Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

How to get GFS data as initial conditions for MPAS-A(solved)

M

Monica

New member
Hi all,
I have successfully compiled MPAS-A and successfully completed a global simulation following the tutorial practice guide provided on the MPAS-A homepage (MPAS Tutorial — Practice Session Guide). When I was planning another simulation, I was faced with the problem of preparing the initial GFS data and SST data.
I download the GFS data on this page(https://nomads.ncep.noaa.gov/pub/data/nccf/com/gfs/prod/gfs.20221020/12/atmos/), such as gfs.t00z.pgrb2.0p25.f000. Then I changed the file name to GFS:2022-10-20_12. When I try to generate the x1.163842.init.nc file, there are some ERRORs like:

ERROR: Didn't recognize format version of data in GFS:2022-10-20_12.
ERROR: Found version 2 but expected either 3, 4, or 5. This could be an endian problem.
ERROR: ********************************************************************************
ERROR: LANDSEA field not found in meteorological data file GFS:2022-10-20_12
CRITICAL ERROR: ********************************************************************************

The various software versions I use are:
pio-2.5.9
pnetcdf-1.12.2
hdf5-1.10.9
zlib-1.2.11
netcdf-c-4.7.4
netcdf-fortran-4.5.3
MPAS-A-7.3
All of the above was compiled by mpiifort.

Also, There are "namelist.init_atmosphere", "streams.init_atmosphere","log.init_atmosphere.0000.out" as shown in the attachment.

I wonder if I'm downloading the GFS data incorrectly, or if I should do some processing on the GFS data. Or there are other errors？

Thanks
Best，
Monica
 

Attachments

  • log.init_atmosphere.0000.out.txt
    8.6 KB · Views: 4
  • namelist.init_atmosphere.txt
    820 bytes · Views: 5
  • streams.init_atmosphere.txt
    408 bytes · Views: 2
I am using GFS data from the same source as my initial/boundary conditions. I am using WPS to create WPS 'intermediate" data file as described in section 7.2.2 in the MPAS User Guide successfully
 
Monica said:
Thank you for your reply. I did it based on your suggestion.
Thans again.
Click to expand...
If you can change the title of this post from "How to get GFS data as initial conditions for MPAS-A" to "How to get GFS data as initial conditions for MPAS-A (solved)" that would help other future users find the solution if they come across the same problem.
 
We're planning to offer some better guidance on how to create "intermediate" files from GRIB datasets in the near future in the MPAS-Atmosphere User's Guide (as well as in tutorial materials). Since this might save someone from having to go through the WPS documentation and tutorial material, here's a quick summary of what should work:

  1. Clone the WPS repository with
    Code: 
    git clone https://github.com/wrf-model/WPS.git
  2. Configure the WPS to use the compilers that you have on your system (and that you probably have used for building MPAS):
    Code: 
    cd WPS; ./configure --build-grib2-libs --nowrf
    In the configuration menu, choose a "(serial)" option.
  3. Compile just the ungrib component of the WPS, along with supporting GRIB2 libraries:
    Code: 
    ./compile ungrib
  4. Symbolically link the correct "Vtable" into the top-level WPS directory; for GFS data, the correct Vtable is Vtable.GFS:
    Code: 
    ln -s ungrib/Variable_Tables/Vtable.GFS Vtable
  5. Symbolically link the GRIB files to be processed into the top-level WPS directory:
    Code: 
    ./link_grib.csh /path/to/GRIB/data/gfs*
    In general, the arguments to the link_grib.csh script should include all GRIB files to be processed by ungrib.
  6. Set the start_date, end_date, and interval_seconds variables in the namelist.wps file. Only the first start_date and end_date values are relevant for the ungrib program, and the interval_seconds value should match the interval between your GRIB files (if more than one GRIB file is to be processed).
  7. Run ungrib.exe.
The end result of the above steps should be one or more "intermediate" files (with names of the form FILE:yyyy-mm-dd_hh) that can be used by the init_atmosphere_model program to produce MPAS-Atmosphere ICs and LBCs.
 
mgduda said:
We're planning to offer some better guidance on how to create "intermediate" files from GRIB datasets in the near future in the MPAS-Atmosphere User's Guide (as well as in tutorial materials). Since this might save someone from having to go through the WPS documentation and tutorial material, here's a quick summary of what should work:

  1. Clone the WPS repository with
    Code: 
    git clone https://github.com/wrf-model/WPS.git
  2. Configure the WPS to use the compilers that you have on your system (and that you probably have used for building MPAS):
    Code: 
    cd WPS; ./configure --build-grib2-libs --nowrf
    In the configuration menu, choose a "(serial)" option.
  3. Compile just the ungrib component of the WPS, along with supporting GRIB2 libraries:
    Code: 
    ./compile ungrib
  4. Symbolically link the correct "Vtable" into the top-level WPS directory; for GFS data, the correct Vtable is Vtable.GFS:
    Code: 
    ln -s ungrib/Variable_Tables/Vtable.GFS Vtable
  5. Symbolically link the GRIB files to be processed into the top-level WPS directory:
    Code: 
    ./link_grib.csh /path/to/GRIB/data/gfs*
    In general, the arguments to the link_grib.csh script should include all GRIB files to be processed by ungrib.
  6. Set the start_date, end_date, and interval_seconds variables in the namelist.wps file. Only the first start_date and end_date values are relevant for the ungrib program, and the interval_seconds value should match the interval between your GRIB files (if more than one GRIB file is to be processed).
  7. Run ungrib.exe.
The end result of the above steps should be one or more "intermediate" files (with names of the form FILE:yyyy-mm-dd_hh) that can be used by the init_atmosphere_model program to produce MPAS-Atmosphere ICs and LBCs.
Click to expand...
Thank you very much ! It really shed lots of light on my research!
 
You must log in or register to reply here.
Top