Update SFS tests to Beta version 1.0 configuration

[DeniseWorthen]

Description

The SFS Beta version 1.0 is currently running on Gaea and Ursa. We need to update the RT configuration to match.

Solution

Compare current RT configuration w/ sandbox and update as appropriate.

Alternatives

Related to

[DeniseWorthen]

@XiaqiongZhou-NOAA Thanks for the sandbox. One major difference is that the sandbox is no longer using OCN and ICE warmstarts but actual restart files. This changes the configuration from a startup to a continue run type. The RT test is set up for 1994 05 01. Are those restarts staged somewhere?

Another issue is that the diag_table is not set up to enable the MOM output logging. I believe that SFS was the first application which G-W anticipated actually relying on the MOM output logging. @DavidHuber-NOAA Can you confirm? See NOAA-EMC/global-workflow#3822 for context.

[XiaqiongZhou-NOAA]

@DeniseWorthen Denise: I made the sandbox from a warm start from SFS. It is not from the start from the beginning. I believe only some namelist input changes should be enough for RTs.

[DavidHuber-NOAA]

Another issue is that the diag_table is not set up to enable the MOM output logging. I believe that SFS was the first application which G-W anticipated actually relying on the MOM output logging. @DavidHuber-NOAA Can you confirm? See NOAA-EMC/global-workflow#3822 for context.

@DeniseWorthen yes, that's correct. The SFS will need the MOM output logging.

[DeniseWorthen]

I see, thanks.

I do think the diag_table changes are still applicable though. The G-W may never have been updated to use the diag_table which adds the minutes field to the output file, "@[MOM6_OUTPUT_DIR]/ocn%4yr%2mo%2dy%2hr%2mi"

[NickSzapiro-NOAA]

I believe only some namelist input changes should be enough for RTs.

Isn't there also a simplified Ozone parameterization being used like NOAA-EMC/global-workflow#3386 @XiaqiongZhou-NOAA ? We want to ensure application configurations reproduce in debug/restart/... modes across platforms

[XiaqiongZhou-NOAA]

I believe only some namelist input changes should be enough for RTs.

Isn't there also a simplified Ozone parameterization being used like NOAA-EMC/global-workflow#3386 @XiaqiongZhou-NOAA ? We want to ensure application configurations reproduce in debug/restart/... modes across platforms

There is no parameter change for UFS RT related to GW#3386. Only a new fix file "global_o3prdlos.f77" is used. The current RTs has been using the new one.

[DeniseWorthen]

@XiaqiongZhou-NOAA These are the differences I see between your sandbox and the current RT test. The points are

• atm is writing 6h output and the output grid has changed.
• nstf_name = 2,1,0,0,0, meaning you are not starting using the NSTF spinup from the ICs
• skeb and sppt parameter changes

Do these all seem correct?

Model_configure:

sandbox	rt test
write_nsflip: .false.	write_nsflip: .true.
output_grid: global_latlon	output_grid: gaussian_grid
imo: 720	imo: 768
jmo: 361	jmo: 384
output_fh: 6 -1	output_fh: 24 -1

Input.nml

sandbox	rt test
rf_cutoff = 300.0	rf_cutoff = 100.0
res_latlon_dynamics = ""	res_latlon_dynamics = 'atminc.nc'
fhzero = 6	fhzero = 24
cscale = 0.2	cscale = 0.5
nstf_name = 2,1,0,0,0	nstf_name = 2,0,0,0,0
do_skeb = .true.	do_skeb = .false.
iseed_ca = 442690846	iseed_ca = 1115337723
new_lscale=.false.	new_lscale=.true.
skeb = 0.08,-999,-999,-999,-999	skeb = -999.
iseed_skeb = 19920501000401	iseed_skeb = 0
skeb_npass = 15	skeb_npass = 30
sppt = 0.5,0.2,0.1	sppt = 0.56,0.28,0.14,0.056,0.028
sppt_tau = 2.16E4,2.592E5,1.296E6	sppt_tau = 2.16E4,2.592E5,2.592E6,7.776E6,3.1536E7
sppt_lscale = 500.E3,1000.E3,2000.E3	sppt_lscale = 500.E3,1000.E3,2000.E3,2000.E3,2000.E3

[XiaqiongZhou-NOAA]

Currently for SFS, we are using CPC's analysis data so no NST variables, but I assume the RT test is using replay ICs, it should nstf_name = 2,0,0,0,0,. I suggest keeping RT test since we will change to use replay ICs in the future.
Anything else are need changed as you listed.

In addition, there is a parameter (GUST_CONST) change in MOM6 input.

[DeniseWorthen]

Thanks. I will also set GUST_CONST = 0.02 in MOM_input