s hard to figure out’ excuse, then it starts to look like, you know, an excuse”. You don’t know what I think about the science. I didn’t tell you. What am I excusing? Your example has nothing whatsoever to do with science: common sense suffices for cutting through the storm of bullshit that surrounds us. But that’s precisely the problem. The battles increasingly have nothing to do with the science. It simply is not good enough to sneer tell people to get an education. I can get a pretty good one in my neck of the woods — second to none. And it really is a bit confusing, unless you read the stuff trying to simplify things to make stuff happen in policy-making. What proportion of adult people who have heard of ‘Global warming’ do you /really/ think have approached the matter with a completely open mind, scientifically, from first principles, and thus by the proper approach agreed that the IPCC the best possible account of where we are? Did you? Really? In the same way that you read every single word of every party manifesto before you vote?

    BTW, sorry for the repeated dose of the previous meandering. Accidentally sent prematurely, and then reprised in vexation…

    [edit, moved to open thread]

  7. Ira Glickstein:

    From Ray #300:

    Ira, Have you not heard of a lognormal distribution? When a distribution has a positive skew, there will be more probability to the right of the mode than to the left. A negative skew (e.g. a Weibull with shape parameter >3.6) yields more probability to the left of the mode. The probability distribution for climate sensitivity is strongly skewed right. …

    You, I, and the IPCC agree that, given limits of 2.0ºC to 4.5ºC; 3.0ºC is the most probable value. If we take the exact midpoint between 2.0 and 4.5, we get 3.25ºC. Therefore the peak of the distribution, at 3.0ºC skews to the left of 3.25ºC (i.e., to the left of the midpoint between the limits).

    That result is easy to calculate. LOG10(2.0) = 0.30103 and LOG10(4.5) = 0.653223. Average them and get 0.477121. The inverse log of 0.477121 = 3.0.

    Thus, it appears that values closer to 2.0ºC are slightly more likely than those closer to 4.5ºC, yet you say the probability distribution “is strongly skewed right”, which seems incorrect because the skew is neither to the right nor strong. Please explain.

    Continuing Ray #300:

    Frankly, that is one thing that is very interesting about the latest estimate from Science–although that conclusion seems to derive from their ocean sensitivity estimates. …

    You appear to be referring to a paper in the prestigious journal Science. Hank Roberts (#289) linked to an interview with a co-author that includes the graph I think you are referring to. (It is the third figure down in the link.)

    The Green (Land) curve peaks around 3.5ºC and extends from about 1ºC to 5ºC. The Blue (Ocean) is multi-modal (multiple peaks) with the highest around 2.2ºC and extending from 1ºC to a bit over 3ºC. The Black (Land & Ocean) has two large peaks around 2.2ºC and 2.5ºC, with minor peaks at around 2.9ºC, 1.4ºC, and 1.7ºC. All three curves skew to the right, which supports your contention that higher values are more likely than lower.

    The co-author states CO2 sensitivity “…is “likely” (66% probability) to lie between 1.7 and 2.6 degrees, and “very likely” (90% probability) to lie between 1.4 and 2.8 degrees, with a best estimate of around 2.2 or 2.3 °C.”

    As I said earlier in this thread (#284) “a range of values usually indicates the +/- 1σ (Greek letter sigma, for standard deviation) for that variable.”

    The +/- 1σ range of the normal curve contains about 68% of the area of the curve, which includes the “likely” limits of the range, and the +/- 2σ range contains about 95% of the area, which includes the “very likely” limits. (The co-author of the Science paper is quoted as saying 66% and 90%, close but not identical to +/- 1σ and +/- 2σ, but he may have been misquoted.)

    This result, if correct, tells us that CO2 sensitivity is significantly different over land and ocean and there also seem to be different peaks associated with NH and SH (more land in NH than SH) and perhaps tropics, temperate, and polar regions.

    Continuing Ray #300:

    … And on ethanol, that was never a climate or energy program, but rather a disguised agricultural subsidy. …Comment by Ray Ladbury — 25 Nov 2011 @ 9:35 PM

    On this you are absolutely correct! When government gets involved, the strongest special interests will usually prevail, in this case Big Agriculture.

  8. Hank Roberts:

    Ira Glickstein, from what you wrote just above, it seems you looked at the picture on the page I cited, but didn’t understand the accompanying text.

    Look at the text again — it’s not intuitively obvious what “the ‘fat right tail’ of the climate sensitivity distribution” means, and they assume the reader has some idea what they’re talking about. You don’t yet.

    It’s not intuitively obvious why those various curves appear in “the IPCC’s climate sensitivity range” discussions, or what they mean. Again the authors there assume the reader has a lot of background that’s still ahead of you. You can get there, but lecturing about how it’s done where you came from isn’t helping.

    James Annan’s several posts and papers offer a critique he and others hope will improve the _next_ IPCC Report’s explanation.

    It’s not your normal distribution.

  9. Hank Roberts:

    For Ira G.:

    You won’t understand if you insist words have your personal meanings. Climate sensitivity isn’t a normal distribution. 2xCO2 isn’t twice today’s value. Likely other assumptions you’re making are also inconsistent with the subject you’re trying to discuss.

    Find out what the words mean in the subject you’re trying to learn. It’s the same chore for all of us, the FAQ section linked at top of page helps.

    The terms used don’t mean what you think they mean:

    Application to climate:

    “… you will often see estimates of the uncertainty in T2x expressed as a range between a low value and a high value. For example, the most recent assessment of the International Panel on Climate Change states that the “likely” range is from 2 to 4.5 degrees Celsius. But here’s where things get more complicated and more worrisome: That range is itself not well-defined. More extreme results (outside this range) cannot be ruled out and therefore just focusing on those two numbers can be misleading.

    A more comprehensive way of expressing the uncertainty range in T2x is in terms of its so-called probability function …. The figure below provides a rough illustration of what that plot typically looks like — the larger the value of the plotted probability density for a given value of T2x, the more likely that T2x is….” http://nicholas.duke.edu/thegreengrok/graphics/feedbacks-graph

    Many of us regular nonscientist readers here try to hang on to help new readers find the basics, and to help silent readers watching learn from the questions and assertions others put into the text — many of them repeated over and over.

    Reading furthers.

  10. David B. Benson:

    Ira Glickstein @307 — You’ll need some understanding of probability density functions which are heavy tailed:
    These are not normally distributed about the mean.

  11. Ira Glickstein:

    Hank Roberts #309: Thanks much for the links, particularly the second one about the possibility of a “fat tail” to the right. That helps me understand what Ray was getting at a couple rounds ago.

    Please straighten me out on the effect of a different starting point for T2x. I was under the impression that it applied to any reasonable starting point. In other words, the expected equilibrium temperature rise going from the pre-industrial 270 to 540 ppmv is about the same rise as going from the current 390 to 780 ppmv, all else (natural variations, etc.) being equal. Is that correct, and, if not, why not?

    Also, when I see a multi-modal curve, such as the third figure in the interview with the co-author of the Science paper you linked to, I assume it means that more than one “population” is involved. For example, if someone did a random survey of heights of people, and ended up with a peak at around 3 feet and another one at around 5.5 feet, I would assume the survey was done on a mixed population, such as grade school students and the parents who were picking them up.

    Does that mean that the Science Land & Ocean curve, which has five fairly well-defined peaks, is telling us that T2x varies significantly between Land and Ocean, and within Ocean between the tropics, temperate, and poles? If so, how does that change our predictions as Mauna Loa CO2 levels continue to increase 2 ppmv or more each year?

    I appreciate that each domain of application of statistical analysis has its own lingo, and that my experiences in Avionics systems may lead me to make assumptions that are unjustified in the domain of climate science.

  12. MARodger:

    Jon @306
    I would agree with you that most people take AGW on trust although to say they “…don’t know what they believe in” suggests there trust should be better grounded. Your comment (and @397 on previous thread) seemed to apply this to the general public as (@397) you go on to liken this ‘not knowing’ to the disinterested and the deniers. The general public’s ignorance of AGW is not unique. The public are also very good at ignorance in most sciences and many other areas of quite basic learning. So in terms of the public, I fear your comment is not saying much and your criticism of scientists in this regard is misplaced.

    More narrowly, your comment is perhaps less true but more pertinent. With folk who make AGW their business – activists, educators, the media – there is often more ‘enthusiasm’ than knowledge on offer. I wouldn’t say this applies to “most” of these people. It probably applies to “too many” of them.
    But in their defence these folk have been dealt an awkward card. AGW is about climatology which deals with an exceedingly high level of complexity and an uncertain future. And that is not all. AGW also deals with the present human activities causing AGW, the technology available (or not available) to combat AGW and their impact on our future society/economy, plus the political processes that can make it all happen.

    So if I encounter an activist who totally misinterprets, say, the economic impact of the UK’s Renewable Obligation Certificates I do not lay into them for being out of their depth (although I likely would if they were of the “skeptical” school of thinking). AGW is such a large pond we are all out of our depth somewhere or other so some level of trust, of not actually knowing what we “believe in”, this is required by all of us.
    Of course there are many activists etc who entirely ignore being out of their depth and worse, entirely ignorant that the people they are believing/trusting in are also out of their depth – that is far less forgiveable and that is why I will lay into such folk.

  13. Ray Ladbury:

    Ira, I take it from your post that you don’t have much experience with probability distributions other than Normal. That is not unusual. I am in the process of preparing a lecture on various distributions encountered in radiation and parts engineering.
    Do you remember from probability class the concept of a distribution’s moments. Integral xP(x)dx is the first moment–aka the mean, xbar. The variance measures width and is related to the second moment centered on the mean:
    integral P(x)(x-xbar)^2 dx.

    The skew is related to the third moment centered on the mean and normalized to the variance/standard deviation. It measures which of the distribution’s tails are thicker. Why do we care about thick tails? In a thick tailed distribution, the probability of x goes to zero much more slowly than would a normal distribution. Now if the right tail is thicker than the left, that means the mode of the distribution must be to the left of the mean–and much more of the probability may be on the right side of the distribution than the left. This corresponds to positive skew. The situation is reversed for a negatively skewed distribution–you get the mode to the right of the mean and more of the probability is to the left of the mode. CO2 sensitivity has a pretty strong rightward skew–I think David Benson and I fit a histogram of the estimates one time and got a lognormal standard deviation of about 0.35–sound about right, David?

    The fourth moment is related to the kurtosis, which measures the relative amounts of the distribution in the peak and in the tails–positive means heavy tails (relative to the normal)and negative implies the tails go to zero more rapidly than a normal.

    There is a whole branch of probability that looks at the behavior of the tails of a distribution–Extreme Value Theory. It proves that the extremes either go to zero at some finite value, approach zero exponentially or approach zero as a power law.

    Nerd humor alert. The Cauchy distribution approaches zero approximately as x^-2, so none of the moments exist (e.g. they all diverge). What is the Cauchy distribution’s least favorite question? “Got a moment?” ba-dum-dum

  14. Ray Ladbury:

    Hey Guys, let’s cut Jon some slack. I think he’s trying to play good cop. I don’t think he meant to imply that the average RC denizen doesn’t know why they believe the science–but RC’s clientele is hardly average for the population. A lot of folks may not fully understand the science, but assume that the people who have made it their life’s work to understand climate might. They realize they aren’t experts, so they are at least not subjects for a Dunning-Kruger study.

    On the denialist side, we have mainly Dunning-Kruger superstars and ideologues who reject the science because their ideology doesn’t have an answer for it.

  15. Ray Ladbury:

    On multimodality: The causes of multimodality are complicated. Sometimes, you do have two distinct populations. Sometimes, however, the response of a single population is determined by competing effects. Subtle changes can affect which effect dominates and result in very different (e.g. multimodal) response even though the characteristics of the population are changing continuously.

    I saw this recently in some test data for radiation degradation in some transistors–as long as the collector-to-emitter current was high, you had a single population. However, when the C-E current dipped down into the hundred-microamp range the population split into two (at least) modes.

    The Schmittner study looks interesting. It is clear that the lack of a thick, high-end tail depends on the ocean reconstruction, and that is very complex. The ocean estimate is also where I’d expect the largest errors, since the response of the oceans is much slower than that of land. Since 3/4 of the planet is oceans, it is not surprising that it dominates. However, since the predominant negative feedback in the climate system is thermal IR radiation, it seems odd to me that you’d get drastically different sensitivities

  16. Hank Roberts:

    > different starting point for T2x

    That’s the hypothetical case — where a megalomaniac climate scientist experiments by instantaneously doubling the amount of CO2 in the atmosphere, holding all else the same, and then waits out the resulting centuries of temperature change to see where the resulting temperature rise levels off.

    So you’re beginning to question your assumptions rather than state them as facts, and to realize you’ve been asking FAQs here while lecturing elsewhere based on mistaken assumptions.

    This isn’t unique to you. Most of the folks who answer the basic FAQs here over and over aren’t working climate scientists, we’re readers who got here a few years before you did. For myself a lot of what I knew was wrong, and a lot of the science is beyond me, and so I try to find simple clear words to paraphrase what I read, and point to sources, and keep up with the research abstracts at least, and try to remember that new research doesn’t overthrow, it extends (maybe) what we know.

    That’s what the IPCC is doing to, on a somewhat larger and more professional scale, eh?

    You’ve picked one of the big questions. The more detailed the discussion the more you need a page of your own (by custom this November ‘Unforced Variations’ will roll over to a new one for December after a while).

    If you start reading and checking your assumptions it’ll take you a while — took me several years and I doubt I wasn’t particularly slow learning (and unlearning what I thought I knew).

    You won’t find me trying to explain probability and climate sensitivity; at most I can point to writers I’ve learned to trust by verifying what I could follow and trusting others’ trust where I couldn’t.

    Again, I suggest looking at the basics.
    Read this: http://www.ipcc.ch/publications_and_data/ar4/wg1/en/ch10s10-5.html

    Read James Annan’s blog on the subject and his papers and discussion criticizing that same chapter and think about how it ought to be written for the AR5.

    Read topics and FAQs here — use the “Start Here” button and the search box.

    Threads like this are good for answering wrong assumptions people come in stating. Many of us have typed much of this repeatedly. Once you know to check your assumptions, and how to find the published work, you can go beyond these open threads, I’d suggest, and ask better-informed questions.

    You are getting answers from real working scientists here; pay attention to their answers. Me, I’m just some guy pointing to the FAQs and sources.

  17. adelady:

    MARodger. The arguments about activists and educators not being fully up to speed on climate science apply equally to all general advocates and educators on items like public health, for example. We don’t expect Ms Jones from the Local Health Group who goes around doing presentations on heart health or allergies or skin cancer or managing diabetes to be fully trained as an endocrinologist or allergist or dermatologist.

    So we shouldn’t expect that an environmentalist or a solar power advocate or a sustainable gardens presenter needs advanced qualifications in the various sciences underlying the material presented.

    It all comes down to trusting expertise. Unfortunately, to trust expertise you first have to acknowledge it. This is where people like the anti-vaccination crowd and climate/ acid rain/ ozone depletion deniers come in. They balk at the very first hurdle.

    They come up with all sorts of rationalisations and excuses, but that’s what it boils down to. They can’t face acknowledging that someone, many someones, may be qualified by virtue of training, experience and proven performance to have their statements taken as an accurate description of facts. When you realise that some people refuse to accept that deaths from whooping cough occur (and increase) in areas where vaccination rates fall, you have to face it that some people are willing to suspend all claims to rational thought when facts contradict their ideological preconceptions.

    And a health advisor doesn’t need to be an epidemiologist or a statistician to present the bald proposition that decreasing vaccination rates will certainly result in illness, disability and death for some children. They only need to cite the figures prepared by others.

    Same thing goes for people presenting information on climate issues. They may have a particular point to make about effects on crops or animals, or on the virtues of various renewable power technologies, or even on political options for taxes or fee and dividend or whatever. They don’t need to be climatologists or economists or ecologists or agriculture experts themselves to present valuable information to others. They just need to collect the best information they can from the real experts.

  18. Hank Roberts:

    PS for Ira, when there’s no FAQ, I try to ask a question in a likely place, in a way interesting enough that someone knowledgeable may be tempted to take time.

    For example that double peak in Fig. 3 you ask about above. I’d thought I understood it, but your posting above made me wonder again. So I asked (the old Usenet way, post what you think and await correction).

    I asked at one of the sources:

    I will await correction ‘over there’ on that. It’d be a digression here.

  19. Hank Roberts:

    PPS for Ira, re where to ask more questions:
    Nathan Urban wrote at the above-linked interview page:

    “… I don’t want to give a tutorial on climate sensitivity here; for that, see my earlier interview on the Azimuth Project and Planet 3.0 blogs.  In particular, I recommend reading the discussion of “feedback effects”, which can combine with greenhouse or other warming to cause more (or less) warming than the greenhouse effect alone.”

  20. Radge Havers:

    adelady @ 317

    “It all comes down to trusting expertise. Unfortunately, to trust expertise you first have to acknowledge it. This is where people like the anti-vaccination crowd and climate/ acid rain/ ozone depletion deniers come in. They balk at the very first hurdle.”

    Which is where a little due diligence comes in. It would be more problematic if there were rational grounds for controversy. But in these cases some simple fact checking and asking some hard questions about the logic of the rhetoric, should fairly quickly lead one at least into the ballpark of where the more reliable professionals are.

    Unfortunately too many people have a knee jerk reaction to “Authority”. So ironically they defer to authorities who puff and blow about being mavericky outsiders, down to earth real people who challenge the innate meanness of authority driven institutional bureaucracy (which in pop culture is always evil, so you instantly know who to root for and who the black hats are). It’s the ultimate short cut and therefore very attractive.

  21. dallas:

    Ray Ladbury said, “However, since the predominant negative feedback in the climate system is thermal IR radiation, it seems odd to me that you’d get drastically different sensitivities.”

    Why would it be odd? CO2 can only return a portion of the energy it absorbs and only return in its spectrum. The energy model that Schmittner appears to be using accounts for the available energy, which provides the energy for CO2 return. The energy available in the ocean for CO2 return varies much less that land. The Northern hemisphere with its greater potential change in albedo and a Gulf Stream energy would have a greater temperature response than the Antarctic which is much more stable. CO2 can only return energy, not manufacture it.

  22. Chris Colose:


    with the big caveat that I haven’t actually read the paper yet (I have it in request, for some reason my university access won’t give it to me)…

    I don’t think the land-ocean diagram (figure 3 of the Nathan Urban interview) means different climate sensitivities of land vs. ocean; rather it means different global climate sensitivities implied when you use land or ocean data (and thus, the two curves really should overlap much more to be believable). The whole point of a lot of these perturbed physics ensembles is to generate a simulation with a lot of (usually subjectively chosen) parameter variation (that affects things like clouds, etc) and thus the ensemble is meant to contain many members which probe a large range of uncertainty. Then, observations (of the LGM or industrial climate) are used to constrain which subset of those models are allowable.

    In some studies (e.g., Stainforth et al., 2005) you get a distribution that is very wide (accepting sensitivities as high as ~11 C) indicating that they probably admit members a bit too easily.

    This type of analysis is different than some of the studies that try to use just observational data (e.g., some of James Hansens papers, where you try to get an estimate of dT and the forcing, then take the ratio). The problem with these is that you need to assume a constant climate sensitivity between base states when trying to extrapolate the results to a 2xCO2 world. The ensemble method gets around this but then assumes that the change in feedback between two states is well simulated by the model.

  23. Ray Ladbury:

    Captain Dallas,
    I think my reservation stems from the seeming inconsistency of saying that land and ocean sensitivities are sufficiently different to merit different pdfs but then saying you can recombine them to get a meaningful pdf for overall planetary sensitivity. If in fact the oceans (75% of the planet)warm significantly less than the land (25% of the planet), then won’t the majority of the increased IR radiation (and warming) be coming from the land?

    Another issue I have with it is that we know that one sure way to get a low sensitivity is to assume a small time to re-equilibration (a la Schneider). We know that the time for the oceans to equilibrate is far longer than the atmosphere/land, and exactly what that time is is less certain than for land. Could we be seeing an underestimate of the time to equilibrium.

    The thing is that either effect would tend to force land areas to warm much more than oceans–and effectively, if you look at their land-only pdf, this is what we see. There’s a whole helluva lot more probability above 3 degrees per doubling than below. For land, where most of us live, the high thick tail is still there.

    I think that it is a mistake to sell this study as diminishing concern. It either says that land will warm much more than oceans permanently, or that it will warm much more and then the land temperature will decrease as the oceans catch up (assuming they don’t belch out a bunch more CO2 in the process).

  24. Joe Cushley:

    Hey, Dallas, comic and scientific genius that he is, is toying with us people…


  25. thingsbreak:

    @332 Chris Colose:

    I have it in request, for some reason my university access won’t give it to me

    Check your email.

  26. Pete Dunkelberg:

    Chris Colose, google on the title (it does not show up in scholar yet)


    for instance. The supplementary material may also be needed. Who has a good link for that? I think you’ll also be interested in

  27. Hank Roberts:

    for Pete D. — the supplementary material is not paywalled, even though the main paper is. This seems usually how it’s done:

  28. Hank Roberts:

    See also this note:

    “Nathan Urban says: November 26, 2011 at 5:19 pm

    Another correction, or at least clarification:

    In our study, “climate sensitivity” is a global average quantity, and we don’t consider separate land or ocean sensitivities.

    The green curve in the figure shows the global climate sensitivity if we analyze only land temperature data. This is different from “climate sensitivity on land”.

    However, the land and ocean do warm by different amounts in response to doubled CO2, so you could define separate land and ocean sensitivities.

    Ok, I don’t know if this is actually clearer, but the point is that we’re estimating a global sensitivity using different subsets of data, rather than estimating local sensitivities that apply to different parts of the Earth.”

    The main topic there (the Azimuth blog) has a good ongoing collection of pointers to other conversations about the same paper:

  29. Ray Ladbury:

    Joe Cushley, I’m sure the two people who read Capt. Dallas’s blog will enjoy the yuks.

    Wow, the dude doesn’t even understand the contributions of Arrhenius–what a bonehead.

  30. Ira Glickstein:

    From Hank Roberts #316

    > different starting point for T2x …

    What I expected, certainly naively, was something like:

    T2x(270) is likely between 2-4.5ºC, so T2x(390) is likely to be in the same range +/- 0.5ºC or so.

    I’ve looked through the links you and others provided, and read some through, and now I feel like the guy who asked a clockmaker “What time is it?” and received a detailed explanation of “How to build a clock” – and I still don’t know what time it is!

    I know the climate system is extraordinarily complex, way beyond what I ever dealt with as an Avionics System Engineer, but I suspect that global warming, while it needs to be informed by exacting climate science, is, to the larger human society, more of a system engineering problem. (Yeah, give a kid a hammer and everything starts to look like a nail, so my engineering background makes everything look like a system :^) but hear me out, please.

    A system engineering team has to come up with alternative system concepts and design implementations, estimate performance, cost, and schedule, and, to win the contract, propose a comprehensive solution the customer set judges to offer the most favorable cost/benefit.
    Accepting the IPCC predictions, we (humanity – as represented by national governments and international groups like the IPCC, World Bank, etc.) have a limited range of alternatives for dealing with past, present, and coming warming and the consequences that flow therefrom:

    1. Do nothing and muddle through as best we can, adapting to the floods, droughts, storms, crop failures, and other resultant disruptions.

    2. Hope that #1 will be ameliorated as cheap sources of carbon fuels deplete, causing prices to rise and self-interest to lead humanity to avoid wasting energy, being more efficient, and adopting alternatives that are carbon-neutral (biomass) and/or carbon-free (water, nuclear, wind, solar, …)

    3. Shame industries and people into voluntarily adopting green lifestyles.

    4. Speed adoption of #2 via a Hansen/Krauthammer (and Ira :^) carbon tax at the mine, well, and port with 100% of the revenues distributed on an equal per-capita basis to citizens and legal residents. This will make carbon-neutral/free energy relatively less expensive, so smart industries and people will adopt them for their own selfish interests.

    5. Mandate drastic, international cap & trade on carbon energy and provide taxpayer-funded R&D, subsidies and loan guarantees to favored alternative energy industries, until carbon usage declines.

    #1 will inevitably cause millions of people to die (along with other animals). But those of us with money and guns will adapt. #2 may ease the pain somewhat. As for #3, yeah, right.

    #4 is politically difficult because many of us think the money will be kept by governments and that powerfully connected industries will be granted exceptions. And, what are the chances that BRICs (Brazil, Russia, India, China) and other rapidly modernizing countries will agree?

    #5, even if it was politically possible on an international scale, which it certainly is not, may not be technically feasible without, at least in the short-term destroying modern, fuel-dependent industry and society as we know it.

    So, what will it be?

    [Response: This really isn’t the right blog to be discussing policy options. That’s not what we’re experts in. My two cents here is that this is a politically vexing issue but that there are plenty of people with suggestions about how to do something along the lines of 4, 5, but without doing anything ‘drastic’. And indeed you are forgetting 6 which is to invest in innovation and 7 which is to stop drastically subsidizing the fossil fuel industry. Suggesting as you do that our only options are in effect to ‘do nothing’ or to ‘end society as we know it’ are, well, alarmist. Vexing problems are vexing. That’s no reason to either deny them or ignore them. –eric]

  31. Ray Ladbury:

    Ira, to a first approximation, a doubling from 270 to 540 will raise temperatures by ~3 degrees–assuming we started at equilibrium at 270 ppmv. Likewise 390 ppmv.

    As to the system-engineering approach–first you have to get people to accept that there is a problem that needs fixing.

    I am not as sanguine as you that guns and money will carry the fortunate through what is to come if we do nothing.

    As to the BRIC response–they stand to lose as much or more as we do if we wreck the climate as human population reaches its crest. What is more, these economies are strongly export driven–tarriffs of the threat thereof could be strongly persuasive.

    If we had started to address the unsustainability of global human civilization back in the ’70s when it became glaringly obvious, we might have had a chance with your option 3) + Gavin’s options 6) and 7). As it stands now, it will have to be a balls-out race to develop and deploy new technology before the response of the planet renders all our actions moot. If we rise to the challenge, we could be in for some very exciting times, and a reasonable probability–though not certainty–of heading off the worst outcomes.

    Unfortunately, I still see more people interested in debating the validity of science that has been accepted since before Darwin, Maxwel and Boltzmann. That is not a hopeful sign.

  32. Chris Colose:

    Thanks for the link (326) and e-mail (325)

    My original interpretation was correct; the authors are not looking at two different sensitivities (land or ocean), which made little sense when I seen some people talking about that. I haven’t read any of the blog commentary apart from the Nathan Urban interview, so I apologize if any of this is repeating itself:

    The type of analysis done by this paper has already been done before (e.g., Schneider von Deimling et al., 2006, Climate sensitivity estimated from ensemble simulations of glacial climate , Climate Dynamics). It appears to me that the main motivation for publishing this one was based on using a different “observational” dataset, one that people seem initially skeptical of due to the fact it allows for seemingly bizarre low amounts of temperature change between the modern and LGM (closer to 2-3 K cooling than 5-6 K that most people believe). In fact, the “best estimate” of the Schmittner paper still underestimates cooling in many regions, notably in Antarctica and in various mid-latitude regions.

    The climate sensitivity results are also not robust to the data one believes the most. The problem is in essence similar to the deep-time greenhouse climate issue of the very small pole-to-equator temperature gradient. One of the big issues here is that if you force a model with a lot of CO2, you warm the tropics and the poles, but you don’t warm the poles enough to match observations. If you put more CO2 in, then the poles get warm enough, but then the tropics are too warm (some improvements in deep-time proxies have made the tropics warmer, but this is still an open issue).

    Somewhat similarly, the idea of this study is to see which sensitivity (out of a wide ensemble of possible sensitivities, based on spanning a large parameter-space) will best generate agreement with observations. Thus, for example, if you “allow” a model with a very high sensitivity, then the agreement is very poor since the changed boundary conditions then results in something similar to a snowball earth (which is clearly not in agreement with LGM observations). If you use a sensitivity too low, then the LGM is not cold enough to match observations.

    The problem in the Schmittner paper is that there are really no “good” agreements with observations- even if you match observations globally with your model, you don’t match a lot of the spatial variation (this is a problem of a simple model, but the observations are incomplete and almost certainly problematic).

    Of course, how the boundary conditions during the LGM (relative to modern) were different is also not well known. We can estimate GHG concentration changes well, but other things (like aerosol and vegetation change) is not well known. Even things like the correct height of the ice sheet have impacts on local temperature, because you need to correct for the fact that an extra mile high ice sheet might give you 10 C local cooling because of adiabatic cooling of air being advected over the region. (Note that the presence of ice sheets is not typically considered an ice-albedo feedback, but rather an imposed boundary condition, and is thus the primary forcing- along with GHGs- driving the LGM to modern temperature difference. Hansen’s preferred method of using observations does the same thing.)

  33. Chris Colose:

    In my above post, I also forgot to touch on using these studies as a template for the future. It’s also a good idea to keep in mind how they define what is a forcing and feedback. The Schmittner paper for example defines vegetation changes as an internal feedback, and thus they don’t need to worry about the uncertainty in surface (vegetation) albedo forcing, that other papers do (like Kohler et al 2010). It’s not obvious whether the change in feedbacks from one climate state are easily translatable to a future CO2 world though

  34. thingsbreak:

    @332 Chris Colose

    The problem is in essence similar to the deep-time greenhouse climate issue of the very small pole-to-equator temperature gradient. One of the big issues here is that if you force a model with a lot of CO2, you warm the tropics and the poles, but you don’t warm the poles enough to match observations. If you put more CO2 in, then the poles get warm enough, but then the tropics are too warm (some improvements in deep-time proxies have made the tropics warmer, but this is still an open issue).

    Jeff Kiehl (via Lee Kump) has also identified a possible solution to this- that models can actually produce realistic tropics-to-pole temperature gradients if they are run using non-modern aerosol loading (warning, link to audio file).

    [Response: This whole ‘can’t capture the gradient’ stuff is somewhat dead as a research topic now, because the latest evidence says that the tropics actually did warm a lot, and getting the gradients to match the observations is just not that out of the ordinary for mid-range sensitivity GCMs. See the paper here (open access): The early Eocene equable climate problem revisited by Huber and Caballeroa (2011) in Climate of the Past.–eric]

  35. Pete Dunkelberg:

    > the tropics actually did warm a lot

    For instance Williams and Funk 2010 A westward extension of the warm pool leads to a westward extension of the Walker circulation, drying eastern Africa. The Indian Ocean warmed more than had been expected, then higher evaporation led to plenty rain which falls back into the sea. North east Africa gets the hot dry air after that so it is hot there as well.

  36. Pete Dunkelberg:

    The Horn of Africa may be about to get heavy rain from Tropical Cyclone Five (Arabian Sea).

  37. Chris Colose:

    Dr. Steig- thanks for the citation! I hadn’t seen that paper before, but it looks like the sort of thing that will become a standard reference. I will need to read it this week. I wish I had more time to follow up on all the latest developments in that field, it’s very neat.

    Thanks thingsbreak– I did a blog post on the Kump and Pollard paper a few years ago, which talked about aerosol responses that could help “the problem” (if there is one anymore). That paper, at least, relates to the Cretaceous, though there’s no reason it couldn’t apply to the Eocene too, and I’ve seen it cited for more general application. There is, of course, a lot of interesting papers that discuss these types of theoretical and “plausible-but-unconstrained” feedback effects that might be “missing from models.” Some of this might affect the relative pole to equatorial differential. Kerry Emanuel has some stuff on how hurricanes might act as a sort of “tropical thermostat” in a deep-time application, but a lot these things are no better than good hypotheses. I find them very interesting to think about though, even if they prove to be impossible to test.

    The people who develop proxies are bright though, and some of the stuff they think of dazzles me, so I wouldn’t discount much as being untestable. I still think there’s a lot of interesting questions related to the pole-to-equator temperature gradient. Of course, just because you can get a model to work doesn’t mean you understand the physics behind why it works and that is where I still think there might be cool research to be done.

  38. MARodger:

    Dallas @321 might have his little joke (@324 & @329) but I don’t think future generations will be laughing when AGW begins to bite. Of course, skeptics are also having a good chuckle at present in the aftermath of the second batch of the leaked e-mails now two years old.
    Strangely no sign from skeptics about the freshly leaked UEA e-mails. Apparently they aren’t as serious (not by a long chalk) but some type of gag notice is preventing their mention in the media. The link below gives the info that can be published. I think this account is going a bit too far but I’m told that does strengthen their legal position.

    Info on fresh leaks

    Legal details

  39. Steve Fish:

    MARodger. Because CaptDallas’ site still states “Join this site” and that “There are no members yet. Be the first!” his chuckle is probably getting a bit strained.

    Re the resignation of Carbon Dioxide, I am very disappointed with the photo because I thought he was older and more dignified. Steve

  40. Edward Greisch:

    Good Science Always Has Political Ramifications


    Book: “FOOL ME TWICE; Fighting the Assault on Science in America” by Shawn Lawrence Otto; Rodale Books

    “When speaking about science to scientists, there is one thing that can be said that will almost always raise their indignation, and that is that science is inherently political and that the practice of science is a political act. Science, they will respond, has nothing to do with politics. But is that true?”

    …..article continues…..

    “But beyond that, science constantly disrupts hierarchical power structures and vested interests in a long drive to give knowledge, and thus power, to the individual, and that process is also political.”
    …..article continues…..

  41. wayne davidson:

    ——–The latest war cry propaganda masterpiece from skeptics;
    “no warming since 2002″……….
    …….. really? Are you sure? Peek at what talking about cooling looks like: http://eh2r.blogspot.com/.
    I fear the warming is faster than the time we have to convince some action against it.

    High Arctic small glaciers are melting at an unimaginable pace, at a place with no industries, under a quiescent relatively inactive sun, where its sunless, in deep darkness for 3 months with no proximate heat from any warm ocean.

    Hank is right about Google, is a wonderful tool.