Links And Ties

The affair of the stolen HadCRU e-mail should tell us a couple of things. The first is that this is why you should worry about privacy. If you do enough naive traffic analysis, not only will you find a pattern - people communicate in patterns - but you'll be able to find something that you can misrepresent. This is the inevitable outcome of pareidolia, the false positive problem, and the infinite possibilities that open up when you don't have to show your working out.

It's a methodology that is common to cranks who stare at the patterns in their heads, to stupid politicians who don't understand or don't want to understand the maths of false positives, and to spooks, red-baiters, and other political thugs who understand them all too well. Viz:

I don't quite understand the HP Sauce concept of 'links' either. Pretty much everyone in the world is damned as an extremist if you take the idea of 'links' as far as they do.

Give me a representative 62MB of your archived e-mail and I'll give you a reason to really hate me.

Secondly, it's about expectations. The scientists involved are condemned for, among other things, being angry about a crappy denier paper getting published in Climate Research and strategising together about how best to protest it. We are expected, in other words, to be shocked that they aren't like the liberal in the joke who is too even-handed to take his own side in a knife fight.

The purpose of the exercise, of course, is to get people sacked if they aren't like that - to impose this stereotype. This is why George Monbiot is wildly wrong. You can't appease the authoritarians; weakness provokes them still further. (Mark Lynas is right, by the way.)

Thirdly, it's based on epic stupidity. The famous "hiding the decline" actually consists of including the actual observed temperatures in a group of parallel data series, rather than, say, removing the anomalous ones. The series in question, Keith Briffa's Yamal tree-ring proxy for temperature, tracks with the observations from the beginning of observed data to the postwar era, and with other proxies before the observation era. Then, for reasons we don't understand, it diverges. Nobody makes any secret of this: they published it in Nature! But you'd have to be incredibly stupid to pick the diverging proxy series over the observed temperatures. (If you want detail, try here.)

Actually, one explanation is that Yamal is in northern Siberia, one of the fastest-warming parts of the planet, and the trees may not be able to respond quickly enough to more warmth. Hilariously, it's also one of the biggest gas fields on the planet.

Fourthly, it's pathetically trivial. If, in fact, the e-mails showed that literally every paper on climate published since whenever was drivel, it wouldn't matter a damn unless the fundamental laws of nature were vastly different. A mixture of gases containing more carbon dioxide absorbs more infra-red radiation than one with less - it's a trivial lab demonstration. And something that absorbs more heat than it radiates will get hotter. It's Newtonian thermodynamics.

And fifthly, Eric Raymond is not the son of God - he's a very naughty boy, who appears to have missed that the code he keeps ranting about is commented-out of the program and never used. (And the less said about this HOWNOTTO and the associated queeny snit, the better.)

Anyway, far from wanting anyone to resign, I'm going to write to the UEA vice-chancellor for research, Trevor Davies, and compliment him on standing up to the red-baiters so far. And I recommend you do too, before he goes floppy.

If anything at UEA wants investigating, it's their IT security practices.