wgrib2/3.6.0 in spack-stack-1.9.2 does not support NetCDF output option

[BijuThomas-NOAA]

Description

The wgrib2(Ursa: /contrib/spack-stack/spack-stack-1.9.2/envs/ue-oneapi-2024.2.1/install/oneapi/2024.2.1/wgrib2-3.6.0-nu5xmwz/bin/wgrib2) fails when I choose a NetCDF output option.

To Reproduce

On Ursa:
cd /scratch3/HFIP/hwrfv3/scrub/Biju.Thomas/wgrib2_test
module use /contrib/spack-stack/spack-stack-1.9.2/envs/ue-oneapi-2024.2.1/install/modulefiles/Core
module use /contrib/spack-stack/spack-stack-1.9.2/envs/ue-oneapi-2024.2.1/install/modulefiles/intel-oneapi-mpi/2021.13-haww6b3/gcc/12.4.0
module load stack-oneapi/2024.2.1
module load stack-intel-oneapi-mpi/2021.13
module load wgrib2/3.6.0

wgrib2 ./gfs.t00z.pgrb2.0p25.f006 -match ':TMP:surface' -netcdf ./gfs_global_2024092400_f006.nc
netcdf package not installed

Expected behavior

wgrib2 in the previous stack version supports NetCDF output option

System, compiler, code, ...

Ursa, Gaea C6, ..

Additional context

Add any other context about the problem here.

[BinLiu-NOAA]

Wondering if this can/will be fixed/addressed relatively easily/soon. We were thinking of using spack-stack/1.9.2 for 2025 HAFS real-time parallel experiments. Or, we will have to stick with an earlier installation. Thanks!

[rickgrubin-noaa]

Wondering if this can/will be fixed/addressed relatively easily/soon. We were thinking of using spack-stack/1.9.2 for 2025 HAFS real-time parallel experiments. Or, we will have to stick with an earlier installation. Thanks!

Ideally this week, pending other work. Apologies for the delay.

[BinLiu-NOAA]

@rickgrubin-noaa, Thanks your quick response! Much appreciated!

[rickgrubin-noaa]

@rickgrubin-noaa, Thanks your quick response! Much appreciated!

ursa

module use /contrib/spack-stack/spack-stack-1.9.2/envs/ue-oneapi-2024.2.1/install/modulefiles/Core
module use /contrib/spack-stack/spack-stack-1.9.2/envs/ue-oneapi-2024.2.1/install/modulefiles/intel-oneapi-mpi/2021.13-haww6b3/gcc/12.4.0
module load stack-oneapi/2024.2.1
module load stack-intel-oneapi-mpi/2021.13
module load wgrib2/3.6.0

cp /scratch3/HFIP/hwrfv3/scrub/Biju.Thomas/wgrib2_test/gfs.t00z.pgrb2.0p25.f006 .

wgrib2 ./gfs.t00z.pgrb2.0p25.f006 -netcdf ./binliu.nc
1:0:d=2024092400:PRMSL:mean sea level:6 hour fcst:
2:998434:d=2024092400:CLMR:1 hybrid level:6 hour fcst:
3:1090512:d=2024092400:ICMR:1 hybrid level:6 hour fcst:
4:1322978:d=2024092400:RWMR:1 hybrid level:6 hour fcst:
5:1576482:d=2024092400:SNMR:1 hybrid level:6 hour fcst:
6:1647427:d=2024092400:GRLE:1 hybrid level:6 hour fcst:
7:1696103:d=2024092400:REFD:1 hybrid level:6 hour fcst:
8:2544184:d=2024092400:REFD:2 hybrid level:6 hour fcst:
9:3392769:d=2024092400:REFC:entire atmosphere:6 hour fcst:
10:4319356:d=2024092400:VIS:surface:6 hour fcst:
11:5021905:d=2024092400:UGRD:planetary boundary layer:6 hour fcst:
12:5620199:d=2024092400:VGRD:planetary boundary layer:6 hour fcst:
13:6218973:d=2024092400:VRATE:planetary boundary layer:6 hour fcst:
[...]

[BinLiu-NOAA]

@rickgrubin-noaa, Thanks a lot for fixing this on Ursa! I repeated the above commands, which worked fine for me now.

Meanwhile, if you have not yet done so, could you please help to check and fix this on other platforms which installed this spack-stack/1.9.2 version (e.g., Hera, Orion, Hercules, Gaeac6, etc.)? Much appreciated!

Note: I do need to use the -nc4 option like below though (otherwise, the file size will exceed the classic netcdf file size limit ~2.4GB):
wgrib2 ./gfs.t00z.pgrb2.0p25.f006 -nc4 -netcdf ./gfs.t00z.pgrb2.0p25.f006.nc

[BinLiu-NOAA]

@rickgrubin-noaa, if/when you get a chance to fix this issue on other platforms (e.g., Hera, Orion, Hercules, Gaeac6, which installed this spack-stack/1.9.2 version as well), could you please let @BijuThomas-NOAA and myself know so that we can quickly test/try it from our end? Thanks! Much appreciated!

[BinLiu-NOAA]

Sorry to follow-up again on this thread. We are waiting for this fix so that we can kick off our UFS-HAFS 2025 HFIP real-time parallel experiments which need start before or on August 1st. Since @rickgrubin-noaa might be busy with other high priority work, wondering if @AlexanderRichert-NOAA or @RatkoVasic-NOAA (or someone else) can help to fix this issue on Hera, Orion, Hercules, and Gaeac6, etc. so that we can test and kick off HAFS real-time parallel experiments on time. Much appreciated!