Many thanks for scaring this stuff up. Most of these I have not seen and I will look at them. My prior reading on this left me with the impression that the early consensus was around 1.5, and that as the issue got politicised (that’s purely my subjective characterisation, obviously), the forcing got jacked up. But, hey—it could be the science, and the papers you’ve found. I owe you one, and again, many thanks.

Constraints on climate sensitivity come, among other places, from the observed 20th century warming, cooling after volcanic eruptions, the climate of the Last Glacial Maximum 20,000 years ago, and the Maunder Minimum when solar activity was low around 300 years ago. In general, temperatures are more easily measured than radiative forcing, so the uncertainties tend to come from characterizing the forcings (how much CO2, how much sulfate aerosols, how to translate their concentrations to a forcing in W/m^2). Another difficulty is that the observed temperature isn’t the equilibrium temperature response, as the oceans take up heat and one has to adjust for what the eventual atmospheric temperature would be once that heat is released.

Some of the literature:
Forest et al., Science 295, 113 (2002): use recent climate measurements to estimate 5% to 95% confidence band of 1.4 to 7.7 K for climate sensitivity. Use ocean temperature observations to help reduce the uncertainty from ocean heat uptake. (The 7.7 K end of that estimate is disturbingly large, but see below.)

Gregory et al., Journal of Climate 15, 3117 (2002): put 5% confidence lower bound of 1.6 K, using surface temperature data and ocean temperature observations. Their probability distribution has a long upper tail.

In fact there was a paper by Roe and Baker, Science 318, 629 (2007) that recently argued that these long tails in the direction of large temperature increase are an inherent property of the climate system, and are very difficult to constrain further.

On the other hand, Annan and Hargreaves, Geophysical Research Letters 33 L06704 (2006) put together various other estimates from observed warming, volcanic cooling, and the LGM to conclude that there is less than 5% probability that climate sensitivity is above 4.5 K. Urban and Keller, Geophysical Research Letters (in press) also argue that the fat upper tails can be reduced by measuring ocean heat uptake and aerosol forcings more precisely.

In short, it seems that observational constraints tend to agree with the 1.5 K to 4.5 K estimates of climate forcing from GCMs. As far as I can tell from this quick look at things, the probability distributions emerging from the data say very clearly that the climate sensitivity is highly unlikely to be smaller than 1.5 K, and as I mentioned above, “climate sensitivity” by definition omits slow feedbacks that potentially make the problem worse. Also, it seems that progressively better measurements and statistical techniques are chipping away at the high end of the fat tail. A most likely value somewhere in the neighborhood of 3K seems like a good bet.

Please know that we who witnessed this, we understand that your evasions, provocations and unpredictable illogic were relentless. You wield a non sequiter the way that Frazier employed his powerful left arm in the fifteenth round against Ali in their first fight.

Be assured that you have earned an honored place in the Troll Hall of Fame. A magnificent performance. Well done.

http://www.lies.com/wp/2009/03/27/blog-post-subhead-large/

Noted without comment.

294: Salient, facts only muddy the waters. Cut it out. Besides, you didn’t do a controlled study.

I think my view of the world of news is more accurate, and it leads me not to rage, cynicism or fatalism.

The antidepressant approach to reality. If it works for you…..

No, your point has been (for example) that people should not encourage the WaPo to fire George Will (245); that writers of opinion pieces have a license to lie and should have a license to lie and that editors ought to bear no responsibility for correcting these lies (158.3); that we should see license to lie in opinion pieces as a good thing (158.4); that newspapers like the WaPo have no substantial effect on discourse (158.5). These are the points you have made, that I have taken issue with.

But I’ll respond to your latest point, #297. My reply: So what? Points of the form “if X disappeared, we’d all get on” are usually not worth making. Random example: If CT went permanently offline for some reason, we’d go find something else to do. If the U.S. monetary system spontaneously collapsed this moment, we’d find some way to reconstruct some kind of economy that keeps us alive and variously employed. Why should I let that kind of fatalism guide my reactions to events in the world?