Namelist settings controlling output in NorESM3

[oyvindseland]

What is the feature/what would you like to discuss?

While history_aerosol and do_aerocom reduce the number of output variables, there are still too many output variables for long simulations, in particular for individual tracers, e.g. concentration and column burdens) Should these be included in history_aerosols or should there be a new option?
Also sometimes the variables are diagnosed before writing them out. Should we also try to remove some of the calculations? My suggestion for now is to make a new namelist variables, e.g. mixture_output foe everything tracer related and physics_diag for physics output that may or not be needed and and for now just frame by if statements output variables that we suggest to take out needed for long simulations. I also think that there are some aerosol variables that should be diagnosed and included in longer simulations, e.g. total wet deposition, but that need to be done in another issue

Is there anyone in particular you want to be part of this conversation?

@gold2718 @DirkOlivie

Will this change (regression test) answers?

No

Will you be implementing this enhancement yourself?

Yes

[oyvindseland]

I will start making checking the impact for the existing namelist variables for additional output and then remove them piece by piece to make lists

Full set at starting point
history_aerosol = .true.
history_amwg = .true.
history_clubb = .true.
history_budget = .false.
history_chemistry = .true.
history_chemspecies_srf = .true.
do_circulation_diags = .true.
use_aerocom = .true.

[MichaelSchulzMETNO]

It would be nice to also have code that could be implemented in NorESM2.3 , eg aggregate relevant mmr variables to total dust, oa, so4, ss, bc burden.

[oyvindseland]

Aggregated MMR and column burdens of aerosol compounds already exists and are written out for individual compounds in both NorESM2.3 and NorESM3_alpha.
The same should be done for surface concentrations; and emissions?
Also while wet and dry deposition of tracers now requires history_aerosol=.true. I think the totals should be aggregated and have a somewhat higher priority.