Guest post by Alec Rawls
Andrew Orlowski at the UK Register has an anecdotal account of Downing College’s skeptics-vs-believers mash-up. Ace of Spades pulled the juiciest bit:
In short, the day lined up Phil Jones, oceanographer Andrew Watson, and physicist Mike Lockwood, the latter to argue that the sun couldn’t possibly have caused recent warming. He was followed by the most impressive presentation from Henrik Svensmark, whose presentation stood out head and shoulders above anyone else. Why? For two reasons. The correlations he shows are remarkable, and don’t need curve fitting, or funky statistical tricks. And he has advanced a mechanism, using empirical science [image above], to explain them.
At the other end of the scale, by way of contrast, the Met’s principle research scientist John Mitchell told us: “People underestimate the power of models. Observational evidence is not very useful,” adding, “Our approach is not entirely empirical.”
Yes, you could say that.
Lockwood’s failed argument against a solar explanation
Orlowski on Lockwood:
The strongest argument, according to Lockwood, for the sun not being a driver in recent climatic activity is that “it has been going in the wrong direction for 30 years.”
Hmmm. So as soon as solar magnetic activity passed its peak, when it was still at some of the highest levels ever recorded, these very high levels of solar activity could no longer have caused warming?
As I have noted a number of times, this argument depends on an unstated assumption that, by 30 years ago (by 1980 or so), ocean temperatures had equilibrated to whatever forcing effect the 20th century’s high level of solar activity might be having. Otherwise the continued high level of forcing would continue to create warming until equilibrium was reached, regardless of whether solar activity had peaked yet. (The actual peak seems to have been solar cycle 22, from 1986-96, not 1980, as Lockwood claims.)
When I pressed Lockwood on his implicit equilibrium assumption he justified it by citing evidence that ocean temperature response to solar activity peters out (as measured by decorrelation) within a few years:
Almost all estimates have been in the 1-10 year range.
But decorrelation between surface temperatures and solar activity is very different from equilibrium. All decorrelation is measuring is the rapid temperature response of the upper ocean layer when solar activity rises or falls. That rapid response indicates that the sun is indeed a powerful driver of global temperature, but it says next to nothing about how long it takes for heat to carry into and out of deeper ocean layers.
This was brought out by AGW believers like Gavin Schmidt who are concerned about the energy balance implications of equilibration-speed. In a simple energy balance model, rapid equilibration implies (other things equal) that climate sensitivity must be low. Since belief depends on high climate sensitivity, the rapid equilibration claim cited by Lockwood had to be shot down, which was managed quite successfully (ibid).
In sum, Lockwood’s rapid equilibrium assumption is dead and buried, leaving him no grounds for dismissing a solar explanation for post 70’s warming. I’ll keep an eye out for video of Lockwood’s presentation, but I doubt he mentioned the rapid equilibrium assumption upon which his argument depends.
More punk students
Remember these graduate student “climate scientists,” going all Clockwork Orange for the planet or something:
Sounds like they made an appearance at Downing College too:
The audience had been good enough to heed Howard’s opening advice that “if anybody mentions Climategate, they’ll be evicted”. Nobody ambushed the CRU crew all day – it was all very polite. I noted that the skeptics made a point of listening politely to the warmists, and applauding them all. A group of students and a few others, simply giggled and mocked the skeptics, however from start to finish. One of their tutors (I presume) was in hysterics all day.
Give ’em an A. They learned their “observational evidence is not very useful” lesson well.
Acutally, the kitchen can continue to warm even if the stove heat is reduced from its previous maximum.
Additionally, the northern hemisphere continues to warm during the summer months despite the fact that the sun reached its maximum output on the hemisphere on June 21.
@-Smokey says:
May 17, 2011 at 7:03 am
“Empirical [real world] evidence is never inferior to modeling arguments.”
Depends on the quality of the empirical evidence.
Do you have ANY empirical evidence that the thermal inertia of the oceans is such that they are continuing to rise in temperature BECAUSE of the level of solar output that has been static for ~30years?
Or is this hypothesis better supported by modeling arguments??
I’m under the impression that we use modeling in place of observation for GISSTemp and others because of the lack of surface stations in the arctic. If this is correct, then we are already substituting models for reality.
Insert cheesy pop-culture sci-fi reference here. Extra points for not reference ‘The Matrix.’
izen says:
May 17, 2011 at 6:47 am
“…unsupported by any empirical evidence, and must therefore be an example of where empirical evidence is inferior to modeling arguments…”
________
This is an interesting comment IMO because it clearly illustrates where belief is inserted into the CAGW mantra. I refer to the words, “must therefore”. The case here where izen gives a more verbose and “technical” reason for Mitchell’s wholly unsupportable assertion then adds the belief words: “must therefore”. If you start with a false presupposition (earth climate is too noisy/ it is what it is: climate reality) your result will also be false (the models are better than empirical evidence / scientific method). Somehow reality is too noisy to descern a readable signal, and the models do not contain this noise and so, “must therefore” be superior to direct observation and empirical evidence. Adding those two little words does not change reality. And the reality is that the models fail at having any predictive value either when forecasting or hindcasting.
Take out, “must therefore” and the whole thing reverts back to the original ridiculus comment that John Mitchell made which has been fully deconstructed in comments above.
________
“Smokey says:
May 17, 2011 at 7:03 am
“Empirical [real world] evidence is never inferior to modeling arguments.”
What he said…
*Well, on reflection, I guess that modeling arguments might be superior to empirical evidence for the astrologer or alchemist.*
Frank: excellent idea!
I’ve been working on a new model that shows no matter how hard I press the accelarator on my car, it cannot exceed 55 MPH. Therefore officer, you may have observed me doing 75 MPH, but obviousely my model proves your observation was incorrect. It has to do with all that “noise” of all those other cars and trucks on the road with me.
“And while the mechansim may be ‘empirical’, as yet there is no empirical evidence of it, and substantial evidence that cloud nucleation is provided by other empirical processes.” – izen
izen,
Are you being ironic here, or are you just that out of touch? Here is a nice overview with links to articles.
http://motls.blogspot.com/2011/05/new-danish-experiment-confirms.html
Smokey says:
May 17, 2011 at 7:03 am
…
Empirical [real world] evidence is never inferior to modeling arguments.”
Correct. Arguments based on models are fallacious: reification : (also known as hypostatisation, concretism, or the fallacy of misplaced concreteness) is a fallacy of ambiguity, when an abstraction (abstract belief or hypothetical construct) is treated as if it were a concrete, real event, or physical entity.[1][2] In other words, it is the error of treating as a “real thing” something which is not a real thing, but merely an idea. For example: if the phrase “holds another’s affection”, is taken literally, affection would be reified.
Another common manifestation is the confusion of a model with reality. Mathematical or simulation models may help understand a system or situation but real life always differs from the model. In extreme cases, the butterfly effect causes the model to rapidly diverge from what is occurring in real life.
Mitchell’s statement is unabashed propaganda. The models have always failed regardless of all the little tweaks, to forcings, feedbacks or flatulence. And he/they know it. Now we are told that we have been underestimating them (?), and that modeling arguements are superior to emperical evidence (?). Not in this lifetime and not on this planet. No amount of redirect with specific technical dissertation on solar flux or sea levels, on correlation or causation can detract from the over-arching reality that the models fail, have failed, and will continue to fail to predict anything to do with climate. Period.
It would be easier if we could just agree that we do not have all the information / data / technology, etc, required (and may never have) to predict a chaotic system. We will never know everything about the climate, sorry if that is an affront to anyones ego, get used to it.
Observational evidence is not always useful. Actually this is a correct statement. The observations we obtain may not assist the building of hypothesises or theories but without observational evidence, we are stuck in the realm of speculation. Without good evidence we cannot build a theory. Without a theory we cannot predict. Unless you are of the religious persuasion; and then you can predict anything you like but it’s certainly not science.
izen says:
“Depends on the quality of the empirical evidence.”
Now the goal posts have been moved to ‘quality’? If so, then the ‘quality’ of the model must be matched, no?
As a delegate at the Downing College conference I concur with the accounts above. I can add a few oddments.
Phil Jones’ lecture was pretty lame, he commented that no one had downloaded his data now that he’d put it in the public domain. Maybe because it’s worthless anyway.
Eric Wolff (Scott Polar Inst.) seemed to think that althought heating precedes CO2 rise in the ice cores, nonetheless its the CO2 that causes the warming (I think he means escalates the warming). Back to time travel issues maybe.
He was also apparently a bit disingenuous when asked if the ice core CO2 records were reliable. He explained that the last 50 years matched the ML levels very well. I rather think the questioner wanted to know if ice a mile down could be relied on to maintain ‘correct’ levels of CO2. Contacting Wolff later by email, he asured me that the ice 1 mile down was in that respect quite unchanged… hmmm.. under 20,000kN/m2 ?
Andrew Watson’s efforts were interesting in that the CO2 absorbed in the biosphere was an acknowledge ‘unknown’ – which of course upset all his careful calculations. Strangely though he was keen to tell us that the ocean were warming, yet they were apparently absorbing even more CO2: shomeshing wrong shurely?
John Mitchell agreed with me during the lunch break that the ‘greenhouse’ concept was a total misnomer… even though he had trotted out the Mars, Earth, Venus canard earlier. I explained that the moon had a similar ST elevation (30-40K in fact) without any ‘GHGs’. I’m not sure whether he was aware of that.
Morner was excellent, flamboyant but devastating: the evidence of ‘no sea level rise’ was there in reality for all to see – not in some computer graphic.
Vaclav Klaus quite quiet but equally devastating – what a breath of fresh air! He’d be very welcome in the UK as Prime Minister (technically possible as we’re in the EU!!) as a substitute for our current pink model.
He blew away the rather pathetic assertion that climate models were better than economic models made by AGWers earlier on in the day. LOL.
Smokey says:
izen says:
“Depends on the quality of the empirical evidence.”
Now the goal posts have been moved to ‘quality’? If so, then the ‘quality’ of the model must be matched, no?
Seems to me that if the quality of the observational evidence is poor, then the resultant model must be either equally poor (if modeling a linear system) or worse by orders of magnitude (if modeling a chaotic system – i.e. climate).
Unless, of course, you are simply making up your initial data.
This whole thing goes back to what I keep saying – these are people who don’t live in the real world.
“People underestimate the power of models. Observational evidence is not very useful.”
Is there any doubt that if observational evidence began to agree with their models then it would suddenly become useful to them again?
As a side note, one of the benefits of Global Warming is that as the temperature increases, models wear fewer clothes. But that is just an observation, so it can’t be very useful to climate scientists.
Experimental confirmation of the solar-cosmic-cloud seed paradigm is in. Nobels for Svensmark and all concerned?
http://calderup.wordpress.com/2011/05/17/accelerator-results-on-cloud-nucleation-2/
Orlowski notes a contrast of opinion:
RockyRoad says:
May 16, 2011 at 8:08 pm
The strongest argument, according to Lockwood, for the sun not being a driver in recent climatic activity is that “it has been going in the wrong direction for 30 years.”
Ok, let’s look at a more discernible situation: Has anybody out ever found that the hottest day of the summer is June 21st? Based on the above logic, it certainly should be (greatest amount of sunlight in the N. hemisphere). But we all know it never is. Never! The hottest day of summer is generally the latter part of July/first part of August–considerably later that the “longest day”.
Ok – but there was high solar activity up until ~1790 – doesn’t this mean that the Dalton Minimum (1790-1820) was warm?
So much for their argument!
“People underestimate the power of models. Observational evidence is not very useful.”
Part of the essence of religion …
What really happened to the scientific view that we are in an interglacial period and still are recovering after the last ice age? That is, it will first be hotter and then get colder by gradually approaching a new ice age. Have we reached the tipping point, regarding the past decades or are that it just a temporary “time out”? ( Rhetorical question, since no one knows the correct parameters to respond to this … Since many parameters are of the volatility nature ( without historical traces ), we will never be able to answer the question. .. Reasonable, it can never be scientifically acceptable to guess …)
At the other end of the scale, by way of contrast, the Met’s principle research scientist John Mitchell told us: “People underestimate the power of models. Observational evidence is not very useful,” adding, “Our approach is not entirely empirical.”
Maybe he meant the power of hypnotic suggestion, like throwing a big party at the beachhouse while a Hurricane draws near.
a group of students and a few others, simply giggled and mocked the skeptics, however from start to finish.
“Students” on an after-school field trip comprised by Kindergartners, no doubt? Perhaps someone should have honored them by having them and their “teacher” stand up and take a bow?
Coincidentally, I was just thinking again about this particular anti-scientific, anti-free thought tactic as used by some other “students” who were led by a Duke University Professor at a talk David Horowitz gave at Duke Univ. some years ago.
Once again an impressive infantilism combined with an intent to obstruct the rational consideration of an issue rises up to demonstrate our age’s own version of the pre-Enlightenment Brain and its Post Normal “Science”.
Another one of this Duke University Professor’s tactics was to disrobe down to bra-level. This was right before some Duke Teachers and Students next helped create the non-existent Duke LaCrosse Team’s “rape” of Crystal Mangum, who it seems might have recently done away with one of her boyfriends iirc. – my point here being that the Duke teachers and students didn’t help Crystal Mangum, either, starting right from the time of the non-existent “rape”.
Poster says: “I couldn’t disagree more. Mitchell is giving very sage advice here on the state of observational data in climate science (and no I am not being sarcastic). By running models, climate scientists should be able to test their theories without having to wade through the high noise levels and/or short-term high amplitude variations in global temps that make the data record so complex (and of limited use).
Sadly, it seems, many of these scientists are looking for disaster scenarios so we take every quote from a climate scientist with maximum skepticism – occasionally to the point of missing one.”
This reasoning is so flawed I don’t know where to begin. I work with geophysical forward models of the excitation of metallic objects by pulsed EM induction. We have good physics that support the model and, in test circumstances, data taken is well fit by an inversion of the forward models. Even in this context, we spent years trying to get good inversions out of the models to the data. The problem was noisy data. Until we developed an entirely new generation of sensors, the problem of inverting an overdefined model to data sometimes produces good results and sometimes not very good results. The key though is good physics and (today) good data.
The approach described in the quoted language makes mincemeat of this approach. It says that, in effect, the model is self validating because of noisy data. I can tell you with certainty that in our domain, models just produced bad results on noisy data. You couldn’t tell they were bad results until you dug up a detected metallic item at great expense.
In the AGW world, the models are not solid physics models–they are better described as physics speculations about how things might work. These models are far more complex than the simple dipole models we use because they include positive feedback loops and lots more parameters to set. That these models may, no, probably do, produce bad results when fit to noisy data is no surprise. Fitting to data is not difficult when you have a lot of parameters to play with. The difficulty is getting a good fit. Yet, the accomplishment (ta da) of actually getting a fit is seen as a validation of the model. That’s entirely erroneous. With enough parameters, you can fit any data with any model if you spend enough time.
The complaint about noisy data seems to me to say that this is an area in which models and their predictions are largely useless or at a minimum, very dangerous for the modeler’s reputation. The failure of the AGW models to predict the temperature leveling off of the 2000’s is an example of the type of failure I would expect from the AGW process. The failure of the AGW prediction regarding equatorial troposphere temperatures is another.
After the trend of the 2000’s became clear, I predicted that AGW groups would discover a new factor that accounts for it but that preserves the basic global warming hypothesis. Sure enough, about two years later, group after group released them. This is not predictive modeling science. It is scrambling to make complex models fit the modeler’s preconceived notions.
John Mitchell told us: “People underestimate the power of models. Observational evidence is not very useful,” adding, “Our approach is not entirely empirical.”
Power? Computational power, yes, predictive power? zero.
The idiocy of this statement by Mitchell underlines the absolute and undying validity and authority of Karl Popper’s “Laws of scientific inquiry” (they indeed have the status of laws).
The scientific method is deductive, not inductive. There are no inductive inferences. If it cant be falsified, it aint science.
According to Richard Feynman…http://www.youtube.com/watch?v=8qAi_9quzUY
According to John Mitchell… our models show that the Challenger never exploded.
And to me as a layman it would produce bad results in the domain of “common sense.” I just don’t understand how scientists (or anyone) can fail to see this.
If the rock concert is too loud and I can’t hear what the person nearby is saying, I can certainly try to guess what they might have said, but I know I am just guessing.
@-moptop says:
May 17, 2011 at 8:24 am RE:- empirical evidence of ionizing radiation causing changes in cloud cover. –
“izen,
Are you being ironic here, or are you just that out of touch? Here is a nice overview with links to articles.”
As you link indicates, yes there is some evidence that ionizing radiation can cause cloud nucleation centers, or at least that it can provide the initial precursers of cloud nucleation centers. But given the very poor empirical observations of cloud cover with no reliable long-term trend or correlation with any solar or other hypothesized driving factor I stand by my assertion that there is no empirical evidence that GCR flux has any effect on actual cloud cover in the real world. As the recent research mentioned, and the Svensmark work at CERN describes the hypothesis that cloud cover changes is a MODELING assumption, not evidenced based.
“Ok – but there was high solar activity up until ~1790 – doesn’t this mean that the Dalton Minimum (1790-1820) was warm? ”
The oceanic lag times seem to be variable.
A change in the level of solar activity seems to have a rapid effect in the air to alter the surface pressure distribution but in addition to that solar top down effect there is a bottom up oceanic effect and we do not yet have a grip on the timing of oceanic variability beyond the most basic and merely regional 60 year ENSO/PDO cycle.
The oceanic effect can either supplement or offset the solar effect over variable periods of time hence the difficulty in acquiring correlations that stand up well over time.
The best thing to do is start afresh from now since we have much bettor sensors in the air and oceans. Watch how both sun and oceans vary over time and see whether the net effect of the two variables combined has an effect on the atmospheric heights and surface pressure distribution.
We can see the effect that the active sun and warm sea surfaces combination had in the late 20th century, namely zonal jets and warming troposphere.
Now with the current combination of a quiet sun and a newly negative PDO we have more meridional jets and troposphere not currently warming.
The test will be how things develop from here on. That will tell us whether the pattern holds.
I anticipate continuing low solar activity and a continuing negative PDO which should translate into continuing jetstream meridionality, more clouds, higher albedo, less energy into the oceans and in due course noticeable tropospheric cooling.