Landfast-ice settings

[anton-seaice]

Dan Atwater has found that turning on Tensile strength (i.e. the plastic response to strain) in CICE works well, and recommends we set the Ktens = 0.2.

To get fast-ice, we would also need to turn on probablistic sea bed stress.

This wouldn't have much impact in the Antarctic, but would grow Arctic fast-ice. There are NSIDC observations (which?) of Arctic sea-ice which include ice-age or ice-type to compare to.

[adfraser]

In Antarctica we don't need to turn on either of the sea bed stresses - well, we could, but neither of them should do much at all, since our bathymetry is much deeper than the deepest keel for most of the shelf. At least setting Ktens to 0.2 would give tensile strength which would lead to more fast ice in embayments, but realistically we won't get much without grounded bergs for now. That's OK - Dan is working on an alternative parameterisation which will be relatively easy to get into OM3 without needing grounded bergs blocking the ocean.

In the Arctic, to get realistic fast ice there, we should turn on either of the sea bed stress schemes, but the LKD one is better-tested according to J-F Lemieux. The probabilistic one is more sophisticated but hasn't been tested as much.