| A current debate in the climate modelling community relates to the question of how to balance computational resources between higher resolution models, more complex models, and larger ensembles. While efforts to increase resolution and complexity still takes the lion share of resources there are nevertheless growing numbers of ensembles of climate change simulations, including the CMIP3 ensemble for the IPCC, the climateprediction.net project and others. As such ensembles proliferate it becomes increasingly important to question a) how to best design such ensembles, and b) how they should be interpreted by and for decision and policy makers. These questions will be addressed in this talk in the context of a categorisation of uncertainties in this extrapolatory problem. |
|