How well can volumetric particle clouds create believable cloud simulations?
What techniques are required to model the atmosphere believably? e.g. http://staffwww.dcs.shef.ac.uk/people/M.Gamito/atmosphere.html
To what extent do clouds that are extracted from satellite data need to portray thermodynamic behaviour for the purposes of simulation?
How can a ground based first person view of the clouds be generated from satellite data, showing correct light on clouds?