|
|
Post by jimcripwell on Jun 22, 2011 18:39:47 GMT
steve writes "They address the questions all the time."
I have never seen one single paper to do with CAGW where a model has been used and that model has been verified against observed data. Please pro ide a reference which proves I am wrong. And please, if you do, dont just give the reference, but pinpoint the page, paragraph, sentence, etc of where the validation was done. And a few words from the reference about the validation process.
|
|
|
|
Post by socold on Jun 22, 2011 21:48:37 GMT
Id also be interested to hear what you mean by "programming shortcomings"
|
|
|
|
Post by magellan on Jun 22, 2011 22:13:06 GMT
Id also be interested to hear what you mean by "programming shortcomings" Jim Cripwell summed it up nicely, and also that getting the right answer for the wrong reason is also a "programming error". That will do for now. |
|
|
|
Post by icefisher on Jun 23, 2011 0:53:43 GMT
icefisher
Do you really believe that climate models do not represent convection and clouds? Or are you making a point about parametrization? If the latter, you do know that *some* models do represent convection and clouds - while they are too expensive to run globally out to 100 years you don't actually need to run globally to 100 years to experiment with cloud feedbacks and variability.
Silly question Steve!
As G&T said in their response. . . . put something up that supports your belief that climate models properly consider convection so it can be examined, replicated, and either accepted or rejected.
You are playing a shell game occasionally allowing a glimpse of the bean while continuing to shift the shells around.
Give us some concrete examples of the detailed and documented handling of convection hand in hand with radiation that is capable of testing.
Otherwise what you have is a one-legged dog. Its like advertising for sale a 2011 Cadillac and delivering a steering wheel. Then say gee guy we have a set of wheels in the garage out back what are you complaining about?
|
|
|
|
Post by steve on Jun 23, 2011 7:28:00 GMT
Icefisher,
Apparently it is not a silly question because you originally said they didn't do convection and now you say they don't do convection "properly". So how much do I have to push you before you'll finally admit that G&T is provably pseudoscientific nonsense combined with unnecessarily obfuscated mathematical equations.
Jim
Nicely qualified statement Jim. Or is it ignorance about the scientific process? Model validation is done before the model is used to make forecasts and projections. So no you won't see a combination of a CAGW paper and a model validation paper - after all how could you given that there are few or no examples of observed CAGW (whatever observed CAGW is!) to validate the model against!
|
|
|
|
Post by scpg02 on Jun 23, 2011 7:46:40 GMT
Icefisher, Apparently it is not a silly question because you originally said they didn't do convection and now you say they don't do convection "properly". So how much do I have to push you before you'll finally admit that G&T is provably pseudoscientific nonsense combined with unnecessarily obfuscated mathematical equations. Jim Nicely qualified statement Jim. Or is it ignorance about the scientific process? Model validation is done before the model is used to make forecasts and projections. So no you won't see a combination of a CAGW paper and a model validation paper - after all how could you given that there are few or no examples of observed CAGW (whatever observed CAGW is!) to validate the model against! If you don't do something right, you are not doing it. You can't say the models come close. They are either modeling or they are not. If they are not modeling convection properly then they are not modeling climate. |
|
|
|
Post by steve on Jun 23, 2011 8:54:56 GMT
scpg02
This is a statement borne of a desire to ignore evidence because you don't like it as clearly you are not unaware of all the areas of your comfortable life that are dependent on imperfect modelling (medicines, computers, cars, buildings etc.). Anyway, where did I say "the models come close"? I didn't.
|
|
|
|
Post by icefisher on Jun 23, 2011 9:35:43 GMT
Icefisher,
Apparently it is not a silly question because you originally said they didn't do convection and now you say they don't do convection "properly". So how much do I have to push you before you'll finally admit that G&T is provably pseudoscientific nonsense combined with unnecessarily obfuscated mathematical equations.
Where on earth did you get the idea that I did any such thing?
The way it works in the real world Steve is the science community are trained specialists at smelling out money sources. Each one of them tries to figure out a unique approach to get some of it.
But that comment was purely in response to your bringing up the topic of some climate models having convection parameters. I can hardly criticize the parameters of anything hidden in a locked blackbox. If one exists and it has been relied upon, then prove it!
When I spoke of not using convection at all I was explicitly speaking of the calculation you provided us that compounded warming rates back to the surface through multiple layers to arrive at the 1.2 deg C for surface warming from a doubling of CO2 and the 3.7watts TOA estimate of theoretical attenuation of radiation.
And "they" was specified as you, GLC, Socold, James Annan and any number of other scientists that have declared a doubling of CO2 "will" result in at least a one degree temperature increase at equilibrium. This was the sole response, claimed to be certain, of the climate science community to McIntyre's call for an engineering analysis.
Are you claiming your calculation had a convection element in it? If so show me where. If not, and you are familiar with any calculation anywhere that does then share that with us!
Otherwise the only thing we have are the calculations that don't include it. What is this thing of which you speak Steve. . . . an invisible God that whispers in your ear in the middle of the night?
|
|
|
|
Post by icefisher on Jun 23, 2011 9:46:25 GMT
This is a statement borne of a desire to ignore evidence because you don't like it as clearly you are not unaware of all the areas of your comfortable life that are dependent on imperfect modelling (medicines, computers, cars, buildings etc.). Anyway, where did I say "the models come close"? I didn't.
Holy smokes batman!
If any company sold those products as only modeled and completely untested to the public they would be out of business.
Steve! Ever hear of FDA testing?, beta testing?, road tests?, and about 10,000 different real world tests applied to building products?
|
|
|
|
Post by steve on Jun 23, 2011 10:01:41 GMT
icefisher
This is what you originally said.
I don't think it is a "silly" question to check if this is what you actually meant. You said the above in response to a discussion about climate models, not a discussion about the simple calculations used to determine the so-called "no-feedback" warming.
With regard to the effects of convection, in winter Devon is cold and in summer Devon is usually warmer. Now if the Sun can overcome the power of convection then why can't the excess CO2?
|
|
|
|
Post by steve on Jun 23, 2011 10:10:15 GMT
Nope. Never heard of FDA testing, but I think I know what you mean. Up to now at least drug testing is primarily controlled by the company developing the product who has the option of selectively releasing good studies and withholding bad ones. Car crash tests were forced on the auto industry by consumer groups (in Europe) and even 10 years ago were pretty limited. The world about us is *not* perfectly tested. Stop arguing the unarguable.
|
|
|
|
Post by jimcripwell on Jun 23, 2011 10:54:15 GMT
steve writes "Nicely qualified statement Jim. Or is it ignorance about the scientific process? Model validation is done before the model is used to make forecasts and projections. So no you won't see a combination of a CAGW paper and a model validation paper - after all how could you given that there are few or no examples of observed CAGW (whatever observed CAGW is!) to validate the model against!"
steve shows his utter ignorance of modelling and the use of models. He seems to be under the impression that one can write a model, independently of the question that the model is required to answer. This is just plain wrong. When any model is used to answer a specific question, it is a requirement to show that the model is, indeed, suitable to solve that problem. Anyone who does not understand this is utterly and completely ignorant of what modelling is all about.
But it is his final quote that I like the best. "how could you given that there are few or no examples of observed CAGW (whatever observed CAGW is!) to validate the model against!"
Precisely!!!!! People use these models to predict the future, when there is no observed data to show that the models have been validated so that they are, indeed, capable of predicting the future. If you cannot show that the model is capable of predicting the future, then you have no basis to claim that the model can predict the future. The defense rests.
|
|
|
|
Post by steve on Jun 23, 2011 11:14:09 GMT
Jim
The problem is in your insistence that a) a model is only measured by its ability or otherwise to predict "catastrophes" and b) the prediction of "catastrophes" is only based on models. Both of these positions are faulty.
Even if your model is *only* capable of predicting a likely average temperature rise of 2-4.5C in a century then there is plenty of observational evidence that supports the proposition that catastrophes will occur as a result. Models have already reached that position. They did so decades ago and their projections have been positively validated so far.
|
|
|
|
Post by jimcripwell on Jun 23, 2011 13:08:33 GMT
steve writes "The problem is in your insistence that a) a model is only measured by its ability or otherwise to predict "catastrophes" and b) the prediction of "catastrophes" is only based on models. Both of these positions are faulty.
Even if your model is *only* capable of predicting a likely average temperature rise of 2-4.5C in a century then there is plenty of observational evidence that supports the proposition that catastrophes will occur as a result. Models have already reached that position. They did so decades ago and their projections have been positively validated so far. "
With respect to the first paragraph, I make no such claim. What I say is that the question and the model cannot be seperated. You cannot use any old model to solve any old question. There is a requirement to show that the model being used is suitable to solve the question being considered. You have not addressed this. What I am saying is that unless a model has been validated to prove that it can predict the future, for the particular question that is under consideration, then you may not use that model to solve that question.
In the second paragraph, you state "then there is plenty of observational evidence that supports the proposition that catastrophes will occur as a result". This is the reference I am looking for. Where is all these "observational evidence" that supports the "proposition that catastrophies will occur"? I cannot find it. Please prove me wrong, as I noted above, with the reference, the page, the paragraph, the sentence, and a few words that show this observational evidence exists, and which proves that models can accurately predict the future.
|
|
|
|
Post by steve on Jun 23, 2011 14:19:25 GMT
Jim,
You are misunderstanding me.
As far as I can see, the principle model result that I think has been well demonstrated is that if greenhouse gas and pollution levels rise at current predicted rates then the planet will warm by 2-4.5C. So that is my question and answer dealt with.
If you want to ask me about catastrophes, then I'll put the model to one side and look for evidence of what has happened in the past relating to rapid changes in the climate, and point you to the PETM, the rapid sea level rises during the Eemian, the Younger Dryas and so forth. If things that happened then happen now or in the next 100 years, they would be catastrophic. More recently there is evidence of what happens with heat waves (Europe 2003, Russia 2010) which reasonably would be hotter, longer and more frequent.
Now obviously people *do* use models to answer more specific questions such as Amazon dryout, changes in precipitation patterns, hurricane intensity and so forth. I would agree with you that the models do not provide good evidence for probabilities of any specific "catastrophic" change, but I don't agree that the lack of certainty is sufficient reason to dismiss the idea of any proposed catastrophe happening because we know from past evidence the climate is fickle and sensitive. If someone gets run over by a bus no amount of modelling will tell you for certain which bones will break.
|
|
