Monday
Apr082013
by 
Bishop Hill
Bishop Hill Tamsin Edwards versus Jonathan Jones
Apr 8, 2013 Climate: Models Climate: Sceptics BH regulars Tamsin Edwards and Jonathan Jones are facing off in a debate about the usefulness of climate models at the Cheltenham Science Festival on 7 June.
Computer-generated models are used to predict future climates, but how much faith should we put in them to guide future actions? Should we treat their predictions as fact or fiction? With the hot topic of climate change ever current, can we wait to find out? Join climate scientist Tamsin Edwards, sceptic Jonathan Jones and policy adviser Claire Craig. Temperatures could rise in this session…
Details here, although tickets don't go on sale for a couple of weeks.
Reader Comments (59)
Tamsin described as a "climate scientist", Jonathan Jones as a "sceptic" -couldn't they have mentioned that he is a physicist?
Jonathan signed off on this copy by the way. I asked if they could put our titles in (to show he is a Prof) but they said they never do, to save space.
They also asked us for biographies for the website, though they don't seem to have linked to them.
How do you actually discuss 'faith in models'. That immediately makes it sound a pretty pointless religious debate.
Surely that would be an ecumenical matter?
I did sign off on the copy after a couple of minor modifications which I requested. As I am saying on twitter I am definitely a sceptic of the "policy usefulness" of climate models, so I was happy for them to use the phrase in that context, rather than my usual "climate agnostic" self-description.
My bio (which they haven't yet used) makes very clear that I am a scientist with wide ranging expertise, but no special expertise in climate science. A central part of my claim is that professional climate scientists are over confident about the quality of their work, but this is largely because all scientists are over confident about the quality of their own work. Outsiders, without the emotional commitment that working in a field inevitably brings, can take a much more objective view of things.
"Faith in models" is a natural extension of "belief in climate change".
Do you believe in climate change?
If so, do you have faith in climate models?
Quick definition from Google - "faith: Complete trust or confidence in someone or something." Isn't that ok to be discussing? Come to the event (and ask a question!) rather than pre-judging it from a couple of sentences in the program...
Mankind would be better served if all the money & resources spent on trying to predict future climate, was instead spent on trying to predict future weather.
I should add that if anyone is expecting me to go in all guns blazing, denouncing climate science and all its creations, they are going to be severely disappointed. The recent Economist article provides a reasonable guide to the line I currently plan to take.
I am sceptical about the policy usefulness of models, and I am more than sceptical of much of the nonsense published in paleo proxy studies. I also think that scientists are far too willing to add "catastrophic" to almost anything which happens in the world (the recent book by Rupert Darwall is an excellent introduction to this phenomenon). None of that alters the fact that there is a core of truth in AGW, albeit decorated with all sorts of fads and fancies.
.....and what about Clair Craig?
Prof Jones
Regarding the addition of 'catastrophic' to every topic that can be remotely linked to climate change / global warming, this is clearly just a case of grant-chasing by the researchers. This is not intended as a harsh criticism of the researchers, but more of how research funding is administered and what gets published.
My academic experience was around geology departments in the UK for most of the 1990s, at a time when catastrophic extinction events were the 'in vogue' area of research. Didn't matter what you were actually researching (geochemical analysis of precious metals as it happens), you had to try and get some reference to catastrophe and extinction events into your grant proposal or paper no matter how tenuous the link was. My assumption is that the current situation is largely similar, but with climate change as the 'must have'.
"...albeit decorated with all sorts of fads and fancies."
Rococo climate change! What a delightful image that conjures up Jonathan.
Jonathan Jones
And from the "core of truth" the embellishments spun. With the real cause for concern being how rapidly they morph from a hypothesis with high a degree of uncertainty into absolute life changing certainties. Is there such a thing as due diligence in Climate Science?
Yes, what is a "policy adviser"? And what added value does a policy adviser bring to the debate?
I should declare up front that I am a policy adviser (no charge made). However, all my policy advice falls on deaf ears.
@Jonathan Jones
Climate models are very useful for policy. They justify grandstanding by politicians, rent-seeking by "entrepreneurs", and empire-building by bureaucrats.
Jonathan Jones
I'm inclined to think this too, but following the comprehensive defenestration of most the sacred texts of the CAGW acolytes, as far as I can see there's precious little left that isn't contentious - maybe a degree Celsius of warming per doubling of CO2, but observations suggest even this is becoming a faith-based stance. Is this AGW even GHG-based? I'm sure we can change our environment by deforestation and urbanisation, but is there really any signal from GHG that isn't completely overwhelmed by non-anthropogenic sources?
Jonathan,
If you have the time, I suggest you read through this thread on Tamisin's blog:
http://allmodelsarewrong.com/limitless-possibilities/
It convinced me that what climate scientists are trying to do is impossible.
Good luck!
Tamsin apparently put different boundary conditions into a model from elsewhere to find out what happens.
Pity the models can't predict the future because they're based on incorrect physics.
Not sure my last comment was submitted - this replaces it.
What about Claire Craig? She is Deputy Head of the Government Office for Science, and works with the (recently changed) Chief Scientific Adviser. She has a background in geophysics.
I invited her because she seemed to talk sense (and was quite funny) at this Bristol university event: http://storify.com/flimsin/trust-me-i-m-a-scientist.
My favourite quote: "Trust isn't binary. I trust my husband, but I wouldn't trust him to fly an aeroplane I was in."
@ Tamsin Edwards, Apr 8, 2013 at 10:19 AM
"Jonathan signed off on this copy by the way. I asked if they could put our titles in (to show he is a Prof) but they said they never do, to save space."
"climate scientist" is a lot more letters than "sceptic" or "physicist". Saving space?
@Richard Tol,
Completely agree. I think climate scientists need to understand that when they get dragged into policy debates they are serving the interests of politicians, civil servants and quangocrats, not vice versa.
@Tamsin Edwards,
Having briefly tried to fly a light plane I wouldn't trust myself to fly a plane, but I did get the hang of operating the radio and the radar transponder.
@Jonathan
Glad we agree.
State of the art climate models are indeed not particularly useful for policy analysis cq hypothetical policy makers who seek the greatest good for the greatest number. Analyses of climate policy typically use simple climate models, but few of the key insights depend on the climate model.
This promises to be a civil, and worthwhile event. I am very much looking forward to hearing about it afterwards.
Given the very rapid development of computing power and skills over recent decades, it would be unwise to dismiss computer modelling of the climate system completely. Who knows, perhaps some good will come of it one day, and perhaps Dr Edwards will be one of the pioneers.
So far, though, I get the impression of net harm rather than good thanks in large part to climate model outputs being used in political arenas as if they were like observational evidence, and against the warnings of modellers themselves - warnings that can be found buried within IPCC reports - that they are not suited for prediction.
Until they are, more humility in their promotion would be in order, and clear distinctions made between their role in climate over months and decades, and in weather over minutes and days. Also in anticipation of this happier future in which climate models might help us a little, I suspect we should be putting a great deal more effort into the far less glamorous business of collecting data throughout the climate system. How much better off would we be now if instead of the politically-driven IPCC squeezing what mileage it could out of CO2, we had had a major international collaboration over improved data collection throughout the climate system? We might have had decades of good data rather than the decades of distraction we have had to endure thanks to so much heed being paid to the alarums of zealots for one cause or another.
Climate models seem to completely ignore Exponential Error, the few times I've seen it actually mentioned they falsely claim a few models can "average" out the exponential growth of errors. It was this that led me to the fact that climate "research" was a scam.
I meant to include this reference in which some climate professionals appeal for more data collection, but I got carried away with my own blethering instead. Here it is now, along with a short extract: http://www.pnas.org/content/110/12/4435.full
If "models" were any good, Ladbrooks would be offering odds.
"Given the very rapid development of computing power and skills over recent decades, it would be unwise to dismiss computer modelling of the climate system completely."
Oh dear me. They're the worlds most expensive dice.
John Shade
Absolutely right. If you have the connections/reputation you can get funding for "new stuff". What is imposssible to get is funding for baseline studies, and especially continuation. If it's been done before and isn't novel, then you have no chance. Climate "science" is exemplary for this: when everything is new, there is no context and therefore anything can be interpreted any way.
"Climate models seem to completely ignore Exponential Error"
They don't ignore it, they remove it by implementing low pass filters and variable resetting between integration time steps. This is necessary to maintain "stability" - bypassing violation of physical laws (conservation of mass, momentum and energy) and exceeding boundary limits.
One point that I think should be mad ein this debate and shouldn't prove too controversial is that climate models are models first and test bed for climate science second.
By that I mean that have knoweldge of complex models is more important to the debate that wha tis being modelled.
This is true of all models, be they weather models, climate models, macroeconometric models or anything else.
Judging by the previous comments, it is likely to be an enjoyable affair, but whether it will bring the protagonists any closer together is likely to be another matter entirely.
As long as people attach more importance to being right than being accurate, nothing much is going to be accomplished (in my opinion).
I don't recall anyone ever saying that they wished that they were wrong but they don't think that they are.
I do recall Professor P Jones saying something to the effect that he hpoed he was right even though it was bad news for the world. I found that to be profoundly frightening.
Tamsin Edwards asks "faith: Complete trust or confidence in someone or something." Isn't that ok to be discussing?
To which I would simply point out:
If there were good evidence to corroborate climate models, would you need faith? And urge climate scientists to spend more time trying to gather corroborating evidence and less time trying to instil faith.
Who said I was trying to instil faith? I called my blog All Models Are Wrong!
Even the event description says "but how much faith should we put in them" - Chelt Sci Fest are not trying to instil faith either.
This (http://www.newsdaily.com/stories/bre936084-us-climate-oceans/) seems at a first glance to be a new example of models adding too much to some limited observations. Anyone know more?
Regards, Tony.
Strikes me that Tamsin's first question is precisely the right one - "how much faith should we put in them to guide future actions? Should we treat their predictions as fact or fiction?"
A model will have some use, it will be an approximation of greater or lesser skill. I doubt that is particularly contentious. Debating quite how much reliance should be placed on a given set are is a very worthwhile exercise, particularly given the degree policy and resources are claimed by the outputs.
The second part of the question is a bit binary though. Given the first premise, then the question is how much fact or fiction is entrained, not one or t'other.
Won't be able to get there unfortunately, but hope it goes well.
Roger Longstaff
That "process" doesn't sound like it "works". Makes a mockery of the "smaller and smaller cells" if they just "average" them together at the end of it.
I've said this before.
A model is a small imitation of the real thing.
AC1 - agreed.
And it is simply staggering how few people seem to understand your point about exponential error accumulation in a complex system!
Tamsin : I am not pointing a finger at individual scientists - there are always good ones (although most of them keep silent).
But when a group of climate scientists look at the output of climate models and conclude that the Earth's climate system is unequivocally warming (which I am prepared to accept if you also specify a time horizon), and then go further to say that it is more than 90% certain that humans are causing most of it through activities that increase concentrations of greenhouse gases in the atmosphere - is the second part not a giant leap of faith?
And when they go further and exhort others to accept this because there is apparently a concensus behind this, are they not trying to instil their faith in others?
You have to buy a ticket!
"A central part of my claim is that professional climate scientists are over confident about the quality of their work, but this is largely because all scientists are over confident about the quality of their own work."
All scientists? Maybe you are thinking about the scientists at CERN who supposedly named the object of their search the "God Particle." It seems to me that the CERN scientists showed some humility when they told funding agencies that their science could not move forward without a collider that costs a bazillion bazillion euros. By contrast, climate scientists told funding agencies that the science is settled and they needed a mere billion euros to perfect "science communication to the public."
I would really like to be present when Tamsin Edwards and Jonathan Jones debate climate models. I have great faith in both of them.
However, I am looking forward to the day that this topic captures the interest of some high level computer programmer who is not in the employ of a climate concern and who is willing to explain the operation of models, not just climate models, down to the level of the heuristics that are used in the computer code for the models. I think it will be very important for everyone to learn that once the "science" is "programmed" then the programmer has to figure out a way to make the program solve. The programmer is contributing computer heuristics. Different heuristics give different solutions.
@Tamsin <My favourite quote: "Trust isn't binary. I trust my husband, but I wouldn't trust him to fly an aeroplane I was in.>
Well you're easily pleased Tamsin. Trust certainly is binary and whether 1 or 0 depends on the characteristic referred to. The truism that trust once lost (in a relationship) can never be regained is one of life's hard lessons. Or to say that I now distrust the BBC (as an organisation) admits no qualification, however praiseworthy some of its output might be.
It sounds like it should be a fascinating discussion. It's great that the three of you are taking the time to do it. I live in the Cotswolds so I'm tempted to go along.
@simon abingdon: "The truism that trust once lost (in a relationship) can never be regained is one of life's hard lessons."
That is what climate scientists are slowly having to appreciate, because when so few are prepared to speak up about the less trustworthy claims of others', they jeopardize trust in the entire profession. And no amount of re-marketing will regain that trust in the short term.
Apr 8, 2013 at 5:42 PM | matthu
It seems that The Team has a new, improved marketing/communication trick™.
I first noticed this a few weeks ago, when much to my surprise I found that IPCC Lead Author and BC Green Party Candidate and Deputy Leader (and climate modeller, so this is almost on topic), Andrew Weaver decided that he needed to protect his delicate little tweets and conversations from my view and/or protect his circle of twitteratti from being exposed to the few and far between tweets/questions I've sent in his direction.
My guess at the time was that this was a trick™ he had picked up from Mann (who will block any who fail to be sufficiently sycophantic, I gather); but this trick™ now seems to have gone "viral" ... Gleick, Laden, Verheggen all seem to have adopted this rather unique mode of "communication".
What a way to win friends, influence people and "gain trust", eh?!
Apr 8, 2013 at 10:48 AM, Tamsin Edwards said
I've always defined faith as a belief in something or someone in spite of a lack of supporting evidence, or facts to the contrary. If your confidence is based on fact, it's not faith.
cheers,
gary
I think I am right in assuming that both Tamsin and Jonathan hold much respect for the expertise and critical objectivity of climate scientist Judith Curry. In this respect her blog provides a treasure trove of debate on the usefulness/shortcomings of climate models. Here for starters is Judith's conclusions on one such blog thread.
' JC summary
Each of these papers makes an important contribution to understanding how we should use climate models and make inferences from climate model simulations. The manner in which climate models have been tuned makes them of dubious use in 20th century attribution studies. In most regions, the climate models have little skill on regional projections of climate change on time scales of 50 years.
Nevertheless, we have the recent recommendations from the NRC National Strategy for Advancing Climate Models that assumes that climate models are useful for these applications and the focus is on supporting decision making. These papers provide further evidence that climate models are not fit for these purposes, and that we are not currently on a path that is likely to improve this situation.'
http://judithcurry.com/2012/11/09/climate-model-discussion-thread/
I dont know what angle the third debater, Government policy advisor Claire Craig is coming from, but for my part I am deeply suspicious that the so-called policy relevant project funding carries with it tacit political pressure on scientists to deliver 'what they want to hear', and that this places unsavoury pressure on scientists to support rather than interrogate and challenge existing memes. Perhaps she can assuage my suspicions.
I hope the debate goes as respectfully and critically as I expect it will.
None of that alters the fact that there is a core of truth in AGW, albeit decorated with all sorts of fads and fancies.
Apr 8, 2013 at 10:53 AM | Jonathan Jones
===================================================================
No sceptic of any intelligence denies that. The real question is how much whether the A part of GW is of any import. Given this century, it is looking that the natural variability of climate has swallowed up the A part so that it of no import.
Re modelling - I am right, I think, in noting that Phil Jones stated that we can't even model the natural variability of climate. In which case, what do we gain from trying to model exception conditions to that?
Please?
And another thing. What is the optimum global temperature? And if we do not know that, are we not pissing in the wind anyway (frankly, I think the idea of a global temperature is of no service whatsoever).
Not related at all to the discussion between Jonathon and Tamsin. But I was amused by the talk by the vegan society at the same festival discussing the carnivores dilemma in whether they need to change how much meat they eat as 'science has shown our love of meat contributes to carbon emissions and climate change.' I think I might know what the answer will be.....