Seaice over land after the analysis insertion

[guillaumevernieres]

Description

This was mentioned here:
ufs-community/ufs-weather-model#2562 (comment)
But the discussion above is based on a ~2 yo version of the soca to cice variable change. It's possible that it has been fixed, but better double check.

To Reproduce:

Run soca2cice and check that we have no sea ice over land.

[DeniseWorthen]

The other item to check is whether any ice volume exists where there is no associated ice coverage (aicen=0 but vicen /= 0)

[shlyaeva]

I'm hoping this is resolved now by calling icepack's cleanup_itd in initialization, but we'll need to check.

[guillaumevernieres]

I'm hoping this is resolved now by calling icepack's cleanup_itd in initialization, but we'll need to check.

🤞

[shlyaeva]

I started working on it, and couldn't make any ice over land in my test as hard as I tried. Looks like someone smart already added code to make sure we're outputting zero over land:

soca/src/soca/VariableChange/Soca2Cice/soca_ciceutils_mod.F90

Line 155 in b437805

if ( pe == mpp_root_pe()) glb%aicen(1:self%ni,1:self%nj,n) = var2d*glb%iceumask

.

I'll double check that the second concern

The other item to check is whether any ice volume exists where there is no associated ice coverage (aicen=0 but vicen /= 0)

is covered by cleanup_itd before closing this.

[DeniseWorthen]

@shlyaeva Thanks for the followup. I found a third condition also, which is to check for whether there is snow volume (vsnon) where there is no aicen. The opposite is fine I think---ice can have no snow coverage.

[shlyaeva]

I just confirmed that if I change the code that puts analysis into the CICE restart files and artificially set vsnon and vicen to be non-zero when aicen is zero, cleanup_itd that is called at the end resets those values back to zero. So I think the current code covers all the issues reported here. OK with everyone to close?

[DeniseWorthen]

From my perspective, it seems like whatever was previously the problem has been fixed in the intervening years. So, all is good.