Climate models are fudged (sorry, tuned) to match the apparent observed climate behaviour, that’s why they claim them to be accurate; and yet that is the very definition of curve-fitting (and of course, they are continually being modified and tweaked to try to maintain that fit.) It’s that simple.

I’m afraid the notion that such models have any predictive skill is not plausible: Their only value is as propaganda tools to try and make it appear that the AGW theory has been validated. It has not.

- “It is true that there are parameterisations in these models, for things such as convection and clouds. These are necessary because current computers are not fast enough to run the models at a high enough resolution to resolve the necessary detail to explicitly model the physical processes.”

…but also because there are major uncertainties in the actual physical effects of these processes - even something as basic and major as the feedback effect of clouds. And even a modest and reasonable tweak of just one such parameter forces a wildly different result to be spewed out from the computer.
If it were acceptable to use such simple abstract constants in place of properly understanding and modelling key elements of such a chaotic and infinately complex system, then we could all run climate models at home on the Commodore 64s in our attics.

So I repeat, these systems do not model the climate. They very crudely implement some aspects of it, and fudge the rest to get the desired result. They represent nothing but the distilled ignorance and prejudices of the people that configure them.

http://www.scribd.com/doc/338170/svensmark-2007cosmoclimatology
http://prola.aps.org/abstract/PRL/v81/i22/p5027_1
http://www.realclimate.org/index.php?p=42
http://www.realclimate.org/index.php/archives/2004/12/index/index.php? p=504

Cosmic rays can indeed alter the physics of cloud formation in the lab, and low clouds are part of the research and theory.

What hasn’t been found is any reason to believe that this provides a valid alternateive explanation for the CO2 Greenhouse, or that it renders Greenhouse physics irrelevant.

If we assume that it is true the scale of coincidence demanded for this to be happening at the same time as the CO2 increases and Greenhouse reminds me of the phrase. “All things are possible, but some are more likely than others”. Mouldwarp hangs his hat on “anything but AGW” because he has an ideological investment in the result. He’s wrong to make that investment because Libertarian fears of an excuse to provide increased government powers are misplaced…. government already has the power to tax and destroy and coerce through force… it needs no excuse to do whatever it wishes. Libertarians make several well-meaning mistakes, but this one has been the most egregious and damning of all of them.

respectfully
BJ

Climate models are not just curve fitting models. In fact, many climate models are very similar to the numerical weather prediction models which are used very successfully to predict the weather for the next week or so (for example, the UK Met Office unified model can be run as a climate model or an NWP model, and much of the code used for climate modeling is shared by the NWP model). It is true that there are parameterisations in these models, for things such as convection and clouds. These are necessary because current computers are not fast enough to run the models at a high enough resolution to resolve the necessary detail to explicitly model the physical processes. However, it is incorrect to say these parameterisations are simply curve fitting models, they are based on a lot of physics and results from observations.

I’m not saying climate model are perfect; they aren’t (and they are a very long way from being perfect). However, the claims you make about climate models being useless are simply wrong.

With regards cloud formation, clouds form as a result of microphysics and large scale dynamics. You can have the necessary cloud condensation nucleii for cloud droplets to form, but if there is simply not enough moisture, a cloud won’t form. For example, over the Southern Alps, clouds often form because warm moist air from the Tasman Sea is lifted up (and cools and becomes saturated in the process) as it goes over the mountains. In other places, different atmospheric microphysics and dynamics are responsible for cloud formation. However, the basic processes behind cloud formation have been understood by meteorologists for a long time (many decades).

To the best of my knowledge, cosmic rays are not a major factor in cloud formation (and definitely not for low level clouds).

re: the link you mentioned is subtitled: “A new way to understand disputes about global warming..

which clearly makes dispute the emphasis.. and to my reading a focus whose objective lacks relevance to any constructive needs-based debate..

one of my sidekicks puts it another way: what’s new about any old fools starting an argument?—could it be that the doctor seeks to assert that learning from a fool in this case will be profitable..?

On the other hand, when the British army simulated the behaviour of a barrel of Guy Fawke’s gunpowder using the army’s weapons developement modelling software they underestimated the energy release by 30%. The modellers had had to estimate the behaviour of an oak barrel and had got it badly wrong.

So maybe mouldy is right but probably he aint.

None. Nada.

- “The current climate models are not useless. It is “easyâ€? to run them retrospectively, to “predictâ€? the climate of the 20th century; this has been done. These retrospective predictions can then be validated against observations. If the models were useless, they would not be able to model the climate of the 20th Century.”

This is just curve fitting - fudging everything until your model matches a desired pattern. It certainly does not mean that the climate is being modelled in any meaningful way. These blatant fudges are termed “flux adjustments,” to try and make them sound scientific, and even those models which claim to no longer use them apparently still make use of “ad hoc tuning” to get the desired result (i.e. the same fudging process, by another name.)

- “You need to revisit your meteorology. Clouds are not “seeded by particles of solar radiationâ€?.”

No, you’re right. I meant to say cosmic radiation, and there is good evidence for this:-

http://www.spacecenter.dk/research/sun-climate/cosmoclimatology/a-brie f-summary-on-cosmoclimatology
http://www.spacecenter.dk/research/sun-climate/other/getting-closer-to -the-cosmic-connection-to-climate

Essentially, the state of the Sun’s activity determines how much cosmic radiation reaches the Earth, and thus strongly influences the amount of low cloud cover: Hence the strong correlation between solar activity and recent climate variation.

The current climate models are not useless. It is “easy” to run them retrospectively, to “predict” the climate of the 20th century; this has been done. These retrospective predictions can then be validated against observations. If the models were useless, they would not be able to model the climate of the 20th Century. In fact, many of the climate models recreate the observed warming during the 20th Century quite well. Notice that I am not saying anything about the cause of the 20th Century warming, just refuting your claim that the climate models are useless.

You need to revisit your meteorology. Clouds are not “seeded by particles of solar radiation”. In fact, clouds are formed when water vapour condenses upon tiny aerosol particles (salt from the sea, dust from the land, pollution etc). This is one of the reasons why research into aerosols is so active at the moment. You are however correct that clouds are one of the largest sources of uncertainty in climate models.