TC and DW

Our old friend Tim Curtin has published a paper in what is supposed to be a peer-reviewed scientific journal. I’m skeptical.

He regresses temperature time series against a variety of predictor variables, concluding that there is no real influence of “non-condensing greenhouse gases” (i.e., GHG except water vapor) like CO2. He achieves this by rejecting regression of temperature in favor of regression of the first-differenced temperature data. You get that by taking the difference between each data value and its predecessor.

I’m puzzled by some claims Curtin makes about what’s required for a valid regression:

In general, the various rules or conditions that must be satisfied for a valid regression are the following:

(1) the predictor samples xt1,2…n and yt must be representative of the population that they are sampling;
(2) the unknown ut must have zero mean;
(3) the predictors must be linearly independent;
(4) the unknown ut must be uncorrelated;
(5) the unknown ut must be samples from a random variable population with constant variance, or homoscedastic.

It’s not so. Predictors don’t have to be linearly independent. Extreme lack of independence complicates regression and can lead to mistaken results, but it can be dealt with, and mild lack of independence isn’t really a problem. The error terms (what he calls “ut”) don’t have to be homoskedastic — that’s what weighted regression is for — or uncorrelated — that’s what generalized least squares (GLS) is for. You can even do ordinary least squares (OLS) with heteroskedastic or correlated errors, but you need to compensate for those factors, and it’s not as precise as GLS.

What really puzzles me is the way he rejects regression with the un-differenced variables. His first regression is temperature (GISStemp) against greenhouse-gas climate forcing:

I first regress the global mean temperature (GMT) anomalies against the global annual values of the main climate variable evaluated by the IPCC Hegerl et al. [17] and Forster et al. [28] based on Myhre et al. [29], namely, the total radiative forcing of all the noncondensing greenhouse gases [RF]

Annual (Tmean) = a + b[RF] + u(x) …. (3)

The results appear to confirm the findings of Hegerl et al. [17] with a fairly high R2 and an excellent t-statistic (>2.0) and P-value (<0.01) but do not pass the Durbin-Watson test (>2.0) for spurious correlation (i.e., serial autocorrelation), see Table 1. This result validates the null hypothesis of no statistically significant influence of radiative forcing by noncondensing GHGs on global mean temperatures.

The Durbin-Watson (DW) test is based on computing the statistic

d = \Bigl [ \sum_{j=2}^n (u_j - u_{j-1})^2 \Bigr ] / \Bigl [ \sum_{j=1}^n u_j^2 \Bigr ] .

It’s an estimate of 2(1-\rho_1), where \rho_1 is the autocorrelation at lag 1. If the noise is white then \rho_1 = 0 and the Durbin-Watson statistic is about d = 2.

The actual value of d is compared to critical values of the Durbin-Watson statistic. There are two critical values for each sample size and number of regressors, dL and dU. If d is less than dL, we reject the null hypothesis of no positive autocorrelation. If d is greater than dU we do not reject the null hypothesis. If d is between dL and dU the test result is inconclusive.

But Curtin says “do not pass the Durbin-Watson test (>2.0)” as though all values of d less than or equal to 2 confirm the existence of autocorrelation. That’s just plain wrong.

The DW statistic for his first regression is d = 1.749. For his sample size with one regressor, the critical values at 95% confidence are dL = 1.363 and dU = 1.496. Since d is greater than dU, we do not reject the null hypothesis of uncorrelated errors.

This test gives no evidence of autocorrelation for the residuals. But Tim Curtin concluded that it does. He further concluded that such a result means no statistically significant influence of greenhouse gas climate forcing (other than water vapor) on global temperature. Even if his DW test result were correct (which it isn’t), that just doesn’t follow.

Curtin’s next regression is of the first-differenced versions of his previous variables. This time he reports a DW statistic of 2.760 and states that it “passes the Durbin-Watson test statistic.” Apparently he concludes this because d exceeds 2. He further concludes that this regression is therefore valid, so the lack of significant response of (first-differenced) temperature to (first-differenced) climate forcing indicates no relationship.

Now for the really interesting part. The DW test can also be used to detect negative autocorrelation (in which successive values tend to be anticorrelated rather than correlated). To do so, you use the value 4-d for the test rather than d. For his result we have 4-d = 1.24. Now the critical values are dL = 1.352 and dU = 1.489 (they’re different because there’s one fewer data point in the differenced time series). Since 4-d is less than dL, we reject the hypothesis of no negative autocorrelation.

In other words, the regression which Curtin said fails the DW test actually passes, while the regression which he said passes, actually fails.

And — the presence of autocorrelation doesn’t invalidate regression anyway.

I have to wonder what kind of “peer-reviewed” scientific journal would publish this. Who were the referees for this paper?

It’s amusing to look at the Appendix to Curtin’s paper. He seems to be confused in his use of units. Apparently his readers might be too, since he feels the need to define the Joule as a unit of energy and Watt as a unit of power. Real scientists are supposed to know that already. So, who’s reading this “scientific” journal anyway?

120 responses to “TC and DW

  1. It sounds like Curtin is stuck in a mindset of doing linear regression by matrix methods, and has no idea that it’s possible to fit a general nonlinear dynamic model to data. Nor does he understand that water vapor is a feedback, not a forcing, and therefore is endogenous in some of the models he accuses of omitting it.

    His discussion of first differencing is also weird. The primary rationale for first differencing is to undo an integration that’s intrinsic to the structure of the problem, not the noise structure. However, that is likely to be defeated by measurement error, requiring more sophisticated methods like Kalman filtering. In other words, Curtin fails Bathtub Statistics 101 ( ).

    • LazyTeenager

      I found that article very informative. It resonates with the suspicions I already had about people who base analyses on correlations of trends or derivatives. E.g. Tamino’s favorite whipping boy.

      It also fits in with my general discomfit with the climate gullible’s general arguments about cause and effect in complex systems like the earth’s climate.

  2. jasonpettitt

    Love open access Journals – I hope it’s the future. But I don’t love the pay to publish business model.

    Not that I’m a scientist or am ever likely to publish anything ever. But I do appreciate science and therefore care that the quality and the veracity of science counts for something.

  3. MieScatter

    On my first read through of his paper, it looks like Tim Curtin has successfully discovered that water evaporates, but failed to say anything about global warming.

    This glorious intellectual contribution to the climate ‘skeptic’ movement should be welcomed for the progress it is.

  4. Horatio Algeranon

    Confused Units
    — by Horatio Algeranon

    In science it’s important
    To get the units right.
    Joule is a unit of energy
    The meter, of the height.

    Dynes are units of force
    And Watts of grand illusion
    GigaWatts are really lots:
    Units of mass confusion.

  5. Quiet Waters

    The Scientific World Journal is an Open Access journal, and as such makes all published articles freely available online without the need for a subscription. Moreover, the authors of an article that is published in The Scientific World Journal are allowed to retain the copyright of their work, and they are free to reuse and redistribute their work in any medium without having to seek any permission from the publisher. In order to cover the costs of publication, authors of articles that are accepted for publication in The Scientific World Journal are required to pay an Article Processing Charge of $1,000, which is typically paid from either their research budget or from departmental or institutional funds.

    I wonder what grant TC published this on…

  6. Guess who said this:
    “Increasing strength of Atlantic hurricanes, disappearance of Arctic
    sea ice, melting of the Greenland ice sheet, six meters flooding in
    coastal cities; are these impending climate catastrophes supported by
    observations, or are they just results of imperfect climate modeling
    and the imagination of overeager climate politicians? I will present
    recent analysis of North Atlantic hurricane activities to show that
    there is no justification for claims that hurricane intensity or
    numbers have increased drastically with increasing atmospheric
    concentration of CO2. Similarly, Greenland temperatures in the 1930s
    and 1940s were as high as they are today. Finally, I will argue that
    the current warming of the Arctic region is affected by multi-decadal
    climate variability more than by an increasing concentration of carbon
    dioxide. Thus we may spend hundreds of billions of dollars on curbing
    CO2 emissions without having a noticeable effect on the ongoing
    climate change in the Arctic.”

    And wrote this:
    We find the empirical climate sensitivity to be between 0.29 and 0.48 K/Wm−2

    Hint: His name can be found on this list

    • TB,
      The above diatribe was of course was highligted and parroted by Pielke Sr. on his blog.

  7. Something in the back of my mind is asking uncomfortable questions. Temperature is a continuous field, is it not ? There is interpolation and assimilation into discrete data points, but this is only to permit trend analysis (one of the Phil Jones papers talks about how many data points you need to assert a temperature over a given area). Is it appropriate to apply point-follows-point tests of any kind ? And should temperatures (bunched into numbers over space and time for convenience) be expected to adhere to a particular statistical distribution (especially if they’re changing over time) ?

    [Response: I don’t see a real problem. In fact we have no choice for variables which are continuous because it’s *impossible* to sample a variable over a continuous interval (except theoretically).]

  8. Ian Forrester

    The Scientific World Journal is a “pay to publish” “journal”. Here is a quote from them:

    Article Processing Charges

    The Scientific World Journal is an Open Access journal, and as such makes all published articles freely available online without the need for a subscription. Moreover, the authors of an article that is published in The Scientific World Journal are allowed to retain the copyright of their work, and they are free to reuse and redistribute their work in any medium without having to seek any permission from the publisher. In order to cover the costs of publication, authors of articles that are accepted for publication in The Scientific World Journal are required to pay an Article Processing Charge of $1,000, which is typically paid from either their research budget or from departmental or institutional funds.

    So Curtin paid $1,000 to get his rubbish published. That is expensive fishwrap.

    • Rattus Norvegicus

      PLoS One is a “pay to publish” journal also, but it is respectable. As Hank points out, TSWJ does not appear to be a respectable journal.

      And I was wondering how it got the 1.5 impact factor. Hanks link explains how it was done.

      • Rattus:

        TWSJ’s impact is not very high, and Hank’s link does not show that they manipulate it, as it only refers to authors who inflate their other journals’ ratings by self-citations in their TSWJ papers.

  9. Especially with the UK Daily Mail costing 55p. Makes no sense at all…

  10. john byatt

    I emailed this to the editor of WSJ as a peer review of the paper

  11. john byatt

    The claim that the journal is peer reviewed, sorry it is tswj not wsj as above

    • Byatt: thanks for all your help. There is much more to my paper than my error re D-W, and I would like to see your peer review of the substance of my paper. Thanks.

  12. Here is a howler: In the paper, his source for interpretation of the Durbin-Watson statistic is Wikipedia!

  13. I appreciate Tamino’s comments and will respond asap, but for now make it clear that I did NOT pay TSWJ anything let alone $1000. They invited me to submit, unsolicited by me. My paper had FOUR peer reviewers, and was edited by a professor of chemistry.

  14. When I regress dT on ln CO2 for 1880-2011, or 1850-2011 (Hadley), I get the same result–R^2 = 76%. But there is a fairly high rho-hat. So I applied Cochrane-Orcutt iteration. When it settles down, R^2 is still 60%. So most of the temperature variation for the last 160 years appears to be down to carbon dioxide.

    The details are here:

    [Response: It looks like you’ve regressed T (not d(T) or first-differenced T) on CO2. Is that correct? I believe Curtin regressed d(T) on d(CO2), not on CO2.]

    • No, I used temperature anomalies, which I’m referring to (incorrectly?) as “dT.” The “d” would be from the period mean, rather than from the previous measurement. Point is, here’s prima facie evidence that even when autocorrelation is compensated for, CO2 is still the dominant forcing on temperature variation for the last 160 years.

      • Anomalies from a 30 year average (multiplied by 100 as they are by Hansen’s Gistemp) are designed to make temperature increases look more scary than they actually are (but the anomalies have the same trend as the actuals).

        As this programme cannot paste diagrams, go to

        Murphy on Nordhaus

        (or Gistemp)

        to see the absolute temperatures.

      • “Anomalies from a 30 year average (multiplied by 100 as they are by Hansen’s Gistemp) are designed to make temperature increases look more scary than they actually are (but the anomalies have the same trend as the actuals).”

        Nonsense. It makes no difference if you are using actual temps or anomalies. What the person at your link did was to change the scale of the graphs (so that it could show 20*C or 30*C of temperature change) in order to make people think that .8*C of warming was no big deal. Why not go all out and start it from absolute zero? Then even the depths of glacial periods wouldn’t show up. What Robert P Murphy did was deliberately deceptive. If you keep the range on the y-axis to temps that have occurred within the last few million years, .8*C of warming will indeed seem significant. The 3*C projected for a doubling of CO2 is properly alarming. I’m almost ashamed to share the same name as Murphy.

      • Actually, TC is right, this has been shown months ago by Dr. Inferno:

  15. They invited me to submit, unsolicited by me.

    That’s even better, as anyone soliciting Curtin for anything related to climate science clearly has an agenda in doing so.

    (now it could’ve been an honest agenda of asking him to submit so the four reviewers could slice him to pieces while regaling each other with copious quantities of beer, but in this case apparently the agenda was something else)

    • dhogaza, wrong again. TSWJ invited me to submit a paper without specifying topic.

      • Rattus Norvegicus

        Tim, were you aware that this was a rag — and I use the word with purpose — which resorted to questionable tactics to increase it’s visibility? Do look at the link I posted about to Hank Roberts’ post and the links that he posted (particularly the first one). This is not a reputable publishing organization.

      • So, it was touch and go between this paper and your overview of 12th century canon law?

      • So, it was touch and go between this paper and your overview of 12th century canon law?

        Or, say, something touching on string theory?

        Even if you’re telling the truth, the notion that a credible journal would say “submit an article on anything, and we’ll waive our $1K fee, peer review and publish it” is just … stupid. You’re saying you could’ve published your views on the evolution of snails in the carribean and they would’ve accepted this as overturning modern evolutionary biology?

        Tim Curtin digs himself (and the journal) into deeper quicksand, no surprise.

      • Yes, but what topic would they care about?

        It’s not like you’re einstein, dude …

      • “TSWJ invited me to submit a paper without specifying topic.”

        I get such invitations every week, mostly from new open-access journals. The first one or two I was flattered by but soon I realized they were spam, sent out to mailing lists.

    • Maybe, but because their reputation is so weak, they probably thought they’d roll the dice with Tim. If it actually turned out he was right, they’d trumpet. If he was wrong, well how much worse could it get for them? And they’re right, of course, because how many people are going to check the math or look for a critical response? (answer: people who already have a strong suspicion, based on the intersection of the paper’s conclusions and the quality of the journal, that something–or many somethings–will be wrong.).

      I wonder if we could get the reviewers to weigh in on their decisions. That would be fascinating, eh? Like prying open a rail car just after a train wreck.

    • trccurtin
      You link to a commentary about a commentary about a critique of an op-ed. You also stray off-topic.
      That said, the tosh you peddle here is the sort of nonsense that Lindzen has been peddling for a few decades – that Average Global Surface Temperature is an irrelevant artifact that would have to shift +/-10 deg C or more to be otherwise. Is this what you are suggesting? Is this why you link to graphs of global temperature with 30 deg C vertical scale?
      Why not plot the PETM or the glacial minima on your non-worrisome graph? Let’s check how far temperatures can deviate before Lindzen’s house dissapears under a km or two of ice sheet? Or family cats start evolving into bats?

  16. I’m far less concerned with the reputability of TWSJ, than I am with Tim Curtin’s explanation of his application of the DW test and his statements about the requirements for regression to be valid.

    • I can understand that, but, there’s evidence that TWSJ is trying to become a more cited version of E&E. The TWSJ people are much more sophisticated than E&E (which seems to have been dying, at least in terms of getting serious scientific attention). Think of it as E&E round two, much less open as to their goals, just as destructive.

  17. Harold Brooks

    I’ve gotten emailed to submit to those kinds of journals a lot. I always assumed they were bulk e-mails and ignore them.

  18. First things first. I want Tim Curtin’s explanation of his application of the DW test and his statements about the requirements for regression to be valid.

    • Tamino claims “But Curtin says “do not pass the Durbin-Watson test (>2.0)” as though all values of d less than or equal to 2 confirm the existence of autocorrelation. That’s just plain wrong.”

      I agree that my wording could have been better by writing d~2, but here is what Tamino’s own linked text says:

      “The Durbin-Watson test statistic tests the null hypothesis that the residuals from an ordinary least-squares regression are not autocorrelated against the alternative that the residuals follow an AR1 process. The Durbin-Watson statistic ranges in value from 0 to 4. A value NEAR 2 indicates non-autocorrelation; a value toward 0 indicates positive autocorrelation; a value toward 4 indicates negative autocorrelation.”

      This is consistent with von Storch and Zwiers (1999:158): “A ‘rough and ready’ approach that can be used when the samples are large [as with my data] is based on the observation that … d = 2*estimated lag-1 correlation coefficient of the residuals … The value of the D-W statistic in our SOI example is 2.057…”

      [Response: This is total, unadulterated bullshit.

      It defies belief that you would even try to excuse your error by appealing to a vague statement about values of the DW statistic. That’s not how statistics works.

      For some bizarre reason you think you can get away with calling your DW statistic “toward 0” and therefore indicating autocorrelation. Are you really so ignorant and arrogant that you think you can just decide arbitrarily what’s sufficiently “toward 0” and what isn’t? Again, that’s not how it works. After hearing your lame excuse, I doubt that you have ever in your life done it right, because clearly you didn’t know how. Any ignoramus who read this blog post knew more about it than you did, because I described it myself and linked to the necessary tables of critical values.

      The stuff you threw in about von Storch and Zwiers is completely irrelevant — just a diversion to cover your butt. All it really says is that d is approximately 2*estimated lag-1 correlation. That has nothing at all to do with your failure to do the test right. What’s really hilarious, literally a scream, is that you couldn’t even get that right! It’s 2*(1 MINUS estimated lag-1 correlation) — as you would know if you had paid attention to THIS POST.

      The document I linked to not only describes the test in sufficient detail to execute it, but gives the necessary tables for critical values. It’s a fact, plain and simple, that you bungled the DW test AND that you got the wrong result. We know because I did it for you. This isn’t an “oops” that everybody makes sometimes. It’s utter failure to have a clue what it is that you’re doing. About something that’s basic — undergrad stuff. I repeat: you didn’t try to do it right and make a boo-boo in your calculation — you simply didn’t know how.

      The greatest surprise of all is that you won’t admit it. It buggers belief (no, I don’t mean “beggars”) that you have deluded yourself into believing that it’s not obvious. Guess what? Everybody knows it. Everybody.]

      The more pertinent issue raised in my paper is why that same Zwiers at no point in his chapter (with Hegerl in AR4 WG1, Solomon et al., 2007) performs or reports any LSR and the required D-W statistics. My paper does, and all its tests show failure of the undifferenced time series to meet the D-W.

      [Response: Bullshit again. I already showed that you got it wrong because you didn’t know how to do the DW test right — so none of your results can be trusted. No, I’m not going to do them ALL for you. I did the first two already, and you were wrong both times.]

      Tamino’s other main issue is with my use of differencing.

      [Response: No. My other main issue is that you made so many false statements about what’s required for valid regression. It so happens that I do think first differencing is a mistake — but I never mentioned that, all I said was that you did it and that your justification for doing so was just plain wrong.]

      As von Storch and Zwiers note (1999:114), “meteorological time series are generally autocorrelated…with the consequence that the absolute value of the t-statistic tends to be too large” (as is clearly the case in the regression of Gistemp anomalies on GHG radiative forcing reported in my Table 1). So von Storch and Zwiers’ textbook (Statistical Analysis in Climate Research) does not confirm Tamino’s claim that “the presence of autocorrelation doesn’t invalidate regression anyway”. I never said it did,

      [Response: You just won’t stop the bullshit. Your list of conditions that MUST be satisfied for valid regression includes: (4) the unknown ut must be uncorrelated.]

      what I did say like von Storch and Zwiers is that the presence of autocorrelation invalidates LSR of undifferenced time series by exaggerating the t-statistic and thereby failing the D-W test.

      [Response: Again you really don’t have a clue what you’re doing. Autocorrelation inflates the t-statistic but does not invalidate LSR, whether of differenced or undifferenced time series. You have to compensate for the autocorrelation — as I said in THIS POST — but it is by no means invalidated. And just as a painful reminder to you, the very first regression did NOT fail the DW test.]

      Their solutions include applying LSR to their ARMA equation 10.31 as that differences both the independent and dependent variables which is what I did.

      Likewise Stern and Kaufmann (1999) state “The lagged first differences of the dependent variables provide a correction for possible serial correlation. The number of lags p is chosen using the Akaike Information Criterion (Akaike, 1973) … The univariate tests show that while the greenhouse gas series are I(2), the temperature series are I(1). The only possible relationship between such series is one where the first differences of the greenhouse gases drive temperatures …”. Feinstein and Thomas (2002:362) also advocate first differencing of time series data in order to avoid autocorrelation.

      [Response: Maybe some day I’ll elaborate my disagreement with those authors. But to stay on topic, your reference to them is just more smoke and mirrors. None of this has anything to do with the fact that you did the DW test wrong because you didn’t know how. You just wanted to quote some references to make it look like maybe you knew what you were doing.

      I’ve seen some pretty pathetic excuses for analysis in my day — but you take the cake.]

      • Hi Tamino: you are right, I was too hasty in replying, so did not pick up the typo in the formula you mention nor check the actual D-W dL and dU.

        But you are also too hasty. Using your linked Tables, for which I am most grateful, we find (based on the text explaining the Tables):

        “Cross-referencing the indicated row (n=31) and Column (k=1), the bounds are dL = 1.147 and dU = 1.274. The observed value of [my D-W] test statistic is 1.749, so it is higher than the tabulated upper bound, and therefore we should reject the null hypothesis of non-autocorrelated errors in favor of the hypothesis of positive first-order autocorrelation.”

        [Response: Is this a sick joke? Or have you actually lost your mind? The linked document states:

        If the observed value of the test statistic is LESS than the tabulated LOWER bound, then you should reject the null hypothesis of non-autocorrelated errors in favor of the hypothesis of positive first-order autocorrelation

        It goes on to say

        If the test statistic value were greater than dU, we would NOT reject the null hypothesis.

        By the way, you’ve quoted the critical values from table A-1 for 99% confidence. I had used the values from table A-2 for 95% confidence, since I didn’t want to impose too strict a standard.]

        So my Paper’s finding in its Table 1 stands even if poorly explained.

        Thus the fact remains that there has never been any LSR falsification of the nul that GHG does not account for “most” global warming.

        [Response: There’s just one thing I can’t figure out — whether you’re trying to fool people in order to save face, or you are genuinely that addled. In either case, be very sure of this: the only person you have even a remote chance of fooling is yourself. Everybody sees it. Everybody.]

    • Tamino, you are of course quite right, I was not at my best yesterday and misread the text and tables you sent me. I apologise, and will make due corrections to my paper when I get the chance.

  19. “I want Tim Curtin’s explanation…”

    Good luck.

    This is the man who, over on Deltoid blog, once found a set of interpolating functions relating sea surface temperature, salinity and total alkalinity for individual ocean basins, interpreted total alkalinity as a measure of pH, extrapolated the function far, far outside its bounds, and declared that ocean acidification would cause the oceans waters to become fresh water.
    And defended that in a conversation that went on for much, much more than 200 posts.

    The man who fit a 5th-order polynomial to [CO2] time series, claimed it best captured ENSO behavior, both in range and when extrapolated into the past and future. And who continued to defend that claim even when it was shown that it predicted negative [CO2] in the short term past, and [CO2] greater than unity in the short term future.

    Who based an entire analysis ands published slide presentation on future [CO2] growth on his inability to distinguish the michaelis-menten equation, from the equation of its limiting value, and used the wrong one.. And then blamed it on Excel dropping a pair of brackets.

    All in one damn thread.


    Don’t hold your breath

  20. Can Global Warming Be Contained? Can The Adverse Effects Of Climate Change Be Reversed?

  21. Bernard J.

    I actually have some (small amount of) time for Tim Curtin…

    Almost uniquely among the Denialati, he will actually put his testes in the burdizzo. Admittedly, he is now carrying around a scrotum bursting with scar and adipose tissue, but he just keeps on whacking the shadow of his jewels into that damned burdizzo.

    Most of Curtin’s mates who practice pseudoscience are all obfuscating talk and no action, which ironically makes it somewhat more difficult to show where they go wrong – as they always do. Deltoid’s Scandinavian Troll Collective is a good example…

    When it comes down to it, ol’ Curtin is amusing, in an “at you” rather that a “with you” way. The bottom line though is that he typifies the depauperacy of the denialist understanding of science and statistics, and anyone who bothers to scratch below the surface of the claims of the Denialati is rapidly confronted with the intellectual absenteeism that characterises their arguments.

    Curtin deserves some small thanks for participating in – and figureheading – this revelation of denialist ignorance.

    • I wonder why Bernard J does not get moderated while I am for ever?

      [Response: You weren’t moderated at all. Every word you posted is intact and unchanged. You were refuted. Completely. You admitted that you didn’t even check the critical values for the DW statistic (which you should have done before submitting your paper). When you finally got around to it (in your last comment) you got it wrong. Again. You got pwned.]

  22. Tim, tamino is giving you the benefit of the doubt. That is, he’s allowing for the possibility that you are actually ignorant of some of the fine, but critical, details. I find it hard to believe that someone with your general understanding of statistical methods could naively come to a conclusion that, after looking at the history of your engagement, you clearly so desperately desire. What you have done–or attempted to do–is effectively confuse non-experts (like me) with regards to a very important issue. If that is true, why would you desire my confusion? I can only think of a few reasons. 1. You don’t care about my confusion. You simply want to be published, either because you want to be respected as an important person or because you need to publish to retain your job. 2. You just know that the theory of AGW must be wrong, and so you’re trying to fit data to that belief. 3. You want AGW to be wrong because you’re philosophically committed to the belief that mitigation will raise the spectre of big government and the loss of cherished liberties. (of course, if that’s the case, you’re bound in a paradox, because intentionally misinforming people is a kind of abrogation of their freedom. To secure your own freedom (if only of thought), you’re willing to diminish the economic freedom of others.) 4. Someone is paying you for these analyses, and you have sold your integrity to the highest bidder.

    Admittedly, none of these conclusions are very flattering, but I lack the imagination to come up with other possibilities (other than the remote possibility that you are an alien life form helping your species take over the Earth). If you are truly innocent and your work re climate is born of ignorance, then why do you consistently–innocently and ignorantly–find the conclusions you want to find?

    • Horatio Algeranon

      the remote possibility that you are an alien life form helping your species take over the Earth.

      Well, if that were the case, he’d prolly be working with Al Gore rather than against him (cuz everyone knows taking over is what Al Gore actually desires)

      …unless, of course, it is a war between 2 alien life forms (the Gorians vs the Inhofians?) and Earth is just the latest battleground (Venus being the previous one: chock up one win for the Inhofians)

  23. Susan Anderson

    DSL, you nailed it.

    It’s refreshing to acknowledge that bad temper is not a bad thing. Since it is endemic in science (and other communities of people who excel at the work and are interested in it) that’s helpful. Tim Curtin fails to understand that the thing that raises the ire of experts more than anything else (rivals, fashions such as string theory, etc.) is falsehood or fudging masquerading as truth.

    It’s easy to be accepted if you don’t claim to know what you don’t.

    Scarecrow: I haven’t got a brain… only straw.
    Dorothy: How can you talk if you haven’t got a brain?
    Scarecrow: I don’t know… But some people without brains do an awful lot of talking… don’t they?
    Dorothy: Yes, I guess you’re right.

  24. Cautionary, further on what I posted at RC mentioned earlier, specifically documenting this TSWJ journal Tim Curtin’s published in:

    “… There is another citation gaming tactic that is much more pernicious and difficult to detect. It is the citation cartel.
    In a 1999 essay published in Science titled, “Scientific Communication — A Vanity Fair?” George Franck warned us on the possibility of citation cartels — groups of editors and journals working together for mutual benefit…. this behavior has not been widely documented; however, when you first view it, it is astonishing.”

    And you thought it was only financiers who could package and sell toxic waste? Nope, publishing companies are creating new ways to do the same thing.

    Why would anyone publish unreliable crap as “science” you ask?
    Because people will buy it, of course.

    People who believe, politically, that “science progresses when everyone can buy the type of science they like” ….
    The Rise of the Dedicated Natural Science Think Tank

    It’s easy to publish bad work these days.
    Almost as easy as blogging your opinions.

    Authors — they’ll be soliciting some reputable work to bundle in with their toxic waste. Don’t lend them your reputation.

  25. EFS_Junior

    “The author (one Tim Curtin) … is responsible for all views expressed and for any remaining errors.”

    Well, it has been shown (rather completely and conclusively, I might add) that “remaining errors” have indeed been found (see above).

    So if someone admits that there are “any remaining errors” and then denies that there are, in fact, “any remaining errors” then what does that say about the author (one Tim Curtin)?

    To me, it says that the author (one Tim Curtin) cannot confess to making “any remaining errors” when the point blank evidence of “any remaining errors” is put straight to his own face even.

  26. Horatio Algeranon

    From Curtin’s paper

    we assess the ….hypothesis, that year on year changes in temperature are determined by year on year changes in those independent variables [radiative forcing]

    “year on year changes” in GMT are primarily due to weather noise** (eg, from el nino, la nina, volcanic eruptions) so it’s a bit unclear (to say the least) why anyone would ever propose /hypothesize that such “differenced temperatures” even might be significantly correlated with changes in radiative forcing due to green house gases.

    Horatio seems to recall some others (carter, McClean and de freitas) who used “differencing” (albeit in a slightly different guise) to thoroughly confuse the issue (if not themselves)

    **GMT can swing by 0.1 C (sometimes more, sometimes less, but that’s a pretty good indication of magnitude) from one year to the next (due to el nino, etc), but the change due to increasing GHG forcing (according to most climate scientists) is an order of magnitude less (~ 0.02C/yr)

  27. The only times you’re supposed to use differencing at all are

    1) When your theory predicts that a differenced variable will be significant. Note, PREDICTS, as a logical conclusion of your theory. This does not refer to a series you pick after trying several possibilities.

    2) When the variable of interest is integrated; i.e., non-stationary, and differencing it results in a stationary time series.

    • BPL: Thanks – but note “until a really satisfactory procedure is available, we recommend taking first differences of all variables that appear to be highly autocorrelated. Once more, this may not completely remove the problem [of autocorrelation] but should considerably improve the interpretability
      of the coefficients”. That is why I differenced – and in my paper I cited SPURIOUS REGRESSIONS IN ECONOMETRICS
      Journal of Econometrics 2 (1974) 111-120.

      In my earlier paper (at ACE2011 and the longer time series Table 1 Regression of Gistemp GMT anomalies (oC) against Radiative Forcing of [CO2] (W/sq.m) 1959-2010.
      had Adjusted R2 of 0.76 and Standard Error 13.97 with Durbin-Watson
      1.31 (Data Sources: NASA-GISS; NOAA-CDIAC; Solomon et al. 2007:141).

      The D-W at 1.31 falls well below the dL of 1.5 (for n=50 and k=1)) in the Tables that Tamino very kindly linked me to, hence the need for differencing.

      • Granger and Newbold 1974 was indeed a ground-breaking paper. But in the 38 years since then, knowledge of how to compensate for autocorrelation has improved a bit. I’d recommend looking at a recent textbook on the subject, such as Walter Enders’s “Applied Econometric Time Series,” or Banerjee et al.’s “Co-Integration, Error-Correction, and the Econometric Analysis of Non-Stationary Data.”

        [Response: I would also recommend Lee & Lund (2004), Revisiting simple linear regression with autocorrelated errors, Biometrika, 91, 240.]

      • Cointegration was just the economists catching up to what the engineers (e.g., Fred Schweppe) had figured out years before.

      • “Burning coals of fire…” ;-)

  28. Tamino, thanks for your excellent work in assessing the statistical methodology here.

    Unfortunately, statistics cannot really be used in place of the underlying physics of radiative transfer to “test” the claims of Lacis and Schmidt et al concerning the relative importance of various gases to the terrestrial greenhouse effect, or to the feedback processes that determine the magnitude of global warming. The simple fact is Curtin doesn’t understand the first thing about any of this, and none of his methods are applicable to the claims of Lacis or Schmidt et al.

    • Chris: really? how about this, and have you replied to it?

      “The very tight affine relationship between temperature and the rate of change of CO2 proves that temperature is the only influence necessary to consider to calculate CO2 concentration. It accounts for virtually ALL of it. There is virtually no room for anthropogenic forcing at all. AT ALL.
      With that information, the information about human emissions becomes superfluous. As far as significant human forcing of atmospheric CO2 concentration goes, we have no need of that hypothesis. QED.”
      Much more in previous comments by Bartemis on the Climate Sensitivity discussion thread (h/t; PL).

      I don’t actually buy into that myself, but what do you say?

      • It’s completely wrong, makes no sense, and has nothing to do with my original comment, which only reinforces my view that you are trying to cause a distraction. I’m sure that will work for the WUWT-Curry-Bishop Hill type crowd, but no one else is going to pay attention nor will it be something talked about in the actual peer-reviewed literature, at conferences, etc.

        If you want to decompose the terrestrial greenhouse effect into its relative contributions, the observational record is actually not a meaningful start point, nor does the observational record yield any insight into what would happen if you took all the CO2 from the air (as in the Lacis experiment, but several others have done this too). You need to actually look at the radiative transfer physics, which invariably requires a physical model. And actually, the fact that water vapor acts a feedback (which is well known) is not terribly relevant either for the attribution because there’s always some water vapor in the air and it is always acting in concert with CO2 at any given time to reduce Earth’s OLR relative to a no-greenhouse OLR value (at the same temperature).

      • Alex the Seal

        News flash…. Some guy on the internet claims anthropegenic CO2 is sucked harmlessly into space. More news at 11.

      • trccurtin
        You say you do not “buy into” it yourself yet you bring it here as though it was worthy of being ‘bought into’. It is palpable nonsense. It is off topic. No further direct answer to your question is required.

        In ancient times, storytellers would talk of fabulous beasts like the dragon, the griffin, the wyvern or the chimera with its body of a goat, head of a roaring lion & tail of a scaly serpent.
        Today fabulous beasts have come alive in cyberspace. The astute observer can see the trail laid by, say, the trccurtin with its presence of a troll, tenacity of a fool and argumentation of a contrarian.

      • Back to the beginning, in the denialist game of snakes and ladders.

      • So, correlation is proof of causation when it’s temp->CO2, but not the other way around? His use of the term “affine” should be a giveaway – in this context it means “scaling the axes to hide all the important stuff.” Bonkers (and not just for these reasons).

  29. LazyTeenager

    trccurtin says
    I don’t actually buy into that myself, but what do you say?
    LT says its wise of you not to buy into it, because the person who made that claim does not considered long and short time scale effects in complex systems with multiple drivers and feedback effects.

    That person assumes short timescale effects should be the same as long time scale effects. A dubious assumption.

  30. Susan Anderson

    C***** on a Crutch! Judith Curry now, who demands scientists toe her line (Montford, for heaven’s sake) to the baying of her fan base, but dare anyone suggest she do her homework, attacks instead of doing science. Something wrong there; this behavior would be understandable from a teenager (but would get straightened out fast by a good teacher) but after that, not. As has been said elsewhere, the non-skeptic “skeptics” will buy anything they hope supports their do-nothing obstinacy, and attack anything they don’t like. That is not skeptical, and at this point it is approaching criminal.

    The inquisition depended on popular support too, but was neither nice nor even Christian, and this too is nasty and wrong.

  31. Tim-san,

    The above would perhaps be true if the laws of electromagnetism, atomic theory, and quantum theory were repealed.

    • Does anyone here see a problem with that? I would be simpler to just repeal all the laws of physics…that would likely include all the various “laws” I work with in geology, so then I don’t have to worry about which specific laws to repeal…

  32. A very effective statistical pin-down, Tamino, but I must correct one statement: “…he feels the need to define the Joule as a unit of energy and Watt as a unit of power.”

    It is wise of Mr. Curtin to presume ignorance of common scientific nomenclature amongst his intended audience (and perhaps his SWJ reviewers), but In current usage a “Watt” is a measure of obstinate ignorance, usually reported in milli-Watts.

    My instruments suggest that Mr. Curtin, some kind of carpet-bagger economist who describes himself as an “unpaid associate [at] the Australian National University”, has a reading quite close to unity.

  33. “…a “Watt” is a measure of obstinate ignorance, usually reported in milli-Watts.”

    Wow! Then ‘nega-Watts’ really are best…

  34. To understand the flim flam, go to wood for trees Notice that the derivative of the CO2 concentration changes by about ~0.5 /50 years or about 0.01/year, pretty small. Now take a look at what happens if instead of averaging over 24 months, you average over 1 month so that you are looking at the actual annual variation which is about forth times bigger.

    If we also look at the change in the average [CO2] mixing ratio over the same period it amount to about 70 ppm. So what is happening. What Chris said, the rapid annual variation at ML is an effect of NPP (Net Primary Production) caused by the annual greening of the Northern Hemisphere as well as changes in circulation from the summer to the winter.

    So what is the original graph showing? The greening of the NH. The longer term greenhouse gas increase is results in the average warming over the longer period. BTW, just to nail this a bit more, it turns out that El Nino results in less emission of CO2 from the ocean because it slows overturning and you can see interesting effects by changing the averaging period to 12, 24, 36, 48 month.

    • Eli, how do you explain that the BoM (Oz) records of [CO2] at Cape Grim in Tasmania show NO acceleration in the monthly changes between 1976 and 2009?

      y = 9E-05x + 0.1213
      R² = 0.0021

      Cape Grim does show small reductions in [CO2] between November and April every year, but far below those at Mauna Loa in the NH summer.

      [Response: Acceleration of the changes? Is that what’s called the “jerk“?]

      • Eh? Now we have to endure rebunking of Tim’s much-debunked crap? And yet another zombie denier argument, too.

        Tamino deserves a better caliber of troll.

      • Tim, Perchance, have you ever done a Taylor expansion of an exponential? If so, what were the first few terms?

      • Is the trccurtin telling us this stuff because as a troll, he is visually impared? Or is it because as a fool he has no brain? Or perhaps as a contrarian, has he to be continually contrary w.r.t reality?

        Lines on graphs wot bend up from straight kinda means more rate of change wot some ud call “acceleration!”

      • The Southern Hemisphere does not green. Take a look at the measurements in Antarctica for an extreme

      • What is the “jerk”?

        [Response: In one sense of the word, it’s the acceleration of the rate of change, i.e., the third time derivative of a quantity. In another sense, It’s someone who is proven beyond doubt to have totally bungled his research, then tries first to deny it, later the diversionary tactic of a gish gallop of posts.

        No, I’m not willing to host your six replies this morning. I deserve a better class of troll.]

      • Also used for people who post questions in preference to reading extant links.

      • Maybe you need better quality con-troll.

  35. Sorry, that should be forty times bigger. Fat fingers

  36. re Barton Paul Levenson | May 14, 2012 at 9:15 pm | + Tamino

    Thanks for those refs. I did not find Lee and Lund helpful, as they did not compare outcomes from their method with the differencing of Granger & Newbold USING THE SAME DATA.

    That is a basic requirement.

    [Response: No. It’s not. The basic requirement is that you acquire basic competency before your results can be trusted.]

    I do not have Walter Enders’s “Applied Econometric Time Series,” or Banerjee et al.’s “Co-Integration, Error-Correction, and the Econometric Analysis of Non-Stationary Data.”

    However I do have Keller’s Statistics for Management and Economics 2005, and he supports what I did report apart from my admitted error at my Table 1.

    Gelman & Hill (2007) (Data analysis using regression) have never heard of Durbin-Watson. Much better but still limited is William H Greene (2008).

    None of the above like you compares the outcomes of the alternative techniques applied to the SAME data sets.

  37. “The Mathgic Box”

    –by Horatio Algeranon

    “Plug and Play” Is what I say
    Don’t worry whether it’s “OK”

    To fit a curve to the nth degree
    And plot a graph of what will be

    Or take a trend on just two years
    Just fudge ahead, and have no fears

    The Mathgic Box is really swank
    Dump data in and turn the crank

    And what comes out, so I am told,
    Is worth it’s weight in Fort Knox gold.

    • I consider myself to possess a reasonable level of scientific literacy…….but oh to have the creative artistic genius of Horatio Algeranon!

      • Susan Anderson

        Yes, he’s got the mathgic right down!

      • Susan: perhaps you can explain why Tamino’s Open (sic) Mind suppressed this Comment of mine (below)?

        Anyway, Tim Lambert’s Deltoid (
        is less Goebbelistic than Grant Foster, so my response is now up there

        Here is what the lovable Tamino suppressed:

        “Again, many thanks to Tamino for facilitating this discussion.

        First of all, I must express my appreciation to Tamino both for his valuable lesson on D-W and for his publicizing of my TSWJ paper. There are some matters arising which I feel are not without interest:

        1. The RF variable in my Table 1 was aggregate RF (CO2 and non-CO2) in W/sq.m. so was not raw data, unlike the [CO2] data in ppm that I used in my ACE2011 paper, which unequivocally failed the D-W test.

        2. The IPCC’s RF data that I used has itself been tuned to the temperature data.

        3. Regressing the Gistemp anomalies used in Table 1 on separated RF variables, namely [CO2] and [non-CO2], including [CH4] and [NO2] etc, we obtain:

        D-W: 1.859.
        Note the negative coefficient on the non-CO2.

        4. Evidently the aggregation of the RF from both [CO2] and [non-CO2] in my Table 1 did much to eliminate autoregression and secure an acceptable D-W.

        5. Allowing for an intercept, the coefficient on the [non-CO2] remains negative but ceases to be statistically significant. The D-W is fine at 2.06.

        6. Regressing the sum of [non-CO2] on [CO2] produces R2 of 0.7, and t=15.64, but the D-W at 0.053 now shows strong autocorrelation.

        7. However, unlike regressions of Gistemp anomalies on [CO2] in ppm from Mauna Loa since 1958, the regression using data only from 1979-2009 yields statistically significant results that satisfy D-W. That suggests something happened around 1978, known as a climate shift to some, which appears to have ended around 2000. What gives?”

      • With that, Tim plumbs the depth of depravity that deniers are so enamored of. In the process of self-seppuku, Tim forever cedes the field of scientific and personal integrity.

        Is there no absolute low that deniers will stoop to? Bueller?

    • Below is my last response to Tamino, which seems to have got lost in blogosphere.


      [Response: Not interested.]

      • Alex the Seal

        Well here’s something else which seems to have gotten lost in the blogosphere.
        How do you reconcile:
        a)”… reducing emissions of CO2 is unlikely to be the effective climate “control knob” claimed by NASA’s Hansen, Schmidt
        b) “.. reduction of CO2 emissions by up to 60% of the 2000 level by 2050 has the capacity to institute a new Ice Age.”

        Click to access Einstein.pdf

        Will the real Tim Curtin please step forward?

  38. Bernard J.

    Oh please, can we tell Curtin to post on the new eponymous thread at Deltoid? I enjoy the cackles too much to miss out on his ‘analyses’.

    And as much as I feel Tamino’s and others’ pain at Curtin’s silliness, he really is one of the Denialati’s “better class” of troll. Says a lot, huh?

    It’s a sad indictment of our culture though, that only a sprinkling of Stupid is required to gum up the workings of sensible decision-making…

  39. Grant Foster: A fellow student when I was at LSE in 1961-63 is now the head of a successful New York law firm.

    Before I contact him in regard to your defamatory statements about me, I ask you to post my wholly technical responses to your Blog’s “Open (sic) Mind” attack on me.

    [Response: Do you really want me to prove, yet again, that your analysis is completely incompetent — in a court of law? You already admitted that you didn’t do the Durbin-Watson test right, and after I showed you how, you got it wrong again — which you also admitted.

    When I decide not to allow your comment which is irrelevant to the topic of this post, you decide to be a bully.

    I’m posting your reply to Susan Anderson so that it will be on the record that you have referred to me as “Goebbelistic.” Be sure to mention that to your attorney friend. Further comments from you will not be allowed, unless they provide further evidence which can be used against you.

    I will not be bullied.]

    • Pathetic.

    • Pathetic and unfortunately, typical…

    • Horatio Algeranon

      “The Goebbeleo Complex”
      — by Horatio Algeranon

      Their tentacles are everywhere
      In scientific journals
      In hallowed halls of government
      (And also in the urinals)
      They even penetrate the bags
      of Redenbacher kernels
      With propaganda methods
      Like those of Joseph Goebbels.

    • Chris O'Neill

      “Before I contact him in regard to your defamatory statements”

      One thing a lot of people don’t understand about defamation is that the truth is not defamation, it’s just the truth.

    • Bernard J.

      Remember when Curtin threatened to report me to the policefor ‘attempting to murder’ him? Ah, funny days…

      Still, if the old fellow was foolish enough to believe that reducing the pH of seawater to below 7 rendered it potable, one can understand how it is that he can’t manage to conduct statistical analyses. And if he did not so ‘endanger’ himself, then he was rather loose with the truth.

      Curtin, if you’re reading this there’s something I’ve been meaning to draw to your attention that was never raised during the original conversation – it’s possible to bring the pH of seawater to below 7 simply by blowing in a glass of the stuff using a straw. I’ve done it many times with students, and strangely (or not…) the water was always salty after the experimnent. Out of perverse curiosity, do you still maintain that seawater can be freshened simply by acidifying it? And if you do concede that someone better informed than yourself was correct about seawater chemistry when you yourself were not, why can you not concede that a professional statistician is better able to identify your inabilities with statistics than you are yourself?

    • This reminds me of the joke about the guy who murdered his parents and then asked the court for mercy because he was an orphan.

  40. I’m really perplexed as to why Tim Curtin would want to take Tamino to court and run the extremely high risk of having the gross errors in his climate writings and methodology not only entered into official records for perpetuity, and even having a Judge potentially make another cracker of a “breathtaking inanity” (aka Dover-Kitzmiller trial) comment in an official judgement!

    Why on earth would you take such a risk?

    • Probably because Curtin thinks he’s right. DK, and all that…

      • Sure, but opinions aren’t defamation. He’s intimated that he wants to sue for defamation, but an opinion of someone’s intelligence or competency isn’t defamation. We’re entitled to our opinions. And we’re entitled to voice them.

        I’d suggest to one and all thicker skin. Learn from the elephantine patience of Mike Mann. Denialists have defamed Mann, saying that he committed fraud. That’s defamation like Grandma used to make. His disregard of his defamers reminds of the Marianne Moore line about ” … presidents who have repaid sin-driven senators by not thinking about them.” Mann doesn’t even say, as I would, “Go roll a hoop in the park.”

        The only statements I’ve seen here regarding Curtain are that he doesn’t understand the math that he’s used. He has a non-judicial remedy: use the math correctly. Or show up his critics by demonstrating his own superiority. The way science does all the time. Non curat lex de minimus. (Or Mickey Mouse either, to beat you all to the pun.)

        And speaking of puns, I wish there were a denialist named Prospect so that I could use the gag, “Where Mann alone pleases though every Prospect’s vile.”

      • I would have quipped in return, something perhaps about, “You are not merely a Prince among men, but a Pope”–but then I realized the allusion is to a line by Reginald Heber, not one by old Mr. Alexandrine:

        Boy, we’re about hip-deep in hypotheticals here, aren’t we?

    • Mike, are you forgetting the nature of D-K?

      If rational assessment of risks were to be expected, this whole conversation would have been pre-empted long ago.

  41. Horatio Algeranon

    “The Troll is always better…”
    — by Horatio Algeranon

    “The troll is always better
    On the other side of the ridge”
    Or so I’m told, by folks who’ve trolled
    …Or maybe it was “bridge”.

  42. Michael Brown is registered to “Hindawi Publishing Corporation” which is on Beall’s list of predatory open-access publishers.

    Hindawi spams thousands with invitations to write articles for their journals. I still have records of receiving such invitations on 3/2/2012, 15/12/2011, 28/9/2011, 5/9/2011, 8/7/2011, 19/4/2011, 4/42011, 10/2/2011 and 11/1/2011 (and many other invitations probably ended up in the trash).

    The entry on Hindawi from Beall’s list follows;

    Based in Cairo, Egypt, this publisher is now on its own after its collaboration with the publisher Sage ended in 2011. This publisher has way too many journals than can be properly handled by one publisher, I think, yet supporters like ITHAKA boast that the prevailing low wages in Egypt, as well as the country’s large college-educated, underemployed workforce, allow the company to hire sufficient staff to get the job done. Still, this publisher continues to release new fleet startups of journals, each group having titles with phrases in common: Advances in … (31 titles) and Case Reports in … (32 titles). It appears that Hindawi wants to strategically dominate the open-access market by having the largest open-access journal portfolio.

  43. Michael Brown

    In my last post I should have stated Hindawi is on Beall’s predatory open-access publisher watchlist rather than list.

  44. Mark Harrigan

    For what’s worth I have had the following exchange with Mr Curtin which reveals how much he really knows about basic physics

    Tim said (in a thread on The Conversation when ridiculing the impact of radiative forcing that I explained from the famous Trenderth energy diagram)

    ” What tosh! How much heat is delivered by 0.9 W per sq. metre? Not enough to boil a kettle!”

    obviously to try and belittle how much the energy imbalance impacted the planet.

    I answered

    “Well Tim the correct answer is talk about how much power is involved and the answer is 459 Terra Watts of power or 459,000,000,000,000 Joules every second.”

    That IS the power delivered to the planet by 0.9 Wm-2

    Not satisfied with this demonstration of his brilliant understanding of physics units Mr Curtin later replied “In brief, it requires 50 Watts to get a kettle to boil in 3 minutes”

    To which I replied “For a perfectly insulated kettle with 1kg of water at STP, it takes 419Kj to increase a litre of water’s temperature from 0 degrees to 100 degrees. That would take 50Watts of power about 2 hours and twenty minutes :)

    If the water were at room temperature it would take 50 Watts about 1hr 50 minutes.”

    He waffled angrily and sort of went silent after that :(

    There was more but at the time of this exchange he was huffing and puffing about his forthcoming “great” paper that would be published in a respectable peer reviewed journal with a credible impact factor (I think it has an impact factor 0f 1.54?)

    I think the main point though is that this paper is doing the rounds of the blogosphere. Perhaps it nees to be properly rebutted?

    Other concerns I had with the paper (I lack the statistics skills to comment on those) are
    P2 – seems to confuse short term changes in CO2 with T
    P3 – makes the flawed statement that CO2 is invariant across the globe (odd given he chose a site in Alaska for his work) when it is known to vary annually by 1-3% and by locale 6-8%
    P7 – compares global annual evaporation of H20 (which of course is part of a cycle) with annual fossil fuel emmissions (of which a large proportion stays there)
    P9 – he compares net radiative forcining at TOA with solar surface radiation (in an attempt to belittle the impact of net radiative forcing) yet appears to ignore all other aspects of the radiative balance at the surface
    P17 – quotes Hougton in relation to how much CO2 is exchanged between the surface and the atmosphere at 20% and therefore concludes the lifetime of atmospheric CO2 can only be 5 years – again failing to understand it’s the net increase that matters

    I look forward to someone in the peer reviewed literature, when they have the time, to rebutting Tims paper. Then again – if no-one ends up citing it because it is riddled with stats errors which apprently Mr Curtin himself admits here – perhaps no one will bother?

    • Lotharsson

      FWIW, he’s still plying his lack-of-understanding at Deltoid.

      The latest gems include:

      Going one better than the infamous 5th order polynomial fit of a year or two ago by fitting a 6th order polynomial. (Next year on to the mighty 7th!) Apparently this 6th order fit indicates that specific humidity will significantly decrease over 2012. Or something.

      Pointing to data sources for his global analysis, and then specifying a longitude range that only covers half the globe.

      Arguing that evaporation is the only variable affecting climate:

      “Evaporation via the sun is not a constant, if it were our climate would be in permanent stasis or equilibrium.”

      …and doing so in a comment that implies that other factors affect climate.

      Claiming that atmospheric water vapour is a forcing that AR4 WG1 “expunged”, and after suitable WG1 quotes were provided claiming that it was expunged as a primary forcing by being relegated to a feedback.

      Asserting that GHGs cool the planet and that N2 and O2 are the real greenhouse gases. This is because the so-called GHGs radiate, which means they radiate to space, and the planet would be warmer without them because when you heat one end of a test tube full of them they don’t radiate the heat to a sensor located outside the other end. This is quite similar to a claim Jo Nova once made to me on an ABC forum:

      …non-GHGs like oxygen and NOx which neither absorb nor radiate heat are the real GHGs while atmospheric CO2 and water vapour are what save the planet from being barbecued because they both absorb AND radiate, as my regressions show. Tyndall and Arrhenius knew that, Tamino … never will.

      No explanation is proffered as to why TC asserted that H2O was previously a greenhouse gas (that explains warming because it’s a “primary forcing”, except the nefarious IPCC suppressed that explanation by treating it as a feedback), whereas now H2O (with its sidekick CO2) is responsible for saving us from the runaway warming effect of the actual greenhouse gases – N2 and O2. Nor is an explanation given as to why the moon, entirely lacking in cooling greenhouse gases, hasn’t “barbecued” itself.

      On then being presented with a quote from Spencer to reiterate the point (“The SECOND misconception is that because greenhouse gases allow the atmosphere to cool to outer space, adding more GHGs can’t cause warming.”), complete with link to an explanatory post, TC claims that the post shows that “because greenhouse gases allow the atmosphere to cool to outer space, adding more GHGs can’t cause warming.”

      You can’t make this stuff up! (But TC can…)

  45. Marcel Kincaid

    At Deltoid, Tim Curtin has now explicitly stated what was implicit in many of his previous claims: that he does not believe that radiation propagates through a vacuum. He also apparently mixes up phlogiston and the ether:

  46. The money quote (HT to Marcel):

    “Then you say; “Radiation happily radiates without the presence of an absorptive/re-radiating gas, even in a vacuum, because it is self-propagating.” Back to phlogiston!

    And then you say even more absurdly; “Given that vacuum doesn’t absorb or re-radiate either, how on earth do you think radiation propagates through a vacuum?” I don’t, as it does not. But that is why an atmosphere consisting only of N2 and O2 is indeed like a vacuum as Murphy has claimed, and that is why N2 and O2 are the real GHGs.”

    • Hmmm, how does the sunlight get here then, since it can’t go through a vacuum. Or maybe its the aether, and Einstein was wrong all along…

      • We’re still awaiting an answer to that question. He’s Gish Galloped on by harrumphing about how he asks the questions “here” (i.e. Deltoid) and everyone else has to answer and galloping back to an earlier claim that the correlation between (the annual seasonal decline in) atmospheric CO2 between May and September and global temperatures means that CO2 cools the planet.

        It’s developing into one of the most impressive delusions I’ve ever seen.

  47. Lotharsson

    And now he appears to be claiming that either (a) the space between sun and earth is not largely vacuum, or (b) radiation can magically jump across a vacuum if it is going to a destination that contains gases that absorb and re-radiate it.

    Option (b) does at least somewhat match with his concept that radiative fluxes only ever proceed from hot sources to cooler sinks, which implies some really astonishing predictive abilities for photon emitting molecules.

  48. He’s all over the place on that thread. One minute he accuses the IPCC of not counting water vapor as a GHG, the next he’s saying that water vapor and CO2 aren’t GHG’s and that they cool the planet. He cites Tyndall and says because N2 and O2 don’t interact with LW radiation, they therefore keep some of it from escaping to space. He says that photons are mostly a fiction. EM radiation can’t travel through a vacuum. The space between the Earth and the Sun is not a vacuum. It’s some of the most astoundingly scientifically illiterate rubbish I’ve ever come across on the internet – and I’ve read a lot of Free Republic.

  49. Bernard J.

    Yes, that Tim-Cube™ thread on Deltoid should be bronzed.

    Nay – it should in fact be smelted for the fools’ gold it contains, courtesy of Curtin, and shaped as a pin for Monckton’s next Nobel…

  50. I thought, when the current debacle started, that nothing would ever eclipse the unforgettable potable seawater thread. Now the full flowering of this unpromising seed has catapulted us into a whole new world.

    If we were in a pub, there’d be just one word shouted in unison with raised glasses.


  51. Michael Brown

    This may be of interest.

    Hindawi’s Scientific World Journal Loses its Impact Factor