Discussion > Let's get real about climate models
Martin:
Is there even the slightest reason to believe that 'assumed' and 'actual' will 'average out'?Spot on!
No reason whatsoever. The relationship between the climate and model will continue to diverge by orders of magnitude during repeated iterations. The only way to limit this behaviour is with arbitrary, non-physics based constraints which are judgement-based/human-derived and wholly unrelated to the dynamical processes of the climate system being modelled.
Simon Hopkinson
Martin, if I were to write a model of climate (not that I am competent, just for argument's sake) I would want to include the solar cycle amongst other natural forcings. But I don't know what they are. I can only estimate based on past experience (measurements, observations). It might be that my estimates are wrong because I don't have enough data (the statistical properties of the variation in these forcings are unknown). Or they might be wrong because although I had enough data, something has really changed. I obviously can't tell between the two - either way, I have my estimated forcings, they might be wrong, but they are as good as I can hope for.
So I run my model and ten years later it turns out that the results were wrong. It is not reasonable for you to criticize me because the estimates I made on the available data were wrong. So you should restrain you criticism to a comparison of the model with corrected forcings against reality.
Raff
It is not reasonable for you to criticize me because the estimates I made on the available data were wrong.I hope no-one would disagree with you on that, raff.
Where we would criticise you is if you clung like a drowning man to that model, swore black is white that the model was in some way "evidence" or that running your model was an "experiment" and that it was "the best we had " and that it ought to be used as the basis for milking the taxpayer of several trillions of dollars/pounds/euros to prevent the catastrophe which your model says is just around the corner, even as the evidence of your own eyes is telling everyone around you that you are mistaken.
And when you compound that piece of idiocy by telling us that we don't know what we're talking about because we're not experts .....
I mean, how "expert" do you need to be to see what is going on around you?
That is what makes us sceptics, raff. We are watching warmists make every excuse for models that patently don't work and they have been told a dozen times or more that they don't work and that they can't work in the current state of knowledge.
When I said the other day that climatologists should come back in 1,000 years or so and we might listen, I was only half-joking. I repeat what else I said in that post: climatology as a scientific discipline is barely out of the womb. Time to stop pretending otherwise.
Mike Jackson
Raff, here's some comfort for you: It's not just YOU that can't write a model of the climate. Absolutely nobody can. Modelling the climate is beyond our capabilities and it always will be. The more we learn about climate mechanisms, the more we learn of our woeful inadequacies in comprehending its complexities and the impossibility of simulating its minutiae. You might be able to write a very crude (comparatively) approximation that behaves with some similarities to the climate system but it will not be an accurate model, more certainly not with precision, infinitely more certainly not sufficiently well to be trusted to inform.
We don't need to wait 10 years for you to discover that your model was wrong. We can tell you before you start, it will be wrong. It is impossible for it to be right, since you have no measure of the uncertainties, the unknown unknowns and unforeseen events and influences which will have ultimate control over the validity of your result.
There is no earthly reason to restrain criticism of you for wasting your time in formulating the model. Rather, if your model's projections were to impact on our economic future (eg de-industrialisation, poverty's "heat or eat" dilemma etc) we have a moral duty to criticise without restraint or mercy, call you an idiot or whatever is justifiable (and I am confident that it is).
Simon Hopkinson
Even the IPCC itself admits that climate models do not actually model the real climate though I would need to go on a longer hunt than I have time for at the moment to find the precise reference. And as I have said before, the IPCC also agrees that climate is non-linear and chaotic and that it is impossible to say what the climate will be like in 50 years time, so how can you possibly construct a model that is in any way meaningful?
I will add that while taking a quick look for that reference above, I found a link which turned out to be to SkS which included the quote "we can't afford to wait 30 years ..." We can afford to wait 30 years because the only "evidence" that anything drastic is going to happen in the foreseeable future exists in the models and in the minds of those activists who are making hay out of this whole thing — money men, politicians, environmental activists and the "world government" crowd at the UN.
And that's not conspiracy theory; that's just opening your eyes to what is going in around you and paying attention to all sides of the argument and remembering that environmentalists are quite happy to stretch the truth and the facts to breaking point to get their way. (Take that from someone who has dealt with them for 20 years in other contexts.)
So don't believe that Bangladesh is in danger of sinking beneath the waves. It isn't. Don't believe that coral atolls will disappear because if rising sea levels. They won't. Or that they will all die because of "ocean acidification". That's not going to happen either.
If people (politicians especially) listened a bit more to the real experts on sea levels and coral and oceans and glaciers instead if environmental activists with an agenda and little regard for the truth and climate "scientists" with damn all in the way of qualifications in any if these fields, we might be better placed to make some sensible decisions about how to deal with things like floods and storms and droughts and heatwaves and cold spells — all of which are weather and all of which added together make up what we call "climate".
Enough already for one afternoon!
Mike Jackson
Mike Jackson, the problem with real experts on sea levels, is they used to use pens, paper, typewriters and cranky old fashioned stuff, that is not compatible with hi-tech computer adjusted modelling technology.
golf charlie
Raff, thinking about it, I am not sure that I can see how your project to re-run the models after the event, with known "forcings" is possible. For the reason that some are not understood (eg the effect of cosmic rays or the Earth's magnetic field on climate) and others are not observable with sufficient precision (eg incoming solar radiation).
We hear of "radiative forcing of climate by trace gases" but the atmospheric CO2 concentration is in fact a parameter of the climate system; it is not itself a "forcing".
Entropic Man has profound faith in the holy formula of climate science that translates changes in atmospheric CO2 to 'radiative forcing' changes; delta F = 5.35 ln (C/C0). But, so far as I can see, this formula remains no more than the result of fitting a log formula to a few numerical results from a model for radiative effects (paper by Myhre and co). It is not a thing that can be physically measured.
Martin A
MJ: my opinion, nicely encapsulated (which seems to be more than I can do).
Radical Rodent
Mike, Simon, your degree of certainty is something to behold. But how do you know all of that? Of Simon, I don't know, but I'm certain that Mike lacks a background that would provide such confidence.
Martin, not possible? I don't see why not. They computed and used what they considered actual forcings up to 2005 and predicted them thereafter. To update the graphs they need only compute and use actual forcings between 2005 and now. The forcings they compute may well be not quite right, but they'd be better than the originals. What do you make of fig 3 in this updated comparison between models and reality that I linked to before?
Raff
Wonderful!
Now raff knows enough about me and my 'background' to tell me what I know and don't know.
Tell me, sir, whence came this wonderful insight into someone you have never met and about whom you know virtually nothing?
And which bits of your comfort zone am I pricking holes in, exactly? Sea levels? Corals? It's all there in the literature. You just need to find it and read it, like I did.
By the way, the second paragraph in that reply is going to have Martin A and RR and Simon Hopkinson rolling around in hysterics or I'm very much mistaken. You just don't get it, do you? You can compute what you like between 2005 and now as much as you like. It will still tell you diddley squat about what is going to happen in the next 10, 20, 50 years,
Which bit of 'non-linear' and 'chaotic' is it that you're having a problem with? Tell us and we'll try to help.
Mike Jackson
Oh good grief! Much as I would prefer a direct answer, it seems that you choose to answer my question by more… well… obtuse methods. One’s “background” is utterly irrelevant when one is merely an observer of events; that would look good in a court of law: “The evidence of this witness cannot be allowed, m'lud, as ’e is obviously not trained as a witness…” You might expect us to give you a bit of leeway in your conclusions, Raff, when you declare that you may be lacking in some functions as a “climate expert” (“(not that I am competent…”), but you then decide that no-one who might have almost as little knowledge as yourself can have no conclusions, particularly when their personal certainty is that there is too much uncertainty to jump to any conclusions. You are displaying a very twisted logic, there, I’m afraid.
And what is with this term, “forcings”? This implies an alien element within the system, which can be removed, which is unlikely. In other words, I moot that there are no “forcings”, merely that within the system there is at least one unknown element of the system – or, more probably, that there are many inter-relationships between all the known elements, as well as the unknown elements, involving at least one, but more likely many unknown reactions, many of which we are as yet unaware of, let alone able to calculate, or to calculate for. No “forcings” involved, just things that we have yet to find – or, as many, many others have said: we need more data!
As Nial has intimated, you are claiming that the models can predict the future, but, like the “lottery win” of Derren Brown, the results are only to be revealed after the event.
Radical Rodent
Dang it, MJ! Why do you have to give better replies than I can manage? Do you want to join TheBigYinJames on my hate list?
Radical Rodent
Honoured to be compared to the BigYin, RR.
As for 'hate lists" I'm on enough already. One more won't make that much difference.
Mike Jackson
The link Raff provides @9:08 is worth reading, especially the comments of Jeff Id, though that may not be the star name Raff wants read.
golf charlie
Mike, there may be some ex-journalists who claim to have deep understanding of climate science and modeling. If one of them recommended as a "corrective" an article that suggested that the temperature axis for a "normal view" of global warming should be -10..110F, as you did, I'd have all the "background" on him to know that he actually understood close to nothing. All of what you say you "know" about climate science is cancelled by that single reference. Anything you say you "know" is just recycled dross from WUWT and the like that has fooled you.
Raff
Raff, the point I tried to make but evidently did not get across effectively, is that, even after the event, the 'forcings' remain unknown. (Because they are not recorded or even measured.)
Interesting that figure 3 includes in its caption
Dashed lines denote updated calculations based on the actual evolution of radiative forcing.
That does not ring true to me. I think that CO2 has risen faster than foreseen - certainly not slower than foreseen as indicated by the graphs.. I assume that the "radiative forcing" was estimated from CO2 by use of the Holy Climate Formula. If CO2 has risen faster than foreseen, I don't understand how the resulting radiative forcing (so called) has risen slower than foreseen.
Martin A
Raff:
Mike, Simon, your degree of certainty is something to behold.
And there I was thinking the degree of religious conviction you have in climate models predicting a future that cannot be known was a wonder to behold. It is not merely uncertain whether a GCM can accurately describe the future of the climate, it is absolutely known that it cannot. And yet here you are, proclaiming your faith.
Witness away, I'm agnostic.
Simon Hopkinson
Raff, you are trying to flog a product with a track record of not working. Just because it is a fraction less unreliable than before, does mean it is now reliable.
As there is another thread started 14/1/2015 on this site giving a new improved version of the climate record, against which version of the past climate record should the latest models be compared with anyway?
If climate scientists deny that they have made any predictions, only projections, and none of them have been remotely close to reality, why to you expect trust now?
golf charlie
Martin, the text refers to updated forcing for "solar activity, volcanic ash in the atmosphere and the concentration of greenhouse gases". In 7 years (the period mentioned since the CMIP5 forcings) wouldn't the change in radiative forcing from CO2 be quite small (let alone the change in forcing from the unexpected acceleration of the change in CO2) relative to the depressed solar cycle? So the solar activity from 2005 to 2012 is certainly better known now than it was in 2005. Ditto the volcanic ash (it can't be less known) and CO2 forcing. So the update seems entirely reasonable.
Simon, how do you know the models are as bad as you claim? I don't know what exact claims are made for models by the climate modeling community, but I assume they think there is some value in them. What experience or knowledge do you bring to the table that those involved in modeling do not have and that lets you be so positive about the quality of the models? Try to answer without accusing anyone of anything.
Raff
Raff, why do you still have faith in the models? Have they served any purpose?
Your questions, and lack of any answers are looking like another Lewandowsky styled 'survey'
golf charlie
Media Hoar
Raff
Simon, how do you know the models are as bad as you claim?
I've already answered this in pretty much every comment I've made in this thread.
How do you know that two lottery tickets, both with different number sequences, cannot BOTH be jackpot winners in the same lottery on the same day? How can you say, today, that you have a winning lottery ticket for tomorrow's draw? Tomorrow evening you'll easily be able to review your numbers and say to yourself "Oh, I see it now, if I'd just used my nephew's birthdate rather than my niece's for that number.." etc. How does that help? Tweaking numbers to get a good result in hindsight does not speak to predictive prowess on future eventualities. In predicting the result of a chaotic system you are bound by chance. "Chance", of course, is not truly chance but a synonym for an inconceivably vast series of significant, unpredictable and interconnected events leading to an instant which, by the nature of its make-up, cannot be foreseen.
I know what you're probably thinking, Raff. "Simon's talking blah, blah, blah, models are wrong, blah, blah, blah."
I can't help with that.
Simon Hopkinson
raff
I fear you and I have now reached a parting of the ways.
That last posting is the most offensive piece of writing aimed at me that I have seen in this site, or indeed anywhere else in a long time.
Your description of me as an "ex-journalist", apart from showing your own witless bigotry, demonstrates as clearly as anything else that you really do not have a clue about what I am or was in my working life. Or my non-working life, come to that.
I had started to come to the conclusion that it was possible to engage with you at a sort of 'Year 4' level (about the intellectual and behavioural standard of an 8-year-old) but that little bit of nastiness is more like a 3-year-old throwing a tantrum.
I see no future in trying to engage with anyone whose only reaction to any sort of approach which they don't understand is to sneer and whose response to an alternative view to their own is to insult.
As I said earlier, "by their words and deeds ye shall know them." Which, I'm afraid, makes you an ill-mannered lout.
Have a nice day.
Mike Jackson
Mike,
That last posting is the most offensive piece of writing aimed at me that I have seen in this site, or indeed anywhere else in a long time.
I do find this kind of response remarkably funny. You're a regular commenter here and you object strongly to the supposedly offensive nature of Raff's response to you. Do you read the other comments on this site? Do you think about them when you do so? Is it only offensive when it's aimed at you, but not at others? Are you somehow special and deserve to be treated with kid's gloves, but others should just take it on the chin? Do you know what the word offensive actually means? I could go on, but you probably get the point....okay, no, I'm probably going to be wrong about that. I'm also ignoring the obvious point, which is that Raff's comment barely register on the offensive scale.
...and Then There's Physics
Jan 14, 2016 at 2:33 AM | Unregistered CommenterRaff
...why that would be a good idea?
To get a better idea of how well the model worked? If you compare real outcome with projected outcome where the projection used incorrect forcings, it is not a like-with-like comparison.
I can see if you are doing research into how to improve* modelling technique I can see there can be some point in doing that.
*[I say 'improve' although that implies that I accept that the models work at all; I think is far from obvious that they do.]
If it were for the purpose of claiming some sort of success for the model, then such a claim would be along the lines of the 19th century surgeon's boast that "the operation was a success (although the patient died)" ie a hollow claim.
The point of the models is not really to project short term temperatures but instead to study the medium to long term where variations between assumed and actual forcings average out.
Hold on. Hold on.
Is there even the slightest reason to believe that 'assumed' and 'actual' will 'average out'? "Averaging out" requires very benign and favourable statistical properties for both 'assumed' and actual - plus very long periods over which to perform the 'averaging' (ie millenia). But the future evolution of so-called 'forcings' is essentially unknown, even more so their long term statistical properties. without which 'averaging out' is just a forlorn hope.
Continually increasing divergence between them amounting to a random walk would probably be the best that could be hoped for.