RE: Grib-to-NetCDF problems

• To: Janßen Holger <Holger.Janssen@xxxxxx>
• Subject: RE: Grib-to-NetCDF problems
• From: Robb Kambic <rkambic@xxxxxxxxxxxxxxxx>
• Date: Thu, 26 May 2005 16:40:30 -0600 (MDT)

On Wed, 25 May 2005, [iso-8859-1] Janßen Holger wrote:

> >> Hi,
> >>
> >> i currently try to read existing Grib-1 and Grib-2 data and
> >> create files in the NetCDF format from these data. I ran
> >> into some problems:
> >>
> >> 1.) Most of my Grib files are not supported yet, as the center
> >> etc. is not configured (missing table entries). I tried to add
> >> the tables myself, but then i got an IOException:
> >>
> >> GribServiceProvider open 2004092000_000_2011_200.grib
> >
> > hi,
> >
> > could you make available or attach 2004092000_000_2011_200.grib
> > if it's only a couple of megs?
> > not knowing 2004092000_000_2011_200.grib format it's
> > impossible to answer your question.
>
> Hi Robb,
>
> i made 2 files available (both quite small in size). The 1st should be a
> standard GME grib file. As i already wrote, the code table etc. is not
> available in grib.jar, but i just added the table and configured it in the
> tablelookup.lst with:
> 78: -1: 2: dwdtab_2.tab.

hi,

could you send me your table dwdtab_2.tab so that i can include it with
the other tables. also, i got the grib files.

thanks,
robb...


>
> 1.)
> http://www.o2000.de/grib/2004092000_000_2011_200.grib
>
> WGrib output:
> 1:8:D=2004092000:TMP:2 m above gnd:kpds=11,105,2:12hr fcst:"Temp. [K]
>
> The 2nd file is a ECMWF file which i thought, should be readable/parsable
> without any further work.
>
> 2.)
> http://www.o2000.de/grib/ecmwf_00-003.grib
>
> WGrib output:
> 1:0:D=2005041400:LSP:sfc:kpds=142,1,0:anl:ECMWF stream?:"Large scale
> precipitation m
>
> > the source code is still in beta development so any bugs
> > that you find would be helpfull to me to know about. there
> > are some extra files in the tar, don't be concerned about
> > them. the code is GribJava.tar.gz and it is located at:
>
> Thank you very much for the source code. Seems like this is alrady an
> updated version of the Grib sources, because now i dont get these
> Exceptions, but the message "java.io.IOException: no valid products were
> found". Currently i'm trying to find out, whats going wrong with our data.
> It is also possible for me, to support the Grib-to-NetCDF process in the
> next time. I hope, i cant deliver some input and maybe help debugging it
> after i got a better idea of how the whole process currently works.
>
> > in the top dir there is a run file with examples on how to run the grib
> > decoders programs. the getData needs to get the offset, ... other
> > parameters from the index to get the data. there's a IndexFormat.txt file
> > with a explanation of the index format. these are the routine that nj22
> > uses in it formulation of netcdf file.
>
> I have a look at these information. At the moment it look like our gribs to
> not contain all meta information that are needed.
>
>
> Best regards,
>  Holger
>
>

==============================================================================
Robb Kambic                                Unidata Program Center
Software Engineer III                      Univ. Corp for Atmospheric Research
rkambic@xxxxxxxxxxxxxxxx                   WWW: http://www.unidata.ucar.edu/
==============================================================================