This is a great series of posts on how climate models can underestimate global warming because of what they’re missing.

But I thought the ECS experiment didn’t specify how the atmosphere gets to 2xCO2. I think the experiment is posed as “If the atmosphere reaches 2xC02 levels, after all anthropogenic emissions and feedbacks have occurred, what will be the equilibrium temperature?”. Your overall point is still valid: given that we don’t know how the feedbacks will play out and we haven’t yet halted our additions of CO2, stopping at twice pre-industrial levels, about 560 ppm, may already be out of the question.

Rob

But I thought the ECS experiment didn’t specify how the atmosphere gets to 2xCO2. I think the experiment is posed as “If the atmosphere reaches 2xCO2 levels, after all anthropogenic emissions and feedbacks have occurred, what will be the equilibrium temperature?”

This is probably the most accurate way of stating the same idea.

The “held constant” expression would amount to the same thing if one assumes that there is a one-to-one correspondence between the equilibrium CO2 level and the equilibrium temperature such that “holding CO2 constant” (whatever that would actually mean, or as he says “magically frozen”) would force the temperature to rise to the corresponding equilibrium temperature. This isn’t such a problem if one is dealing with a small initial pulse of carbon dioxide as this will involve very little carbon cycle feedback, but the larger the pulse, the more the feedback, and thus the greater the difference between the initial pulse and the equilibrium level of CO2.

Anyway, for those who are interested, a good paper to look up might be:

Climate –carbon cycle feedback analysis, results from the C4MIP model intercomparison
Friedlingstein, et al
15 Jul 2006
Journal of Climate, Vol 19, pp. 3337-53

It mentions the figures of between 0.1 and 1.5 degrees Celsius additional rise in temperature for BAU due to carbon cycle feedback.