Having studied meteorology and doing therein statistics lead me always back to sessions, which are not directly connected to my current research focus. Like today, where I started with statistical methods applied on precipitation data. It was in fact a very diverse session, since precipitation data, due to their properties, are quite complicate to handle. Afterwards, I spend my time in a session about ensemble methods, which mainly focused on supermodels. Therein within a multi model ensemble each model is not calculated independently but the models are allowed to communicate with each other. This seems to be an intresting approach, even if there is to proof that the result is really benefitial over indepently run models on a large scale. Nevertheless, the first results looked quite promissing.
The whole day was dominated by the large sessions on the current IPCC AR and so I took a look into one of their sessions in the early afternoons. Afterwards, it was time for the great medal lecture by Olivier Talagrand. He got quite quickly through his work of his own and his collaborateurs, which was nice to see. All the different data assimilation techniques in one hour was interesting and remembered me of a lot of lectures during my studies. A worthy medal recipent. At the end stand the traditional poster session.
In general I like the medal lectures, since they really allow to cover a whole theme in a little bit more detail and give a good overview over a topic (which depends of course on the abbility of the presenter). The traditional 12 minute slots can be sometimes quite tiring and so it is good to have some alternatives on the menu.