Within this paragraph lies one of the core issues I have with climate science: the concept that you can run "experiments" with models. Personally, I think science must verify its theories in the real world, and a true experiment is only something that is performed on nature.
We can perform almost no such "true" experiments. There is one universe, one world, one nature and we are stuck in one moment of it. Climate science, along with cosmology and evolution, simply can't be, by this standard.
We may think that we learn something by running a model or finding an elegant mathematical theory, but we cannot be certain, we always need a physical system as an unbiased arbiter of truth.
The one thing we can do is attempt to predict the path of climate change. Climate scientists have. Sadly, they've done a good job. Of course, to have the level of certainty we'd like about how much signal is there as opposed to noise, we have to wait until it is too late to do anything about it. In effect, we have one short physical system experiment to run and we are running it.
Regarding computer models ... My gut feeling is that trying to model something as complex as the climate is hubris.
Here's the problem ... My gut feeling is that to modify our actual atmosphere significantly without knowing the consequences is hubris. Trying as best we can to figure out the consequences of our actions, even if it is a hard thing to do, is merely moving in the direction of prudence. This prediction of global warming is not new or particularly complex. You can model the climate on the back of an envelope and the answer is not very different than the fancy models give now. The details are different, but the result is always a significant change in climate, usually in one direction. CO2 is one of the gears of our climate system, one we don't understand very well, and we have gripped it and shoved it hard over, in an effort, as a species, to pull over and have an excellent lunch. The hubris of computer modelling is simply necessary.
In fact, the climate modelling business and the Phil Jones' of the world are the exact opposite of what you allege. Each time the modelling has been done, starting with Ahrrhenius (sp?) and his page or two of numbers, the answer has been fairly clear. If climate guys were fixated on the result, rather than being scientists, they would regard the question as answered and the results as proven. Instead, they try to cover more uncertainties, account for more variables, consider more sources, sinks and reservoirs of energy and predict more particular results. They modify their models based on review, criticism and debate. They in fact altered their statistics at one point based on McIntyre's criticism. The answer has not changed much. Of course there is no certainty and of course they need to keep working on it. At what standard of proof should the rest of us consider adjusting our future behavior? Do you regard it as proven to be 0% chance that climate change can be as destructive and prompt as predicted? What % is safe enough that we consider no change of behavior as prudent?
The stakes in this are enormous. Our current standard of living and size of population is pretty much unthinkable without fossil fuels. If there is a 50% chance we need to replace it we should be working on it now; not being inactive on the basis that the models predicting that are an ill-intentioned hoax. There one climate skeptic whose claims are coherent, in my view, Bjorn Lomborg. I don't agree with him on a lot, but I'd have happily just nodded quietly and agreed if we'd been able to follow his advice five or six years ago after the last IPCC report and raise $25 billion a year or so in a carbon tax and research the heck out of how to maintain our standard of living without changing the atmosphere so much.
