That WSJ column
Feb 5, 2013
Bishop Hill in Climate: sensitivity

When Matt Ridley wrote his column on climate sensitivity just before Christmas, he was given a very hard time by several web commentators. In rather hysterical vein, we had Joe Romm, of whom perhaps the least said the better. Needless to say, he was (ahem) rather critical of the Ridley column.

There was also Keith Kloor, who seemed to think that the article represented a bad case of confirmation bias:

Ridley [argues] that this “important debate” has now been given a revelatory jolt by the unpublished work of someone he describes as “a semiretired successful financier from Bath, England, with a strong mathematics and physics background.”

It is my assumption, after reading Ridley’s column, that he believes the financier (Nic Lewis) is on to something big–namely that climate sensitivity is low and that we can all stop worrying about this thing called global warming...

It’s telling that Ridley didn’t include any skeptical perspective of this “extraordinary claim.” What does it tell us? That on the issue of climate change he has become a victim of confirmation bias.

Kloor's case seems to have a bad whiff of argumentum ad authoritatem about it (What?! A financier?!), but what is rather more surprising is that Lewis's own article had appeared at BH two days before Kloor's was published. Kloor knew that Ridley's article was based on Lewis's work and I know Kloor reads BH, so one might also assume that he had read the article. I'm therefore slightly surprised that he didn't reference Lewis's article with its consideration of the rest of the climate sensitivity corpus, and to the best of my knowledge he has not even noted its existence.

Either way, Kloor knows that Ridley's article was based on Lewis's work and he either knows or will know that Lewis's work considers the rest of the corpus. That being the case can he still make the claim of confirmation bias stand up?

Particularly when Ridley seems to have been proven spectacularly justified in the last six weeks: we have now had one prominent expert in the area say that climate sensitivity is likely low and another say that low is more likely than high. Our only completely empirical estimate says low and, as Lewis notes, we now have a whole series of empirically constrained studies to back that finding up.

Keith Kloor's position is that he "[pays] due to all the pertinent science related to CS, not just that which confirms [his] bias". Personally speaking I think one can accept empirically based studies over GCM-based studies, particularly those embodying estimates of aerosol forcing higher than the IPCC's own best estimates, without being guilty of confirmation bias. This is just the scientific method.

Likewise, since nobody seems willing to make the case for uniform priors, I think reasonable people can agree with Steve Jewson, that the IPCC shouldn't be citing papers that use them.

So what then is the pertinent science related to CS that Kloor thinks that Lewis has failed to consider? Is he a fan of uniform priors, of models over empirical findings, or of models based on out of date aerosol data? Or is there something we have all missed?

Article originally appeared on (http://www.bishop-hill.net/).
See website for complete article licensing information.