Choice of CICE6 parameters

[sanAkel]

CICE6 parameter settings from 2 experiments of @zgarraffo follow:

[Santha.Akella@ufe03 rtofs.20250401]$ pwd -P
/scratch4/NCEPDEV/marine/Zulema.Garraffo/FV3_RT/expt_14.1/rtofs.20250401
[Santha.Akella@ufe03 rtofs.20250401]$ 
[Santha.Akella@ufe03 rtofs.20250401]$ 
[Santha.Akella@ufe03 rtofs.20250401]$ diff ice_in ../rtofs.20250430/
4,6c4,6
<     year_init           = 2025
<     month_init          = 03
<     day_init            = 31
---
>     year_init           = 2023
>     month_init          = 07
>     day_init            = 01
392,394c392,394
<   stream_yearfirst = 2025
<   stream_yearlast = 2025
<   stream_yearalign = 2025
---
>   stream_yearfirst = 2023
>   stream_yearlast = 2023
>   stream_yearalign = 2023

Ignore the differences, just consider path to ice_in.

Goal is to create a setting that is:

• As close as possible to those used in RTOFS v2.5 (uses CICE4).
• Compatible with CICE6.
• Brings in known/tested/proven advance at 0.08-deg horizontal resolution (forced ice-ocean).

[sanAkel]

@zgarraffo for the sake of information, can you please detail how you got to:
/scratch4/NCEPDEV/marine/Zulema.Garraffo/FV3_RT/expt_14.1/rtofs.20250401/ice_in ?

1. Clone ufs-weather-model,
2. Set up a regression test - Which one?
3. ...?
4. ...?

[zgarraffo]

I have been using year_init=2023, initial year of atmospheric forcing. I don't think it makes a difference using 2023 or 2025. It may be clearer to use 2025 03 31 as the first day of the run.

[sanAkel]

@zgarraffo for the sake of information, can you please detail how you got to:
/scratch4/NCEPDEV/marine/Zulema.Garraffo/FV3_RT/expt_14.1/rtofs.20250401/ice_in ?

1. Clone ufs-weather-model,
2. Set up a regression test - Which one?
3. ...?
4. ...?

Appreciate your answers to the ⬆️ questions, not about the difference in dates. Thanks

[zgarraffo]

-I always cloned as
git clone --recursive https://github.com/ufs-community/ufs-weather-model.git
For CICE with insertion, I cloned git clone --recursive git@github.com:awallcraft/CICE-EMC.git
-Always did the regression test for datm_cdeps_mx025_gefs
-Different executables depended on the time of cloning.
-I used input files gathered by Hae-Cheol Kim for the COAPS version of MOM6.
-Depending on the run are the parameters each experiment ush/hycom_var.job and ush/submit_incup_fcst.sh
The version of UFS-WM used in expt 14.1, 14.2, 14.3 are from early 2025. I tested the September version of UFS-WM, needing a change in the atmospheric mesh.
Up to now the only run with ice coverage assimilation is 14.3

[sanAkel]

-I always cloned as
git clone --recursive https://github.com/ufs-community/ufs-weather-model.git
For CICE with insertion, I cloned git clone --recursive git@github.com:awallcraft/CICE-EMC.git
-Always did the regression test for datm_cdeps_mx025_gefs
-Different executables depended on the time of cloning.
-I used input files gathered by Hae-Cheol Kim for the COAPS version of MOM6.

1. What about the CICE6 input file (named ice_in)? Is it the same as what comes from the datm_cdeps_mx025_gefs? (14.1, 14.2)

2. If so, the only change to be expected is: insert_sic? (In 14.3)

-Depending on the run are the parameters each experiment ush/hycom_var.job and ush/submit_incup_fcst.sh
The version of UFS-WM used in expt 14.1, 14.2, 14.3 are from early 2025. I tested the September version of UFS-WM, needing a change in the atmospheric mesh.
Up to now the only run with ice coverage assimilation is 14.3

[zgarraffo]

ice_in that I am using may come from COAPS (from Hae-Cheol)
I copied the regression test ice_in to /scratch4/NCEPDEV/marine/Zulema.Garraffo/FV3_RT/input_files_2025/ice_in_mx025_reg_test
Some differences are, in the 1/12th ice_in:
grid_ice=C (may need to be changed to B)
tfrz_option='mushy'
a few variables only in the regression test (not sure what are the default)
grid is 0.08 deg
etc

[zgarraffo]

From the regression test, I only use the executable and the modules.

[sanAkel]

From the regression test, I only use the executable and the modules.

In that case, Moving forward, please try:

#82

[sanAkel]

From the regression test, I only use the executable and the modules.

In that case, Moving forward, please try:

#82

@zgarraffo any luck trying above mentioned clone/build script?