Add frazil to ice-shelf melt rates

[irenavankova]

New variable frazilIceFreshwaterFlux is introduced to keep track of frazil ice formation instanatenously (and not only commulatively over coupling timestep as before)

New variable landIceFreshwaterFluxTotal is introduced to keep track of total melt rates that includes both interface melt/freeze and frazil ice formation, so that landIceFreshwaterFluxTotal = landIceFreshwaterFlux + frazilIceFreshwaterFlux.

Associted changes in MPAS-Analysis will reflect this to plot Antarctic meltrate maps with landIceFreshwaterFluxTotal instead of landIceFreshwaterFlux.

The changes here affect diagnostics only.

The plot here shows the monthly averaged fields (landIceFreshwaterFlux, frazilIceFreshwaterFlux, and landIceFreshwaterFluxTotal) for the year 6 month 12 of a Gcase with the ECwISC30to60E2r1 mesh. Freezing patterns over the Filchner-Ronne look much more realistic with landIceFreshwaterFluxTotal.

[xylar]

@irenavankova, this is amazing and very quick work!

I looked carefully through the code and I think this is exactly what we want. As always, you have done careful work and the code looks good.

I will run a 5-year CRYO1850 B-case IcoswISC30E3r5 simulation on Chrysalis with this branch. That should work as a "do-no-harm" test in that configuration and should also provide the data I need to update MPAS-Analysis to support these changes.

[xylar]

@irenavankova, also thanks for the figures above! I'm so glad this work turned out to be important! I wouldn't have wanted you to put this time in if it turned out to be a negligible effect.

[xylar]

@cbegeman and @jonbob, could you each have a look at this and let @irenavankova know if there is anything you would like to have happen before this moves over to E3SM? If you would like to wait for my testing and MPAS-Analysis updates, that's fine.

[darincomeau]

Since I'm knee deep in the ocean conservation analysis member, does this work relate to/impact the existing variable accumulatedLandIceFrazilFlux at all?

[irenavankova]

Since I'm knee deep in the ocean conservation analysis member, does this work relate to/impact the existing variable accumulatedLandIceFrazilFlux at all?

It doesn't affect it. It just keeps track of the nonaccumulated version of that variable.

[xylar]

It is only indirectly related to accumulatedLandIceFrazilFlux. That variable is the sum over the whole simulation of the frazil flux, masked to under land ice. frazilIceFreshwaterFlux is for the full ocean, not just under land ice and it is the instantaneous value (so basically the time derivative of accumulatedLandIceFrazilFlux) that would be passed to MALI as a freshwater flux in coupled simulations along with the melting/freezing at the interface.

[cbegeman]

Approving based on visual inspection.

[cbegeman]

Thanks @irenavankova! This will be great to have!

[xylar]

Analysis from my 5-year B-case run is here:
https://web.lcrc.anl.gov/public/e3sm/diagnostic_output/ac.xylar/analysis/20240226.CRYO1850.ne30pg2_r05_IcoswISC30E3r5.include_frazil.chrysalis/yrs_5-5/ocean/index.html

I was a little surprised that we get very little frazil in that case but it could be that conditions are just too warm or that 5 years isn't long enough.

[jonbob]

I think this is ready to move over to E3SM. Before we can merge it, we'll have to run the scripts to make the bld files consistent with Registry -- if only to pick up on the newer field descriptions

[irenavankova]

Thanks, @jonbob, I will move it over to E3SM now then. I am not sure which scripts you are referring to - is that something to do on my end?