-
document progress in model performance from CFSv1 to CFSv2, including the
impacts of reanalysis, initializing atmosphere/ocean/land, and the change in
the model itself;
-
identify key model biases and deficiencies in the CFSv2;
-
help identify most promising research directions for the development of
CFSv3, for example, to design experiments to understand how to incrementally
improve upon CFSv2, and to test coupling of the latest GFS with
state-of-the-art model components.