It caused problems for a post-processing script that I use. I alluded to this in this week's section meeting, but it took awhile to figure out the source of the problem for my script.
I dug into FMS and found that if we add
&mpp_io_nml
cf_compliance = .true.
/
to input.nml, then time_bnds:units is the same as time:units.
Can we modify input_nml.yaml so that the above is added to input.nml?
The also makes the value of time:calendar lowercase and changes the attribute name cartesian_axis in MOM output to axis. I do not expect these changes to cause problems for tools.
The latter change might actually make MOM output more amenable to the cf-xarray library that @dcherian and others have developed.
@alperaltuntas , can we talk about how to get cf_compliance = .true. in mpp_io_nml added to input.nml into MOM_interface?
This unit difference is a headache for my analysis scripts.
In out current MOM6 output, units for the variables
time
andtime_bnds
areThis difference is not compliant with CF conventions, sec 7.1, paragraph 2.
It caused problems for a post-processing script that I use. I alluded to this in this week's section meeting, but it took awhile to figure out the source of the problem for my script.
I dug into FMS and found that if we add
to
input.nml
, thentime_bnds:units
is the same astime:units
.Can we modify
input_nml.yaml
so that the above is added toinput.nml
?The also makes the value of
time:calendar
lowercase and changes the attribute namecartesian_axis
in MOM output toaxis
. I do not expect these changes to cause problems for tools.The latter change might actually make MOM output more amenable to the cf-xarray library that @dcherian and others have developed.