Bishop Hill A saching
Sep 6, 2014 Matt Ridley From time time to time I have noted the tendency among upholders of the climate consensus to hurl strongly worded accusations of wrongdoing or abusive epithets at their opponents, apparently without considering it necessary to provide any evidence in support of their allegations. I'm thinking here of Nigel Lawson or Owen Paterson being described as "deniers" by just about every left-wing journalist in the country, without apparently needing to justify the accusation in any way and despite overwhelming evidence to the contrary.
There was yet another example of the same thing today, with economist and left-wing talking head Jeffrey Sachs aiming brickbats at Matt Ridley on account of his recent article about the lack of any surface temperature rise:
Ridley climate ignorance in WSJ today is part of compulsive lying of Murdoch media gang. Ridley totally misrepresents the science.
Not a shred of evidence is presented, not a hint of an explanation as to how he considers that the science might have been misrepresented. It's almost as if evidence is seen as superfluous. (This is somehow familiar. Now what does it remind me of?)
This kind of wild namecalling makes Sachs look rather deranged in my opinion, although as we have seen this kind of thing is common on the other side of the lines in the climate wars. I really think we need a word for it: a "saching", perhaps. Maybe readers can think of something better.
Reader Comments (147)
Yes, Geoff, it really is falling apart. I was about to quote this from Matt:
As you say, it's now admitted fraud or looming libel action for Sachs. The kind of double bind the climate guys have been able to foist upon so many. But this may just be a manipulation too far.
Evidence-Free Fulminating, 'Efing' or 'Effing'.
NicfromNYC
Well spotted. I did use the wrong dataset. Mea culpa.
Richard Drake
I’ve long thought that the climate monster would be slain by a legal action, and not by rational argument among reasonable people who disagree. (How could it, when the two most prominent scientists in Britain think, like Sir Paul Nurse, that those who disagree with them must be crushed, or, like Brian Cox, that what scientists say must be true because, well, because they’re scientists, or something?)
What makes this fascinating is that Ridley is not a typical climate sceptic, and Sachs is not a typical climate activist. Can a world-renowned economist get away with accusing a prominent parliamentarian writing in the Wall Street Journal of lying? Can the journalist get away with accusing a reputable academic of signing a pile of steaming Findus written by a second-rate PR man?
Is this our Marquess of Queensberry moment?
Geoff, I'm not as sure on the how but I think we agree that the when just got closer, thanks very much Viscount Ridley. I'm off into Sachs and the transition from communism by which he made his name - the desperate attempts in 1990s to get the managed conflict genie back in the bottle, as I would see it, to which I tried to allude in Thatcher and peace last year. However much you hated Maggie was John Holdren really the right guy to be giving that acceptance speech in Oslo in 1995? It was such over-promoted mediocrities that were always going to be the undoing of The Cause. Does it end with a ghost-writer of the calibre of Bob Ward? I mean, really. Undignified or what?
17 year long falsification of the theory of CO2 being the main cause of temperature change.
Geoff, Americans "don't give a hoot" and might occasionally use "lingo." The first term is used in polite company where "don't give a sh*t" might not fly. It could be that both would be less often used by younger cohorts than the over 50 crowd.
Maybe another American a bit shorter of tooth might have different view of this.
E.M. all in a day’s work. I despised Watts and Nova for years of cherry picking bad stations, implying overall bias, minus proof of it. Then Goddard doubled down on innuendo. Thanks E.M., for being real.
I find that 'pause deniers' describes them best. They just don't want to believe what nature is telling them and the reasons for that are nothing to do with science. Faux-green is just the new black:
http://news.bbc.co.uk/nol/shared/spl/hi/programmes/analysis/transcripts/25_01_10.txt
"PORRIT: ..The root problem is the model of economic growth that drives a particular pattern of economic activity that creates the emissions that lead to the climate impacts. And I think you have to go further upstream, if you like, in looking to the root problem in climate change; and going further upstream means going to the heart of the growth economy that we have today.
ROWLATT: ....you do get the sense that concern about global warming is becoming a social status
issue. People are much more likely to bashful than boastful about their foreign flights and adventures abroad nevertheless talking to some environmentalists you get the sense that even if we could suddenly switch to a world where growth without increased emissions were possible, they’d still have a problem with our urge to consume.
PALMER: In the 70s and 80s the environmental movement believed that if it put the scientific facts –
the data – in front of us, we would all wake up and we would reform ourselves and create a utopian,
happy world. What then happened is the classic collapse of that utopian hope and you move into stage two and stage two is the apocalyptic. So for example, the world is going to be swamped by floods, struck by fire, destroyed by plague, everything will collapse, society will fall apart. it’s that use of fear that is the main indicator of this......I think the core of what the environmental movement has done is it has taken sin, guilt and fear from religion and has used those very strongly.
HULME: Some of the deep green movement would buy into this - that actually climate change is the best
opportunity that we have got in order to get our political goal of a more egalitarian, localist, less consumer driven society onto the table. And we’ve seen over 40 or 50 years different tactics I suppose from some of these deep greens, eco-socialists if you like, to drive forward this idea and climate change is the latest and is an opportunity....I think all campaigning organisations, whether in civil society or whether in politics, will use agendas in order to further their particular goals. Whether you admire those other goals or not is a separate set of questions, but it seems to me this is the way in which climate change has emerged over the last 5 or 10 years "
Yes, EM, youa culpa. You say Ridley is either a fool or a liar. So. Hitherto I have paid close attention to what you write. I'm sad to discover this was a mistake. From now on I bracket your contributions with those of ESmiff (as sometimes amusing but not to be relied on).
osseo
Lesbian lizard lady.
A pause which disappears when you move the start date forward a year is hardly a robust phenomenon.
r.
Sep 6, 2014 at 8:08 PM Entropic man
Shouldn't you (and everybody else) be using some sort of data tapering window to reduce or eliminate edge effects that otherwise happen when you chop of a section of a time series prior to seeing if what is left has a trend? Same sort of thing as using a suitable data window prior to doing spectral analysis of a time series.
In other words the results greatly changing when you change the starting point for analysis is a statistical artifact due to using an inappropriate (ie rectangular) data window. That is to say, the result of using an unsuitable analysis technique.
Nicfrom NYC osseo
Put it down to senility rather than mendacity. I mucked up the Woosforthetrees graph settings.
Martin A --
In one sense the OLS slope calculation *does* implement a taper of sorts. The most primitive slope calculation can be considered to be the incremental change from one measurement (e.g., month or year) to the next. A rectangular-windowed-average of this difference series works out to be the same as the change from the first measurement to the last. The OLS slope is mathematically equivalent to a weighted average of the differences, with the weighting pattern being an inverted parabola, maximal at the center of the interval and nearing zero at the end points.
That said, the pause duration computed by McKitrick (and cited by Ridley) is always going to have the attribute that an extended interval will not have the same property, by virtue of the fact that the pause is *defined* to be the longest consecutive period of a statistically insignificant slope. So I don't think that EM's objection on that count is justified. If you hear that a football club has won their last 16 matches, you can be sure that they didn't win the 17th one back.
On the other hand, McKitrick calculated a slope of 0.09 K/decade for HadCRUT4 starting in 1995, and the same (to one significant digit) for 1996. His calculation of the 95% confidence window of +/-0.10 K/decade (which strikes me as overly conservative at first glance) allows those to count as "pause" years. EM's point that the slope is not zero-ish at the declared "start of pause" is well-taken.
On the third hand, the slope from those years remains well below what the models predicted. I find that to be more important than setting a date for the start of pause.
Harold W
Hadcrut4 annual averages are usually quoted with confidence limits of +/- 0.1C. All one can reasonably say of 17 years ( :-) )of apparently flat data is that the system has warmed by no more than 0.2C or cooled by no more than 0.2C. Anyone claiming that "Global warming has stopped" is going a long way past the information content of the data!
There is also the sceptic claim that the models failed to predict the pause. The model results are a trend line with 90% confidence limits defining a prediction band. Current figures are within the band. See Fig 3 here .
Finally, rememember Tamsin Edward's point that if you extend the 1970-1995 trend line it passes through the current temperatures. The pause may be an illusion caused by a few years of elevated temperatures following the 1998 El Nino.
EM, if you are withdrawing your 'fool or knave' characterisation of Ridley, I happily withdraw my own peevish remarks. Much is to be forgiven to 'senility' (I hope!).
Entropic man (Sep 11, 2014 at 12:52 AM), you've referenced that Met Office Decadal Forecast plot several times now as evidence for CAGW and keep quoting their 90% confidence range but I don't think it helps your case.
A simple 'eyeballing' of the plot suggests the empirical data is showing only half the warming predicted by the models (i.e. +0.5C vs +1.0C over 50 years) and this is acknowledged in the text when it says "...the forecast initially remains towards the lower end of the range simulated by CMIP5 models...". More importantly, it also suggests that if the trend of the last 10 years continues for another decade it will provide clear and unambiguous evidence that the models are seriously flawed!
Entropic man -
Perhaps the best comparison between models and observations is Ed Hawkins' here. The comparison is with RCP4.5 runs, and is masked to include only those areas included in HadCRUT4.
I don't have the ability to mask (at least not easily), but I happened to have the RCP6.0 runs, and produced this comparison. As with Hawkins' more authoritative graph, observations are running at the extreme lower edge of the spaghetti. One can count that 24 of 25 runs are above the observed values, as can be seen in this zoomed-in section.
Hawkins' graph places observations somewhere between 5th and 25th percentile of model runs, say 15th. The UKMO graph which you cite also places observations slightly above the 5th percentile (lower edge of green-shaded region). One should be aware that the UKMO chart re-baselines to 1981-2010, thereby including part of "the pause", hence it is likely to indicate a higher percentile rating. The complementary effect of the re-baselining may be seen at the early end of the graph (1960s); anomalies are at (or above) the very top of the 90% confidence region.
So I'm really not sure why you assert that the models predicted the pause. With such a huge spread -- looking at the RCP6.0 runs, 2014 anomaly ranges from 0.43 K to 1.31 K -- it's difficult to avoid the possibility that a couple of the extreme cases exceed observations. And it is the model mean which is (nearly) universally taken as the prediction.
Tamsin's observation is intriguing, and I wasn't aware that she had made it -- was it in a paper? Indeed the jump due to the 1998 El Niño was unusual. Was Tamsin suggesting that global warming will tend to manifest as jumps followed by plateaus, or was this just an observation that OLS trends can be eely when the underlying process is not linear?
Just a quick comment to say that I wasn't aware that I'd said it either...
Tamsin-- :D
Tamsin, HaroldW
I used the idea myself last year on a forum somewhere, then saw it at tamino's site in January.
I thought he'd credited it to Tamsin. I presumed that she'd published it somewhere and I'd picked it up.
Probably the idea was in general circulation and I misremembered the source. Senility strikes again!
More importantly, it also suggests that if the trend of the last 10 years continues for another decade it will provide clear and unambiguous evidence that the models are seriously flawed!
Sep 11, 2014 at 8:29 PM | Unregistered CommenterDave Salt
If the trend resumes and the pause ends within the next decade, will you regard that as clear and unambiguous evidence that the models are working?
Entropic Man -
Ah, Tamino. Not quite the same thing as Tamsin. It could have been Tam O'Shanter, I suppose, but what's Doon is Doon. (Tamino's trendline was from 1979-1997; presumably 1970-1995 was your version.) I will throw no stones at your "senility" from my glass house, I've done much the same.
Anyway, let's drop that bit then. Have you any response to the other parts of my comment at 1:42?
Entropic man (Sep 12, 2014 at 5:18 PM), the common sense answer to that is obvious...
Entropic Man-
You might be interested in this recent paper by Macias et al. I haven't finished digesting it, but this part seems relevant to the claim of whether a hiatus exists:
HaroldW
Discussions of "predicting the pause" tend to have a strong political component,which we can hopefully avoid.
1) Does the temperature record show a pause?
I am quite happy to agree that Hadcrut4 shows no 21st century temperature increase, still showing 1998 as warmest. Other records show a slowing in the rate of rise, but have later years highest, particularly 2010. I tend to default to GiSS because it tries to include high latitudes, which Hadcrut4 ignore. The GISS land/ocean 5-year average stops rising in 2002, and I'll quite happily accept a pause in their data from 2002 on.
Interestingly the GISS station data shows less of a pause than the land/ocean data and BEST land data hardly shows a pause at all. Could this be because sea surface temperatures are rising less than land temperatures? OHC? :-)
2) Did the CMIP5 models predict a pause?
Nice graphs! Clearly a comparison between the mean 21st century model output and Hadcrut4 shows the models running high, as they ran high in the 70s and 80s and low in the 60s and 90s. The models seem less variable than Hadcrut4.
I would regard judgement based only on the mean as oversimplified. I'll probably get more rude remarks, but I regard confidence limits as important. In this case they tell you the limits of the models as predictors.
Your graph shows the distribution of the models, some warmer and some colder, with Hadcrut4 bouncing around within their range. The problem for the modeller is to marry the physics which is predictable, with the variability in conditions which is not.
To get around this the modellers produce a probable range of values for industrial pollution, vulcanism, insolation etc and then let the model randomise them during a run. Runs with low pollution low vulcanism, and a warmer sun; these show rapid warming. Runs with high pollution, high vulcanism and weak solar cycles show slow warming.
21st century reality has behaved like the slow warming models, with high pollution above average vulcanism and weak solar maxima. Several papers last year showed that models in which these variables matched reality matched Hadcrut4 better. I regard this as evidence that the models work but you have to look a little deeper than just comparing the overall model mean with the temperature record. Those models which correctly predicted natural variability in the 21st century correctly predicted the pause.
3) Does the pause mean that global warming has stopped?
I suggest probably not. There was a thirty year pause from 1940 to 1970. There were shorter pauses such as the later 80s. There were also high rate transients like the early 80s and late 90s. I see no evidence that the current pause is going to be any more permanent. The energy budget imbalance persists, OHC continues to increase, ice melts; all indications that the underlying warming trend continues, though presently swamped by short term negative forcings affecting sea surface and station temperatures.
4) Cycles
This is an old hobby horse at Tallbloke's site, and one I've ridden myself. Inspection of GISS readily suggests an 11 year cycle and a 60 year cycle. It is fascinating to see Macias et al showing a similar pattern from a proper analysis. It readily explains the pauses from 1880 to 1910, from 1940 to 1970 and from 2002 to the present as descending phases in the 60 year cycle. It also explains the rapid warming from 1910 to 1940 and from 1970 to 2000. Extrapolated, it would suggest a pause to 2030 and rapid warming thereafter.
It also peovides evidence for an accelerating long term warming trend. Each warming phase adds more heat than the one before, while each pause shows less cooling. The 1880 to 1910 pause produced a significant cooling effect. The 21st century pause is, at best, flat.
That is based on a quick scan. Like yourself I'll read it properly and we can dissect it next week.
.
David Salt
Common sense would tell you the world is flat and the stars revolve around us. It is not a particularly useful tool in science.
EM - you are talking bollocks again. The consensus science of the day just added ever more complex epicycles to the Ptolemaic model in a forlorn attempt to explain its inability to meet observations of planetary orbits. And then thankfully along came Copernicus with his Heliocentric model, which was essentially based on common sense and a little critical thinking, two attributes which many climate scientists clearly lack, or subjugate, in their devotion to the the cause.
"What if climate change appears to be just mainly a multidecadal natural fluctuation? They'll kill us probably..." Tommy Wils, 2007. climategate email 1682 - http://di2.nu/foia/foia2011/mail/1682.txt
Entropic man (Sep 14, 2014 at 1:26 AM), please don't try to lecture me about science... you just sound pompous and/or ignorant.
If you really want to win peoples respect and demonstrate your maturity/magnanimity, why not reply to osseo (Sep 11, 2014 at 2:35 PM)?
Lapogus
If the 60 year cycle was all you needed then 1880, 1940 and 2000 would have had the same temperature.
Instead of meaningless insult perhaps you could join the scientific debate and supply some useful evidence to explain why there is a long term trend on top of the 60 year cycle.
David Salt
I made an error compiling my graph but it does not affect my basic opinion.
I still think Ridley is a fool or a knave. He presented as "truth" a distortion of the evidence. Which of the two options actually applies to him would depend on whether he did so deliberately.
David Salt
Ridley's article was based around Nic Lewis' TCR estimate, a single estimate from a single source, right at the lower extreme of the possible TCR range based on all studies and based on simplistic assumptions regarding variability and linearity.
Ridley fails to mention this, presenting it as though it were universally accepted. At best, he may not have been aware of the bulk of work on TCR; at worst he was presenting a deliberately slanted position.
Do you really regard this as the proper way to present science to the public?
Entropic man (Sep 14, 2014 at 1:46 PM) asks "Do you really regard this as the proper way to present science to the public?"
Well, it may not meet your 'holier-than-thou' level of purity, but it was far better than most of the articles published by the usual suspects (e.g. the 'Grauniad', the BBC, New Scientist, Sci. Am., and other media outlets who claim to specialise in presenting a balanced view of science to the public but who also excel at being 'economical with the truth'), though I never hear you complain about their lack of purity!
By the way, I'm at a loss to find any example of him even suggesting that what he wrote was the "truth" (i.e. there being no uncertainty in the value, let alone the final word on the subject... you know, "the science is settled"), though I can appreciate someone with your 'perspective' could read between the lines to convince themselves that he did.
Davis Salt
Listen to yourself.
"Grauniad', the BBC, New Scientist, Sci. Am., and other media outlets who claim to specialise in presenting a balanced view of science to the public but who also excel at being 'economical with the truth'"
Has it occurred to you that most media do not have any conspiracy agenda. They present what is generally perceived as correct. They present climate science as most scientists accept it.
The bias is from a few media sources whose proprietors have a sceptic agenda. For some reason the Wall Street Journal and the Daily Mail come to mind.
Can't think why. :-)
Entropic man, why is it that whenever I read your words I'm reminded of Burns...
"O wad some Power the giftie gie us
To see oursels as ithers see us!
It wad frae mony a blunder free us,
An' foolish notion"
David Salt
Exactly. Imagine how you, a member of a fringe science lobby spouting propaganda fed to you by lobby groups, look to the rest of the science community.
EM - my comment was in response to your ridiculous 'common sense' analogy. And if you wish to take it as an insult then it can't have been meaningless.
Do you really think that media organisations like the BBC are always perfectly impartial on every issue? That they never have an agenda or bias? If so you are a bigger fool and even more gullible than I thought.
Ho Ho Ho
Here’s a BBC scandal that should really make you disgusted
Entropic man (Sep 14, 2014 at 8:44 PM) said "Imagine how you, a member of a fringe science lobby spouting propaganda fed to you by lobby groups, look to the rest of the science community".
Well, if that 'fringe science lobby' includes people like Freeman Dyson or Judith Curry, I'm more than happy to be associated with it.
On the other hand, I wonder just how content you are to be associated with the likes of Michael Mann or Stephan Lewandowsky? But don't worry EM, we still have respect for you... just not very much!
David Salt, lapogus
Have you looked at Climate etc lately? Judith Curry is no longer a good model scientist.
On "common sense". For thousands of years common sense told everyone that the stars revolved around the Earth. Remember Galileo and the Pope?
As scientific understanding improved common sense changed, but it still took till 1853 for Earth's rotation to be demonstrated empirically by Foucault's pendulum.
Entropic man -
Thanks for your detailed and thoughtful reply (Sep 14, 2014 at 1:22 AM)
Just a few words in response, as I have had a busy weekend and still haven't got round to reading the Macias et al. paper...
I generally agree with you on your 1). I suspect that the HadCRUT4 coverage limitations are not very important -- at least when Hawkins masked the models to match HadCRUT4 it didn't seem to make much of a difference.
In 2), you wrote that you "regard judgement based only on the mean as oversimplified." I don't take issue with that, but note that (a)it is the mean which is most often cited, if not the upper confidence bound, and (b)the uncertainty region is huge! Some of that is unavoidable as natural events such as ENSO are not predictable (at least as of now), but the greater part of the uncertainty region is due to structural differences (spread from model to model) rather than natural variability (spread from run to run of the same model). [I am, of course, not considering the scenario uncertainty, which isn't of great importance when looking at comparisons to date.]
You also wrote that "the modellers produce a probable range of values for industrial pollution, vulcanism, insolation etc and then let the model randomise them during a run." My understanding was that all CMIP5 runs used a specific forcing scenario. [Well, four specific scenarios to be precise.]
3) I fully agree with you on this point. I might quibble with the equivocation on the word "warming" -- the term "global warming" is specifically about the surface temperature trend, although I agree that ocean heat content is a better metric of the underlying physical mechanism.
4) I also agree here, that the quasi-cyclic process, whatever it is, is superposed on a "secular trend" (borrowing the phrase from Macias) which is increasing. Here, for example, are 30-year OLS trends which are very suggestive on this point.
It remains true that the models' temperature predictions match reasonably the trend in the higher-growth periods and overestimate the trend during the cooler periods. One can see the latter in the bottom curve of this earlier-mentioned graph, where the models are generally below HadCRUT4 at the early end of the reference period (60s) and generally above by the end (80s), with the notable exception of the El Chichón eruption, to which the models over-react. When combined with the 21st century comparison, the suggestion that the models are fundamentally correct about CO2 sensitivity, and that the discrepancy with observations is only due to natural variation, seems less and less tenable.
Is the glass half empty or half full? I tend to focus on the positive features, while you focus on the negatives. :-)
I found a paper via this site, discussing the design of experiments using CMIP5. Unfortunately Taylor et al( 2009) is rather dense.
Entropic man -
The Taylor et al. paper which you cite above (5:09 PM) doesn't indicate any experiments with varying forcings along the lines you mentioned at Sept14 1:22 AM, viz. " To get around this the modellers produce a probable range of values for industrial pollution, vulcanism, insolation etc and then let the model randomise them during a run. [...] 21st century reality has behaved like the slow warming models, with high pollution above average vulcanism and weak solar maxima. Several papers last year showed that models in which these variables matched reality matched Hadcrut4 better." Perhaps a different source, then.
As for half-full vs. half-empty, I don't think I'm being unduly harsh on the models' perormance. Even AR5 WG1 found it desirable to reduce the projected warming over the medium term, although the adjustment was of an ad hoc nature. [I was disappointed that they didn't attempt a similar adjustment for the long-term (end-of-century) projection. I suspect that none of the authors was entirely satisfied with the compromise chosen, though.] In addition, the global average temperature appears to be the easiest metric to estimate; e.g. precipitation is a much uglier situation.
If I may propose an analogy, it's as if we're at the stage of Robert Goddard's first rockets. They generally go upward and demonstrate that we know something of the principles involved. And the models will improve over time. But it was a long way from the first rockets to a controlled flight. One could have said of those early attempts that they were based on known physics, much as GCMs are described now. True as far as it goes, yet so incomplete a characterization.
EM, you've 'jumped the shark' with your comment on Judith Curry... you are truly delusional.
Harold W
I'm not sure where the randomisation came from. Like many memories, the original source has become obscure. One of the most recent papers on gcm performance was Meehl et al (2014) .
Perhaps the GCMs are more like the early Nimbus satellites, a lot less sophisticated than their successors, but still capable of generating useful data.
Thank you. It's pleasant to have a civilised discussion with someone of different views.
You see more reflected from a different perspective.
David Salt
I think not. She is conducting a flame war with Michael Mann, pushing stadium waves and proposing Bose-Einstein equations as a mechanism for cloud formation on very little evidence. She gave a a presentation yesterday for the George C Marshall Institute, a political lobby group who have previously misrepresented the damage from smoking, acid rain and CFCs.
She has shot her own credibility as a working climate scientist baldly in the foot.
Entropic man:
"It's pleasant to have a civilised discussion with someone of different views. You see more reflected from a different perspective."
Likewise.
Just out of curiosity EM, who now fits your puritanical definition of a scientist?
'Lord Ridley is a knave or a fool'... 'Judith Curry .. conducting a flame war'?
Really, EM?
"Bloody instructions, which, being taught, return
To plague the inventor..."