Swap hi_min to 0.2 and add aicenmin=1.e-5

[blimlim]

This PR adjusts parameters for the minimum ice thicknesses and concentrations allowed by Icepack. The changes are based to the discussion in ACCESS-NRI/dev_coupling#45 (comment).

The first change, following @ofa001's recommendation, is to increase the threshold concentration for zapping from puny to 1.0e-5, matching the value in CM2.

The second change is to move where hi_min=p2 is set. It was originally set in an if/else block which was only activated for kitd != 1. In our current setup, we have kitd = 1, and so the hi_min setting didn't come into effect.

I haven't replicated the more complex way CM2 sets aicenmin, which swaps the value depending on whether the zero-layer scheme is used. I wasn't sure whether copying that would be useful for us or if it would just add unnecessary complexity.

More recent versions of Icepack move hi_min to icepack_parameters (and might let it be set by namelist), and so we will likely have to adjust these changes when we update the version.

[kieranricardo]

Looks good to me! :)

[ofa001]

Hi @kieranricardo I think the issue is that the #cesmcoupled settings was inside the kitd loop and we had different settings. I agree and some point we should have an #accesscoupled set up. I think @bimlim is rught later icepack editions are going to make hi_min a namelist setting, but it wasnt quite clear from their headings if it would link well with the Met office approach, though the zero layer version has been removed, in these later versions.

[blimlim]

Thanks @kieranricardo and @ofa001! I just had a quick question on the CESMCOUPLED vs ACCESSCOUPLED flags: Would the idea be to first replace instances of CESMCOUPLED with ACCESSCOUPLED?

In this case, I'm wondering if we should leave the settings under the flags unchanged. I'm not sure if I've understood correctly though!

[kieranricardo]

hey @blimlim, I think we'll want to add in additional ACCESSCOUPLED flags and keep the CESMCOUPLED flags (which will be used by OM3). But actually either we're better to leave that as it is - ignore my previous suggestion!

[blimlim]

Ah I see, thanks :)