Thursday was
Uncertainty day. Typically, this session struggles to attract enough abstracts to fill a session, is allocated a small room, and ends up being packed to the gills. It seems that everyone wants to hear about it, but no-one wants to do it - and it's not helped by the perennial absence of one or two significant groups who one might think could both benefit and contribute by attending. This year a slight change in scope (including specific foci on metrics for model evaluation, and impact studies) opened things up to a wider clientele which filled 3 sessions. It had a couple of interesting moments but I was a little underwhelmed - it seemed that a lot of people are doing the same old stuff just because they can without giving it too much thought. Sokolov presented the MIT stuff, and I couldn't resist pointing out (as I'd written
here) that both the high and low values for the ocean mixing parameters he used were incompatible with what is known about ocean circulation - a point he explicitly acknowledges in the MIT report - and he didn't seem to have any adequate response. There were however a couple of interesting ideas in other talks relating to methods for multimodel analysis that I will try to follow up.
I can't remember what I went to in the 4th oral session - nothing worth writing notes about, it seems. The posters for the uncertainty session followed in the evening. My "uniform prior" poster attracted by far the most attention out of the things I was involved in. It seems that there are still plenty of people who are not aware of the issue even though I've presented it several times, which presses home the value of actually getting it published somewhere. At least one person who works in the field promised that they would not unthinkingly use uniform priors again... We had to run off to the Musikverein a bit early to see the excellent
Hilary Hahn.
Friday was a very light day. Lots of people leave during the day, if not before., and there is always much less on. I looked at the
millennium session which is a new area of interest for us now that the IPCC are asking for model simulations of this period. We have plans for some work, but nothing concrete, so it was useful to get up to date on what some others have done. I managed to scrounge my 4h free lunch of the week by attending a meeting that it transpired I wasn't actually invited to. Oops. But no-one really cared. After lunch there was literally nothing to go to. This was rather convenient, as I still had to write my talk for a workshop on the Saturday. This was a select assembly (sponsored by
PAGES) of people interested in paleo data assimilation, that is data assimilation in the context of paleoclimate simulations and proxy data. But first, there was the famous Convenors' Party on Friday night (not that I was a convenor, but I went as jules' guest). The first rule of the party being, don't talk about the party, I can say no more.
The
PAGES workshop was a really good day. It was tough after a long week but well worth it. Since I spend much of my scientific life in a niche within a niche, it is great to find that there are others (albeit few!) with similar interests and understanding, and I left feeling that maybe I'm not quite so mad after all. The new
MARGO data set seems to have sparked several projects relating to climate (at least ocean) state estimation at the last glacial maximum. We (jules was also there) came out with some concrete ideas for stuff to do, which is pretty much our benchmark for a successful meeting.
As a final wind-down, we dragged ourselves to
this Chinese restaurant, which was a very pleasant end to the week. The food is at least decent, and plentiful, and being able to sit out in their quiet garden on a warm evening beside a pond full of koi was a very peaceful and relaxing change from the bustle of Vienna city centre.
All in all, it was a good week, and we'll probably be back next year. Wouldn't mind a change of location - it was strongly rumoured that the meeting was going to shift to Paris, in fact this "decision" can be found on various cached pages including the EGU itself! However, it seems that the vote went for Vienna again, and it's certainly not a bad choice.
