Melt pond questions

[kshedstrom]

In CICE, the melt pond area, volume, pond ice cover are carried in the model as scalars in each cell, not dimensioned by the number of categories. In each meltpond routine, they are then split up according to some local function to get ipondn (by ice category) and friends. What we have in SIS2 is melt pond area and mass by ice category. Do we continue with the plan to call Icepack?

[Hallberg-NOAA]

I think that the plan is to call icepack when and where it makes sense to do so. However, if there are advantages to using a different scheme (as it seems might be the case here), I think that the plan is to incorporate any icepack capabilities into more general SIS2 code. Is this the case where the partially implemented approach in SIS2 is better, or do we want to follow the lead from icepack?