Page tree
Skip to end of metadata
Go to start of metadata



  • Scott Gregory, Clara Draper, Jeff Whitaker will collaborate on 'teaching' GSI to use the netcdf format for input and output.


Presentclara.draperscott.gregoryThomas AuligneGuillaume Vernieresrahul.mahajanJohn MichalakesTara Jensen, ben.rustonYannick Tremolet


GMAO's obs_diag branch will be merged in GSI trunk. UFO work can start before this using the BUFR to NetCDF converter from ESRL but the two formats need to be compatible. The idea for the immediate future is to extend/modify ESRL's converter generate the same NetCDF files as the GSI but with added metadata required for the UFO. The GSI solver should be able to start from the same file format.


  • Make sample NetCDF files available to the group and add link on confluence page (Will.McCarty or rahul.mahajan)
  • Adapt ESRL's BUFR to NetCDF converter to the format of the sample files, for observation types already in the converter (scott.gregory)
  • Make sample files available to MET and post-processing groups for further discussion about a common format (Yannick Tremolet)


Present: clara.draperscott.gregoryThomas AuligneJeff Whitaker , Will.McCarty, John Michalakes

Initial meeting. 

  • No labels


  1. I have attached a first cut at python code that converts conventional and AMSUA bufr files to netcdf, written to the template provided by Will McCarty (Rahul supplied the ftp). The output now has 'nobs_bufr' dimension instead of 'nobs' output from GSI. I am refining the AMSUA code a bit, right now it is partially hard-wired to NOAA15.

  2. I revised the amsua code to convert a bufr file containing multiple satellites to a diag template for each satellite...