Issues for zero layer ice and tidying up diagnostics

[ofa001]

Just going to list like @anton-seaice has some of the issues, we need to look at for the ESM version, they can be handled in a single issue or split.

evp_ice, evp_snow, found one additional place where these were zeroed out in CICE6 icepack, and not in GSI8.1 will be worth testing to see if this is the bug and if successful, we can feed through to the snow and ice evaporation diagnostic currently using same diagnostic in CM2 version.

Set minimum values for aicenmin, hsmin suitable for zero-layer model, not default values.

Update minor errors in diagnostics where timestep dt and density of ice and snow where missing,
Leave weighting in for minimal ice cover in at first step, but do test before removing in a separate run.