grib2 data files in plot-data-plane

[jir245]

Hi, I'm currently trying to use a grib2 data file with plot-data-plane. The variable that I would like to read is an Aerosol Optical Thickness field, which is vertically-integrated, so does not have a level associated with it. When I run plot-data-plane with the grib2 file, how do I set the level information? I've tried level=0, level="", level="(,)" without any luck. I get a" ERROR : VarInfo::set_level_info_grib() - failed to parse level string '(,*)'" error. Any help on this would be greatly appreciated. Thanks!

[jprestop]

Hi @jir245. Thank you for your question. For specifying the level information for a GRIB2 file, please take a look at this section of our documentation searching for "File-format specific settings for the “field” entry" to take you to the correct place. I'll also paste in the information below for easy reference and will bold the “level” entry starting point. Please let us know if this answers your question or if you have any other questions.

GRIB1 and GRIB2:

• For custom GRIB tables, see note about MET_GRIB_TABLES.
• The “name” entry specifies a GRIB code number or abbreviation.
  • GRIB1 Product Definition Section
  • GRIB2 Product Definition Section
• The “level” entry specifies a level type and value:
  • ANNN for accumulation interval NNN
  • ZNNN for vertical level NNN
  • ZNNN-NNN for a range of vertical levels
  • PNNN for pressure level NNN in hPa
  • PNNN-NNN for a range of pressure levels in hPa
  • LNNN for a generic level type
  • RNNN for a specific GRIB record number
• The “GRIB_lvl_typ” entry is an integer specifying the level type.
• The “GRIB_lvl_val1” and “GRIB_lvl_val2” entries are floats specifying the first and second level values.
• The “GRIB_ens” entry is a string specifying NCEP’s usage of the extended PDS for ensembles. Set to “hi_res_ctl”, “low_res_ctl”, “+n”, or “-n”, for the n-th ensemble member.
• The GRIB1_ptv entry is an integer specifying the GRIB1 parameter table version number.
• The GRIB1_code entry is an integer specifying the GRIB1 code (wgrib kpds5 value).
• The GRIB1_center is an integer specifying the originating center.
• The GRIB1_subcenter is an integer specifying the originating subcenter.
• The GRIB1_tri is an integer specifying the time range indicator.
• The GRIB2_mtab is an integer specifying the master table number.
• The GRIB2_ltab is an integer specifying the local table number.
• The GRIB2_disc is an integer specifying the GRIB2 discipline code.
• The GRIB2_parm_cat is an integer specifying the parameter category code.
• The GRIB2_parm is an integer specifying the parameter code.
• The GRIB2_pdt is an integer specifying the product definition template (Table 4.0).
• The GRIB2_process is an integer specifying the generating process (Table 4.3).
• The GRIB2_cntr is an integer specifying the originating center.
• The GRIB2_ens_type is an integer specifying the ensemble type (Table 4.6).
• The GRIB2_der_type is an integer specifying the derived product type (Table 4.7).
• The GRIB2_stat_type is an integer specifying the statistical processing type (Table 4.10).
• The GRIB2_perc_val is an integer specifying the requested percentile value (0 to 100) to be used. This applies only to GRIB2 product definition templates 4.6 and 4.10.
• The GRIB2_aerosol_type is an integer specifying the aerosol type (Table 4.233). This applies only to GRIB2 product definition templates 4.46 and 4.48.
• The GRIB2_aerosol_interval_type is an integer specifying the aerosol size interval (Table 4.91). This applies only to GRIB2 product definition templates 4.46 and 4.48.
• The GRIB2_aerosol_size_lower and “GRIB2_aerosol_size_upper” are doubles specifying the endpoints of the aerosol size interval. These applies only to GRIB2 product defintion templates 4.46 and 4.48.
• The GRIB2_ipdtmpl_index and GRIB2_ipdtmpl_val entries are arrays of integers which specify the product description template values to be used. The indices are 0-based. For example, use the following to request a GRIB2 record whose 9-th and 27-th product description template values are 1 and 2, respectively:
• GRIB2_ipdtmpl_index=[8, 26]; GRIB2_ipdtmpl_val=[1, 2];

[jir245]

Hi, thank you for your answer. I think the challenge that I'm having is that the grib2 data field does not have a level associated with it. I have from that data file:
:Grib2_Level_Type = 10; // int
:Grib2_Level_Desc = "Entire atmosphere";

In this case, I can define the level type with the variable “GRIB_lvl_typ” but I think it needs more information than this?

[JohnHalleyGotway]

@jir245, I'd recommend trying level="Z0"; or level="L0";.

If that doesn't work, please just send along your sample GRIB2 file and we can figure out the details. If it's small enough, you can just zip it and attach it to the discussion. Or if GitHub tells you it's too large, please follow these instructions to send it.

Lastly, can you please let me know the institution with which you work? We've started keeping track of the support we provide by assigning a "requestor" label to each discussion... which helps with future funding.

[jir245]

Thanks again for your reply. That still didn't work for me. Would I be able to email you the data file? I'm from NRL and working with air force data and would rather not post the file here if possible.

[jprestop]

@jir245 Thank you for letting us know. If you'd like to utilize DoD SAFE you can send the file that way. Here is a link that includes my email address: https://ral.ucar.edu/directory/julie-prestopnik

[jir245]

ok, thanks. sent via SAFE

[JohnHalleyGotway]

@jir245 thanks for sending your sample data file.

I ran it through the wgrib2 utility and see that it doesn't know about the metadata for this file:

> wgrib2 US057GMET-GR2mdl.0018_0200_00000F0RL2025010900_0001_000000-000000aero_aod_visib.gr2 
1:0:d=2025010900:var discipline=0 center=57 local_table=1 parmcat=20 parm=203:entire atmosphere:anl:

And if wgrib2 doesn't have a record for this combination of identifiers, it's very likely that the MET GRIB2 tables don't either.

First, the quick and dirty, but definitely not recommended longterm, you can tell plot_data_plane to just plot the first record it finds in the file:

plot_data_plane \
US057GMET-GR2mdl.0018_0200_00000F0RL2025010900_0001_000000-000000aero_aod_visib.gr2 \
plot_R1.ps \
'name="VIS"; level="R1";'

As long as the name VIS appears in at least one MET GRIB2 table (e.g. in share/met/table_files/grib2*), then this will run and simply pull data from the first record in the file (e.g. R1).

Second, the preferred solution is to define a new GRIB2 table entry for this unknown dataset. Look in the ls share/met/table_files/grib2*.txt files for examples.

Let's create a simple one to define metadata for this single variable.
And running wgrib2 with the -varX option make this easier:

wgrib2 -varX US057GMET-GR2mdl.0018_0200_00000F0RL2025010900_0001_000000-000000aero_aod_visib.gr2 
1:0:var0_14_1_57_20_203
#  -varX            inv          raw variable name - discipline mastertab localtab center parmcat parmnum

Here's the file to create:

echo 'GRIB2' > grib2_local.txt
echo '0 14 1 255 57 1 20 203 "AOD_VISIB" "Aerosol Optical Depth" "Numeric"' >> grib2_local.txt

Then set the MET_GRIB_TABLES environment variable to point to this file.

export MET_GRIB_TABLES=`pwd`/grib2_local.txt

And then run plot_data_plane referencing the new name we just defined in our local table.

plot_data_plane \
US057GMET-GR2mdl.0018_0200_00000F0RL2025010900_0001_000000-000000aero_aod_visib.gr2 \
plot_name.ps \
'name="AOD_VISIB"; level="L0";'

And that works to create this image:

The best longterm solution would be updating the contents of this USAF GRIB2 table file directly in the MET repository to reflect the conventions the USAF uses when creating GRIB2 files. That would prevent the need for creating a custom GRIB2 table and guessing about the metadata. If you're in touch with anyone with the USAF, please encourage them to share updated GRIB2 tables with the METplus development team.

Hope that does the trick!

[jir245]

Thank you so much, this worked for me!

[JohnHalleyGotway]

Great, thanks for confirming. I'll go ahead and close/lock this discussion. But feel free to start a new one if more questions arise.