The IC files are not copyied from its source location to target location

[emilyhcliu]

What is wrong?

The IC files are not copied from their source location to the target location while running the setup script - setup_expt.py The COMROT is created, but it is empty.

HOMEgfs=/work/noaa/da/eliu/ORION/Global-Workflow/global-workflow
workdir=/work2/noaa/da/eliu/ORION/para
ICSDIRBase=/work2/noaa/da/eliu/ICs/UFO_eval/data/para
ICSDIR=/work2/noaa/da/acollard/UFO_eval/data/para/output_ufo_eval_feb2024_9Aug
IDATE=2024021900
EDATE=2024021906
RESDETATMOS=768
CDUMP=gdas
PSLOTBase=gdas_eval_2024
PSLOT=${PSLOTBase}_conv_ps_JEDI
PSLOT=${PSLOTBase}_iasi_JEDI
CONFIGDIR=$HOMEgfs/parm/config/gfs
COMROOT=$workdir/comroot
EXPDIR=$workdir//expdir

$HOMEgfs/workflow/setup_expt.py gfs cycled --idate $IDATE --edate $EDATE --app ATM \
    --start warm --gfs_cyc 0  --resdetatmos $RESDETATMOS  --nens 0 \
    --run gdas --pslot ${PSLOT} --configdir $CONFIGDIR \
    --comroot $COMROOT --expdir $EXPDIR --icsdir $ICSDIR

What should have happened?

The IC files should be copied over to the target location (ROTDIR) from the source location (ICSDIR)

Steps to reproduce

1. Get a set of ICs for GFS (define and create your ICSDIR for step 2)
2. run the setup script - setup_expt.sh with the input arguments (see example above)
3. check the ROTDIR

Bugzilla issue

No response

Additional information

Nonw

Do you have a proposed solution?

Not yet. There are new updates related to the stage of initial conditions, but I have not looked at them in detail.

[emilyhcliu]

@KateFriedman-NOAA I'm not sure who I should assign this issue to.

[RussTreadon-NOAA]

@emilyhcliu , I believe the behavior you are observing is due to PR #2651. This PR was merged into develop yesterday. @KateFriedman-NOAA sent an email with the subject line [Global Workflow] Initial conditions staging job refactor.

[KateFriedman-NOAA]

@emilyhcliu To add to @RussTreadon-NOAA 's comment, the PR that went into develop yesterday to refactor the staging job for ICs removed the IC staging/linking from the setup_expt.py step. So what you are seeing is intentional. When you set up a new experiment you should now see one or more stage_ic jobs in your xml which will either run in the first half cycle (e.g. gdasstage_ic and enkfgdasstage_ic) or at the beginning of each forecast-only cycle (e.g. gfsstage_ic or stage_ic for gefs). The new stage_ic jobs will run first and pull staged ICs into your COMROT. If you encounter any issues with the refactored stage_ic jobs, please open an issue to let us know, thanks!

[KateFriedman-NOAA]

@emilyhcliu Also, I see you are setting ICSDIR in your setup: ICSDIR=/work2/noaa/da/acollard/UFO_eval/data/para/output_ufo_eval_feb2024_9Aug That path should land in your EXPDIR's config.stage_ic and the stage_ic job will pick them up from there. Ensure that your ICSDIR has the same folder structure as COM, the stage_ic job will expect that. Let me know if you have any questions.

[KateFriedman-NOAA]

Closing. Please feel free to open a new issue if you encounter any errors in the refactored stage_ic jobs. Thanks!

[emilyhcliu]

@emilyhcliu Also, I see you are setting ICSDIR in your setup: ICSDIR=/work2/noaa/da/acollard/UFO_eval/data/para/output_ufo_eval_feb2024_9Aug That path should land in your EXPDIR's config.stage_ic and the stage_ic job will pick them up from there. Ensure that your ICSDIR has the same folder structure as COM, the stage_ic job will expect that. Let me know if you have any questions.

Yes, I do see that in my config.stage_ic.

[emilyhcliu]

Closing. Please feel free to open a new issue if you encounter any errors in the refactored stage_ic jobs. Thanks!

@KateFriedman-NOAA and @RussTreadon-NOAA I got it. I will look into the refactored way. Thanks.