15 September 2009

Are You a Climate Skeptic?!

Further evidence that climate science is deeply troubled. In a conversation on the difficulties of sustaining intensity of public opinion on climate policy, I cited the Knight et al. BAMS piece in conversation last night in a small group of professional colleagues, mentioning that it said that global temperatures had not increased in a decade, and that this sort of thing worked against sustaining intense public support. It seemed like a fairly obvious point, and uncontroversial.

However, one colleague, a scientist working at the climate science-policy interface who I met for the first time this week, looked at me incredulously when I mentioned the 10-year period of no warming and asked, "What are you a climate skeptic?!"

I said, "No I'm not, I'm just citing a paper in BAMS I saw today."

He accused, "You said ten years; That is selective."

I replied, "Well, that is just what the paper says. I'm just explaining what it said."

He again accused, "No, you said 10 years!"

Um, OK.

When stating facts reported in peer reviewed journals that are in articles justifying continued concern about warming results in being called a "climate skeptic" by a colleague, then you know that the community has some deep issues.


  1. A colleague of mine who works in renewable energy said that he used to be a climate skeptic, but had changed his mind because of the recent evidence that climate change was occurring far more rapidly than previously predicted.

    I asked whether he was thinking of global temperature (which was below long term trend) or arctic ice (which was recovering from the 2007 low) or sea level rise (which was following the long term trend or just possibly flattening off), as none of these major indicators seemed to show any signs of acceleration. He accused me of cherry picking recent years: the fact that he had picked these years as evidencing rapid change didn't seem to concern him.

    It's a funny old world in climate science.

  2. Yes. It seems perfectly valid to extrapolate from an unverifiable anecdotal account of something said by an un-named person to "the community". I can't see any problems with that at all.

  3. -2-RW

    You are right to imply that it is not a statistically robust sample. However, this sort of thing occurs enough in my interactions in this community that I find to be troubling. Perhaps your experiences are different.

  4. RW,

    A number of scientists have said that there is a pervasive groupthink mentality in the scientific community on this issue. Some have even said that they were not able to speak freely until after they retired because so much research money is tied to global warming. In an atmosphere where expressing skepticism is career suicide, Roger's anecdote is pretty mild.

    "What are you -- a climate skeptic?" is precisely the kind of ad hom comment that should raise eyebrows in an academic community.

  5. Btw, since Roger was present as a witness (in fact, a participant), what do you mean by "unverifiable"? Are you questioning his veracity?

  6. "an unverifiable anecdotal account of something said by an un-named person"

    Nice way to call the man a liar. I can't see any problem with that at all.

    It certainly fits the pattern of ad hominem-first attacks when any true-believer statements are questioned.

  7. Roger

    Do you think the pressure of the public/political spotlight is responsible for the 'issues' ? Did you perhaps feel that you were being suspected of being a 'traitor' ?

  8. Do you mean Black et al? The paper linked seems to be Knight et al.

  9. I was charitably taking the view that RW was referring to my anecdote, not Roger's. As nobody knows who I am , it is indeed unverifiable.

  10. That's the scientific community's idea of an insult? That's like a Church-goer hearing you saying "I don't know if god exists" and responding "What are, an atheist?"

    As an agnostic, this has happened to me...

    Although that would actually be a little insulting.

  11. Hi Roger. I admire your honest approach to these matters. I'm not sure why you act as if you are allergic to the "skeptic" label. I don't see why anyone would regard it as a pejorative term in relation to most fields of science, and certainly not in a field experiencing as much flux as climate science.

    As I read your stuff you come across as what I would call a garden-variety skeptic; you are skeptical of what alarmists pitch, and also of what the so-called "deniers" pitch. I would say that positions you quite well, and you should wear the label quite proudly. It's one of the reasons I give your posts the time of day even though I'm not on board with some of your thoughts on policy and I fall in a different spot on the spectrum regarding what current science has to say about the effects of CO2.

    If I'm misunderstanding you (again) feel free to correct me (again).

    I'm concerned about your statement that this 10-year trend is not a "statistically robust sample". Not sure what you mean by that. A decade-long trend is not a "sample" of current trends. Surely "current" means no more than what is happening, trend-wise, in the most recent 10 years or so.

    In statistics we distinguish between the notion of a "poll" and that of a "census". A poll samples a subset of a population and infers something about the population. A census, in contrast, collects numbers from the entire population and reports them. While a poll estimates population behavior, a census is the population behavior.

    If it is an objective fact that the earth has cooled (by a particular measure) over the last decade, and if by "current" we mean a time-fram subsumed by this decade, then it is tautological that the earth is "currently cooling" (by this measure) -- this is census data, not poll data. There is no question of robustness -- a concept that only applies to inference from poll data.

    Another meaning for "robustness" concerns the sensitivity of the data to the choice of initial and terminal points. You must know that when one reports a "trend" over a time period in this field it is not the slope of a secant line (which is highly sensitive to choice of initial and terminal points). It is generally a least-squares or similarly obtained regression line. Such regressions are about as robust as possible: change the initial/terminal points by a year or two, and the resulting slope will not change by very much, which is what I take you to mean by "robust". Make it 12 years or make it 8, or anything between, and the conclusion remains the same. That's robustness!

    If you are referring to long-term trends, then of course a 10-year census is not a "robust" sample in the loose sense of being non-arbitrary or non-representative. But neither is any contiguous 20 or 30-year census, particularly if we wish to infer 100 year or longer trends. Indeed, the actual time scales we should be dealing with are millennial at minimum, and possibly geologic. On such scales no climate trends from the 20th century look at all out of place. It is precisely the "non-robust" selection of a tiny decade-scale period in late 20th century, with appallingly inaccurate or outright false comparisons to centuries-long "proxy data" that artificially begs an alarming conclusion.

    The most recent 10 year data is a perfectly "robust" and appropriate counterpoint, as it demonstrates how seriously flawed the GCMs relied on by the IPCC are, with CO2 continuing to rise while temperature falls, and with the complete failure of the tropical "hot spot", that is the model-inferred signature of the hypothesized CO2-based global warming, to appear.

    Scientific theories (or models) cannot be proven by statistical correlation, no matter how robust. But they can be falsified by robust data that contradicts their predictions. 10 years of cooling and failure of the hot spot to appear gives, in my mind, ample "robustness" to justify rejection of these models.

  12. 1o years selective? How about 150 years on the scale of 10.000 years of climate changes?

  13. After spending some time today trying to find a current working definition of the term “climate skeptic” it seems that it’s a changing, inexact term. The term is mostly applied to people who are skeptical of the science justifying believe in AGW. With that usage Roger is not a climate skeptic.

    However, it seems that as the global warming debate gets more rancorous the term “climate skeptic” is being used as a pejorative to be fired at anyone who disagrees with the true believers. By that measure it fits Roger with his contrarian view of cap and trade.

    Sometimes it’s just too hard to deal with conflicting viewpoints. For example; since I failed to understand how Roger could be called a climate skeptic I just concluded that the person throwing the pejorative was a naïve, unthinking believer in the church of Global Warming. I could be wrong but now I can stop thinking about it. ‘-)

  14. Imagine what this conversation could be like in 2012 if the sun remains quiet...


    ...and the years 2010 to 2012 are actually cooler than 2009.

  15. FWIW, scenario B in Hansen's 1988 testimony has a ~12 year period between 1970 and 1982 where the temperature remains flat and it's pretty damn flat between 2010 and 2020.


  16. I actually wrote about this issue in my blog in July. Here is the conclusion part:

    If all scientists are skeptics, then why would skeptic be an insult? The obvious reason is that the people who use the term are not skeptics, but "climate believers". These are people who have accepted the meme of AGW without the skeptical science that created the hypothesis. For the believing mind, skepticism is not part of their mental outlook. Once something is incorporated in their belief system, questioning it, testing it, trying alternative explanations are not normal scientific inquiry, but heresy to be punished.


  17. Roger,

    What is the citation for that BAMS paper? I'm finding nothing in BAMS with Knight in the author list for 2009.

    As a climate scientist myself I share the concern over the use of "skeptic" as a pejorative.

  18. -18-WillH

    It is part of this online supplement


  19. Roger - my experience certainly is different. I am an astronomer. I work in a department where a lot of people work on atmospheric physics. One of the atmospheric physicists is giving a seminar here in a few weeks called "A Global Warming Sceptic's Case". I'm sure everyone will be very happy to listen to what he has to say.

    On a point of substance, obviously I don't know what your colleague was actually arguing. But is the 1999-2008 trend in global temperatures statistically different from the 1979-2008 trend?

  20. Eli
    Back-projections before 1988 don't count since that was the reality-based tuning constraint. And if the clear rising trend of scenario B (CO2 linear increase) can be described as "pretty flat" then you'll have no trouble describing reality up to 2009 as very flat, since it follows scenario C (no CO2 increase beyond 2000).

  21. Roger,

    That’s of course too quickly of a judgement, though I don’t find the reaction entirely surprising either, in the context of the popular debate about climate change. Not offering an excuse here, but merely a possible explanation.

    The ‘global warming has stopped since 1998’ canard (Lucia?) is so often used as a pretext for not addressing the long term change in climate, that scientists have understandably grown very wary of even the mentioning of these short term trends. 99 out of a 100 times this is done in order to bash ‘AGW’. I could equally imagine an evolutionary biologist getting defensive when missing fossils are brought up; good chance that it’s a creationist talking. Better not to start accusing someone immediately, but with so many creationists/”skeptics” around, it’s at least understandable (though not excusable).

    And a possible analogy:

    Over at the Examiner, Thomas Fuller had a post outlining a ‘new generation of skeptical arguments against the theory of anthropogenic global warming’ (AGW), which he felt had a lot of merit. I got to his post via a comment thread at RealClimate, where he asked for input. He did get quite a lot of feedback, but unfortunately, a lot of it was packaged in a rather negative tone. However, he framed his questions as “skeptical arguments advanced against the theory of anthropogenic global warming”. He also acknowledged that scientists are getting frustrated “answering the same ‘primitive’ objections repeatedly, only to see them resurface shortly thereafter, something that I am sure is frustrating.” I think a logical consequence is that his framing of the topic aroused a defensive reaction from supporters of the scientific consensus.

    I recount this event here: http://ourchangingclimate.wordpress.com/2009/07/12/next-generation-questions/


  22. Bart/ ourchanging --
    Talk about carnards; Three in one!
    * I have always cautioned anyone from making any comparison starting in 1998.
    * I haven't ever said global warming stopped at all.
    * I also have never said we should not address long term changes in climate.

    I've consistently told my readers that global warming is real both based on the physics and based on statistics etc. I'm for promoting alternate energy for a variety of reasons including limiting carbon emissions. I specifically want to see nuclear including in the mix.

    What I have said I've said the AR4 multi-model mean overpredicts warming while always pointing out the data show warming clealry exists and has not "ended". I have said the second many, many times.

    However, for some reason, some people (you?) seem to want to translate any suggestion that some specific set of model runs might be a tad above the actual trend as being the same as saying there is no warming. Then you use your mis-representation of what I have said as an excuse to mis-represent or over react to what other people say.

    Frankly, I think the tendency of some scientists (you?) to provide examples of precisely the behavior Roger discusses in the blog post tends to make readers believe Roger's reports that these reactions are common enough.

  23. "The ‘global warming has stopped since 1998’ canard (Lucia?) is so often used as a pretext for not addressing the long term change in climate,..."

    Hypothetically, how long into the future would the world have to go without a year that was, say, 0.1 degree Celsius warmer than 1998, before you would agree that "global warming has stopped"?

    To 2013? 2018? 2023? 2028? Longer?

  24. Lucia,

    Me mentioning your name was purely a tongue in cheek comment regarding me using the word "canard", which you pointed out to me I had used in the wrong way previously. It had nothing whatsoever to do with what I think your views are regarding the (future) state of the warming. No need for the defensive reaction (which happens indeed to be the red line of this blog post as I read it).


    There are several papers out that point to the normalcy of 10 year periods that show no warming, without the underlying warming trend actually having changed. I'm not aware of the same conclusion to hold for 15-20 year time periods, so my guess is that in such a time frame, warming will resume. If not, we'll have to closely examine the reasons why it hasn't.


  25. Yourclimate: "Several Papers" presumably refers to Easterling and Wehner (ONE paper!) which said that some models in some scenarios don't warm sometimes for ten year periods at a time and the rest of the stuff you said. The problem of course is that models may have something interesting to say reality is much more pertinent. So Easterling and Wehner looked at that to. And they found a couple of periods like the recent period, and concluded that such things can happen in the middle of substantial warming. BUT all the periods in the actual data were associated with volcanic eruptions, so there is no evidence that such periods on the way to warming are "normal" at all in the real world.

    Now ask yourself, if there is an extended period of no warming-that is, if the recent situation continues for a few more years, does that not at the very least mean that projections of much more rapid warming are far to pessimistic? No? Yes? I can't seem to get anyone to be clear on this and I figure you think you know quite a bit.

    By the way, in the troposphere, the period of no warming is actually in excess of twelve years now, not ten.

  26. Bart/Ourchanging--
    Ahh.. Ok. Even on re-reading, it still seems to suggest that I advocate the things in that sentence. But, I can now see where that was not your intention.

    People accusing me of advocating those things or wishing to block action on warming is not unprecedented on the web. So.. yea.. I read it that way.

    I apologize for my misunderstanding that. You were using the word "canard" correctly this time. :)

  27. Bart,

    You write, "There are several papers out that point to the normalcy of 10 year periods that show no warming, without the underlying warming trend actually having changed. I'm not aware of the same conclusion to hold for 15-20 year time periods..."

    Are you counting the Knight et al. BAMS piece among those papers?

    It's reasonable that you would, because the Knight et al. BAMS piece says this:

    "Near-zero and even negative trends are com­mon for intervals of a decade or less in the simulations, due to the model’s internal climate variability."

    "The simulations rule out (at the 95% level) zero trends for intervals of 15 yr or more, suggesting that an observed absence of warming of this duration is needed to create a discrepancy with the expected present-day warming rate."

    But notice how they say that the simulations rule out (at a 95% level) zero trends for 15 years or more, but they make it seem like 10 years is nothing special.

    So what about 11 years? 12 years? 13 years?

    I think their "10 years or less...are common" is really just spin. They simply don't wish to acknowledge that even a 10-year interval of non-warming creates a greater-than-50-percent likelihood of a discrepancy with predicted warming rates.

    P.S. Notice they write periods of "a decade OR LESS" (emphasis added) are common. Good spin!

  28. That BAMS Knight paper... there really not much there there, is there? It's pretty short. But based on what the paper says:

    * after the temperature trends were observed, they collected together 10 model runs every one of which had a lower long term trend than the average from IPCC A1B runs, then found that that group of lower warming runs exhibits some number of zero or negative 10 year trends.

    Things not revealed to the reader:
    1) How that group of 10 models hindcasts.
    2) The variability of "weather noise" in that group of 10 model runs relative to the earth's variablity
    3) What the comparison would revela if they did not correct for ENSO. (This actually matters-- depending on how well ENSO explains anything in the models. Maybe model ENSO's aren't very explanatory? We don't know unless it's shown or discussed.)

    I'm sure I could think of more. But, quite honestly, I don't think Knight is an example of a paper that shows 10 year periods with zero trends were "expected". As it stands, I don't even think they showed readers of the paper that 10 year periods with zero trends would be expected based on any reasonable standard other than, possibly, "EMERGENCY. WE NEED A PAPER THAT SAYS THIS!"

  29. jg,

    FWIW, Hansen's scenerio B has a linear increase in the CO2 FORCING, not the CO2 mixing ratio (that is exponential, see down at the bottom of the link).

    The global temperature rises in Scenerio B rises from about 1980 to 2010 and then goes pretty much flat for ten years (see the first figure in the link). Moreover the 1970-1980 period in Scenerio B, which Hansen considered the most likely, was somewhat flatter than the measured trend.

    So yeah.

  30. Eli--
    Yes. Hansen's projections published in 1988, did not do a very good jop hindcasting the dip due to the eruption of Fuego during the 70s, and consequently, it's hindcast of the trend during that period was not very good. This has nothing to do with Knight or even "Scenario B".

  31. The Earth has recently recovered from a period of unusual cold (LIA). The speed of recovery and present level are fairly typical of the last few thousand years, and the present level is near the average. The last decade or so has flat to slightly dropping global temperatures, and presently many (including AGW supporters) think it will continue for a decade or more. There is no tropic hot spot in the predicted level. The Arctic, which melted a lot recently seems to be recovering, and Antarctic is generally cool. Would someone please tell me what predictions have been made and supported that strongly support AGW as opposed to clearly falsifying it.

  32. "The global temperature rises in Scenerio B rises from about 1980 to 2010 and then goes pretty much flat for ten years (see the first figure in the link)."

    What is the slope of a linear regression line through the years 2010 to 2019?

    "Moreover the 1970-1980 period in Scenerio B, which Hansen considered the most likely,..."

    The years 1970 to 1980 had already passed when James Hansen did his analysis. So all three scenarios should have had the same same forcing for the 1970-1980 period, which should have been the actual forcing. As pointed out by jgdes, the fact that a model mimics the past doesn't prove much about the model's ability to predict the future. As he noted, the actual temperatures in 2009 are closest to James Hansen's Model C, in which all greenhouse forcings stopped increasing in 2000 (i.e. climate change was solved in 2000).

  33. From accused denier Steven Levitt:


  34. In my opinion any scientist who isn't a sceptic isn't doing his job right. Always questions your results/conclusions, until you can no longer find a reason to doubt them.

    Then they are 'true' only as long as noone finds any evidence to counter them.