There have been questions over the years about the thermodynamic parts in the ROMS ice model. I finally took a hard look at things and made some changes. It went from not melting enough to melting too much, so well, expect changes. I added an ICE_BOX test to emulate that in Maykut and Untersteiner, 1971, and I do now get proper cycles of ice melt and growth in it.
Why am I still messing with this model when there's CICE, you ask? We've got some obligations to run an offline-nested domain and CICE doesn't support that at this time.
Changes to ice model
Re: Changes to ice model
I've found and fixed more bugs in the ice model and it is finally behaving sensibly. I've also fixed some bugs in the CICE coupling (it's still slow, though). This is just a heads-up that if you care about the sea ice, you should fetch the updates now.
ETA: Still issues with CICE coupling. Sigh.
ETA: Still issues with CICE coupling. Sigh.