In this talk I introduced the CANARI large ensemble to a general physics audience.

Presentation: pdf (16.4 MB)

The talk started with an introduction to numerical models and concepts of resolution, before explaining how large models use large compute. I introduced some ideas of uncertainty and storylines, before introducing ensembles in general and large ensembles specifically. I discussed the CANARI project itself, and then the CANARI model and ensemble configuraion (including boundary condition production for CORDEX and other users). I discussed the reality of keeping a large ensemble in the machine, and the volumes of data (6 PB) produced.

I then introduced some of the early science results from colleagues, including some analysis of cyclone tracks and European storms.