Modify NUOPC cap to accept separate glc runoff fluxes

[alperaltuntas]

This PR addresses issue #281

Changes made:

• Added Forr_rofl_glc and Forr_rofi_glc import files corresponding the glc runoff fluxes via the rof component.
• Currently, these new fluxes get added to fluxes%lrunoff and fluxes%frunoff.
• For now, these new fluxes are imported only if cesm_coupled is .true.. If EMC folks decide the import these fluxes in the same way in the future, the import statements can be taken out of the if-blocks.

[alperaltuntas]

@mvertens Could you confirm if this PR addresses the issue #288 ?

[gustavo-marques]

We should keep these new fields separate in fluxes and not under fluxes%lrunoff and fluxes%frunoff.

Suggestion:
fluxes%lrunoff_glc(i,j) = fluxes%lrunoff_glc(i,j) + kg_m2_s_conversion * IOB%lrunoff_glc(i-i0,j-j0) * G%mask2dT(i,j)

fluxes%frunoff_glc(i,j) = fluxes%frunoff_glc(i,j) + kg_m2_s_conversion * IOB%frunoff_glc(i-i0,j-j0) * G%mask2dT(i,j)

[gustavo-marques]

Thanks for this PR, @alperaltuntas!

Are the changes in src/initialization/MOM_shared_initialization.F90 intended?

In addition to what I mentioned above, we also need to:

1. Add two new fields for the heat content associated with lrunoff_glc and frunoff_glc (heat_content_lrunoff_glc and heat_content_frunoff_glc). The coupler should provide these fields, and here is a PR that can be used as an example of adding them in CMEPS.

We should only account for heat_content_lrunoff_glc and heat_content_frunoff_glc when ENTHALPY_FROM_COUPLER = True.

2. The new mass and heat fluxes should be added to src/core/MOM_forcing_type.F90 and src/diagnostics/MOM_sum_output.F90. Please follow what was done in Option to provide enthalpy fluxes via NUOPC coupler #214

[alperaltuntas]

@gustavo-marques I just pushed a commit here and a PR to CMEPS (ESCOMP/CMEPS#489) to address your comments. One outstanding task is to determine what to do with the diagnostics associated with lrunoff and frunoff. Now that we have separated lrunoff_glc from lrunoff and frunoff_glc from frunoff, I wonder whether we should split the diagnostics as well, and if so, whether the split diagnostics truly correspond to cmor definitions of those diagnostics. See, for instance, the frunoff diagnostic whose cmor name is set to be ficeberg as defined in src/core/MOM_forcing_type.F90

[gustavo-marques]

These are the new available diagnostics:

"frunoff_glc"  [Unused]
    ! modules: {ocean_model,ocean_model_d2}
    ! long_name: Frozen glc runoff (calving) and iceberg melt into ocean
    ! units: kg m-2 s-1
    ! standard_name: glc_water_flux_into_sea_water_from_icebergs
    ! cell_methods: xh:mean yh:mean area:mean
"lrunoff_glc"  [Unused]
    ! modules: {ocean_model,ocean_model_d2}
    ! long_name: Liquid runoff (glaciers) into ocean
    ! units: kg m-2 s-1
    ! standard_name: water_flux_into_sea_water_from_glaciers
    ! cell_methods: xh:mean yh:mean area:mean
"total_frunoff_glc"  [Unused]
    ! modules: ocean_model
    ! long_name: Area integrated frozen glc runoff (calving) & iceberg melt into ocean
    ! units: kg s-1
"total_lrunoff_glc"  [Unused]
    ! modules: ocean_model
    ! long_name: Area integrated liquid glc runoff into ocean
    ! units: kg s-1
    ! cell_methods: xh:mean yh:mean area:mean
"heat_content_frunoff_glc"  [Unused]
    ! modules: {ocean_model,ocean_model_d2}
    ! long_name: Heat content (relative to 0C) of solid glc runoff into ocean
    ! units: W m-2
    ! cell_methods: xh:mean yh:mean area:mean
"heat_content_lrunoff_glc"  [Unused]
    ! modules: {ocean_model,ocean_model_d2}
    ! long_name: Heat content (relative to 0C) of liquid glc runoff into ocean
    ! units: W m-2

[gustavo-marques]

One minor suggestion: only make the new diagnostics available if GLC is active.