The monthly-seasonal-decadal prediction session was quite interesting. The pick of the bunch was Weigel's medal lecture, which explored various issues relating to ensemble systems, such as why adding even bad models to a multimodel systems will generally improve overall performance (mainly by increasing ensemble spread and reducing overconfidence, it seems). He also talked about weighting different models - a hot topic in climate change prediction - and proposed a new intuitive way of evaluating prediction skill (which met with some objections, to be fair). There were other possibly contradictory presentations concerning the possibility of skillful predictions much beyond one year, which still seems a challenge. Of course these people are aiming at prediction of the seasonal anomaly, which is not the same thing as predicting changes in the mean climate.
In the ocean temp and salinity session Syd Levitus presented a new analysis of ocean heat content, based on various corrections to the observations. He claimed it all made little difference to the overall trend, but I reckon there are more subtle effects (relating to the decadal variability) that will actually be quite important to users. In the feedbacks session, Bates's medal lecture was a bit of a disappointment to me. He spent a long time going on about how there were numerous incompatible definitions of "feedback" in the literature. Although I'm sure the basic issue is well known it seems to be a bit more subtle than I had realised, and I'm relieved to see it is not just a climate science issue. He then went on to talk about Lindzen's iris model, praising it as elegant although a bit limited. He described how he had added some more realistic elements to represent dynamical effects...and had to rush through the last few slides so fast that I don't actually know what results he ended up with! Grrr. I got the impression the sensitivity was higher, though.
Lunchtime was the business meeting for the climate division. A fair bit of routine stuff (CL is still going well, growing fast and now the 3rd largest division in the EGU in terms of abstracts submitted) but there was also the Exxon kerfuffle that I mentioned previously. I really don't know how this is going to work out. I find it hard not to be deeply suspicious of the whole thing - why Exxon alone, and what is the EGU actually gaining from it? I also couldn't resist a quick poke of the embers regarding the journal Climate of the Past. This is officially the journal of the division but the division's full title is "Climate of the past, present and future" and a careful reader might observe an incompatibility between the two that excludes a fair proportion of the division members' interests, including most of mine! What makes it more amusing, this year they are planning a special issue with papers from the medal lecturers, including today's talk which is firmly in the numerical weather prediction camp. But this issue came up a couple of years ago, and I don't have a good solution. CP is rapidly becoming one of the best paleoclimate journals, so I can certainly sympathise with the current editors not wanting to tinker with it, and I'm not sure if there is really a niche for a "Climate of the present and future", let alone sufficient energy to run it. But something like that would be a nice forum for more open discussion of issues in climate change and prediction. We'll see how things go...
Dinner was a slice of pizza in our room, cos jules had to prepare her talk for the following day. The life of a jet-setting scientist is all glamour!