Discussion > Do computer models provide 'evidence'?
Raff
Your point about no one getting worked up in science, engineering or maths if a colleague referred to a model output evidence. In a physicist and an aerospace/space engineer. I've developed and used models sometimes to help validate designs. If I were to say that models I used are evidence and left it at that I would get laughed at in any of the jobs I was in ( maybe even sacked on some ). If instead I say that they predict this that and the other my colleagues and I can make a judgement call as to whether we want to believe it or use it. That maybe what you are talking about.
That distinction where you choose to accept models as evidence is different than asserting models as evidence which was what has been pointed out.
Paul sums it up above. To most people evidence is observable data.
Micky H Corbett
Micky, if in your sentence above you used the e word in place of 'predict' (i.e. with suitable equivocation) I don't believe for a minute that anyone would bat an eyelid (unless they had been primed to be sensitive to the word). On the other hand I understand well that if you said "ok lads, my model gives us all the evidence we need, so let's bet someone's life on it" you would be lucky to get out with your dignity intact. Clearly context is everything.
Out of interest, if your work needed a good estimate of something that could not be measured, let's say the greenhouse effect of doubling CO2 in the atmosphere, would you and your colleagues use the best available analytical methods to estimate the probably value and use it with appropriate caution or would you all just find another job?
Raff
Raff,
if your work needed a good estimate of something that could not be measured, let's say the greenhouse effect of doubling CO2 in the atmosphere
I can't speak for Micky H, but personally I'd look for something that gave me an indication of what the value might be, for instance historical measurements and research. Then I'd look at what proxy data was available. If the evidence I gathered was contradictory (like Ancel Keys had in his research) I might decide that there was no correlation and put what I thought might something measurable and potentially a major problem further down the risk list and look for something else to investigate. If I selected only the data that agreed with my theory/hypothesis then I'd expect to have problems down the line when reality and predictions/probabilities/scenarios started to drift apart. By that time I might be a rich and famous and even a minor celebrity and not care.
What I wouldn't do is make a guess at what I thought it might be and work back from there.
SandyS
Mickey H Corbett: ‘What I get from the article is that it is interesting that if you take the models and the consensus then the outlook is actually not as bad as the doom and gloom.’
What I see in the article is an example of highly persuasive writing, intended to bring about a change in the reader’s views about climate change. I don’t think ‘interesting’ cuts it, except in an ironical way.
On the matter of evidence, I agree that there’s a degree of semantics involved, but what I think is also missing from the debate is: evidence of ‘what’?
Most people have done some sort of household budget, which is made up of a mix of past income and expenditure and current and projected income and expenditure. I can go through that exercise to arrive at a projected financial position 12 months out.
If I were asked to show my evidence for my conclusion, l would need to show both the actual past and the present as well as projected future income and expenditure.
So in some important aspects of our lives, we in practice regard the workings and results of models as ‘evidence’. It’s just not the same type of evidence as produced by direct observation and measurement (keeping in mind that observation can also be faulty).
This also ties in with the notion of planning based on models. As noted right at the beginning of these comments, we do accept the evidence from some models, for example wind tunnel testing.
Where I think Ridley has overstated his case is in making an absolute distinction between evidence/not evidence rather than stronger evidence/weaker evidence.
If evidence is regarded something that can be used to support an assertion/claim, then you could simply claim that climate models are evidence of what climate scientists think might happen in the future.
In my view, that would provide a starting point for debate, whereas debates over definitions can get bogged down in semantics and circularity, as you say.
Brendan H
Radiative physics models that calculate the effect of GHG in the atmostphere by calculating the effects on many layers is a standard way of treating such problems. This is not an abomination of climate science and nobody serious would argue that it cannot deliver a soltion which can be relied upon. Equally, everybody knows that the accuracy of the solution delivered depends upon many factors including the level of approximation (how thin are the layers, for example) and of course fundamentally upon the correctness of the implementation. As the approximation is made more realistic the accuracy improves (narrowing in, not changing radically). The model is delivering evidence (oops, sorry, strong grounds etc).Jul 27, 2014 at 8:19 AM |
"This is not an abomination of climate science and nobody serious would argue that it cannot deliver a soltion which can be relied upon."
Not an abomination of climate science?
'Radiative forcing' is a notion dreamt up by 'climate scientists' ** and can exist only in a computer model. It does not exist in reality. It cannot be measured physically and models used to compute it are not capable of being validated by comparison of what they compute with physical reality. So anything calculated using the concept remains hypothetical.
The problems with climate models are far more serious than little details like the thickness of the layers used in radiation computations.
"...nobody serious would argue that it cannot deliver a solution which can be relied upon."
I think a *lot* of serious people would say that computer calculations of the effect of GHG's in the atmosphere cannot be relied upon. The evidence of their unreliability is there for us to see.
____________________________________________________________________________________
**
Radiative Forcing‘the change in net (down minus up) irradiance (solar plus longwave; in W m–2) at the tropopause after allowing for stratospheric temperatures to readjust to radiative equilibrium, but with surface and tropospheric temperatures and state held fixed at the unperturbed values’
IPCC AR4
Martin A
Brendan H
I would agree with you about Ridley's article being persuasive as upon reflection my interpretation is biased to not worry about the details and to look at the general scientific picture. But other readers may not do this and think that Ridley is being a bit scattered in his overall approach.
The stronger evidence/weaker evidence thing is an interesting way to look at it. As I said earlier I think that asserting that models are evidence was a bit of a faux pas in the Lawson debate - maybe that was why Matt Ridley reacted. The fact that we are all discussing the nuance of models I think by now shows that we probably know how to gauge it. The non scientific person maybe not so much and certainly the BBC should have shown a bit of restraint. Also Matt Ridley maybe should have clarified what he meant. I'd agree with that.
Raff, when you asked "if your work needed a good estimate of something that could not be measured, let's say the greenhouse effect of doubling CO2 in the atmosphere, would you and your colleagues use the best available analytical methods to estimate the probably value and use it with appropriate caution or would you all just find another job?"
I've had to do this sort of thing for interaction of ion beams from plasma thrusters and for erosion of grids over life for example. I'd say that if all you had were analytical methods, that you didn't have any test data or even previous flight data as a basis to estimate a ball park figure you would use the phrase "it's beyond the remit of the project". Basically it'll be too expensive and take too much time, and also models on their own are just theory. You have to back it up with something. I wouldn't feel right providing only models - it would to me at least feel ethically wrong (that's just me though - I can't speak for others).
Often a careful analysis of assumptions before any work has been done is all that is needed to show that it will be very hard to produce a meaningful or useful result. Another aspect is that maybe the equipment to measure to that accuracy hasn't been invented yet. And yes this meant a project was indeed terminated so either it was time to get a new project or a new job/contract!
Micky H Corbett
Martin, I phrased it badly but I was referring to the use of numerical methods to calculate the effect of GHGs as not being an abomination. I've not seen serious criticism of the use of such methods or of the result, around 1.2C.
Micky, the thing is we have no other option if we want to understand climate. We cannot just say that the problem is out of scope or too difficult.
Raff
Raff
I appreciate that but the method can't be short-cutted. I had a discussion post up here long ago about experiments to perform in order to gain more understanding of CO2 effects. No one took it up but that's by the by.
I think for Co2 you would need to look at the lab experiments and try to characterise as much as possible. Could convection effects be removed or minimised in some way so as the radiative effects could be seen? That would be a start.
Could a 100 m box be built to simulate an atmosphere? That kind of thinking. Start with something known and increment.
Micky H Corbett
Mickey H Corbett; ‘The stronger evidence/weaker evidence thing is an interesting way to look at it.’
When I read the comment about ‘evidence’ from models, I was quite sceptical. But then I began to think about what would constitute modelling and what would constitute evidence, and hard and fast distinctions sounded less convincing.
The wind tunnel as a model has already been mentioned. A chemistry experiment in the lab may not be considered modelling, but because of its controlled nature, nor is it strictly real-world interaction.
So I think there’s a place for considering the results of climate modelling as evidence, as long as the inputs are based as much as practicable on real-world interactions rather than wishful thinking, and the results are suitably hedged.
So I think you can have a range of evidence types. This ties in with communication, and the distinction between prediction and forecast/projection. In everyday discussion – even by scientists – it’s not easy to maintain the distinction clearly, since even the best communicators tend to use near-synonyms indiscriminately and morph towards expressive language and lack of nuance.
Brendan H
Raff finally conceded that evidence is not the correct description to describe model output, and for that I'm grateful. It wasn't a very well-tempered admission, but then we're all big boys here, we can take it. Unfortunately that major turning point got lost in the cross-fire of the posts.
The main thesis of the thread is agreed upon.
TheBigYinJames
"A chemistry experiment in the lab may not be considered modelling, but because of its controlled nature, nor is it strictly real-world interaction. (sic) " Brendan H
Huh?
In what way are things done in the lab not 'real world'?
I suppose you mean that what you observe in the lab, where you can control temperatures, impurities, pressures, or whatever, does not tell you precisely what happens in, say, an industrial process where those things are less well controlled.
It's a giant leap to go from that to saying that it therefore follows that unvalidated computer models of climate systems (built on a foundation of unverified assumptions, 'parameterizations', and so on), produce evidence about the physical world.
Big Oil
I agree BigOil. Brendan H is unfortunately opening up the definition of 'evidence' again to include everything which suggests a hypothesis might be true. Since we spent about 700 posts trying to nail the scientific definition of evidence, this is not helpful.
Evidence is a physical effect of something happening. You can't have a physical effect of something that hasn't happened yet. You can only have a prediction, which is not evidence, no matter how well understood the basis of your prediction is. Me saying the sun has risen a billion times in history is not evidence that it will tomorrow, and the basis of rotation and gravitation is well understood.
Case closed!
TheBigYinJames
Evidence is a physical effect of something happening.
That's it, in a nutshell.
Martin A
TheBigYinJames
Just to add a touch of humour, I read a comment somewhere in the last couple of days and can't remember where to credit the originator, it was along the lines of the fact that the sun rises everyday proves there's an endless supply of suns.
SandyS
The ancients used to believe that the sun died every night. The Aztecs in particular at one time believed the blood of sacrificed men was needed to revive it, so ended up slaughtering thousands of them.
It's funny the sorts of crap that the gullible will believe, and the amount of harm they will inflict on others to support their delusions.
TheBigYinJames
TBYIJ
It's funny the sorts of crap that the gullible will believe, and the amount of harm they will inflict on others to support their delusions.
A new one on me, mentioned (I think) earlier in this thread (SandyS?) was the 19th century Xhosa cattle killing in southern Africa. Far more dire for the Xhosa than the Climate Change Delusion will have been for us, but with something in common.
In 1856, a fourteen year old girl named Nongqawuse had a vision on the banks of the Gxarha River in southern Africa. Entranced, she saw dearly departed ancestors, their cattle hiding in the rushes, and she heard other cattle underground waiting to come forth. She was told that if her people would but kill all their cattle, their ancestors would arise from the dead, the cattle lowing in the subterranean passages would come forth, and all the whites would
be swept into the sea. Nongqawuse’s prophecy provoked the colonially embittered Xhosa (cōe-săh) people to rise up and kill their cattle. As the movement drew to a close, around 400,000 cattle had been slaughtered and an estimated 80,000 Xhosa died of starvation. Those that remained were reduced to working as laborers throughout the Cape Colony after being pushed off some 600,000 acres of their ancestral lands.
Martin A
Those are really odd beliefs aren't they? How can those foreigners and peasants have been so stupid? But there are still, I'm told, numerous people in developed societies who believe that some guy who got himself nailed to a cross 2000 years ago was the son of a god and that he somehow did it on purpose for their benefit! And get this, they even believe that this same god cares for us all, yet at the same time allows unspeakable suffering, cruelty and pain. And to cap it all they believe that we are incapable of destroying our environment because this god won't allow it. the wacko things some people believe, eh?
Raff
Raff,
I take it back,. You're an offensive cretin.
You still seem to believe that we're quasi-religious right-wing nuts here, this is in spite of a quite heartfelt explanation about my own origins a few posts ago. The further implication that I am somehow racist by pointing out the Aztecs believed some funny things is objectionable and I demand an apology.
You are currently in a discussion thread with a bunch of atheists with impeccable scientific qualifications, and even a bona fide academic scientist currently working in the field. We have been wasting our time with you, and your petty prejudices, which you appear to cling to long after they have been debunked. I am sorely disappointed with you.
I will be ignoring your input raff until I receive an apology.
TheBigYinJames
" And to cap it all they believe that we are incapable of destroying our environment because this god won't allow it."
Come on Raff, you made the last bit up.
As usual, I find it hard to make out what you are on about.
That not believing in the climate change religion goes hand-in-hand with believing in the sky-fairy?
Martin A
Yes, I agree, it was ill-judged and I apologize. Martin, see http://en.wikipedia.org/wiki/Cornwall_Alliance.
We believe Earth and its ecosystems – created by God’s intelligent design and infinite power and sustained by His faithful providence – are robust, resilient, self-regulating, and self-correcting, admirably suited for human flourishing, and displaying His glory. Earth’s climate system is no exception. Recent global warming is one of many natural cycles of warming and cooling in geologic history.
Some of the signatories might surprise you.
Raff
Raff,
Apology accepted.
Yes there were and are some people on this side of the fence that I wouldn't trust as far as I could throw them.
But I cannot control what they say they believe, even if it muddies the water with proper scientific scepticism.
There are some real cranks this side of the fence.
By the same token, the catholic church allies itself with strongly the climate alarm movement. Should I judge you and your beliefs by that?
TheBigYinJames
Some of the signatories might surprise you.
Well I only recognised two names under scientists etc I think that at least one is widely known as a god botherer who is sceptical of CAGW so hardly surprising he'd sign it.
Plus Pat Boone. Haha
But on their website I found nothing like your "And to cap it all they believe that we are incapable of destroying our environment because this god won't allow it."
So Raff, as I said, *you made that up*.
The nearest I could find was the following:
WHAT WE BELIEVE1. We believe Earth and its ecosystems—created by God’s intelligent design and infinite power and sustained by His faithful providence —are robust, resilient, self-regulating, and self-correcting, admirably suited for human flourishing, and displaying His glory. Earth’s climate system is no exception. Recent global warming is one of many natural cycles of warming and cooling in geologic history.
2. ...
Taking out the references to the supernatural, I myself would have no problem signing up to:
1. We believe Earth and its ecosystems —created by God’s intelligent design and infinite power and sustained by His faithful providence —are robust, resilient, self-regulating, and self-correcting, admirably suited for human flourishing, and displaying His glory. Earth’s climate system is no exception. Recent global warming is one of many natural cycles of warming and cooling in geologic history.
Not exactly a crackpot statement even if you don't agree with it.
Martin A
"Should I judge you and your beliefs by that?" - if it amuses you, go ahead. It would be a bit forced though. I think you would have trouble really believing that any significant part of the population that represent AGW science were acting from religious belief, whereas a belief that many opponents of the science are religious does not seem so far fetched (to me). I started wondering whether this site had its motivation in religion when talking to Radical about evolution here - how deep does it go? As of now I know of two atheists here, one probable religious and the rest unknown so it is too early to judge.
Did I make it up (about god not allowing things)? No, I have read it somewhere or other. It is of course possible someone else did. But my quote above highlights the word "sustained" which implies divine *intervention*, as does "powerful sustaining" which is highly indicative.
Our examination of theology, worldview, and ethics (Chapter One) finds that global warming alarmism wrongly views the Earth and its ecosystems as the fragile product of chance, not the robust, resilient, self-regulating, and self-correcting product of God’s wise design and powerful sustaining.
And then there is the suggestion that GW alarmism fails the test of theology, which can be taken to imply that the 'creator' would not allow his creation to be endangered:
We believe that idea – we'll call it "global warming alarmism" – fails the tests of theology, science, and economics.
I don't have the patience to read it in detail (a sentiment I imagine you share), it is a truly odd mixture of religion and free market economics. But I don't think my assertion (about god not allowing things) is unreasonable. I agree there is no smoking gun (direct *evidence*).
Signatories - from Wiki there are Lindzen, Carter, Spencer, Mckitrick - I was aware only of Spencer. I haven't read the site to find more.
As for the de-godified statement, describing ecosystems as "robust" is highly debateable and dependent on how you define the word. When species come and go in the geological blink of an eye (those in existence now are thought to be just a tiny fraction of those that there have been) and can be exterminated with relative ease, I would conclude that they are fragile. We only need to move a species to the wrong place and it can run amok and destroy its new home for the original inhabitants.
Resilience and self regulation are similarly poorly defined. Over long periods Earth clearly is. But we are not interested in long periods but rather in short ones (the next few generations or tens of generations).
Admirably suited for human flourishing? Was it always? In 4.5 billion years we have been here a hundred thousand or so. Would we have flourished at other times?
And in saying that recent global warming is "one of many natural cycles of warming and cooling in geologic history" you seem to be saying that AGW doesn't exist, that recent warming was natural. I find that slightly incosistent to say that you know for sure that the warming was natural (not AGW) and yet to proclaim cluelessness about future movement of temperatures.
Raff
"In 4.5 billion years we have been here a hundred thousand or so. Would we have flourished at other times? "
Wassamatter, don't you believe in evolution?
It's not relevant what people think about other subjects, one should address their BEST arguments in the subject at hand. Not their weakest, certainly not things they have not even expressed and which you have assumed ot inferred.
And this one has gone, jumped the shark.
rhoda
I've been following this debate on and off for the past week or so and the tortuously of the arguments used by Raff in which he veers from models, to analytical and numerical solutions etc. is mind numbing. I thought at one point agreement had been reached but apparently not.
The answer is straightforward. Models cannot provide evidence. Philosphically a model is simply the embodiment of a hypothesis. It matters not that the model might be a loosely assembled bunch of ideas, or be coded in an algorithm that may or may not have an exact analytical solution, or may need numerical methods with approximations associated with gridding, convergence etc.
Evidence is the body of observations and facts that one brings to test, or evaluate the output of the model. Such evidence can only be provided by observations of nature, measurements, or the output of experiments designed to interrogate nature. Evidence cannot be provided by the output of any model.
The best we can say is that the outputs of a model are consistent with the available evidence. We cannot prove a model, but we can disprove one when evidence does not agree with the outputs.
These same comments apply to models of physical and chemical processes, models of economic and social behaviour etc.