Change model_data to model

[KateFriedman-NOAA]

By NCO's request, model_data should be just model in COM paths.

The following instances are observed within the GSI:

sorc/gsi_enkf.fd/regression/global_enkf.sh:     $nln $datens/$memchar/model_data/atmos/history/${prefix_ens}.atmf00${fhr}.nc sfg_${global_adate}_fhr0${fhr}_${memchar}
sorc/gsi_enkf.fd/regression/global_enkf.sh:            $nln $datens/$memchar/model_data/atmos/history/${prefix_ens}sfcf00${fhr}.nc sfgsfc_${global_adate}_fhr0${fhr}_${memchar}
sorc/gsi_enkf.fd/regression/global_enkf.sh:    $nln $datens/ensstat/model_data/atmos/history/${prefix_ens}.atmf00${fhr}.ensmean.nc sfg_${global_adate}_fhr0${fhr}_ensmean
sorc/gsi_enkf.fd/regression/global_4denvar.sh:dathis=$COMROOTgfs/$dumpges.$PDYg/${cycg}/model_data/atmos/history
sorc/gsi_enkf.fd/regression/global_4denvar.sh:$nln $datens/ensstat/model_data/atmos/history/${prefix_ens}.sfcf006.ensmean.nc         ./sfcf06_anlgrid
sorc/gsi_enkf.fd/regression/global_4denvar.sh:  $nln $datens/$member/model_data/atmos/history/$sigens ${ENS_PATH}sigf${fh}_ens_${member}

Companion issue in workflow: NOAA-EMC/global-workflow#2686

[RussTreadon-NOAA]

Thank you @KateFriedman-NOAA for opening this GSI issue. Changes are needed to GSI scripts and test case data directories.

Attention @ShunLiu-NOAA , @hu5970 , and @CoryMartin-NOAA

[RussTreadon-NOAA]

Work for this issue will be done in RussTreadon-NOAA:feature/model.

[RussTreadon-NOAA]

regression/regression_var.sh specifies the following paths for directory which contains data to run global and regional GSI regression (ctests):

• Gaea: export casesdir="/gpfs/f5/ufs-ard/world-shared/GSI_data/CASES/regtest"
• WCOSS2: export casesdir="/lfs/h2/emc/da/noscrub/russ.treadon/CASES/regtest"
• Orion & Hercules: export casesdir="/work/noaa/da/rtreadon/CASES/regtest"
• Hera: export casesdir="/scratch1/NCEPDEV/da/Russ.Treadon/CASES/regtest"
• Jet: export casesdir="/lfs1/NESDIS/nesdis-rdo2/David.Huber/save/CASES/regtest"
• Discover: export casesdir="/discover/nobackup/projects/gmao/obsdev/wrmccart/NCEP_regression/CASES"

The following has been done on WCOSS2 (Cactus and Dogwood), Orion (Hercules), and Hera:

1. cd $casesdir/gfs/prod/enkfgdas.20240222/18
2. in each directory ensstat, mem001, mem002, up to mem010
   • mv model_data model
   • ln -fs model model_data
3. cd cd $casesdir/gfs/prod/gdas.20240222/18
   • mv model_data model
   • ln -fs model model_data

Once PR #762 is merged into develop, the link model_data will be removed from the WCOSS2, Orion/Hercules, and Hera casesdir.

We need to also mv model_data model on Gaea, Jet, and Discover. @DavidHuber-NOAA and @DavidBurrows-NCO , can you make this change on Jet and Gaea, respectively, when the time comes?

@ShunLiu-NOAA, @hu5970 , @CoryMartin-NOAA , and @wx20jjung : Do we need to maintain a GSI ctest capability on Discover? Who runs the GSI on Discover?

[wx20jjung]

Russ, Thanks for the note. I no longer use Discover. Jim

…

On Mon, Jun 24, 2024 at 1:09 PM RussTreadon-NOAA ***@***.***> wrote: regression/regression_var.sh specifies the following paths for directory which contains data to run global and regional GSI regression (ctests): - Gaea: export casesdir="/gpfs/f5/ufs-ard/world-shared/GSI_data/CASES/regtest" - WCOSS2: export casesdir="/lfs/h2/emc/da/noscrub/russ.treadon/CASES/regtest" - Orion & Hercules: export casesdir="/work/noaa/da/rtreadon/CASES/regtest" - Hera: export casesdir="/scratch1/NCEPDEV/da/Russ.Treadon/CASES/regtest" - Jet: export casesdir="/lfs1/NESDIS/nesdis-rdo2/David.Huber/save/CASES/regtest" - Discover: export casesdir="/discover/nobackup/projects/gmao/obsdev/wrmccart/NCEP_regression/CASES" The following has been done on WCOSS2 (Cactus and Dogwood), Orion (Hercules), and Hera: 1. cd $casesdir/gfs/prod/enkfgdas.20240222/18 2. in each directory ensstat, mem001, mem002, up to mem010 - mv model_data model - ln -fs model model_data 3. cd cd $casesdir/gfs/prod/gdas.20240222/18 - mv model_data model - ln -fs model model_data Once PR #762 <#762> is merged into develop, the link model_data will be removed from the WCOSS2, Orion/Hercules, and Hera casesdir. We need to also mv model_data model on Gaea, Jet, and Discover. @DavidHuber-NOAA <https://github.com/DavidHuber-NOAA> and @DavidBurrows-NCO <https://github.com/DavidBurrows-NCO> , can you make this change on Jet and Gaea, respectively, when the time comes? @ShunLiu-NOAA <https://github.com/ShunLiu-NOAA>, @hu5970 <https://github.com/hu5970> , @CoryMartin-NOAA <https://github.com/CoryMartin-NOAA> , and @wx20jjung <https://github.com/wx20jjung> : Do we need to maintain a GSI ctest capability on Discover? Who run the GSI on Discover? — Reply to this email directly, view it on GitHub <#761 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AMPASA3YRAEOEGS4R6QDCWLZJBHD7AVCNFSM6AAAAABJ2A563CVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCOBXGAZTGNRZGM> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[DavidBurrows-NCO]

@RussTreadon-NOAA No problem. I can take care of Gaea when the time comes.

[DavidHuber-NOAA]

@RussTreadon-NOAA /lfs1 is no longer a valid mount on Jet, so the cases directory will need to be recreated elsewhere. I will rsync the data from Hera into /lfs5/NESDIS/nesdis-rdo2/David.Huber/save/CASES/regtest and notify when complete.

[DavidHuber-NOAA]

@RussTreadon-NOAA The data have been synced from Hera to Jet and the links with them. I have made changes to regression_var.sh and will open a PR to feature/model shortly.

[DavidBurrows-NCO]

Morning @RussTreadon-NOAA. I see this PR was closed. Gaea is down today, so I will get those paths updated when it's back and let you know.

[RussTreadon-NOAA]

Thank you @DavidBurrows-NCO . Yes, we can get path changes for Gaea and Jet global regression test data into future GSI PRs.

[DavidBurrows-NCO]

@RussTreadon-NOAA I updated the Gaea directories for GSI regression tests and tested the global runs:

Test project /gpfs/f5/ufs-ard/scratch/David.Burrows/gw_submods/gsi_my_fork/build
    Start 1: global_4denvar
    Start 6: global_enkf
1/2 Test #6: global_enkf ......................   Passed  664.40 sec
2/2 Test #1: global_4denvar ...................   Passed  2042.73 sec

100% tests passed, 0 tests failed out of 2

Total Test time (real) = 2042.79 sec

[RussTreadon-NOAA]

Thank you @DavidBurrows-NCO for updating the global paths on Gaea. I appreciate your running global ctests to confirm that the tests work as they should.