Wed Jun 23 09:59 CEST 2021
anjaka on mattermost:  wir wuerden gerne aus Vergleichsgruenden einmal slab ocean
 mit einer realen Landverteilung laufen lassen.
 Ich kann mir vorstellen, dass es da schon passende Input Files zu gibt?

mkdir CM2.1p1-slab
cd CM2.1p1-slab
mkdir INPUT
cd INPUT
ln -s ../../CM2.1p1/INPUT/* .
# Remove (symlinks to) the restart files to do a cold start,
# because the warmstart state is for the full ocean.
# Caution: fv_rst.res contains the surface topography, thus must be preserved,
# despite its name it must be treated as initial condition, not as restart file :-(
# Note also, that it is in raw binary format, not NetCDF :-(
# And since that file was produced in big-endian byte order, we must set
#    export F_UFMTENDIAN=big
# before invoking the model.
rm atmos_coupled.res.nc \
   fv_srf_wnd.res \
   ice_model.res.nc \
   mg_drag.res.nc \
   ocean*.res* \
   physics_driver.res.nc \
   radiation_driver.res.nc 
   radiative_gases.res.nc \
   soil.res.nc \
   strat_cloud.res.nc \
   vegetation.res.nc

Copy input.nml, data_table, diag_table, field_table from StripedPlanet-slab setup.
I.e. no aerosols, no dust.
Set year for input of CO2&friends to 1860.

To compute roughness from real world topography, set in input.nml
 source_of_sgsmtn = 'computed'