Higher resolution: reinventing the sphere and other elements of model E
	Maxwell Kelley

As the resolution of a climate model increases, and its increasingly 
sophisticated representations of physical and chemical processes 
respond accordingly, some parts of the model may need redesign. I will 
outline the rationale for replacing the latitude-longitude computational 
grid in model E by a "cubed sphere", and present results from some initial 
tests. I will then discuss some challenges that may arise in using (a high-
resolution version of) model E to diagnose the importance of phenomena 
at relatively short timescales and/or small spatial scales, or to calculate 
non-traditional atmospheric circulation statistics.