Bishop Hill A chat with Leo and Doug
Jan 11, 2012 Climate: other I had an interesting exchange of views on Twitter last night with the Guardian's Leo Hickman and the Met Office's Doug McNeall. I was pressing them on the issue of climate models and their reliability and more specifically the fact that their ability to predict temperature is unvalidated.
Leo rather ducked the issue, noting that he wasn't qualified to comment, but I think that Doug and I came to a meeting of minds. Doug draws some comfort from the fact that other aspects of the global climate can be validated - we mentioned tropospheric fingerprints and I'm assuming ENSO would be another one. We didn't go into the question of just how much of a model/data match there is in these areas and I realise the fingerprints are hotly disputed. Let's therefore park these questions for the minute.
What was more interesting was that Doug and I agreed that unknown unknowns are a big problem for climate models. I think this recognition does take the debate forward so I'm gratified for Doug's input. Now I think we have to ask ourselves whether this ignorance has been properly relayed to the public and to policymakers.
On a related question, Doug and Leo asked what it would take to convince me of the validity of the models. Scientifically, this is an easy question. You need a lot of out-of-sample data. This will clearly take time, but it's the only way to do it. Pretending otherwise is simply to fool yourself. That said, I think if we had had some sort of a match between data and models since the IPCC projections in 2001, I would probably have moved a long way towards Leo's "convinced" position. When you think about it, it's remarkable that this failure of the models doesn't appear to have dented Leo's confidence in them at all.
Bishop Hill
Leo tweets:
I didn't say I was confident in models or not. I was talking about about basic principle tht rising CO2 causes warmng and that we'd b foolish to ignore that risk.
Interesting stuff.
Reader Comments (63)
BBD, your statement makes no sense. Part of warming attribution is based on the hockey stick and paleoclimate. Paleoclimate is based on the hockey stick as well.
Doug McNeall -
In your post at 3:09 PM, you contend that long-term climate prediction is a boundary-value problem, rather than an initial-value problem. This view is not universally held, see e.g. Collins & Allen or Pielke. I also disagree that your citation proves this; rather it seems to be an assumption in their analysis.
The reason why I incline against the description as a boundary-value problem, is that this formulation rests on the assumption that we know the incoming and outgoing energy fluxes accurately; this allows us to predict the total energy in the terrestrial system at points in the future. Given that, the claim is that unpredictability in the local distribution of energy ("weather"-level variation) is not particularly significant. However, the terrestrial climate system is not closed, and there is more than enough uncertainty in the amount of energy fluxes (in my opinion, at least) to disbelieve that we can predict the trajectory of total energy over long time scales, in any skillful manner.
the models largely do the right kind of thing on those timescales is one reason why we think it's reasonable to assume they do the right kind of thing on other timescales.
This is rubbish.
Unfortunately, no evidence is available to support that contention other than assertion, and realistic representations of natural variability render spatial or temporal averaging useless in terms of convergence of results. Scale averaging cannot overcome the loss of accuracy caused by divergence from initial conditions.
Another way of putting that: climate is an initial value problem, too.
Mike
I'm not yelling at you. Emphasis does not equate to yelling. Try and play it straight, eh?
Just stop misrepresenting Cook by quoting that email out of context. Now you know the context, you know that doing so is dishonest and misleading.
So, with regards to the models, if I've understood things correctly (always doubtful) what we're saying is that at any moment in time the following is true:
The latest models that we can meaningfully test are nth generation models. However, we are currently using n+mth generation models, which are much better than those awful nth generation models that performed so poorly. And the n+mth generation models tell us we are doomed.
Is there any way out of this?
Bish,
Did you ever ask Betts or McNeall to comment on the fact that the MET Office (together with others of the usual suspects - including UEA) are "funders / partner organisations" with ZeroCarbonBritain2030.
From a highly competitive field, I still consider the ZeroCarbonBritain2030 "report" to be the MOST egregiously tendentious, incompetent and dishonest report on policy actions that should be taken to fend off cAGW doom, of which I am aware.
Do Betts and McNeall endorse this report? Do they feel that this bunch of ultra thermageddonist loonies make suitable bedfellows for a taxpayer funded Department of the MoD? If they have any hesitation because they aren't conversant with energy issues, tell them to check out the Royal Academy of Engineering's "Generating The Future" report. Still written by thermageddonists but at least mathematically competent and engineeringly competent thermageddonists. The contrast between the conclusions is stark.
So far as I am concerned, the MET office, with their support for the lunatic fringe of eco-warriors, are completely without credibility in any issue relating to climate.
'Leo rather ducked the issue, noting that he wasn't qualified to comment, ' Nor is anyone else at the Guardian and yet that never stops them from making great claims of doom based nothing more than political outlooks and BS their spoon feed by the 'Team' and Bob 'fast fingers ' Ward .
@James Evans,
The nth generation models told us we were doomed,
The n+mth models tell us we are still doomed.
The crappy old nth generation models, no one now remembers but a few nitpickers, were essentially right (we are doomed) but they got it slightly wrong. The much more sophisticated n+mth generations are telling us we are to be doomed and we'd best believe it because they are so much more sophisticated, and the cycle continues. The test of these models of whatever generation is always years in the future. I see much scope for continually moving the goalposts.
I suggest this is a propaganda device for telling us we are doomed unless we mend our wicked ways and pay up.
Maurizio,
If there really is a Global Shrinkage problem, then we need to immediately require all males to start on an exercise program to develop our ability to lick our eyebrows. It won't help with shrinkage, but may aleviate some of the impacts.
Its not unknown unknowns that is the problem, but ignored knowns. The GCM's would almost certainly fit the global temperature record pretty well if they included solar magnetic effects (eg. as quantified by the 'previous solar cycle length' proxy). Since solar activity generally increased over the 20thC and pCO2 also rose, the statistical result of leaving out the indirect solar effect is to plump all that variance into the calculated CO2 climate sensitivity. That is what happens when you leave out a significant covarying parameter in a statistical model - which GCM's basically are, given they are fitted to and validated (supposedly) using the 20thC climate data.
The problem is as soon as they do actually include such a factor the calculated climate sensitivity will drop to a value close to L&C 2011, and the whole climate scare will evaporate, along with massive amounts of climate money.
I suggest Leo and Doug do a good course on multiple regression then try it out for themselves on some covarying test datasets. You'll immediately see what I'm talking about - if you leave out a big significant variable the derived climate sensitivity will come out 'way too high.
James Evans (6:29 PM) -
"The latest models that we can meaningfully test are nth generation models. However, we are currently using n+mth generation models, which are much better than those awful nth generation models that performed so poorly."
I agree that seems to be the case. If this premise were generally accepted, there would be much less contention; but there are many who believe that the current generation of models are finally right. And therein lies the rub. In fact, there are still those who yet defend even the IPCC FAR model of 1990.
You ask what is the way out. At some point, I expect that a model (not "the models" -- many will fall by the wayside) will become more accurate and no longer deviate so far from reality that we refer to it as "awful". I make no book on what generation "n" of climate model this will apply to; I rather doubt it is the current generation. However, it is logical to expect that a climate model is first able to predict, within reason, global average temperature for, say, a 30-year period (which seems to be a common metric for delineating "climate" vs. "weather"). Only later should one expect accuracy in regional estimation over such a period, or for reliable predictions of the average over a longer period, say, a century. [I note in passing that there have already been projections for periods of as long as a millennium, which seems useless in the extreme. Even under the assumption that climate prediction is a boundary value problem (already contentious), it is required to project the course of human energy consumption, more specifically the usage of fossil fuel energy. Not to mention the biosphere's long-term reaction to higher CO2 levels. It is at best an exercise in guesswork.]
Theoretically, there's another, less happy, way out: we finally throw in the towel and admit that climate is not predictable. Based on the history of human endeavour, I don't think this is likely (nor admirable).
little known fact the Carbon Briefs editor - Christian Hunt (former greenpeace) was also involved in the Zero CarbonBritain Report (part of the editorial team)
A former massage therapist and philosphy graduate, was credited in the Climate Change science section of the Zerocarbon Britain report (it's ok, so was Sir John Houghton /sarc off and Jo Abbess, in the renewables section)
One of these days I wil get around to writing about it, as the latter was commisioned by a parliamentary group to right report on personal carbon rationing.. Tim Yeo, and others quite keen on it.
I wonder as a mere BSC, and MSC and I over qualified to be a peer reviewer on the journal Climate POlicies (Houghton, PAchauri on the board) as said former massge therapist/philsophy gardutae is a perr reviewr onsaid title..
If only that Exxon funding could find its way to me... !! ;-)