20030714: dcgrib help

Seth,

Use dcgrib2 (use of dcgrib is depricated).
I unzipped your file, then ran:
cat pgbexample | dcgrib2 -v 1 -d - YYYYMMDDHH_###_@@@.gem
The above decoded 140 grids on NCEP grib #002 (2.5x2.5 degree global).

Or....you can run nagrib, specifying the ncarncep1.tbl reanalysis parameter
tables as shown below.

 GBFILE    GRIB data file name               pgbexample
 INDXFL    GRIB index file name               
 GDOUTF    Output grid file                  testme.gem
 PROJ      Map projection/angles/margins|dr   
 GRDAREA   Area covered by grid               
 KXKY      Number of grid points in x;y       
 MAXGRD    Maximum number of grids           200
 CPYFIL    Grid file whose navigation is to  gds
 GAREA     Graphics area                     dset
 OUTPUT    Output device/filename            t
 GBTBLS    Input GRIB decoding tables        
ncarncep1.tbl;ncarncep1.tbl;vcrdgrib1.tbl
 GBDIAG    GRIB diagnostic elements           
 PDSEXT    PDSEXT    Y or N, add PDS extens  NO
 Parameters requested: GBFILE,INDXFL,GDOUTF,PROJ,GRDAREA,KXKY,MAXGRD,CPYFIL,
 GAREA,OUTPUT,GBTBLS,GBDIAG,PDSEXT.
 GEMPAK-NAGRIB>


In general, using nagrib on the reanalysis files would be preferable since you 
can
specify which parameters tables you want to use.

Steve Chiswell


>From: "seth.mcdowell" <address@hidden>
>Organization: UCAR/Unidata
>Keywords: 200307142007.h6EK7jLd009408

>Hello, I am a student at Western Kentucky University. Our center has gempak 
>for us to use, and we have some NCEP-NCAR reanalysis data sets. I tried 
>converting these with dcgrib, but to no avail. The grib files are not supposed
>  
>to be in gaussian mode, but perhaps they are?
>
>http://geocities.com/sethmcdoogle/pgb.zip is an example file.
>
>Thanks in advance,
>Seth McDowell
>