Update the scripts that handle NCODA output(s), i.e., increments that are passed to the model.

@zgarraffo

With the model swap (hycom+cice4 with ufs-datm), the files and/or scripts that:

Post process NCODA output (ocean and sea ice)
And feed to the ocean and sea ice model(s)

need to be modified.

At the minimum the following scripts need to be modified:

See the following

Paths for sample files from @zgarraffo:

/scratch3/NCEPDEV/marine/Zulema.Garraffo/FV3_RT/expt_14.4_hera/rtofs.20250504/INPUT
/scratch3/NCEPDEV/marine/Zulema.Garraffo/FV3_RT/expt_14.4_hera/rtofs.20250504/RESTART_inc
/scratch3/NCEPDEV/marine/Zulema.Garraffo/FV3_RT/expt_14.4_hera/ncoda/hycom_var/restart/

Use h from: ls /scratch3/NCEPDEV/marine/Zulema.Garraffo/FV3_RT/expt_14.4_hera/ncoda/hycom_var/restart/lyrthk_lyr_1o4500x3297_*_0024_fcstfld

And script:
/scratch3/NCEPDEV/marine/Zulema.Garraffo/FV3_RT/expt_14.4_hera/ush/ncodaz_inc*sh, utility is: ncodaz_inc2mom6nc_lyr.sh

Please point me to the script to create the ice coverage nc file, since it is ready

Please point me to the script to create the ice coverage nc file, since it is ready

It was added via PR #54

Let me know if you need an example of how to use on wcoss/ dwood.

Yes, please, point me to the specific script or example.

@zgarraffo

Yes, please, point me to the specific script or example.

@zgarraffo It in our interest to get done properly instead of one-off scripts!
~~You'll get it ... when~~ I am done working on it via this branch.

I do have a driver script at /lfs/h2/emc/stmp/santha.akella/test

santha.akella@dlogin03:~/my_stmp/test> cat test.sh 
#!/bin/sh

export DATAROOT=/lfs/h2/emc/stmp/santha.akella/test/
export jobid=icecov_sic
export HOMErtofs=/lfs/h2/emc/ptmp/santha.akella/RTOFS_GLO/
export USHrtofs=${HOMErtofs}/ush/
export FIXrtofs=${HOMErtofs}/fix/
export PDY=20251216
export dtg=20251215

CONMIN=/lfs/h2/emc/couple/noscrub/zulema.garraffo/COMDIR/prod/com/rtofs/v2.5/rtofs.${PDY}/
IDM=4500
JDM=3297
var_name="icecov"
#var_name="icethk"
#var_name="icetmp"
#var_name="mixlyr"

SIZN="${IDM}x${JDM}"
typec=${var_name}_sfc_1o${SIZN}

cd ${DATAROOT}

# link to topography
ln -f -s ${FIXrtofs}/depth_GLBb0.08_09m11ob2_mom6.nc depth_GLBb0.08_09m11ob2_mom6.nc

# input file
ln -sf $CONMIN/ncoda/hycom_var/restart/${typec}_${dtg}00_0000_analfld ./${var_name}_${dtg}_analfld

# run
$HOMErtofs/ush/rtofs_glo2d_ice.sh depth_GLBb0.08_09m11ob2_mom6.nc ${var_name}_${dtg}_analfld ${var_name} ${DATAROOT}
err=$? 
echo " error from ${HOMErtofs}/ush/rtofs_glo2d_ice.sh =" $err

@zgarraffo

@zgarraffo you can try ⬆️ using code add/modified via branch: issue_99 which is added via this PR.

Note:
Once merged into develop, I will delete branch: issue_99 (and you can use develop branch).

@zgarraffo

@zgarraffo, @DanIredell-NOAA Can one of you please take upon the other (than sea ice) pieces of incremental update?

creation of ice nc file through exrtofs_glo_ice_update.sh, Santha's python scripts:
I think there is no renormalization in the scripts. The ncoda icecov file has values between -999 and 100 (ice coverage in percentage) . To input in cice6 (or cice4), the ice coverage has to be renormalized (multiply by 0.01) to to have a maximum of 1 for input in cice, and the -999 numbers over land replaced by another bad number.

In addition, I found a bug that was always in RTOFS that doesn't affect the result. I also have this bug in my ice conversion script:
In multiplying by 0.01 to obtain the ice coverage from the ncoda file, the bad value for ncoda should be -999 and not 999.
The multiplication by 0.01 should be done where values are different from -999, and when they are -999, should be replaced by a bad number, which in hycom is 2^100.

The result is that values over land in RTOFS v2.5 (production) are -9.99 while for hycom should be 2^100. Since the bug affects values on land, it doesn't affect the results of the run, I am pretty sure. Ocean and ice values are good.

creation of ice nc file through exrtofs_glo_ice_update.sh, Santha's python scripts: I think there is no renormalization in the scripts.

False! It is done here.

The ncoda icecov file has values between -999 and 100 (ice coverage in percentage) . To input in cice6 (or cice4), the ice coverage has to be renormalized (multiply by 0.01) to to have a maximum of 1 for input in cice, and the -999 numbers over land replaced by another bad number.

Of course, I know that!
When we meet on Tue, we have to look into why you have the scaling factor not being applied. See my numbers that I posted in the PR. Copy/pasted for your convenience:

Summary for sic: Min=0.0000, Max=1.0000

And

        float sic(ny, nx) ;
                sic:_FillValue = -9.99e+33f ;
                sic:units = "fraction" ;
                sic:long_name = "Ice Concentration" ;
                sic:scale_factor_applied = 0.01 ;

In addition, I found a bug that was always in RTOFS that doesn't affect the result. I also have this bug in my ice conversion script: In multiplying by 0.01 to obtain the ice coverage from the ncoda file, the bad value for ncoda should be -999 and not 999. The multiplication by 0.01 should be done where values are different from -999, and when they are -999, should be replaced by a bad number, which in hycom is 2^100.

The result is that values over land in RTOFS v2.5 (production) are -9.99 while for hycom should be 2^100. Since the bug affects values on land, it doesn't affect the results of the run, I am pretty sure. Ocean and ice values are good.

In today's world, we call them "Fill_Values", and its also included, see this.

Again we have to see how you are running the code pulled from branch https://github.com/NOAA-EMC/RTOFS_GLO/blob/issue_99/ since my tests contradict everything ⬆️.

To be 💯 sure of my ⬆️ assertion, I re-ran the same test case: test.sh as I did for the PR and see below plot: all values are bounded within [0, 1].

@zgarraffo

Based on #98 (comment) from @zgarraffo

@zgarraffo

@zgarraffo I re-read:

creation of ice nc file through exrtofs_glo_ice_update.sh,

❓ How that script worked for you! After I saw your usage of it, I added 2 more arguments to it, without them it should throw an error! If you want to try again, please go ahead!

Anyway, for my test, I used the rtofs_glo2d_ice.sh directrly (see above test.sh) instead of calling via exrtofs_glo_ice_update.sh. In production, exrtofs_glo_ice_update.sh will used, so I'm 😄 your test made me look into it deeper and solidify it.

@zgarraffo

@zgarraffo Hopefully on/by Tuesday 01/13 you can commit a new script. Let me know if you get stuck somewhere.

I cloned and compiled.

Since it is not decided how the MOM6 restart files will be named and in what directory they will be, I am the present names as in my runs.

Since it is not decided how the MOM6 restart files will be named and in what directory they will be, I am the present names as in my runs.

Please use whatever you got.
But make sure the script gives the answer that looks like what you got on Hera/Ursa, if not the same.

In branch ncoda_ufs_cycle
updated sorc/hycom_tools_for_mom6.fd/ncodaz_inc2mom6nc_glb_lyr.f
incremental update velocities should be re-staggered (not done)
added parm/rtofs_glo_ncoda_inc2mom6nc_lyr.input

modified scripts/exrtofs_glo_ncoda_inc.sh
Not verified by running it.

Not verified by running it.

Who will do that?

I can do it tomorrow or so. You asked me to commit in whatever state things are.

I can do it tomorrow or so. You asked me to commit in whatever state things are.

👍

There is no need for manual notification of commits; GitHub does it automatically.

The script for creating the incremental update file in the production environment now works (after talking with Dan)

Resolved via #107

Navigation Menu

Repository navigation

Update the scripts that handle NCODA output(s), i.e., increments that are passed to the model. #98

25 remaining items

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Update the scripts that handle NCODA output(s), i.e., increments that are passed to the model. #98

Description

Activity

zgarraffo commented on Dec 30, 2025

sanAkel commented on Dec 31, 2025

zgarraffo commented on Dec 31, 2025

sanAkel commented on Dec 31, 2025

sanAkel commented on Jan 2, 2026

sanAkel commented on Jan 3, 2026

zgarraffo commented on Jan 4, 2026

sanAkel commented on Jan 4, 2026

sanAkel commented on Jan 4, 2026

25 remaining items

sanAkel commented on Jan 9, 2026

zgarraffo commented on Jan 9, 2026

zgarraffo commented on Jan 9, 2026

sanAkel commented on Jan 10, 2026

zgarraffo commented on Jan 12, 2026

zgarraffo commented on Jan 12, 2026

sanAkel commented on Jan 12, 2026

zgarraffo commented on Jan 13, 2026

sanAkel commented on Jan 13, 2026

zgarraffo commented on Jan 14, 2026

sanAkel commented on Feb 10, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions