Click images for more details



Recent comments
Recent posts
Currently discussing

A few sites I've stumbled across recently....

Powered by Squarespace
« Met insignificance | Main | Significantly Met O££ice - Josh 223 »

Met Office admits claims of significant temperature rise untenable

This is a guest post by Doug Keenan.

It has been widely claimed that the increase in global temperatures since the late 1800s is too large to be reasonably attributed to natural random variation. Moreover, that claim is arguably the biggest reason for concern about global warming. The basis for the claim has recently been discussed in the UK Parliament. It turns out that the claim has no basis, and scientists at the Met Office have been trying to cover that up.

The Parliamentary Question that started this was put by Lord Donoughue on 8 November 2012. The Question is as follows.

To ask Her Majesty’s Government … whether they consider a rise in global temperature of 0.8 degrees Celsius since 1880 to be significant. [HL3050]

The Answer claimed that “the temperature rise since about 1880 is statistically significant”. This means that the temperature rise could not be reasonably attributed to natural random variation — i.e. global warming is real.

In statistics, significance can only be determined via a statistical model. As a simple example, suppose that we toss a coin 10 times and get heads each time. Here are two possible explanations.

  • Explanation 1: the coin is a trick coin, with a head on each side.
  • Explanation 2: the coin is a fair coin, and it came up heads every time just by chance.

(Other explanations are possible, of course.)

Intuitively, getting heads 10 out of 10 times is very implausible. If we have only those two explanations to consider, and have no other information, then we would conclude that Explanation 1 is far more likely than Explanation 2.

A statistician would call each explanation a “statistical model” (roughly). Using statistics, it could then be shown that Explanation 1 is about a thousand times more likely than Explanation 2; that is, statistical analysis allows us to quantify how much more likely one explanation (model) is than the other. In strict statistical terminology, the conclusion would be stated like this: “the relative likelihood of Model 2 with respect to Model 1 is 0.001”.

A proper Answer to the above Parliamentary Question must not only state Yes or No, it must also specify what statistical model was used to determine significance. The Answer does indeed specify a statistical model, at least to some extent. It states that they used a “linear trend” and that the “statistical model used allows for persistence in departures using an autoregressive process”.

If you are unfamiliar with trending autoregressive processes, that does not matter here. What is important is that HM Government recognized, in its Answer, that some statistical model must be specified. There is, however, still something missing: is their choice of statistical model reasonable? Might there be other, more likely, statistical models?

(There is also a minor ambiguity in the Answer, because there many types of autoregressive processes. The ambiguity is effectively resolved in a related Question, from 3 December 2012, which discussed “autoregressive (AR1) processes” [HL3706]; other Answers, discussed below, confirmed that the process was of the first order.)

I found out about the Question (HL3050) put by Lord Donoughue via the Bishop Hill post “Parliamentarians do statistical significance”. I then discussed the choice of statistical model with Lord Donoughue. I pointed out that there were other models that had a far greater likelihood than the trending autoregressive model used by the Answer. In other words, the basis for the Answer to the Question was untenable.

Moreover, I had published an op-ed piece discussing this, and related issues, in the Wall Street Journal, on 5 April 2011. The op-ed piece includes a technical supplement, which describes one other statistical model in particular: a driftless ARIMA(3,1,0) model (again, unfamiliarity with the model does not matter here). The supplement demonstrates that the likelihood of the driftless model is about 1000 times that of the trending autoregressive model. Thus the model used by HM Government should be rejected, in favor of the driftless model. With the driftless model, however, the rise in temperatures since 1880 is not significant. In other words, the correct Answer to the Question (HL3050) might be No.

Lord Donoughue then tabled a Parliamentary Question asking HM Government for their assessment of the likelihood of the trending autoregressive model relative to the driftless model. HM Government did not answer. Lord Donoughue asked a second time. They did not answer. He asked a third time. Again they did not answer. He then asked a fourth time.

A Parliamentary Question that has been tabled in the House of Lords is formally answered by HM Government as a whole. In practice, HM Government assigns the Question to a relevant ministry or department. In our case, the Questions have been assigned to the Department of Energy and Climate Change; the designated minister is the Parliamentary Under Secretary of State, Baroness Verma. Verma obtains answers from the Met Office. The person at the Met Office with final authority is the Chief Executive Officer, John Hirst. In practice, Hirst delegates authority to the Chief Scientist at the Met Office, Julia Slingo. Thus, it is actually Slingo who was refusing to answer the Parliamentary Questions, with Hirst and Verma backing her (perhaps without thinking).

I have had a few e-mail exchanges with Slingo in the past. Slingo has never really addressed the issues that I raised. Instead, she has replied largely with rhetoric and a display of gross ignorance about undergraduate-level statistics; for an example, see the Bishop Hill post “Climate correspondents”. Thus, I decided that trying to talk directly with Slingo about the Parliamentary Questions would be a waste of time. Hence, I tried talking with Hirst. My message to Hirst included the following.

Last week, Lord Donoughue tabled Parliamentary Question HL6132, about statistical models of global temperature data. HL6132 is essentially the same as HL5359, which the Met Office refused to answer. The Met Office Chief Scientist does not have the statistical skills required to answer the Question; there is, however, at least one scientist at the Met Office who does have the skills—Doug McNeall. I ask you to ensure that the Question is answered.

Doug McNeall is a statistician. He and I have had cordial e-mail discussions in the past. In particular, after my op-ed piece in WSJ appeared, on 12 August 2011, McNeall sent me an e-mail stating that the trending autoregressive model is “simply inadequate”. Indeed, that would be obvious to anyone who has studied statistical time series at the undergraduate level. Note that this implies that a statistician at the Met Office has stated that the Answer given to the original Parliamentary Question (HL3050) is unfounded.

Lord Donoughue’s fourth Question was, as before, refused an answer. Afterwards, I received the following message from Hirst.

I would like to assure you that the Met Office has not refused to answer any questions. The questions you refer to were answered by Baroness Verma, Parliamentary Under-Secretary of State at the Department of Energy and Climate Change.

I note that in her response to HL5359 and HL6132, and a number of other questions from Lord Donoughue, Baroness Verma has offered for him to meet officials to discuss this and related matters in more detail.

Afterwards, Lord Donoughue asked the question a fifth time. And I sent the following message to Hirst.

I do not know whether your message is serious or just your way of telling me to get lost. In case of the former, some elaboration follows.

The question that Lord Donoughue has been asking requires the calculation of a single number. The calculation is purely arithmetical: there is no opinion or judgment involved (nor is background in climate needed). Furthermore, the calculation is easy enough that it could be done in minutes, by someone with the appropriate statistical skills. You could think of it as being similar to finding the total of a column of integers.

The number that Lord Donoughue is asking for is 0.001, according to my calculation. (Yes, it is that simple.) Lord Donoughue, though, would like the number calculated by an official body. He therefore tabled Parliamentary Questions asking HM Government for the number.

Lord Donoughue has now received Written Answers to four such Parliamentary Questions: HL4414, HL5031, HL5359, HL6132. None of those Answers give the number. Instead, the Answers make excuses as to why the number is not given. The main excuse seems to be that the number is not important. The importance of the number, however, is a separate issue: even if the number has no importance at all, the arithmetical calculation can still be done, and the number can still be given.

HM Government has been relying upon the Met Office, to supply them with the number; the Met Office has refused to do this. In other words, the Met Office has refused to answer the question—contrary to the claim in your message. What reason does the Met Office have for refusing to supply the number? The required time would be less than the amount of time that the Met Office has spent in refusing.

Parliamentary Questions have a history going back centuries. I do not have expertise in this area, but it is my understanding that HM Government is obliged to either provide an Answer to a Question or else give a valid reason for not providing an Answer. The refusal of the Met Office to supply the number would thus seem to be leading to a violation of a centuries-old parliamentary convention. Indeed, I have now talked with other members of the House of Lords and the Commons about this: there is real concern, and apparently also by parliamentary officials.

Lord Donoughue has now asked for the number a fifth time. The tabled Question is as follows (HL6620).

To ask Her Majesty’s Government … whether they will ensure that their assessment of [the number] is published in the Official Report; and, if not, why not.

The Answer is due by April 12th. My hope is that if the Met Office continues to refuse to supply the number, HM Government will get the number from elsewhere.

There was no immediate response to that. I did, however, receive an invitation from Doug McNeall to visit the Met Office and discuss the statistics of trends in global temperatures. I replied as follows.

Kind thanks for this. In principle, such a meeting would surely be valuable. The Met Office, however, is refusing to answer a simple arithmetical question, and moreover, is presenting dishonest reasons for doing so. Given that, I do not have confidence that discussion could be in good faith.

Hence, I respectfully decline. If the Met Office supplies the number, I would be happy to discuss this further.

A week later, the fifth Question (HL6620) was answered as follows.

As indicated in a previous Written Answer given … to the noble Lord on 14 January 2013 (Official Report, col. WA110), it is the role of the scientific community to assess and decide between various methods for studying global temperature time series. It is also for the scientific community to publish the findings of such work, in the peer-reviewed scientific literature.

Thus, in the opinion of the Met Office, Parliament has no right to ask scientific questions of government scientists.

A few days later, I received the following message from Hirst.

I’m sorry for the delay in replying; I have been away from the office.

I’m sorry if my previous e-mail gave you the impression I did not wish to discuss this matter further. That was not my intention. Indeed, if you are not satisfied with the answers that have been given to Lord Donoughue’s Parliamentary Questions, I would be more than happy for us to debate your concerns, as part of a detailed scientific discussion about the statistical modelling of global mean temperatures.  

I understand Doug McNeall has offered to arrange a meeting with you and other Met Office scientists who work in this area. I feel this would be a sensible way forward and, although our views may differ in some respects, can assure you we would approach this meeting in good faith.

I look forward to hearing from you.

Hirst is clearly supporting the obstructionism. I decided that there was no point in replying.

Under the rules of Parliament, the person with responsibility for a Parliamentary Question is the government minister who delivers the Answer. In our case, that minister is Baroness Verma. According to the Companion to the Standing Orders and Guide to the Proceedings of the House of Lords, §4.68 Ministerial Responsibility, “Ministers should be as open as possible with Parliament, refusing to provide information only when disclosure would not be in the public interest” and “Ministers who knowingly mislead Parliament will be expected to offer their resignation to the Prime Minister”.

Lord Donoughue then sent a strongly-worded letter to Under Secretary Verma, citing the section on Ministerial Responsibility, and adding “I trust we will not reach that point since you are clearly not behind the wilful refusal to answer the Question”. Indeed, Verma seems to have been trusting that the Answers supplied to her by the Met Office were written in good faith.

Then Lord Donoughue asked the question a sixth time (HL62). The Answer, this time, included the relative likelihood. The full Answer (excluding footnotes) was as follows.

There are many ways to analyse time series, including the use of physical and statistical models. The relevance of any technique depends on the question asked about the data. The Met Office has compared the likelihood of the two specified models for fitting the three main independent global near-surface temperature time series (originating from UK Met Office and NASA and NOAA in the US), using a standard approach.

The statistical comparison of the model fits shows the likelihood of a linear trend model with first-order autoregressive noise in representing the evolution of global annual average surface temperature anomalies since 1900, ranges from 0.08 (Met Office data) to 0.32 (NOAA data), relative to the fit for a driftless third-order autoregressive integrated model. The likelihood is 0.001 if the start date is extended back for example to 1850 (Met Office data). These findings demonstrate that this parameter is very sensitive to the data period chosen and to the dataset chosen for a given time period, for such a statistical model.

A high value of relative likelihood does not necessarily mean that a model is useful or relevant. The climate is a highly complex physical system; to model it requires an understanding of physical and chemical processes in the atmosphere and oceans, natural variability and external forcings, i.e. with physically-based models. Work undertaken at the Met Office on the detection of climate change from temperature observations is based on formal detection and attribution methods, using physical climate models and not purely statistical models, as discussed in Chapter 9 of the Contribution of Working Group I to the IPCC’s Fourth Assessment Report, 2007.

The second paragraph gives the relative likelihood of the trending autoregressive model with respect to the driftless model. The relative likelihood is 0.08, if we analyze years 1900–2012 , and it is 0.001, if we analyze years 1850–2012 (using Met Office data). In either case, then, the trending autoregressive model is much less likely than the driftless model to be the better model of the data. Hence, the statistical model that was relied upon in the Answer to the original Question (HL3050) is untenable.

Most of the third paragraph is verbiage. In particular, the cited “physical climate models”, which the Met Office runs on its supercomputer, do indeed provide some evidence for global warming. Physical climate models and statistical models are both known as “models”, but they are different things. It is only the statistical models that are relevant to the Question. The physical climate models, though impressive in many ways, do not provide observational evidence for global warming.

The issue here is the claim that “the temperature rise since about 1880 is statistically significant”, which was made by the Met Office in response to the original Question (HL3050). The basis for that claim has now been effectively acknowledged to be untenable. Possibly there is some other basis for the claim, but that seems extremely implausible: the claim does not seem to have any valid basis.

Plainly, then, the Met Office should now publicly withdraw the claim. That is, the Met Office should admit that the warming shown by the global-temperature record since 1880 (or indeed 1850) might be reasonably attributed to natural random variation. Additionally, the Met Office needs to reassess other claims that it has made about statistically significant climatic changes.

Lastly, it is not only the Met Office that has claimed that the increase in global temperatures is statistically significant: the IPCC has as well. Moreover, the IPCC used the same statistical model as the Met Office, in its most-recent Assessment Report (2007). The Assessment Report discusses the choice of model in Volume I, Appendix 3.A. The Appendix correctly acknowledges that, concerning statistical significance, “the results depend on the statistical model used”.

What justification does the Appendix give for choosing the trending autoregressive model? None. In other words, the model used by the IPCC is just adopted by proclamation. Science is supposed to be based on evidence and logic. The failure of the IPCC to present any evidence or logic to support its choice of model is a serious violation of basic scientific principles — indeed, it means that what the IPCC has done is not science.

To conclude, the primary basis for global-warming alarmism is unfounded. The Met Office has been making false claims about the significance of climatic changes to Parliament—as well as to the government, the media, and others — claims which have seriously affected both policies and opinions. When questioned about those claims in Parliament, the Met Office did everything feasible to avoid telling the truth.


PrintView Printer Friendly Version

References (1)

References allow you to track sources for this article, as well as articles that were written in response to this article.
  • Response
    All right you knuckle-draggin', science-ignoring, global warming denialists! Here's your morning read. If you have a progressive friend on Facebook who watches Jon Stewart all the time, she'll be able to help you with the big words. The Parliamentary Question...

Reader Comments (163)

They are cooked, to a turn and it could not happen to a more deserving crew

May 28, 2013 at 4:05 AM | Unregistered Commenterombzhch

From the Ecclesiastical Uncle, an old retired bureaucrat in a field only remotely related to climate with minimal qualifications and only half a mind.


I believe that on occasions when one liners are the only effective means of communicating your position, you have said - no doubt in more eloquent but equally brief phrases - that you think the globe is warming and that this warming is not much due to human activities.

Do the developments discussed in this and the companion post demonstrating that global temperatures for the period from 1880 (or maybe 1850) seem likely to have occurred by chance mean that you are now in a position to alter your one liner to you DOUBT the globe is warming etc? Or something similar?

Or is 1880 too recent?

Your one liners have political significance in a way that the learned arguments that adorn this thread do not.

This old ex-bureaucrat looks forward to your next broadcast!

May 28, 2013 at 5:40 AM | Unregistered CommenterEcclesiastical Uncle

Quite the most astonishing statement in this article is that 'the Met Office does not consider that Government should ask scientists questions'.

As the Government funds the Met Office and as the Government IS accountable for its spending, it is a sine qua non that they DO have the right to ask scientists questions, indeed they have a duty to UK taxpayers and citizens so to do.

For the Met Office to claim otherwise is not only untenable, it is an immediate sacking offence for its entire senior leadership.

If they can't understand that, they are idiots of the highest order.

May 28, 2013 at 8:35 AM | Unregistered CommenterRhys Jaggar

Rhys Jaggar:

Quite the most astonishing statement in this article is that 'the Met Office does not consider that Government should ask scientists questions'.

Both astonishing and not in the article, according to my browser search. Would you mind giving the number of the paragraph in which this statement occurs?

May 28, 2013 at 8:43 AM | Registered CommenterRichard Drake

Doug (and other readers still capable of independent thought): There are many possible ways of answering your initial question: Is the 0.8 degC warming since 1880 statistically significant?

How about averaging the temperatures for 1865-1895 (30 years around 1880) and the most recent 30 years? (Anything less than 30 years is usually considered to be "weather" rather than "climate".). Let's compare the mean and standard deviation for each of these periods by the usual statistical method. What is the probability that we can reject the null hypothesis that the mean temperature difference between these two periods might be zero? I haven't done the calculation, but all thinking skeptics must know what the result of such a calculation will be - it's unambiguously warmer today. No fancy statistics can change this fact.

A second way - but not the only way - to answer your question about whether it is warmer is to ask if the slope of a linear fit to the data is significantly greater than zero. This requires one to understand what type of noise is present in the data. Unfortunately, there is NO reason to postulate that any linear function (plus noise) is a good model for how temperature has warmed. We know that any model that incorporates all of the physics expected to influence climate will include terms for GHGs, TOA solar radiation, and aerosols, and might include terms for oscillations like the PDO. (With only a century plus of data, a 60-year oscillation like the PDO will ruin any model relying on a linear trend plus autocorrelated noise.). Richard Telford is exactly right; beating an inappropriate linear model doesn't prove anything about the statistical significant of warming. Such models are merely straw-men! Admittedly, the IPCC'S stupidly used such models to calculate inappropriate confidence intervals in AR4, but their mistake doesn't make your use of a linear trend plus noise model correct.

Why do you arbitrarily chose to start your linear model in the 1880 or 1850? Why not go back to 250 years ago (the beginning of the Industrial Revolution) or 1000 years ago or 10,000 years using proxy data? That will guarantee that any linear trend you find is indistinguishable from zero. Your choice of starting dates is equally stupid as these earlier dates - radiative forcing due to accumulating GHGs was negligible back then. The IPCC attributes most warming to anthropogenic GHGs ONLY in the second half of the twentieth century. They weren't honest enough to explain that earlier warming represented mostly natural variation, but the attribution papers came to this conclusion. When the signal for GHG mediated warming wasn't clear until the second half of the twentieth century, there is NO reason to look for a linear trend starting in 1850 or 1880.

You are right to criticize the IPCC's use of linear trend plus correlated noise models for calculating confidence intervals around warming trends. The inappropriate use of such models doesn't constitute evidence that it hasn't been warming. The crucial question is how much warming is natural variation and how much is caused by GHGs. Nic Lewis addressed how much warming is likely to have been caused by GHGs with his posts on ECS and TCR. If TCR is about 1.3 degC and if net forcing is equivalent to half way to 2x CO2, then about 0.65 of your 0.8 degC warming is due to GHGs, no matter what inappropriate linear trend plus noise models you fit to the data.

May 28, 2013 at 8:52 AM | Unregistered CommenterFrank

(a) Doug is not questioning whether there has been warming. He's questioning whether the observed warming could due to natural random fluctuations. His answer is yes.
(b) His model is based on the data from HADCRU and GISS. That's why he starts when they do.

May 28, 2013 at 9:47 AM | Registered CommenterPaul Matthews

Frank asks: "Why do you arbitrarily chose to start your linear model in the 1880 or 1850?"

This isn't in any way arbitrary. It is the start date of the two high res. records that are available. By using any of the earlier dates that you suggest, you introduce so much noise and uncertainty that the recent warming cannot be shown to be significant, unless you perform one of the resolution-changing con-tricks so beloved of the paloe crowd.

May 28, 2013 at 9:53 AM | Unregistered Commentersteveta_uk

Congratulations to Mr. Keenan and Lord Donahue. Someone might care to point out that the 0.08 since 1900 is statistically insignificant and also question why MET office data differs so significantly from NOAA data. Does the MET office admit that NOAA are more adept at fiddling?

May 28, 2013 at 10:32 AM | Unregistered CommenterSolomon Green

is it worth agreeing an unambiguous question, or set of questions, to pass on to our respective MPs regarding these and recent findings, so we can applying more widespread pressure?

May 28, 2013 at 10:53 AM | Unregistered CommenterVarco

Paul, Steveta -
I think Frank and Richard Telford have it right. There's no reason to expect that a constant trend should exist all the way back to 1880; anthropogenic forcing has increased greatly since then. And a driftless ARIMA(3,1,0) process may fit better than an AR(1)+drift model, but it is not physically consistent as a model of climate "noise" because the integrative part (the "1") causes its variations to grow unboundedly with time.

For more detail, I'd suggest Lucia's blog "The Blackboard; here's a recent short comment and an earlier post.

Keenan suggests that because ARIMA(3,1,0) fits better over the long period than AR(1)+drift, he has proven that it should be used (and hence that warming is not statistically significant). But those are not the only possible models. Other reasonable models would include an oscillatory component (of duration ~60 years) which accounts for some of the "noise". And would not assert a constant drift rate over a century, when forcings have not been at all constant over that period.

May 28, 2013 at 10:55 AM | Registered CommenterHaroldW

Probably worth reading the full text of Doug McNeall's email to Doug Keenan. Doug K links to it above, but here it is anyway - I'd encourage folks here to read it carefully, as there are important points here about what is meant by "significant".

Dear Doug Keenan,

Thanks for clarifying your concerns somewhat; hopefully I can set your mind at rest on some of your worries. I suspect that parliamentary questions are not the best route through which to discuss some of the technical aspects of statistical theory.

Please rest assured that we do not use a linear statistical analysis of global temperature trends in anything other than a descriptive manner. I believe that the trends were reported as linear as a method of description for consistency with the IPCC Fourth Assessment Report. The report itself acknowledges the deficiencies of the statistical model that it uses, while arguing that the analysis still gives some useful information. I’m inclined to agree.

These trend lines are just there to summarise the data, and shouldn’t be used (for example) to extrapolate a forecast of global temperatures in the future. They shouldn’t be taken as a representing the “true” statistical model for global warming. Perhaps you might suggest other methods for summarising the data?

I think the problem is the proxy war being fought around the word “significant”, and the difference between its scientific and formal statistical use. We should be careful to separate out their use. There appear to be individuals who would like to say that the evidence for warming is “not statistically significant”, and others who would like to say that it is. This is to misunderstand the
nature of “significance”.

Of course, we both know that the correct answer to the question “is the trend of global mean temperature statistically significant?” is not “yes”, or even “no” but, “that is not a valid question”. This is because the appropriate statistical model to use for the timeseries is not known perfectly, and any statistical significance test uses assumptions about that timeseries that may turn out to be invalid. Cohn and Lins (2005) for example, state “significance depends critically on the null hypothesis which in turn reflects notions about what one expects to see.”

This does not mean that there is no evidence of a significant increasing temperature trend ­there is plenty of scientific evidence for this just that using naive statistical tests in the absence of other information is inappropriate.

A significance test attempts to answer the question “given that there was no anthropogenically driven global warming, what is the probability that we would see these temperatures?” This is interesting, but not really what we are looking for. As you and others have noted, this kind of test does not allow you to distinguish between forced trends, and the degree of long term persistence in the system. I would suggest a Bayesian solution to this problem. I think the appropriate question is “given that we see this these temperatures, what is the probability they are anthropogenically driven?” Using Bayes theorem, this combines the likelihood (from the first question), with the prior probability that global temperatures are anthropogenically driven, to some degree. Of course, this prior probability contains subjective judgements and information from elsewhere, including fundamental physics.

Our judgements about the probability that temperatures are anthropogenically driven, contain information not just from the global temperature trend, but also our knowledge of the way that the system works. The global temperature series in isolation simply does not contain the information we are looking for.

The Bayesian solution cuts through the problem that you identified, of being able to tell the difference in the validity of (as in your example), a linear trend with an autoregressive element, and a driftless autoregressive integrated model. The model that we choose has to be consistent with the understanding of the system that we gain from other data, and our basic knowledge of the physics of the system. We encode our knowledge about physics, and the rest of the system in climate models, and their simulations. Again, I stress that a great deal of work has been done in this field, but that it is more likely to be found in the detection and attribution literature (and corresponding IPCC chapter), than in the observations literature.

In conclusion, my suggestion is that when asked if there has been a “significant” change in temperatures since the 1880s, we should say “yes”. If we are asked if there has been a statistically significant change in temperatures since the 1880s, we do not say “yes” or “no”, we say “that is not a valid question”. A difficulty may be in getting people to understand the reasoning sufficiently to accept the latter as the correct answer to the question. In this, I welcome your help.


Doug McNeall

Cohn, T.A. and Lins, H.F. (2005) Nature’s style: Naturally trendy, Geophys. Res. Lett., 32,

May 28, 2013 at 11:24 AM | Registered CommenterRichard Betts

Harold, Huh?
"There's no reason to expect that a constant trend should exist all the way back to 1880"
This is not what Doug is doing.

"he has proven that it should be used"
This is not what Doug is claiming.

I second RB, that we should read carefully the comment from the other Doug. In particular, the question "Is temperature rise significant" is a fairly meaningless one because the answer depends on your statistical model for the data.

May 28, 2013 at 11:49 AM | Registered CommenterPaul Matthews

This thread makes an interesting change from the boring lukewarmist navel gazing present in so many BH threads lately.

May 28, 2013 at 12:23 PM | Registered CommenterRKS

Paul -
The "AR(1)+drift" model assumes that the drift is constant over the entire period of analysis, which is what I intended to convey by the phrase "a constant trend should exist all the way back to 1880". To the extent that this is not so, that model's consistency with observations is reduced. We believe that forcings have not been constant (or even nearly so) since then, hence we should not expect a constant drift term. To apply this model over such an interval, is to create a strawman.

As to what Doug Keenan is claiming, he wrote "the trending autoregressive model is much less likely than the driftless model [viz., ARIMA(3,1,0)] to be the better model of the data. Hence, the statistical model that was relied upon in the Answer to the original Question (HL3050) is untenable." Perhaps I've read too much into that statement, but it seems to me that Doug K. is suggesting that ARIMA(3,1,0) should be used for statistical significance tests.

I appreciate Doug McNeall's statement that statistical significance can only be determined in the context of a particular statistical model. I wouldn't go as far as he, that it a "not a valid question", but it is certainly not as straightforward a question as it appears on the surface.

May 28, 2013 at 12:26 PM | Registered CommenterHaroldW

So the Met Office gave a clear answer. Deny, deny, deny!!

May 28, 2013 at 12:32 PM | Unregistered CommentercRR Kampen

RB - The email does indeed bear close reading, but it also raises further questions. There is a very interesting interplay of the various uses of the word 'significant'. Doug McNeall mentions it's scientific and statistical variants, but not the common usage. The question that the man on the Clapham omnibus needs the answer to is firstly 'has it warmed' - to which the answer is clearly 'yes'. Then he needs to know 'has it warmed "significantly"', where the word this time is used in the common way to mean 'has it warmed to a large and important degree', where the scientific meaning would have the question ' has it warmed to an extent that we can reliably see that in the figures and differentiate it from the noise'.

So far as I read it, we can say that yes, there is warming, but there is a lot of umming and ahhing about firstly whether it is noticeably different to previous such episodes, and secondly that attributing the cause, given imperfect understanding of the variables, is 'problematic' and certainly 'debateable'.

So, the warming signal may or may not have risen above the noise - you gentlemen are paid to prod and poke around that one - but it certainly ain't at the level that our esteemed customer of the London public transport service should have to bother his head about it. Which is a pity, because he's being robbed blind on the basis that what little heat increase we have seen is going to doom the earth and all it's multifarious peoples.

May 28, 2013 at 12:46 PM | Unregistered CommenterCumbrian Lad

"Perhaps I've read too much into that statement, but it seems to me that Doug K. is suggesting that ARIMA(3,1,0) should be used for statistical significance tests."

Yes, you are reading way too much into that statement. I've not read anyone suggest that ARIMA provides a beter model of a climate - all Doug is saying (IIUC) is in effect that even> this model is better, so the one used by the MO is clearly unsupportable.

May 28, 2013 at 1:06 PM | Unregistered Commentersteveta_uk

As a straight A non scientist trying to make ANY sense of this.....the mention of water vapour jogged my memory and I recommend the following to all those who have an open and therefore enquiring mind, but who crapped out in Maths and Science when at school:

Naturally for those of you who are familiar with this work, apologies.

The lack of a timely and swift, honest response should shock no one who has any experience of dealing with large organisations with agendas that are not in the public interest.

May 28, 2013 at 2:46 PM | Unregistered CommenterIPMS

@ Richard Betts, 11:24 AM

The letter that your comment quotes is also available on one of Doug McNeall’s blogs:

The first comment there includes an extract from my reply. This was discussed in an earlier post on Bishop Hill, “Questions to ministers”.

I used to be a supporter of the Met Office. This year, three things have happened to change that: the indefensible press release about increased UK rainfall (partially described in the BH post “Climate correspondents”); the purely-rhetorical answer that I got from Chief Scientist Slingo when I asked about the derivation of statistical assertions sent to the Government Chief Scientific Adviser (post in preparation); the Met Office obstructionism in response to parliamentary questions and the extreme reluctance to acknowledge that the claim of statistical significance is untenable (described in this post).

It seems clear that the primary concern of the Met Office is far from science. If that is indeed the case, then the Met Office needs to change.

May 28, 2013 at 3:03 PM | Unregistered CommenterDouglas J. Keenan

Richard Betts,

The way the Met Office handled this, is a shame and you know it.

May 28, 2013 at 4:19 PM | Unregistered CommenterIbrahim

It is a little frustrating when the Met's stalwarts turn up to argue over semantics and definitions on a statistical question but do not defend the lazy statements of their chief scientist and others when they are supposed to be explaining AGW and say CO2 traps heat with no logical chain or reference. Are we supporting back-radiation at the moment or the lapse rate/ TOA meme? How's that best evidence thing coming? Now that the statistics have let us down and the paleo has left us confused, what's left?

Or. let's make it simple. If I say 'nothing much is happening' what have you got to show that is wrong?

May 28, 2013 at 4:22 PM | Unregistered Commenterrhoda

Thanks, Doug. A very good article!
The most proper statistical model should be used to analyze the data, otherwise we are deceiving ourselves. This deception has been amplified by the alarmist press and the IPCC.
I have placed links to this page in my climate pages at Observatorio ARVAL.

May 28, 2013 at 5:05 PM | Unregistered CommenterAndres Valencia

Richard Betts has made it clear that his position (and presumably that of his employer) is the only valid question is how much warming CO2 causes and that is the only question he is prepared to countenance. We are all witness to the fact that the historic output from the MO is premised on the assumption that CO2 is in some magical way the control knob of global climate.

Doug McNeall hints at this when he says "Our judgements about the probability that temperatures are anthropogenically driven, contain information not just from the global temperature trend, but also our knowledge of the way that the system works." Well Doug and Richard I fear that you guys dont have nearly enough knowledge of how the system works to be able to make any claims about the future path of climate change.

And you have a bigger problem - as Douglas Keenan points out in his comment at 3:03 pm - you are now perceived as activists rather than scientists. A state of affairs that needs to be corrected pdq. On past performance, the prospects dont look good.

May 28, 2013 at 5:21 PM | Unregistered CommenternTropywins

In response to Doug Keenan's article, a famous mathematician who believes in AGW wrote to me:

It's a bit odd that you're willing to accept unfounded proclamations like the one you just sent me, claiming the Earth had not warmed up since 1880. Have you looked at graphs of the Earth's surface temperature, land surface temperature, etcetera, prepared by the various groups of scientists who study this stuff? Do you know who the 3 main projects are to measure these trends, and what their results are?

May 28, 2013 at 6:35 PM | Unregistered CommenterMarvin Jay Greenberg

Does this mean that when someone asks us "has the Earth's temperature warmed in the last century?" We can say "NO!"

May 28, 2013 at 6:39 PM | Unregistered CommenterScott Scarborough

We may have evidence of increasing temperature, but we DON'T have evidence of a systematic cause? All we have is judgements based on shaky global temperature series and a crude knowledge of how we think the system might work.

May 28, 2013 at 6:52 PM | Unregistered CommenterMartyn

Nothing will happen until a government is formed that does not include wet Tories, LibDems or Labour AND we are no longer shackled to the EU corpse. There is too much political capital at stake and none of our current people in government has to courage to admit that they were fooled.

There are billions of pounds / euros / dollars being spent every year in propaganda to keep the people uninformed. The number of people like us is growing, but we are faced with a most formidable adversary who uses all the propaganda tools honed by the Soviet Politburo over decades (compared to them, Goebells was just a naughty boy).

Nothing will happen in the life of this parliament because of the LibDems veto. Nothing is likely to happen in the life of the next parliament either. The only hope is for a game-changing event: a national blackout; the collapse of the EU; national bankruptcy and IMF intervention; etc. Barring all this we'll see the status quo unchallenged for at least a decade.

I wish it were different, but the key word is REALPOLITICS.

May 28, 2013 at 7:28 PM | Unregistered Commenterroman_column


Thanks for that other graph.

May 28, 2013 at 8:34 PM | Unregistered CommenterJames Evans

How about this: If I say the earth has cooled in the last 10 years the Warmist says "No it hasn't, that change is not statistaclly significant." So I reply: "Then it hasn't warmed in the last century." The Warmist says: "Of course it has, look that the temperature chart." Then I say: "But that change is not statistically significant!" Is that how it works?

May 28, 2013 at 8:47 PM | Unregistered CommenterScott Scarborough

It does seem somewhat obscure to the untrained eye and I can’t see the MSM picking up on it any time soon, even if they wanted to.
Full marks to the characters involved for persisting & shining a light on this small corner.

Intrigueing as it is there is more to significance than statistics, which is sensitive to selection of both model and endpoints.
I look forward however to seeing how clearly the significance of this discovery can be communicated to a naive audience.

May 28, 2013 at 9:36 PM | Unregistered CommenterEddie Sharpe

Thanks, Doug, for this post and Shub for focusing on the key issue: "The physical climate models, . . . , do not provide observational evidence for global warming."

Five authors left information from the 1945-46 era to decipher Climategate in the context of world history: Sir Fred Hoyle, George Orwell, Paul Kazuo Kuroda, David Snell and Robert Jungk.

One of these (PKK) became my research mentor in 1960 and assigned a research topic, "The origin of the solar system and its elements," that would be answered by precise experimental data and observations from the Nuclear & Space Ages:

If my interpretation of these measurements and observations is correct, then government-funded science has been primarily concerned with hiding the source of energy stored in cores of heavy atoms, planets, stars and galaxies after the Second World War ended.

Sir Fred Hoyle published two landmark papers [1,2] in 1946 that formed the foundation of
a.) The Standard Solar Model of H-filled stars, and
b.) The Big Bang Model for H-production at time, t = 0

With kind regards,
Oliver K. Manuel
Former NASA Principal
Investigator for Apollo

[1] Fred Hoyle, “The chemical composition of the stars,” Monthly Notices Royal Astronomical Society 106, 255-259 (1946).

[2] Fred Hoyle, “The synthesis of elements from hydrogen,” Monthly Notices Royal Astronomical Society 106, 343-383 (1946).

May 28, 2013 at 10:54 PM | Unregistered CommenterOliver K. Manuel

I will add my voice to those of Latimer Alder and others.

This is a huge story in need of a sub editor, a convenient handle, and a Leveson style Commision of Inquiry.

May 29, 2013 at 1:37 AM | Registered Commentertapdog

Wow. Neat.

I used to be a government analyst, and I'm sad to say I saw similar examples of statistical mendacity all the time. Only the favoured narrative got reported, even if simpler expanations fitted the data better. Sligo [Slingo.BH] is an unfortunate example of a tendency in department "chief scientists"; political appointees without real technical skills who rise to their station from parroting the views of their masters. She's clearly dealing in bad faith in this exchange.

May 29, 2013 at 2:55 AM | Unregistered CommenterAlistair

Interesting analysis, but why is the Earth retaining heat, and won't that cause an increase in temperature. One reason I can see is that heat is being 'used up' melting the Artic Ice cap and the majority of ice in glaciers. The temperature will increase when all the ice is gone.

May 29, 2013 at 8:30 AM | Unregistered CommenterPeeBee

Steveta and Paul: Thanks for the replies, but poor performance of a LINEAR trend plus noise since 1880 is simply a strawman. Accurate collection of data began in the late 19th century, but GHG mediated warming didn't become important for another half century. Why doesn't Keenan tell us about trends since 1930 instead of 1880 and 1850?

Keenan's choice of a linear trend is suboptimal given what we know about the physics of climate. Volcanos increase the natural variability that complicates detection of effects due to GHGs. Keenan's approach ignores what we know about the properties of volcanic aerosols; detection and attribution methods make use of this knowledge. Unfortunately, the D&A scientists have far too much flexibility when choosing the data they use. If they chose a data set with the missing hotspot in the upper tropical troposphere, less warming would be attributed to GHGs.

If you want a sensible answer to the question of how much warming is probably due to GHGs, consult Nic Lewis's posts! His approach is a far more sensible one than Keenan's strawman.

May 29, 2013 at 9:53 AM | Unregistered CommenterFrank

Richard Betts,

Your explanation doesn't wash. Just a simple answer would suffice. Could you please tell us if the model used by MO is correct, valid and is the right model for calculating statistical significance for warming from 1880 till today? If so, could you tell us the number asked for by Lord Donahue? After all it is MO who said that the warming is statistically significant. If so, they should be able to back it up with specific numbers and methods of calculations and confirm it in minutes. After all, they've been beating their drums and wailing about unprecedented warming, statistically significant warming etc. and beating their chests. Based on their statements policy decisions affecting millions of people and billions of dollars are being taken.

So when they fail to answer a simple question 5 times and find every nuance, excuse and reason for delaying, obfuscating and outright lying, pardon me if I don't believe the defence you are providing to us on behalf of your salary paying masters. An even to a layman like me the reply they have given is weasel worded and literally acknowledges that the statistical method they are using for this calculation is not the correct one. A simple graduate student in statistics knows that by now.

So if MO have nothing to hide all they need to come out and say is that" Hey you guys, the warming is statistically significant from 1880 till today, here is the number for significance and here's how we calculated it ". A blameless straightforward person or organisation would do that. As a professional would you agree with that?

Instead, what we get is this endless bullshit and buck passing which we've too often seen from politicians and too often seen from Julia Slingo. Please don't defend that.

May 29, 2013 at 9:56 AM | Unregistered CommenterVenter

Well done Doug.....don't let up, you're family tree shows a long lifeline!

May 29, 2013 at 12:56 PM | Unregistered CommenterJoffre Burger

Global warming is at its most fundamental level the result of a radiative imbalance at top of atmosphere. Such an imbalance results in an accumulation of thermal energy in the oceans, atmosphere, and in the land masses of Earth. The imbalance also results in the melting of land ice and sea ice. Large differences in heat capacity dictate that approximately 93% of the excess heat will go into the oceans, and only 2-3% each into the atmosphere and land masses. The most extensive and accurate measurement system to date of ocean heat content, the ARGO float system, shows an unabated increase of thermal energy into the oceans, and GRACE and mass balance measurements show unabated melting of land ice on Earth. Coupled ocean-atmosphere heat transfers dramatically affect atmospheric temperatures while having little effect on OHC, again due to heat capacity differences. Judging global warming by atmospheric temperatures alone is woefully incomplete science.

May 29, 2013 at 1:05 PM | Unregistered CommenterOwen

You are so wrong on every level.

May 29, 2013 at 1:28 PM | Unregistered Commenterlurker, passing through laughing

Owen, care to tell us how the heat gets into the deep ocean. And how ARGO represents a data set that can tell us anything yet? You seem to have amassed a giant heap of certainty which is not quite warranted.

May 29, 2013 at 1:32 PM | Unregistered CommenterRhoda


See Note also that sea level measurements are consonant with ARGO results - powerful consilience. ARGO is the very best system we have - if it showed a drop in OHC, I assume that you would be its champion.

May 29, 2013 at 2:32 PM | Unregistered CommenterOwen

A paywalled reference whose summary smells of false certainty? How did the heat get down there? Is this supposition or desperation? Oh, a reanalysis using multiple disparate data sets. That's confidence-inspiring. Either way it needs more data to move out of the conjecture category. ARGO? Is not widespread enough, hasn't been going for a long enough time in relation to ocean cycles. Maybe it will tell us something someday. I assure you that I would distrust a drop in OHC, if argo was capable of doing more than indicate one.

May 29, 2013 at 2:48 PM | Unregistered Commenterrhoda

Fantastic! A good does of sunlight.

May 29, 2013 at 2:51 PM | Unregistered CommenterKevin

Yes Rhoda

the bogeymen that used to live 2 meters above the surface of the land now live in the deep oceans. Be very afraid!

or not

May 29, 2013 at 2:54 PM | Unregistered CommenternTropywins

Owen, thank you.
"... woefully incomplete science."
Hundreds of billions on the table, and we are led by failed theology students and Hollywood special effects artists,... led into a fear of our world and for our future.
Man up, Owen: give it to us straight, "Trust those bureaucrats, weaseling around an incomplete science - they know what is best for the entire world." Or not.

May 29, 2013 at 3:02 PM | Unregistered CommenterJohn R T

In the letter from Doug McNeall, he uses phrases such as:

" it depends upon what you expect!!;"

"this prior probability contains subjective judgements and information from elsewhere, including fundamental physics"

"As you and others have noted, this kind of test does not allow you to distinguish between forced trends, and the degree of long term persistence in the system;"

With my simple level of stats I would assume that if you have a dataset, temps for each day over a number of years there must be a series of statistical tests that can be performed on the data to find out the amount of noise (SD), slopes or trends and probabilities.
I did not realise we need to decide what we are looking for before we ran the tests?

May 29, 2013 at 3:14 PM | Unregistered CommenterSteve Richards

People have been talking about fitting cycles to temperature data but I haven't seen anyone actually doing it. So I had a go.

Of the three options (linear trend, flat, or cycles) which do you think makes most sense?

May 29, 2013 at 3:19 PM | Unregistered CommenterJeremy Shiers


You did not respond to the sea level data, which clearly indicates continued warming, Both from thermal expansion and melting of land ice. BTW, the latest 3-month 0-2000 m number is out on OHC from ARGO, showing continued increase.

May 29, 2013 at 4:29 PM | Unregistered CommenterOwen

Doug McNeall wrote:
“I would suggest a Bayesian solution to this problem. I think the appropriate question is “given that we see this these [sic] temperatures, what is the probability they are anthropogenically driven?” Using Bayes theorem, this combines the likelihood (from the first question), with the prior probability that global temperatures are anthropogenically driven, to some degree. Of course, this prior probability contains subjective judgements and information from elsewhere, including fundamental physics.”

My understanding of Bayes’ theorem is that by feeding in one known (or supposed known) probability any other probability can be lessened or increased, even when this appears to defy logic.

If Doug McNeall really advocates this approach then he should also be advocating that his colleagues explore alternative questions such as “given that we see this these [sic] temperatures, what is the probability they are driven by solar variation?”, “Given that we see this these [sic] temperatures, what is the probability they are driven by changing surface reflectivity?”, “Given that we see this these [sic] temperatures, what is the probability they are driven by interstellar dust?”, “Given that we see this these [sic] temperatures, what is the probability they are driven by natural oceanic activity?”…

While it has been proven in laboratory experiments that CO2 is a greenhouse gas, its full effect on the earth’s climate has yet to be demonstrated by experiment. When I leapt to a dubious conclusion from a successful experiment my science teacher at school reminded me of the discovery of antimony by a Germany monk, who tested its properties by feeding it to the monastery’s pigs. Noticing that the swine thrived upon it, he fed it to his fellow monks, poisoning them. Hence the name “anti-moine” or antimony.

May 29, 2013 at 5:58 PM | Unregistered CommenterSolomon Green

Owen, I don't doubt the warming of the ocean, per se. I consider the data deficient, but there we are. However, there is the lack of a plausible mechanism for AGW leading to ocean warming at depth. Just because Trenberth thinks he has found his 'missing heat' there and somehow made the numbers match up does not convince, short of a proper mechanism and some measurements and observations of it happening. Other theories of ocean circulation have the idea that timescales of centuries and millennia are involved. Am I to believe that? And sea level? I think I've explained that I don't know what the oceans are doing and I suspect the veracity or gullibility of those who say they do.

The entire sceptic/warmist problem is that of uncertainty vs certainty. Even if the warmists turn out to be right I can't see how they have reached a position of certainty based on the data we have. I ask them repeatedly here, to no useful effect.

May 29, 2013 at 6:12 PM | Unregistered CommenterRhoda

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>