Wednesday, April 29, 2009

The EGU review part 2

OK, on with the show.

Oh, actually, I have one more thing to say about Vienna. It was amazingly, disgustingly, smoke-ridden. I was astonished at how bad it was, or else how intolerant I have become. It seemed that every public space was thick with the stuff, and the smokers had absolutely no consideration for how they behaved and dropped their fag-ends wherever they wanted (even in smart public parks like the Schonbrunn Palace, when there was a bin a few yards away). It made me realise that Japan has really come a long way in the time I've been here, as it is basically fine here now, with a handful of exceptions. Rant over.

With the EGU being so big and busy (and not providing much in the way of info on the ground), it's helpful to be at least partially organised before you turn up. The program is on the web, with a half-decent interface to pick interesting sessions/talks (only half decent, becuse there was no author search and neither was there an effective means to delete a whole session once it had been picked). So we had spent our last afternoon at work doing that while waiting for our posters to print.

Monday for me started with some nonlinear time series stuff. There's generally some interesting maths in the NP sessions, but nothing really caught my attention here. The only thing I noted was about synchronising models via chaos which is an idea that's been around for a few years, and I'm sure it could be valuable if it works in realistic applications (ie synchronising an imperfect model to reality, using limited and imprecise observations). Next came a session on Holocene climate, which I didn't get much from either.

Straight after lunch, Stefan Rahmstorf gave a well-attended talk on ocean circulation in the open session. He argued pretty convincingly that no-one had really shown the meridional overturning was stable, the tests involving GCMs have been somewhat limited and there are arguments for at least some of them being biased in a stable direction. Although the case was well made, one might think that if GCMs really could reproduce this behaviour, then someone would have found it by now - even if the testing is not fully comprehensive (it can't be, for computational reasons) a lot of people have put a lot of effort into looking for it!

Last afternoon session was Extreme Climate (like extreme ironing, perhaps). Geert Jan van Oldenborg presented his new theory that the anomalous warming in Europe was actually largely due to aerosol reductions, not the ocean mixed layer theory he was pushing last time I saw him. This now seems mainstream, I think - and implies that future warming in this area will probably be relatively slower for the next couple of decades. I wonder if various people will try to pin the hot summer of 2003 on all these awful pollution reduction policies and technologies that were obviously directly responsible for thousands of deaths? I'm not holding my breath!

The evening session was taken up with posters again. This means fewer speaking slots, but it also means the posters are more fun than they used to be, because basically all the climate people are there at the same time. Oh, the free wine helps too (not that it's much good or even very copious, before I get accused of living it up).

3 comments:

Alastair said...

James,

You wrote "[Stefan Rahmstorf] argued pretty convincingly that no-one had really shown the meridional overturning was stable, the tests involving GCMs have been somewhat limited and there are arguments for at least some of them being biased in a stable direction."

Can you explain further. I have read the abstract but it does not give much away.

Stefan Rahmstorf has argued that abrupt climate change (i.e. D-O events including the Younger Dryas were caused by the THC switching. If he now reckons it is not stable does that mean that he no longer believes that abrupt climate change is caused by the THC changing from one stable state to another?

Cheers, Alastair

James Annan said...

"Stable" here means stable at the current state, not bistable with on/off! Ie, he thinks it is plausible that the bistability exists in reality, and even perhaps in GCMs that have simply not been adequately tested.

Alastair said...

Thanks for that clarification.

I just could not believe that a climate modeller was admitting that the models were wrong :-)

They are obviously missing a vital negative feedback loop. That fits with the Bates lecture which you mention in part 3.

I am particularly interested in this problem since reading When is Positive Feedback Really Negative Feedback?Cheers, Alastair.