Friday
May312013
by 
Bishop Hill
Bishop Hill 
Buy
Books
Click images for more details
Currently discussing
-
Nov 19 - Barbie Jacket
on
Need an economic adviser? Here are 5 ideas for finding the very best one -
Nov 11 - US Design Hub
on
Launched my very own online store.
Links
A few sites I've stumbled across recently....
-
Science blogs
-
Political blogs
-
Other blogs
Subscribe
Copyright © 2006-2013, Bishop Hill All rights reserved.
Reader Comments (102)
The MET Office blog post states (point 1):
"The basis for this claim ('...the claim that the increase in global warming is larger than could be explained by natural variability...) is not, and never has been, the sole use of statistical models to emulate a global temperature trend. Instead it is based on hundreds of years of scientific advancement, supported by the development of high-quality observations and computational modeling.
A bureaucratic argument from authority that refers to another political authority (IPCC). Tiresome, repetitive, unscientific and circular. Is computational modeling bereft of statistics? Are we just word-fiddling for fun here?
Just show me the compelling evidence of anthropogenic cause before we're all freezing our socks off, or possibly what warming has occurred + 'adjustments', has reversed.
If the last seventeen years of negligible and statistically insignificant warming in the face of rising atmospheric concentration of CO2 does not falsify the AGW hypothesis, the MET office climate scientists should be asked what observations they would accept that would falsify the hypothesis.
I mean, the hypothesis is falsifiable isn't it? Isn't it?
Steve Mosher,
Which of the IPCC models, if any, do not predict something, or more likely everything, wrong in the future? And how do they do it?
I've not yet met a model that won't leave the rails unless it has some explicit, arbitrary, and unscientific, constraints to prevent it from doing so.
The more interesting question, to me, is how long does it take for any given model to shoot off into Narnia.
"The Slingo paper (did she really write it, do you think?) "
Compiled.
Acknowledgement is given to Dr. Colin Morice, Dr. Doug McNeal, Dr. Peter Stott, Dr. Nick Rayner and
Dr. John Kennedy, all from the Met Office, for their time and expertise in helping to prepare this paper.
@shub says:
'IOW, the question "has it warmed?" cannot be answered by just looking at the instrumental temperature record.'
Is there another form of 'warming' that won't show up in increased temperature? Bu....ed if I can think what it'll be......
But maybe the Met Office are privy to some other form of Thermodynamics not available to mere mortals like us?
May 31, 2013 at 9:54 PM | Martyn
Which one of them spells 'modelling' 'modeling'?
May 31, 2013 at 8:42 PM | Nullius in Verba
So, from the perspective of an Uninformed Lay Person (ULP), perhaps it wouldn't be too far off the mark to suggest that the arguments in Slingo's paper are somewhat shrouded in ... uh ... fog
Not sure about the full paper but, using the above ... uh ... "model", the Executive Summary yielded a rating of 18.41.
Considering that ...
... one might be inclined to ask Slingo if she'd care to divulge the intended audience with whom she was attempting to "communicate";-)
May 31, 2013 at 9:54 PM | Martyn
Which one of them spells 'modelling' 'modeling'?
May 31, 2013 at 10:24 PM rBilly Liar
Their spell checker. They probably left it at its default "US dictionary" setting.
Steve Mosher said
"Keenan chose a model that fit the data better" - I am simply speechless at that statement
Useful to remember that the test of a model is its utility, of which goodness-of-fit is just one dimension.
In this case the use to which the model was being put was to measure "the rate of warming in the temperature series" with goodness-of-fit being being the only criteria (basically the discussion was proceeding on the basis that this time series was abstract from any real world system such as global temperature). On that basis you get one answer.
If on the other hand you want to use this model for other purposes then other selection criteria come into play for your model, and the preferred model becomes different.
[and this one. BH]
Mod -
My comment wasn't worth publishing twice (must have hit "refresh") - can you delete the second one?
[Done. BH]
"Is there another form of 'warming' that won't show up in increased temperature?"
'Warming' is easily detected by subtracting a later temperature from an earlier temperature and seeing if the resulting number is positive. It's not the issue.
What we're really asking is whether there is more warming than expected from natural variation, which is why they talk about 'significant warming'. And it is quite true that you can't tell just by looking at the data. You have to know what the normal background variation looks like. If your only clue to what the normal background looks like is the same bit of data that you're asking the question about, that might or might not have a signal in it, you can't deduce anything. If you build your background noise model assuming there's no signal, you'll find no signal. If you build your model assuming there's a signal of a particular form, you'll find a signal of that form. It's circular reasoning. You need to input some extra information or assumptions from outside.
"one might be inclined to ask Slingo if she'd care to divulge the intended audience with whom she was attempting to "communicate""
Probably the sort of audience that asks Questions in Parliament about "driftless ARIMA(3,1,0) models", and expresses a deep interest in the answers. We know what she means.
I put this quote up a while back on unthreaded. I like its concurrent simplicity and profundity, so excuse me for repeating it in the context of the statistical significance in the temperature records.
'Now for global warming. Of course we are going through a period of warming, but so far as human culpability is concerned I am a total sceptic and I fear we are dealing with political manoeuvring. There was, for example, much greater marked warming at the end of the Maunder Minimum; what about the Mediaeval Maximum, when Britain was hotter than it is now? No doubt, the present period of warming will be followed by a period of cooling, as has happened in the past time and time again. After all, the Sun is to a mild extent a variable star and we cannot control it –the Late Sir Patrick Moore 2007
http://onlinelibrary.wiley.com/doi/10.1002/wea.38/pdf
Kudos to you Nullius in Verba!
You managed to not only read but to follow Slingo's paper and then translate it for our benefit. I read, then re-read sections and still decided that Slingo's paper was classic Nixonian obfuscation at it's worst.
After Martyn's clarification that Slingo's paper was compiled by a committee, the sheer emptiness of it's wording becomes more understandable.
MetO's blog page is a very disappointing post for any organization to claim as theirs. MetO's oblique and direct ad homs against Doug are embarrassing to any professional.
Given the animosity displayed, I suppose MetO will next blame Keenan for the demise of cock Robin and delaying spring?
Time has arrived for England to source their weather and climate prognostications from anywhere but the Met. Defund MetO.
HAS wrote (twice, but I won't hold that against him)
"basically the discussion was proceeding on the basis that this time series was abstract from any real world system such as global temperature). On that basis you get one answer."
Isn't the point that the time series was one of measurements of global temperature and that's all the model was intending to examine? I can't think of what other uses than for statistical analysis you would use a statistical model.
@ATheoK
Time has arrived for England to source their weather and climate prognostications from anywhere but the Met. Defund MetO.
Or maybe Funding By Results?
As an ULP -Uninformed Lay Person (HT/Thanks Hilary) - I suspect that it's more advantageous for a 'scientist' to support a consensus involving 'Known Unknowns"/"We can't think of any other factor' as it protects them with the shield-wall strategy - relying on your neighbour for protection- that was such a potent preservation-mechanism in dark-age warfare rather than taking a chance with trebuchets, gunpowder and advanced excel.
It's called peer-review nowadays.
It's a well-tried Pyramid 'trick' and highly successful for early protagonists. Harm, pre-Internet, was the exclusive prerogative of late entrants - the young, the guillable and the green!
Hopefuly; no longer. The misled may be forgiven for their callowness but the self-interest and dishonesty (thank you -FOIA - 2009) of their mentors will be, forever, remembered and despised.
The latter will become the sacrificial lambs on the altar of an unforgiving MSM and a grateful blame-shifting political class whose very existence depends on their ability to position their fingers so as to best fiddle the Yo-Yo that defines 21st Century Democracy.
Establishment Climatology; You may not all recognise what is happening to your craft and, once respected, predictions but it's fast becoming undone.
Don't seek a source for your impending downfall, further than your own misguided but understandable motives, other than yourselves.
God forgive you, for those who have suffered from your certainty, hubris and frailty that has tainted and harmed an entire generation.
I, for one, am unpersuaded that Keenan's point rises above semantics and counsel readers not to over-invest in this line of reasoning. Lucia has written on this topic and I agree with her.
A couple of years ago, I experimented with AR1 and ARFIMA (d=1) calculating the likelihood of various trends given observations though I don't think that I did a post on it. The maximum likelihood trend for the two variations are rather close. The principal difference is that the spread is larger for the ARFIMA model.
Because the spread is broader in the ARFIMA case, in some circumstances, the 95th percentile encompasses 0 trend. Thus, Keenan's claim that the trend is not "statistically significant". But this is a sort of debating point as the most likely trend isn't altered appreciably by the ARFIMA model.
It is also a two-edged sword. Because the spread is wider, the possibility of higher trends is similarly not precluded. Thus, arguments that there is a statistically significant difference between models and observations may not hold up.
If the Met Office is to be confronted, I would much prefer that they be confronted on more substantive issues.
is someone going to post a response comment at the MET blog? If no one going to make some rumblings about the general fail that is the blog post.
"They're not even within 100 miles [of Baghdad]. They are not in any place. They hold no place in Iraq. This is an illusion ... they are trying to sell to the others an illusion." --Baghdad Bob
"A wide range of observed climate indicators continue to show changes that are consistent with a globally warming world, and our understanding of how the climate system responds to rising greenhouse gas levels." --Baghdad Julia
“Mr Keenan then goes on to argue that you can only use a statistical model to determine whether the warming we have seen is statistically significant.”
The fact that Slingo doesn’t understand that this absolutely must be true is proof that she is unfit to hold her office.
All of the other things that she mentions may be relevant as to why the statistical results turn out to be what they turn out to be. But they are not cumlative reasons along with the statistical results, they are only reasons for those statistical results. If they are good enough and complete enough as reasons then that will be ultimately shown by the statistics.
"I, for one, am unpersuaded that Keenan's point rises above semantics and counsel readers not to over-invest in this line of reasoning."
It depends on whether you are trying to estimate the trend, or simply to dispute the Met Office's previous claim that the trend can be shown to be significantly non-zero.
"It is also a two-edged sword."
Again, that depends on what you're trying to do. If you're trying to show that the trend is low or zero, then I agree (and I'm sure Doug does) that it doesn't do that. If you're trying to show that it hasn't been shown to be high or non-zero, then it works fine.
"If the Met Office is to be confronted, I would much prefer that they be confronted on more substantive issues."
Good idea. I'm sure if you pass on some suggestions, our Parliamentarians will be interested.
But substance is not the only criterion. The aim I think was to define a short and simple question with a mathematically well-defined and unambiguous answer that they couldn't wriggle out of. The more substantive questions tend to be more complicated and ambiguous, and it's rather easy for them to 'creatively misinterpret' the question to avoid having to answer it. You know how hard it is to get a straight answer out of some people.
Sorry, missed a bit...
"Thus, arguments that there is a statistically significant difference between models and observations may not hold up."
If that's the case, then so be it. We're not the sort to judge an argument by its conclusions. Or we shouldn't be, anyway.
However, to test whether there is a significant difference between models and observations, we shouldn't be using arbitrarily-assumed ARFIMA models, we should be using the distribution of natural variation generated by the climate models themselves. So the question doesn't arise.
Nullius in Verba:
I can't imagine who you're thinking of :)
Mosher: "That model says there is no warming. But looking at the data we know it has warmed."
Steve, I don't know if he choose the right model (I don't know if he choose a model at all), but it seems to me that your given reason for saying that it is not the right model is wrong.
The question that we want an answer to is not "has it warmed". The question that we want an answer to is, "has it warmed to a degree that precludes such warming being caused by natural variation". Natural variation is then established by the variation within the data that you are running your statistical model against. Then, if we can say with 95% certainty that the data that is being investigated could not be what it is only based on it's natural variation, then we have significance. At least that is how I read it.
May 31, 2013 at 10:29 PM | Hilary Ostrov
My hypothesis is that she truly believes what she is trying to formulate clearly but has yet to formulate it clearly. Happens all the time.
They've been doing climate for hundreds of years ? ... Observations based on what data ? ... The computational models ? ... which are not validated ? Let's make it up as we go along.
Jun 1, 2013 at 5:55 AM | Unregistered CommenterTheo Goodwin
"My hypothesis is that she truly believes what she is trying to formulate clearly but has yet to formulate it clearly. Happens all the time."
As Lord Rutherford said: "If you can't explain your theory to a barmaid, it probably isn't very good physics"
Now here's the problem for Julia and her cohorts. The theory is simple, CO2 causes heat to be retained in the atmosphere and this causes a rise in temperature. Any barmaid can understand that, just as any barmaid with the data would ask, "Well why hasn't the temperature risen in line with temperature as it did between 1979 and 1995?" As somebody said upstream, "What falsifies this hypothesis?"
I had a brief correspondence with Doug McNeall on his blog where he said it wasn't possible to get statistical significance from the global temperatures time series, but the temperature change was "scientifically significant" between 1880 and 2000. This piqued my interest so I asked what signified scientific significance and had they applied this to the period 1880 to 1940, and was that temperature rise scientifically significant. The conversation ended at that point because Doug was too busy.
I think a lot of people are missing Doug's point. The Met Office are telling the politicos that there has been a rise in temperature and that this has been caused by human emissions. They are claiming this rise is significant, Doug is pointing out to the politicos that for the rise to be significant it has to be statistically significant against a statistical model, and that the choice of model can show it is/isn't statistically significant. No more than that, he's not denying it's warming, he's saying more accurate models (an you can argue about whether their more accurate/useful) show that there is no statistical significance in the rise in temperature.
The temperature has, of course risen, but a rising temperature isn't proof that humans caused it.
"Well why hasn't the temperature risen in line with temperature as it did between 1979 and 1995?"
Should have read:
"Well why hasn't the temperature risen in line with CO2 as it did between 1979 and 1995?"
Possibly the grandest example yet of the 'Post Hoc, Ergo Propter Hoc' fallacy.
============
It seems to me our friends across the pond are embarking on a blue on blue mission.
Perhaps a little more than a superficial reading of the parlimentary question and answer is required. Or a quick read of a few near upstream posts might be enlightening.
If the Met Office are arguing that the warming since 1850 is not natural, and the IPCC consensus (in AR4 I think) stated that CO2 levels were not suffice to cause any of the pre 1950 warming, what the does the Met Office attribute the century of warming between 1850 and 1950 to?
Yet again I think Julia and her merry Met Men should reflect on some real data and for the sake of their careers and scientific credibility accept that there is an elephant in the room rather than continue to try to obscure it's presence with complex statistical models and obfuscations:
CET and 7 other historical datasets (can anyone spot the CO2 signal?)
Alaskan glaciers reveal tree stumps
Glacier Bay - retreat started in 1770s
The elephant in the room is the long slow thaw from the LIA, and CO2 has feck all to do with it.
Steve - It would be very helpful, in view of your climate auditing activities over the years and your detailed observations of the goings-on in Climate Science, if you would suggest two or three of the issues on which the Met Office would best be confronted.
Its the Epsom Derby today Julia.
You putting a bet on.You too scared you cant be accurate or cant be certain.
You dont bet on the horses but you bet on the weather.
Predicting the weather in the next 50 years when you have long retired and lived of a nice public sector pension and everyones forgotten what you claimed all those years ago.
Easy isnt it.Julia try predicting the weather for the next 50 hours instead.
Jun 1, 2013 at 2:24 AM | Unregistered Commenterrobotech master
You could try to comment. Just don't criticise them for their failure to provide an answer to a parliamentary question no fewer than five times or they'll moderate it out of existence. Believe me.
Hilary
I get a figure of 17.55 (close enough) with a Flesch Reading Ease figure of 24.62 (max 100 - higher the better) and a recommendation to re-write eight sentences to improve comprehensibility.I don't think this is in any way deliberate; they just talk like that as part of their normal behaviour.
For what it's worth this posting has a GFI of 8.65 and a FRE of 62.94
+1 to NiV's reply to Steve M above.
Doug Keenan's work in this should not be understood as an exercise in devising an alternate form or model of climatology by itself. It is primarily useful in showing the lack of rigour in definitions, terminology and understanding in a high-level, dedicated unit such as the MO. There is no mystery to why climatologists study paleoclimate, or use GCMs. This can of worms was opened by the Met Office.
Is the temperature rise in the past century over and above what would have otherwise occurred ? - is a difficult question. One deserving of better answers than the facetious ones the MO gave, and continues to give. And an admission that it cannot be answered fully. Climate activists are cleverer than this. When confronted by this line of questioning, they usually flip to the "We have just one Earth, poor us" argument, admitting their ignorance.
What Nullius in Verba said, at 4:18 AM, I am in strong agreement with.
The basics of modern statistical analysis are easy, but sometimes people who have learned an earlier approach are reluctant to accept that that approach is no longer best practice. I suspect that this is what is happening with Steve McIntyre.
Briefly, in statistics, inferences are not drawn directly from data; rather, a statistical model is fit to the data, and inferences are drawn from the model. In our case, the Met Office claimed this inference: “the temperature rise since about 1880 is statistically significant”. To draw the inference, they chose this model: a linear trend with AR(1) residuals. All of this was their doing. I only got involved afterwards.
Demonstrating that the Met Office claim is ill-founded is something that is obviously valuable. To demonstrate that, I showed that another model was about 1000 times more likely to be the better model (where “better” is defined in a certain technical sense). Thus the Met Office model is failing to explain substantial variation in the data, and so it should not be used.
Note that I did not advocate using the other model, or indeed any model. Because I have not chosen a model, I am unable to draw any statistical inferences. My conclusion is merely that the basis for their inferences is untenable.
The other model has no trend. Hence there are no confidence intervals for a trend. This point seems to be difficult, to judge by Steve’s comment above and an earlier comment of his, but it is actually easy. As an analogy, consider a model that consists of a straight horizontal line, with some noise around it. The line cannot have a trend, because we have defined it to be horizontal. Because the line cannot have a trend, there can be no confidence intervals for a trend. Nonetheless, we can ask this question: how does the straight line compare to other models of the data? We can answer the question via relative likelihood.
This method of doing statistical analysis was largely unknown in the 20th century. It has become far more common in the 21st century. The advantages of the method over previous methods would take time to explain; here I will just mention one.
The method based on comparing statistical models via relative likelihood is essentially founded on thermodynamics. Every probability distribution can be represented arbitrarily closely by a thermodynamic system (in principle). The Second Law of Thermodynamics is usually stated as “entropy is always maximized”. And the method based on relative likelihood was originally called an “entropy maximization principle”. In other words, the method is rooted in how nature works.
For anyone seeking to learn more, the standard reference is Model Selection by Burnham & Anderson (2002). That book currently has about 20000 citations on Google Scholar, which seems to make it the most highly-cited statistical research work published during the last quarter century.
The statistical approach that I have been using is the one described in that book. Full details on how I used it for the global temperature data are in the Supplement to my op-ed piece in the Wall Street Journal.
When I was writing the op-ed piece, I corresponded with the second author of the book, David Anderson, who confirmed my use of the method; for that reason, Anderson is listed in the Acknowledgements section. Also listed in the Acknowledgements is David Banks, who is a former editor of the applications section of the world’s leading statistical journal, Journal of the American Statistical Association, and who read a draft. In addition to the kind assistance from Anderson and Banks, WSJ had someone who was familiar with time series check the piece. In the two years since publication, no valid problems have been reported to me.
As a Labour Peer I have opposed some past Conservative privatisations. Now , after my recent dealings with it, I would consider supporting the Tory Chancellor if he chooses to privatise the Met Office. A privatised MO might or might not be more competent, transparent and honest. But it would surely have stricter financial controls, have less scope for pursuing private Green agendas, and try to put more appropriate people at its top. It should be required to concentrate on more accurate weather forecasting and not influencing governments to spend scores of billions on some imperial moral mission to save the planet.
Then the government should establish a new independent institution to monitor the scientific and statistical validity of the analyses and forecasts on which its climate and energy policies are based. Just as the excellent new independent Office of Budget Responsibility ,under Robert Choate, monitors and reports publicly and regularly on the economic and financial statistics which underpin the Chancellors Treasury policies.
That way we might get better weather forecasts and climate and energy policies with safer foundations.
May I here thank all those who have suggested Questions which I might ask in the Lords to advance our understanding of what the Government and the Met Office are up to and why. Very helpful. Please think of more. Make them precise and factual. And accept they may have to be refined to meet the House rules...........and learn to smile if the MO ducks them. In the end the truth will out.
If the trend in temperature is so small that we need to make subtle choices of which statistical approach to use, the the trend we are looking for (rise or fall in temperature) is so small that we have no need to worry about it.
From the Met office blog:
"However, the claim that the increase in global warming is larger than could be explained by natural variability has a clear and well understood grounding in fundamental physics and chemistry. There is very high confidence (using the IPCC’s definition) that the global average net effect of human activities since 1850 has been one of warming. The basis for this claim is not, and never has been, the sole use of statistical models to emulate a global temperature trend. Instead it is based on hundreds of years of scientific advancement, supported by the development of high-quality observations and computational modeling."
The 2007 IPCC AR4 report says: "Most of the observed increase in global average temperatures since the mid-20th century is very likely due to the observed increase in anthropogenic greenhouse gas concentrations." Here, "very likely" means more than 90% confidence.
The Conclusions section of the InterAcademy Council (IAC) report dealt with the IPCC reporting of uncertainties. It said this.“ IPCC’s guidance for addressing uncertainties in the Fourth Assessment Report urges authors to consider the amount of evidence and level of agreement about all conclusions and to apply subjective probabilities of confidence to conclusions when there was “high agreement, much evidence.” However, such guidance was not always followed, as exemplified by the many statements in the Working Group II Summary for Policymakers that are assigned high confidence but are based on little evidence. Moreover, the apparent need to include statements of “high confidence” (i.e. an 8 out of 10 chance of being correct) in the Summary for Policymakers led authors to make many vaguely defined statements that are difficult to refute, therefore making them of “high confidence”. Such statements have little value.”
The IAC also said: “In the Committee’s view, assigning probabilities to imprecise statements is not an appropriate way to characterize uncertainty.”
Should one regard the word "most" as imprecise? if so then its inclusion in the IPCC statement beginning, "Most of the observed increase in global average temperatures...", results in a "vaguely defined statement." Such statements have "little value" according to the IAC.
As an aside, the BBC Trust, Chair of Editorial Standards, Alison Hastings used the IPCC conclusion to support the move away by the BBC from reporting climate science impartially. She misquoted. Under the title " Trusting what you see and hear: the media's role in covering science accurately" she said; " Climate change is 90 per cent likely to have been caused by humans. That is the conclusion of the influential Intergovernmental Panel on Climate Change in 2007." Ho,ho,ho.
climatebeagle May 31, 2013 at 8:04 PM
"The Fifth Estate" is now truly a part of the "establishment" and as such can no longer be swept aside and ignored.
Times they are a changin...
Lord Donoughue:
I think they call that noble cause corruption. But may I also quote from what you wrote on Watts Up With That on Wednesday:
Unjustified, deceptive yet government-coerced transfer of wealth from the poor to the rich. Not such a noble cause after all. Thank you for saying all this. And this:
If it does get hot down there, you can at least count me as a fan.
Richard Drake, thank you for your kind words. If it gets really hot down in Hell, I may need a fan !
Lord D: Always keen to oblige! You may also be interested in a series of tweets I exchanged with a young man called Barry Gardiner on Tuesday. I was pretty close to Barry at school and I assume that's why he took time to reply. Twitter's not so good for nuance, as Lucia Liljegren said today, but it was interesting at the end that Barry took my 'reversal of poverty in our lifetime' to refer only to the future. I meant what Paul Collier called the 'middle four billion' who had escaped from grinding poverty since the death of Mao - in The Bottom Billion in 2007 (based a world population of 6 billion). Of course we'd all love to see more of the same. I hope Barry, who I take to be a sincere Christian, just as he was as a teenager, thinks harder about the implications for the poor in this area, something that I believe many of us on Bishop Hill see more clearly right now.
Nullius in Verba sends me to bed tonight with a smile:
"Probably the sort of audience that asks Questions in Parliament about "driftless ARIMA(3,1,0) models", and expresses a deep interest in the answers."
Exquisite.
Intellectually I agree with you Steve; however because this is the Met we're talking about I greatly appreciate Doug and Lord Donoughue taking the Met to task over basics.
Even a monkey if trained to nod and murmur appreciatively in response to mannerisms can seem knowledgeable in a physics discourse. However ask the monkey to reference exactly a basic foundation to the discourse and that is where the animal exposes their real level of knowledge. This explains why repeating the question every time the Met swung and missed was essential.
It is also interesting that it took a sly reference to finally get the Met to pony up the answer hidden as it was in a sea of meaningless words. The Met's obvious distaste and bad humor because they were forced to answer is also intriguing. Why such disagreeable attitude and responses from the Met for something they should've shared readily?
Every inch gained on the CAGW crowd is more than an inch earned no matter who earns it. Steve, many of us think the world of your common sense appraisals and even more of your succinct detailed data analysis and dogged persistence in data honesty. Please do not be disappointed with Doug and Lord Donoughue's win even if it isn't terribly complex and substantive.
It occurs to me that it might be worthwhile for a statistician to take a brief look at the CET record to determine whether the period up to 1950 is statistically distinguishable from that after. Given the length of the record, that would be the best chance available of determining that recent warming cannot be accounted for by natural variation.
ATheoK asks "The Met's obvious distaste and bad humor because they were forced to answer is also intriguing. Why such disagreeable attitude and responses from the Met for something they should've shared readily?"
I think the answer is that Slingo and her ilk are the Gordon Geckos of science. In their hubristic world view of themselves as Masters of the Universe, they should not have to get their hands dirty dealing with questioning by mere Members of Parliament.
Jun 1, 2013 at 4:08 AM | Tilo
Spot on. There is nothing more to be said.
Back on 28th February, the Met Office forecast that Spring temperatures were likely to be “near average”. They also said colder temperatures were no more likely than warmer ones.
“This forecast is based on information from observations, several numerical models and expert judgement”
Perhaps they would like to tell us which of these were faulty.
http://notalotofpeopleknowthat.wordpress.com/2013/06/02/woops-met-office-fail-again/