[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: grib2 decoder

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: Re: grib2 decoder
  • Date: Mon, 3 Jan 2005 16:41:52 -0700 (MST)

On Mon, 3 Jan 2005, Jim Cowie wrote:

>
> Thanks for the update Robb. I figured the files were going to become
> very large but I was thinking that in netCDF4 when HDF is the underlying
> data storage layer that compression would help out here. Is that likely
> or am I dreaming?


no that will help a bunch. but some of the new grib2 eta files are almost
400 megs, current expansion would result in 8 gigs. wow  don't know how
netCDF4 will handle that large of a data set. so on demand decoding will
be the most reasonable solution.

robb...


Maybe it is best to just degrib on the fly after all.
>
> -jim
>
>
> Robb Kambic wrote:
> > On Mon, 3 Jan 2005, Jim Cowie wrote:
> >
> >
> >>Hi Robb,
> >>
> >>I saw in one of your emails last month a mention that the new
> >>gribtonc decoder (for grib2) was going to be released pretty soon.
> >>Can you tell me when? We are antsy to try it. Thanks,
> >
> >
> > jim,
> >
> > good news and bad news. the good news is the decoder has been developed
> > and tested, the bad news is the resulting netCDF files are huge. the large
> > file size is the main reason for the delay. the expansion for the files have
> > a range from 1:20 to 1:50  on average. for example:
> >
> > NDFD  12.4 megs -> 626.9 megs
> > eta2  16.3 megs -> 668.5 megs
> >
> > a couple reasons, the netcdf file is not optimized and the more missing
> > values in the data the larger the expansion. also the complex packing
> > scheme permits more variables and levels. john and i have been working
> > together to leave the files in the grib2 format and to extract the data in
> > an on demand process.  to implement this, the grib files are scanned to
> > create an index of products and grid informations. his programs reads
> > the index and presents the information to the user.  it works well. here's
> > an url to down load ToolUI program.
> >
> > http://my.unidata.ucar.edu/content/software/netcdf-java/v2.2/webstart/index.html
> >
> > in a nutshell,
> >
> > click viewer  then open a grib2 file
> > click geogrid  then open the same grid file  geogrid boxes should be
> >    checked
> > click view  to access a sanity type view of the variables.
> >
> > ToolUI currently  doesn't do grids with multiple GDS or thin grids, we are
> > working on those features now.
> >
> >
> > i'll probably be only available by email tomorrow if you have
> > questions then i'll be on PTO for a week.
> >
> > robb...
> >
> >
> >
> >>-jim
> >>
> >>--
> >>Jim Cowie
> >>NCAR/RAL
> >>address@hidden
> >>303-497-2831
> >>
> >
> >
> > ===============================================================================
> > Robb Kambic                            Unidata Program Center
> > Software Engineer III                          Univ. Corp for Atmospheric 
> > Research
> > address@hidden                 WWW: http://www.unidata.ucar.edu/
> > ===============================================================================
>
>
> --
> Jim Cowie
> NCAR/RAL
> address@hidden
> 303-497-2831
>

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