As for eORCA025.L75 we used the WOA18 1980-2010 climatology (AKA decab81B0).
This is done with SOSIE using the same namelists just changing the name of the mesh_mask.nc file.
After Sosie processing, files were reformatted to NEMO standard with mknemolike.sh script.
Sosie was taken from last commit on GitHub (6e10db03e3042d5eb297eb). There are 2 set of namelist for L75 and L121. And 3 namelist for relative Salinity, Potential temperature and in situ temperature.
For some nco limitation in ncwa with netcdf4/HDF5 files we use the following procedure to obtain the SSS file from eORCA05.L75_81B0_WOA18_1m_vosaline.nc. (As far as only surface level is involved we can use
either the L75 or L121 files.
ncks -d deptht,1 eORCA05.L75_81B0_WOA18_1m_vosaline.nc eORCA05.L75_81B0_WOA18_1m_SSS.nc
# transform SSS file (with deptht dimension) into netcdf classic format:
ncks -3 eORCA05.L75_81B0_WOA18_1m_SSS.nc eORCA05.L75_81B0_WOA18_1m_SSS.nc3
# eliminate deptht dimension
ncwa -a deptht eORCA05.L75_81B0_WOA18_1m_SSS.nc3 eORCA05.L75_81B0_WOA18_1m_SSS.nc1
# transform SSS file with no deptht into netcdf4/HDF5 with deflation factor of 1
ncks -4 -L1 eORCA05.L75_81B0_WOA18_1m_SSS.nc1 eORCA05.L75_81B0_WOA18_1m_SSS.nc4
# Do some cleaning of intermediate files
rm eORCA05.L75_81B0_WOA18_1m_SSS.nc3 eORCA05.L75_81B0_WOA18_1m_SSS.nc1
# rename final file to its official name :
mv eORCA05.L75_81B0_WOA18_1m_SSS.nc4 eORCA05.L75_81B0_WOA18_1m_SSS.nc
In NEMO, 2D files must not have a degenerated vertical dimension. The latter procedure is the easiest way to eliminate such a dimension. It turned to be a bit more complicated than it should be, just because of the NCO limitations.
Conversion from Potential temperatures to Conservative Temperature and from relative salinity to Absolute salinity.
mk_teos10.sh script is used for such a convertion. It is based on GSW tools implemented in JMMTOOLS
At the end of this process SA and CT files are ready for use as initial conditions for our purpose.