Update MOM_input parameters for tx2_3v2/tx2_3v3 grids

[gustavo-marques]

Updates parameter values in MOM_input.yaml and MOM_input.json for the tx2_3v2 and tx2_3v3 grids. Changes include:

New parameters:

• TAPER_LEITHY, LEITHY_DEPTH, LEITHY_WIDTH — enables depth-tapered Leith+E backscatter
• USE_KH_BG_2D, KH_BG_2D_FILENAME, KH_BG_2D_VARNAME — enables 2D background harmonic viscosity from file
• CHANNEL_DRAG_MAX_BBL_THICK — limits BBL thickness for channel drag (5 m)
• CORRECT_BBL_BOUNDS — enables correct BBL thickness/viscosity bounds

Tuning changes:

• LEITHY_SMAG_LAP_CONST: 75.0 → 85.0; LEITHY_CK: 0.0 → 1.0
• BBL_EFFIC: 0.2→ 0.01
• KH_RES_FN_FACTOR: 0.9 → 0.5; KH_RES_FN_POWER_FACTOR: 0.4 → 0.5
• MEKE_GMCOEFF: 0.06 → 0.07
• MEKE_KHTH_FAC: adds 0.01 value for tx2_3 grids
• INTERPOLATE_RES_FN: enabled for tx2_3 grids
• TRUNCATION_N_SAVE: set to 1,000,000 for tx2_3 grids
• USE_OLD_FC_GAMMA_IC and USE_OLD_KHTH_BUG: disabled (True → False) for tx2_3 grids

Removed: MLE_DENSITY_DIFF (reverts to default)

USE_CR and CR_FILE are now only set when ATM = cam

Addresses #307

[gustavo-marques]

@alperaltuntas and @mnlevy1981
We need to import the following:

/glade/campaign/cesm/cesmdata/inputdata/ocn/mom/tx2_3v3/KH_BG_2D_tx2_3v3_260306.nc
/glade/campaign/cesm/cesmdata/inputdata/ocn/mom/tx2_3v2/KH_BG_2D_tx2_3v2_260306.nc

[mnlevy1981]

The KH_BG_2D files have both been imported

[alperaltuntas]

The KH_BG_2D file has NaN values on 30 ocean cells of the tx2_3v2 grid (I haven't checked the v3 file). That's causing the DEBUG tests to fail. While we should fix those cells in the input file, that alone may not be enough since we take averages on q-points (where there may be adjacent NaN land points). So I think two things should happen:

1. Fix the file. The 30 cells should be corrected.
2. Set land point values to zero in the code.

@gustavo-marques could you address (1)? I can then take care of (2) if it turns out to be needed after (1).

See a diagnostic script:
/glade/work/altuntas/cesm.sandboxes/cesm3_0_beta08/components/mom/check_kh_bg_2d_nans.py

[alperaltuntas]

The KH_BG_2D file has NaN values on 30 ocean cells of the tx2_3v2 grid (I haven't checked the v3 file). That's causing the DEBUG tests to fail. While we should fix those cells in the input file, that alone may not be enough since we take averages on q-points (where there may be adjacent NaN land points). So I think two things should happen:

1. Fix the file. The 30 cells should be corrected.
2. Set land point values to zero in the code.

@gustavo-marques could you address (1)? I can then take care of (2) if it turns out to be needed after (1).

See a diagnostic script: /glade/work/altuntas/cesm.sandboxes/cesm3_0_beta08/components/mom/check_kh_bg_2d_nans.py

Actually, (2) is already in the code. So, I think all we need is (1).

[gustavo-marques]

I've updated both files, and I kept the same name, so we do not need to update MOM_input:

/glade/campaign/cesm/cesmdata/inputdata/ocn/mom/tx2_3v3/KH_BG_2D_tx2_3v3_260306.nc
/glade/campaign/cesm/cesmdata/inputdata/ocn/mom/tx2_3v2/KH_BG_2D_tx2_3v2_260306.nc

[alperaltuntas]

@mnlevy1981 @gustavo-marques Can we override already imported files?

[gustavo-marques]

@alperaltuntas, Mike just told me we cannot. I am recreating these files with today's date.

[gustavo-marques]

I've created new files using today's date, and updated MOM_input and input_data_list. Here are the new files:

/glade/campaign/cesm/cesmdata/inputdata/ocn/mom/tx2_3v3/KH_BG_2D_tx2_3v3_260427.nc
/glade/campaign/cesm/cesmdata/inputdata/ocn/mom/tx2_3v2/KH_BG_2D_tx2_3v2_260427.nc

[mnlevy1981]

I've created new files using today's date, and updated MOM_input and input_data_list. Here are the new files:

/glade/campaign/cesm/cesmdata/inputdata/ocn/mom/tx2_3v3/KH_BG_2D_tx2_3v3_260427.nc
/glade/campaign/cesm/cesmdata/inputdata/ocn/mom/tx2_3v2/KH_BG_2D_tx2_3v2_260427.nc

These files have been imported as well

[alperaltuntas]

@gustavo-marques The tx3_v2 tests are now passing, but I'll also run v3 tests before merging.