Merge changes from MOM6-CICE6

[anton-seaice]

No description provided.

[anton-seaice]

I believe this is correct, but we need to merge #23 first to confirm.

[ezhilsabareesh8]

I think fsd related parameters are required in MOM6-CICE6-WW3 configurations, for example, tr_fsd shouldn't be removed from ice_in unless it is true by default.

[anton-seaice]

Oops. Thanks. Ill fix tomorrow.

I can't see any other mistakes - all the fsd history parameters are unchanged ?

[ezhilsabareesh8]

all the fsd history parameters are unchanged ?

The following history parameters are unchanged, however restart_fsd = .false. is ommited not sure what it actually does

&icefields_fsd_nml
    f_fsdrad       = 'md' 

[anton-seaice]

Ok thanks - yes we need that too to get it to read fsd from the cice restart file!

https://cice-consortium-cice.readthedocs.io/en/main/user_guide/ug_case_settings.html#tracer-nml

[ezhilsabareesh8]

Is it is ok to use the same initial conditions for both the RYF and IAF configurations?, It's the same in MOM6-CICE6 configs as well.

ice_ic = "./input/iced.1900-01-01-10800.nc"

[anton-seaice]

Yeah I think that is fine. Its not perfect. But the ice will be kinda of close, and the ice will stabilise much quicker than the ocean will. So it seems ok from that perspective.

[ezhilsabareesh8]

Looks good to me

[anton-seaice]

Ah. This doesn't run anymore:

 (abort_ice)ABORTED: 
 (abort_ice) called from ice_restart.F90
 (abort_ice) line number          744
 (abort_ice) error = 
 (ice_pio_check)NetCDF: Variable not found, (read_restart_field) ERROR: missing 
 varid fsd001

Because our ic file doesnt include fsd

We must have missed this in the RYF config too.

(Not sure how to make the best fsd initial conditions, or whether to fix this some other way!)

[ezhilsabareesh8]

(Not sure how to make the best fsd initial conditions, or whether to fix this some other way!)

Maybe we can set the ice_ic to default for now?, I think fsd initial conditions cannot be generated from CICE5?

We must have missed this in the RYF config too.

Yes, restart_fsd is set to false in RYF.

[anton-seaice]

I think setting ice_ic to default would be troublesome, it gives a very unrealistic ice area. We could leave restart_fsd as false and create a new issue to resolve it.

[anton-seaice]

Ah. It looks like restart_fsd=.false. was correct. If the runtype is continue, this gets turned on in the driver, otherwise it initialises internally:

https://github.com/CICE-Consortium/CICE/blob/0af031d785d3bc622cd19af48a2e9465b5abe9a0/cicecore/drivers/nuopc/cmeps/CICE_InitMod.F90#L373

[ezhilsabareesh8]

This doesn't run anymore:

(abort_ice)ABORTED: (abort_ice) called from ice_restart.F90 (abort_ice) line number 744 (abort_ice) error = (ice_pio_check)NetCDF: Variable not found, (read_restart_field) ERROR: missing varid fsd001

But when you set restart_fsd = true, you're getting the above missing variable error right?

! floe size distribution if (tr_fsd) then if (trim(runtype) == 'continue') restart_fsd = .true. if (restart_fsd) then call read_restart_fsd

According to the above code even when we set restart_fsd = .true., it should run right? "

[anton-seaice]

If runtype=initial (and restart_fsd=.false.) then it will initialise fsd using the init_fsd call

If runtype=continue or restart_fsd=.true. in ice_in, then it will try to read fsd from the restart/ic file

[anton-seaice]

This works and generates output and history files for each component. Doesn't restart due to still pointing to a binary before COSIMA/access-om3#160 was included