Sea ice mass budget diagnostics for CMIP7

[William-gregory]

Additional diagnostics for computing sea ice area, sea ice mass, and snow mass budgets. Below are the CMIP diagnostic names how they have been derived.

Ice concentration
sidconcdyn : equivalent to XPRTc
sidconcth : equivalent to LSRCc + LSNKc

Ice mass
sidmassdyn : equivalent to XPRTi
sidmassevapsubl: split EVAP to derive ice contribution
sidmassgrowthwat : equivalent to FRAZIL/LatHtFus
sidmassgrowthsi: split S2IC to derive ice contribution and average over all part_size categories (incl. open water)
sidmassgrowthbot: equivalent to LSRCi - sidmassgrowthsi - sidmassgrowthwat
sidmassmeltbot : equivalent to BSNKi
sidmassmelttop : equivalent to LSNKi - BSNKi - sidmassevapsubl
sidmassth : equivalent to LSRCi + LSNKi

Snow mass
sisndmassdyn : equivalent to XPRTs
sisndmasssubl: split EVAP to derive snow contribution
sisndmasssi: split S2IC to derive snow contribution and average over all part_size categories (incl. open water)
sisndmassmelt : equivalent to LSNKs - sisndmasssubl - sisndmasssi

Note that, for backwards compatibility, I have kept the output units of the standard mass budget diagnostics (XPRT, LSRC, LSNK, BSNK) as kg/m^2yr-1

[MitchBushuk]

Thanks Will for this PR! A few comments/requests:

1. In the new snow evaporation diagnostic, I believe there is a sign error on Line 1126 of SIS2_ice_thm.F90. I think this line should actually read:evap_s = evap_s + evap, since in the case that evap is negative evap_s should be negative (corresponding to frost formation).

2. For all diagnostics that have a corresponding CMIP7 CMOR diagnostic name, these should be registered with the diag field. This will allow users to save the diagnostics using either the CMOR name or the SIS2 name. For example, the XPRTc registration should look like this:

CS%id_xprt_c = register_diag_field('ice_model', 'XPRTc', diag%axesT1, Time, &
               'frozen water area transport convergence', 's-1', conversion=US%RZ_to_kg_m2, &
               missing_value=missing, cmor_field_name='sidconcdyn', &
               cmor_standard_name='tendency_of_sea_ice_area_fraction_due_to_dynamics', &
               cmor_long_name='Sea-Ice Area Fraction Tendency Due to Dynamics')

3. Regarding the units of XPRT, LSRC, LSNK, BSNK, a concern with this change is that this will affect comparisons with all earlier runs of SIS2 that used units of kg/m^2yr-1 for these terms. CMIP7 requests units of kg/m^2s-1 for all mass budget terms. One possible route forward could be to use per second units for all the new sea ice mass budget, snow mass budget, and sea ice area budget diagnostics, and retain the per year units for XPRT, LSRC, LSNK, BSNK (noting that these are for the combined ice+snow mass budget, which is not part of the CMIP7 request). Does @Hallberg-NOAA have thoughts on this?

[William-gregory]

hey @MitchBushuk, my latest push contains your recommended changes and I have renamed variables according to the SIMIP definition. I have not yet tested these in a simulation, but will aim to do that asap

[Hallberg-NOAA]

It looks to me like this PR includes CMOR related options that will not even compile until #229 has been successfully merged into SIS2.

[William-gregory]

Thanks for catching those errors @Hallberg-NOAA (and for initiating PR #229). My latest push makes those few adjustments, but for now I guess we wait for the other PR to be merged :) If you spot any other errors in the mean time, let me know!