Everyone talks about the climate leak, but does anyone look at the e-mails? I’m talking about the e-mails stolen from the Climatic Research Unit at the University of East Anglia.
Well, I’m doing it. But rather than trawling for keywords, I figured that I’d start by looking at a random sample of the e-mails. I want to look for evidence of professionalism (or the reverse) on the average. I’m more interested to know if some one (or some group) is normally competent and honest, versus looking for the occasional stupidity.
So, I started with 199286511.txt, chosen by clicking part-way down the listing. It was boring.
Then I tried 1200425298.txt. A perfectly normal discussion about what to write in a paper. Some discussion about how air pollution might affect temperature trends in China, and a little flash of humor (“11 out of the top worst ten polluted cities in the world are in China, or something like that”).
Next came 1139850906.txt. This is an e-mail discussion about a book chapter. They sent a draft of a chapter out to an American geophysicist and he pointed out some problems with it. Pretty ordinary stuff. Writing, writing, writing, looking for mistakes. Presumably fixing them. But, boy, these guys live by e-mail! It’s 295 lines long.
1255558867.txt is interesting. They argue about what to show in a figure for a RealClimate post. One says the proposed figure is “very deceptive”, but someone else replies “…but if showing the full spread from CMIP3 is deceptive, its hard to imagine what sort of comparison [between data and models — ed] wouldn’t be deceptive…”. This is people trying hard to simplify a complex bunch of computations: trying to put them on a web site and have it readable and not misleading.
Thinking about this overnight, it actually sounds like a good dynamic. You don’t criticize someone’s work to their face (as it were) and get away with it, not unless they are more interested in improving the work than in saving face. To really prove this, though, you’d have to look at later e-mails to see if these people were still working together. Especially over e-mail among different sites. It’s very easy to drop someone from the address list if you don’t want to be annoyed by them.
Making things readable by the general public is hard. Not because the public is dumb, but because they haven’t experienced some important things. You don’t have a gut feeling for how a chunk of mathematical model behaves until you’ve run it 100 times. The general public doesn’t really know how these models respond to changes in (for instance) the amount of dust in the simulated atmosphere. Of course, neither do I, but I’ve run enough mathematical models to know what I’m missing. Learning how a good big complex simulation behaves takes a while, and it can have lots of surprises. It’s like having kids: you cannot imagine the impact they will have on your life until you’ve done 1000 diapers.
Anyway, they’re trying to set up a reasonably fair comparison. They are having some trouble, I gather, because they have two models available, and both (CMIP3 and MAGICC) are almost-but-not-quite right for the plot they need to make. Perhaps in an ideal world, they wouldn’t have made the plot, but I read this: “The kind of things we are hearing ‘no model showed a cooling’,… need to be addressed directly.” They feel their work is being misrepresented by outsiders, and they want to respond.
So far, I haven’t seen any red flags, and I don’t yet know what George Monobiot is all excited about.