Hans von Storch and a colleague interview with Eduardo Zorita at the University of Hamburg website. Zorita has some interesting things to say about climate models:
What would be your advice for young researchers who want to work on climate simulations?
Climate modelling is a quite broad and complex area. In my opinion, there are two dangers that a student should avoid. One is to get stuck in a daily routine of programming and launching simulations, and slowly forgetting that simulations are performed to answer some previous question. This question should be the main driver of the work, the model is just a tool. Climate models are nowadays so complex and require so much technical attention that it is easy to get off the track. The second danger is to fall in love with your model and lose sight of the real observations out there. Models are in this sense dangerous and climate models even more so.
I was also interested in his ideas for where some money should get spent.
What would you do with an additional million Euros for your research?
A million euros is nowadays not much. But to answer your question I would setup a project to understand the behavior of tropical clouds in the Late Maunder Minimum, at the height of the Little Ice Age 300 years ago, from proxy records and model simulations. This could give us hints about cloud cover changes in climates a bit different from the present and thus help us say something about the future climate change.