Produce AMIP SST and sea-ice ancillaries for ESM1.6

[MartinDix]

AMIP uses monthly SST and sea ice with a special time interpolation technique that maintains the monthly means. The ancillary files are modified to allow this and are not the monthly means themselves.

E.g. SST (1 degree resolution)
/g/data/qv56/replicas/input4MIPs/CMIP7/CMIP/PCMDI/PCMDI-AMIP-1-1-10/ocean/mon/tos/gn/v20250807 /tos_input4MIPs_SSTsAndSeaIce_CMIP_PCMDI-AMIP-1-1-10_gn_187001-202212.nc and SST boundary conditions
/g/data/qv56/replicas/input4MIPs/CMIP7/CMIP/PCMDI/PCMDI-AMIP-1-1-10/ocean/mon/tosbcs/gn/v20250807/tosbcs_input4MIPs_SSTsAndSeaIce_CMIP_PCMDI-AMIP-1-1-10_gn_187001-202212.nc

Recommended procedure is to interpolate the tos data and then to apply the time adjustment and NOT to interpolate the tosbcs data, e.g., https://code.metoffice.gov.uk/trac/ancil/wiki/CMIP6/ForcingData/SstSeaIce#Documentation.

Although the CMIP7 experiments don't require the full period it might be useful for other experiments so we should create ancillary files with the whole period.

[penguian]

See https://input4mips-cvs.readthedocs.io/en/latest/dataset-overviews/amip-sst-sea-ice-boundary-forcing/ for a description of the datasets.

[penguian]

Do we also need the seaIce files?

$ tree /g/data/qv56/replicas/input4MIPs/CMIP7/CMIP/PCMDI
/g/data/qv56/replicas/input4MIPs/CMIP7/CMIP/PCMDI
└── PCMDI-AMIP-1-1-10
    ├── ocean
    │   ├── fx
    │   │   ├── areacello
    │   │   │   └── gn
    │   │   │       └── v20250807
    │   │   │           └── areacello_input4MIPs_SSTsAndSeaIce_CMIP_PCMDI-AMIP-1-1-10_gn.nc
    │   │   └── sftof
    │   │       └── gn
    │   │           └── v20250807
    │   │               └── sftof_input4MIPs_SSTsAndSeaIce_CMIP_PCMDI-AMIP-1-1-10_gn.nc
    │   └── mon
    │       ├── tos
    │       │   └── gn
    │       │       └── v20250807
    │       │           └── tos_input4MIPs_SSTsAndSeaIce_CMIP_PCMDI-AMIP-1-1-10_gn_187001-202212.nc
    │       └── tosbcs
    │           └── gn
    │               └── v20250807
    │                   └── tosbcs_input4MIPs_SSTsAndSeaIce_CMIP_PCMDI-AMIP-1-1-10_gn_187001-202212.nc
    └── seaIce
        └── mon
            ├── siconc
            │   └── gn
            │       └── v20250807
            │           └── siconc_input4MIPs_SSTsAndSeaIce_CMIP_PCMDI-AMIP-1-1-10_gn_187001-202212.nc
            └── siconcbcs
                └── gn
                    └── v20250807
                        └── siconcbcs_input4MIPs_SSTsAndSeaIce_CMIP_PCMDI-AMIP-1-1-10_gn_187001-202212.nc

24 directories, 6 files

[blimlim]

There's a few places I've seen code which applies the monthly adjustments:

• The first is the Metoffice code mentioned in the above link: https://code.metoffice.gov.uk/trac/ancil/browser/contrib/branches/dev/jeffknight/r4697_amiplbcs4cmip6/SstSeaIce

• It looks like there's updated fortran+python code from PCMDI here: https://github.com/PCMDI/amipbcs
  If this is what's used by PCMDI for the released files it could be good option.

• My first job at Monash was to run this process on some custom SSTs/Sea ice. I had trouble with the metoffice code and ended up adapting a version that had been used with CESM: https://github.com/blimlim/ACCESS_Tayloring

[MartinDix]

@penguian We do also need the sea ice

[penguian]

See https://code.metoffice.gov.uk/trac/ukcmip6/wiki/AMIPBoundaryConditions4CMIP6 for a desciption of the Karl Taylor procedure and why it is needed in the case where the model needs daily data interpolated from the monthly mean data. This matches what @blimlim says above.

[penguian]

I will pursue the PCMDI repo above.

Also are the ACCESS-ESM1.5 files below in the correct format for ACCESS-ESM1.6?

/g/data/vk83/configurations/inputs/access-esm1p5/modern/amip/atmosphere/boundary_conditions/global.N96/2021.07.08
├── amip_seaice_n96_greg.pp
└── amip_sst_n96_greg.pp

0 directories, 2 files

[blimlim]

The format should be correct for ESM1.6