Bishop Hill Detection, attribution, disintegration
Feb 20, 2013 Climate: D&A James Annan hit the climate blog headlines the other day, with a post that not only wrote off the possibility of high climate sensitivities but also revealed that one climate scientist had been so sure of the omerta among his colleagues that he openly admitted to lying to promote political action on climate change.
Annan has now written another potentially blockbuster post, which discusses recent publications in the area of detection and attribution (D&A) - the bit of the climate change science that assigns guilt. What Annan reveals is that the studies that have supported the claim of "the majority of recent warming is manmade" are fatally flawed.
I've been fairly critical of the conventional D&A approach in the past, primarily on the grounds that the null hypothesis of no anthropogenic influence is always false a priori (and therefore a failure to detect an anthropogenic influence is always a matter of insufficient data). These recent papers point to another, arguably more terminal, problem. Attribution will inevitably fail as the anthropogenic effect increases!
Read the (relatively technical) reasons why here.
As Annan explains, this is going to leave D&A specialists with a headache to rank alongside the one being ensured by the climate sensitivity team.
It will be interesting to see how the D&A community addresses this problem. Atribution of the observed changes to GHG and other influences was touted as a major step forward when it was first achieved, so it would surely be rather embarassing to lose the ability to do this. It looks a bit like they are trying to just ignore it for now, but that can't really be tenable as a long-term strategy.
Reader Comments (120)
I have been tracking all this climate modelling from a different direction. From what is being imposed as Systems Thinking in education back through to the Systems Thinkers at MIT and elsewhere. Social planners basically wanting reality to fit the model. And people too for that matter.
It produced a driving insight that we should keep in mind with all these models and visuals like the hockey stick or smoothing out MWP. "We want to supply the orientating metaphors that students and adults will see the world through." Since changing reality is beyond the reach of even the most ambitious, well-funded climate scientist, providing the filters that perceive day to day reality will have to do. Hence all the attempts to make catastrophic weather events about climate and CAGW. The visuals and repeated assertions confuse weather and climate in the minds of lofo voters who then see everything bad in weather as proving the assertion.
Any way that "we want to provide the metaphors and themes and concepts and visuals" goes a long way towards making the nonsensical make sense. Plus grounding all of what's being pushed in emotion, not logic. It's about justification for desired planning power. Economic and social.
It seems to this Oxfordshire housewife that these are no more than guesses built on a cardhouse of assumptions. When Richard Betts was in D&A we asked him here whether his conclusions were not the reuslt of circular reasoning. That the AGW effect was programmed into the model first and you could not claim to have done anything when you found it in the output. All attempts by Roger Longstaff to get him to reveal the scheme whereby radiative effects got into the model were answered with silence. (OK, or missed in the torrents of intended-to-be-awkward questions).
There is no foundation of knowledge which supports realistic detection and attribution at this stage.
And in service of that attempt we see ongoing efforts to "fit up" CO2 worthy of the Met at its best.
Annan:
But was it ever 'achieved'? Seems not.
How could this be? Did not Mr Secretary Davey lately reveal AGW to be irrefutable?
A fine example of Trenberth's reversal of the null hypothesis - CO2 treated as 'Guilty until proven innocent!'.
This says as much about the inadequacy of the peer review process as it does about the biases of the scientists publishing these results.
Hmmm,
I will need to read this further, but on my first viewing the "technical" details of Annan's complaints seem only to extend to concern that either:
a) We use a shorter period of data which leaves confidence intervals too wide to reject "no anthropogenic effect", or
b) We us a longer period of data which leaves confidence interval tightly enough constrained to accept "no anthropogenic effect".
If this is the case, it is only a problem in so far as you are certain that there is a powerful anthropogenic effect and the data must somehow be wrong. A standard enough point of view in climate science circles it seems.
Another step in the "climate climbdown" is this new paper, The upper end of climate model temperature projections is inconsistent with past warming, by Peter Stott ... and Ed Hawkins.
I agree with rhoda, it has always seemed odd to me that a bunch of guys and gals that believe that CO2 has caused the late 20th century warming could build models that test the hypothesis. I don't know, and probably wouldn't understand, how they build their models, but the attribution is obtained by running the model with and without CO2. Now we're back at the beginning, they believed that CO2 caused the warming and input that into the model. They then took the CO2 out and the warming dropped, so CO2 caused the warming.
I'll stop here because the above looks as though it could have been written by Lewis Carroll, himself an able mathematician and logician it seems.
To be honest, although I like what Dr. Annan's saying, and he seems to be something of a real scientist when dealing with the data, I get the feel that he's another one who thinks he's yummy. To be fair he's critical of all sides of the argument, but I believe he carry's his criticism too far when he disses Dr. Curry.
Which leads me to something else I find puzzling, and it's this. Like any of these people, or not, they have all been practising their trade for decades, yet, as Dr. Annan frequently points out, have written absolutely crap papers (if we are to believe Dr. Annan), which is puzzling on two counts. First, I point to their experience in the science, and secondly why didn't the reviewers find the papers crap?
I note that Annan has received the (Woody) Guthrie award for a "thinking blogger". But describing anthropogenic attribution as an "achievement" should trip the confirmation-bias alarm.
You can get anything you want,
At Pachauri's Restaurant.
Feb 20, 2013 at 12:26 PM | Paul Matthews:
Paul, please don't take this the wrong way, but they are on a hook of their own making, and as the magnitude of the error they've made dawns on them, with country after country announcing energy crises and rising energy prices, and we've barely started on the course of action these scientists have been demanding, we should give them space to backtrack and get the truth out. As Colonel Collins said, we should be "fierce in battle and magnaminous in victory." It's going to take time, I've been in Thailand this week and three different people have told me the weather's doing odd things all over the world. I want these guys to be able to stand up and tell them it's not, they won't if we ridicule them, they'll just stay schtum and hope it goes away without anyone noticing the massive damage they've done in their ignorance of the climate system and a willingness to support any left wing theory on the environment.
We've a long way to go. Besides, although Stotty is an extreme alarmist, I kinda like Ed Hawkins.
In 1967, Manabe and Wetherald started modern climate modelling with computer trickery but the basic physics was right - the energy input, the 160 W/m^2 solar SW, was transmitted in the atmosphere via IR.
In 1971, Lacis and Hansen introduced the Sagan aerosol optical physics which purported that small cloud droplet size gave higher albedo - part of the global dimming hypothesis, Sagan was wrong - large droplets give higher albedo.
In 1972, the NIMBUS team reported on the first measured OLR and claimed that the difference between black body emission from the Earth's surface and the OLR was the clear sky atmospheric greenhouse factor. Read about it in Ramanathan's 1997 Volvo prize paper: Ambio Vol 27 No 3 May 1998.
Using the 2009 'energy budget', this increases IR absorbed by a factor of 157.5/23 = 6.85. The extra 134.5 W/m^2 is imaginary and then vanishes after the phony positive feedback is created. The claim of 'back radiation', 333 W/m^2 is based on imagining that pyrgeometers measure a real energy flux - it's the potential energy that emitter would transmit to absolute zero. This is radiative equilibrium physics 101, taught to every physical scientist and engineer, and they got it wrong.
For 40 years, Climate Alchemy has been based on a Big mistake.
Nicely put geronimo.
I think that the modellers are similar to the alchemists of old, who believed that it was possible to turn base metals into gold, and that they could find a methodology to do this. Of course, with our modern understanding of chemistry and nuclear physics we now understand this to be impossible (at least, without a nuclear reactor), but there is no way that the alchemists could have known this. However, in the case of the modellers, I think that the proper application of known physical principles shows that what they are trying to do is impossible.
I do like "omerta". Presumably enforced by order of il capo di tutti clima capi, whoever that shadowy figure may be.
I was wondering how long till the chocolate teapot was brought out on this rather enjoyable thread. A major climate climbdown is in the offing, as Paul neatly puts it, and one can't help smiling a little. Attempts to contaminate genuinely scepticism with the deliberately farcical are surely also in their death throes.
Richard
If you don't like it, why not just ignore it, then the rest of us can carry on enjoying the thread without two distractions.
(Or three, if you count this one!)
I frequently do ignore it. And sometimes I laugh at loud at it. I see no problem with that.
"The next step, if we are following Lewis Carroll, is to consider that the warming was actually caused by the inclusion of CO2 in the models."
Gin and tonic all over my keyboard, it's 2010 here in Bangkok, I'm not an early drinker.
So, you agree with James that
"null hypothesis of no anthropogenic influence is always false a priori"
then?
or to put it another way
"no sane person believes these forcings have zero effect, so what exactly is the purpose of a null hypothesis significance test in the first place?"
;)
Doug
Feb 20, 2013 at 1:19 PM | Doug McNeall
"So, you agree with James that
"null hypothesis of no anthropogenic influence is always false a priori"
then?"
I do Doug. Everything has an effect on the climate, what we don't know is the extent of that effect. My null hypothesis is that anthropogenic influences didn't cause most or the warming in the late 20th century. Now prove they did. Dr. Annan, clearly my intellectual superior is groping his way towards a similar null hypothesis, but being a theoritician hasn't yet got to the stage where a null hypothesis doesn't have to be "null" in the black and white sense that humans caused/didn't cause the late 20th century warming, but can be null in the sense that they didn't cause all of it. I'm sure he's reading our comments as we speak and is taking on board the pearls of wisdom dropping from every post on this thread.
To me it would seem an extremely stupid null hypothesis that elephants have no effect on the climate, they surely do, it's the extent that's important.
Bravo geronimo.
Doug, check the first word in Andrew's twitter bio!
The whole GHG/AGW theories are fatally flawed.
The GHG theory violates the laws of thermodynamics, both 1st and 2nd and ignores the fact that the so called GHG's are saturated with energy from the inbound solar radiation and are unable to adsorb any radiation from the surface.
Our input of CO2 into the system is only 3% of the total but we are blamed for any problems. Look first at insects since their input is far larger than ours.
Also think about the temperature increase claimed for the existing atmospheric CO2 content, some 33C. These few CO2 molecules must be radiating in the far UV to have such an impact. Total rubbish.
Paul: I hadn't spotted that. Since when has Andrew publicly called himself lukewarmer?
Climate "science" is, so far as I know, the only branch of science where models are taken for reality. It's common to read climate science papers that talk about "experiments" meaning running unvalidated computer program models with different input parameters. I think I remember that Graham Stringer MP, to his credit, took exception to this.
The Met Office's line, all along, as I understand it, has been:
- Their models have been validated by verifying that they reproduce the past history of climate. [Although, so far as I can see, a simple lookup table of past climate values would be validated as a model by this test.]
- Leave out the CO2 bit of the model and the model no longer reproduces the past climate. This proves that the warming was caused by human released CO2.
Probably, il capo di tutti clima[?] capi must be Alphonse Gabriel Gore, if the 'capo' title goes to the person who's made most money out of climate alarmist activities.
"To me it would seem an extremely stupid null hypothesis that elephants have no effect on the climate, they surely do, it's the extent that's important."
Yes, assuming your conclusion is always the easiest thing to do.
Andrew
John Marshall: I agree with Mike Jackson that this isn't the place to discuss that. Same with anyone else pushing a variant of the slayer creed. This thread is about James Annan and his significant rowing back on the attribution claims that have been so central to recent reports of the United Nations IPCC. That's important in its own right. There are other places for your stuff.
Martin A on the apparent Met Office line:
How come anyone, scientist or policy maker, ever fell for this? Nobody honest who's tried to model anything in software would do so.
I'm with geronimo re null hypotheses, and with Annan. It is very commonplace for null hypotheses to be known to be false a priori. The device of testing the null hypothesis is merely a way of assessing the strength of evidence against it in a particular data set. Sometimes we know neither the direction nor the magnitude of some presumed effect. We gather data about it, and check to see if it provides convincing evidence that the null (effect) hypothesis is implausible. If it fails to do that, it does not mean that the null is true, merely that our evidence about it is inconclusive. A man found not guilty in a court of law is not necessarily innocent. In general, the best he can claim is that the evidence against him was insufficient.
Despite the turbulence and complexity of the climate system, we can reasonably conclude that all participants in it have some effect, including our good selves in various ways, including the release of CO2. Our challenge is to estimate both the direction and the magnitude of the effects on measurable items such as precipitation, temperature, windiness, and the like. The more 'noise' due to other factors varying in the system, the harder it will be to find convincing evidence for the nature of any supposed effect. A properly open-minded null hypothesis for manmade CO2 on global mean temperature would be that it is zero. We are confident that that is not true, but we want to see how convincing the evidence in a particular data set is to support either a cooling effect, or a warming effect, or a cooling effect in some circumstances and a warming effect in others. Given the relatively poor correlation of CO2 levels with rising and falling global mean temperatures in the past 100 years or so where similar rises in temperature occurred over decades with appreciably different CO2 levels, with cooling in between and a flattening at the end while CO2 continued to rise, the difficulty facing anyone keen to reject the null hypothesis is apparent.
"To me it would seem an extremely stupid null hypothesis that elephants have no effect on the climate, they surely do, it's the extent that's important."
Link to the Elephant Climate Foot/Hoof/Paw Detection literature? ;)
Andrew
"Link to the Elephant Climate Foot/Hoof/Paw Detection literature? ;)"
Sorry Andrew, I don't have a link. My proposition is based on my going to see the Russian State Circus and being invited to see the elephants after the show. I am in a position to tell you that they put enough methane into the atmosphere to have a discernible effect on climate, given that I near had to be carried out of their enclosure. Not enough for purists I know, but give yourself the opportunity to mix with elephants in an enclosed environment and you'll see what I mean.
To a radiometer in the atmosphere, thermal GHG band IR emission appears as if it were at the black body amplitude once concentration exceeds the self-absorption threshold, ~200 ppmv for CO2. This is standard physics from analytical spectroscopy,
Because the Earth's surface approximates a black body IR emitter and the first ~100 m of the lower atmosphere is at the same temperature, the opposing thermal IR streams cancel each other out, standard radiative equilibrium physics all scientists and engineers are taught.
The real emissivity/absorptivity of self-absorbed CO2 is about half the BB level. This has been known in metallurgy since the late 1940s work by Hottel. Climate Alchemists get the imaginary CO2 threat by imaginary ‘back radiation’.
I see AlecM is back, doing his thang.
geronimo,
Perhaps you can just describe your Russian Circus Experience as a Gaseous Anomaly? ;)
Andrew
TheBigYinJames
And why not?
Do some of the preceding comments imply that the "elephant in the room" is in fact the "elephant in the circus"?
I have no problem with elephants - "a whistle a day keeps the elephants away". No elephants around here, so the technique is vindicated, indeed it's validated.
In an argument about whether the lighter electron orbits the heavier nucleus, or vice-versa, Rutherford said “When a flea is on an elephant, it's the flea that jumps, not the elephant.". The high heat-capacity ocean is the elephant, and the relatively very low heat-capacity atmosphere is the flea. Scientists of Rutherford's stature had, and still have, a good sense of scale.
Martin Reed, because it's only tangentially related to the thread, as always.
I rarely disagree with geronimo but I can't support his 12:54 post.
Tim Collins is the epitome of the officer and gentleman: 'remember to be magnanimous in victory' is what I would expect from him. For those scientists and commentators who now realise they backed the wrong beefburger (for whatever reason, noble or otherwise) and now wish to recant, magnanimity is surely the correct response.
However, there will be irrational diehards and fanatical holdouts who will never concede the true nature of the CAGW scam.
For them, I prefer a line from a little later in the good Colonel's speech: "The enemy should be in no doubt that we are his nemesis and that we are bringing about his rightful destruction".
@geronimo I agree "we should give them space to backtrack and get the truth out."
And when they have then stamp heavily all over them in the same way as they have done to sceptics.
Reapplication for tenure would be a start.
Phil D, we should be generous to them in direct propertion to the earliness of their arrival, I say.
Richard Drake Feb 20, 2013 at 2:03 PM |
Martin A on the apparent Met Office line:
"Leave out the CO2 bit of the model and the model no longer reproduces the past climate. This proves that the warming was caused by human released CO2."
How come anyone, scientist or policy maker, ever fell for this? Nobody honest who's tried to model anything in software would do so.
****************
1) The policy makers didn't have much idea about any of this. It came from a supercomputer so had to be right.
2) They all wanted to believe it. There were no marks for not believing it.
Martin A describes a modeler's position:
"- Their models have been validated by verifying that they reproduce the past history of climate. [Although, so far as I can see, a simple lookup table of past climate values would be validated as a model by this test.]"
Does anyone doubt that a model can perfectly reproduce one or more lines on a graph? The so-called hindcast validation test requires only that models achieve that feat. No doubt the model can extend those lines into the future. But extending the lines into the future is betting that the future will be virtually identical to the past. Of what value is that? None.
John Shade
"Despite the turbulence and complexity of the climate system, we can reasonably conclude that all participants in it have some effect, including our good selves in various ways, including the release of CO2."
In practical terms, this might be undebatable. However, the effects of CO2 depend on the feedbacks and at this time the feedbacks are simply unknown. Given the feedbacks, the net contribution of CO2 might prove to be nothing.
Nevermind Secretary Davey saying that AGW science is irrefutable. No less an authority than the renowned scientist and ex-bar steward former deputy Prime Minister John Prescott said the same thing a few years back and implied that anyone who disagreed was a criminal. You can't get a higher authority than that (and he got loads of freebies representing the British government at numerous international shindigs).
With theo.
The belief that radiative imbalance will 'do something' is the physicist's fallacy. The earth's climate is a system. What may seem intuitive may not be how things turn out at all. The first thing is to let go of prejudices.
TBY, please don't police. Off-topic is defn not good. But off-topic spamming and/or purposeful thread disruption are the real problem. Alec is doing neither.
I'll keep my comment brief. Annan's starting point is wrong. You can start off wrong and yet reach a good destination.
"Phil D, we should be generous to them in direct propertion to the earliness of their arrival, I say." --TheBigYinJames
Crafty. But right.
Martin A. wrote:
The Met Office's line, all along, as I understand it, has been:
- Their models have been validated by verifying that they reproduce the past history of climate.
I have always been puzzled by this argument. If the Met Office's models can reproduce the past history of the climate in a reasonably accurate way (I would be rather suspicious if the agreement was too exact!) then, unless I am overlooking something, the Met Office must have a good understanding of the natural causes of climate change. However, if the Met Office really does understand natural variation why did their models fail to predict the 16 year hiatus in global warming?
Feb 20, 2013 at 11:31 AM | rhoda
Well done Rhoda !! one of the very few people here that have managed to see right through Dr R Betts' cover story.