Early ice result
Here is an early result from the ice model- ice thickness fields (hice) every 15 days.
there are several problems. First, there is whole-lake coverage, which happens very rarely and did not happen in '08. Second, ice thicknesses exceed 1m, and while there's not a lot of data on this I don't think that's the least bit realistic. Finally, there is ice near Duluth until June. Duluth is a cold place but it's not THAT cold.
However, we get ice formation where we would expect it- along the southern (shallow) coast. There is very little ice in the eastern basin.
I think that much of the remaining problem is tied up in better understanding how shortwave radiation is put into the model during periods of ice coverage.