Proposed I* compsets.

[mvdebolskiy]

From @rosiealice:

OK so for me the most obvious FATES_NOCOMP configurations for I compsets are:

I1850Clm60FATESnocompCru : 1850_DATM%CRUJRA2024_CLM60%FATES-NOCOMP_SICE_SOCN_SROF_SGLC_SWAV_SESP
I1850Clm60FATESnocompCPLHIST : 1850_DATM%CPLHIST_CLM60%FATES-NOCOMP_SICE_SOCN_SROF_SGLC_SWAV_SESP

I2000Clm60FATESnocompCru : 2000_DATM%CRUJRA2024_CLM60%FATES-NOCOMP_SICE_SOCN_SROF_SGLC_SWAV_SESP

• Need a %NorESM argument for the compsets.
• Should we get rid of the '-nocomp' and make 'nocomp' just the default FATES when we are running %NorESM compsets?
  Pros: Simplifies the NorESM compset names.
  Cons: Not future proof for NorESM3.x
  Maybe add 'FATESfull as an alternative compset to avoid confusion. So, have
  FATESfull, FATESnocomp and FATESsp.
  Conclusion: Have a %noresm and FATESfull, FATESnocomp and FATESsp compset setting.
  Aliases:
  I1850NorClm60FatesNoCompCru

Nor_I1850NorClm60FatesNoCompCru

for n compsets, we currently have
N1850fates-nocomp : 1850_CAM70%LT%NORESM%CAMoslo_CLM60%FATES-NOCOMP_CICE_BLOM%HYB%ECO_MOSART_DGLC%NOEVOLVE_SWAV_SESP
NE1850fates-nocomp : 1850_CAM70%LT%NORESM%CAMoslo_CLM60%FATES-NOCOMP_CICE_BLOM%HYB%ECO_MOSART_DGLC%NOEVOLVE_SWAV_SESP_BGC%BPRP
NF1850fates-nocomp : 1850_CAM70%LT%NORESM%CAMoslo_CLM60%FATES-NOCOMP_CICE%PRES_DOCN%DOM_MOSART_DGLC%NOEVOLVE_SWAV_SESP
NF1850ghg_fates-nocomp : 1850_CAM70%LT%NORESM%GHGCAMoslo_CLM60%FATES-NOCOMP_CICE%PRES_DOCN%DOM_MOSART_DGLC%NOEVOLVE_SWAV_SESP

I don't know the difference between the latter sets. But it would be good to have one that acted like an 'AMIP' (fixed sst's) simulation. So that we could test the impact of parameters on land-atmosphere fluxes. Ideally we would want that in 2000 and 1850.

Then there are other more exotic coupled compsets that we might want to shift to using FATES_NOCOMP by default to avoid people making errors when creating coupled NorESM cases. But I am not sure which of the coupled compsets are in regular use. I would be tempted to shift the majority of them on to FATES_NOCOMP and leave only the full FATES ones that we might need for testing in here. We aren't supporting full-FATES coupled configurations in NorESM3 anyway, so it makes sense to not have aliases for them?