RE: decoding the UKMET (fwd)

• To: "Ross,Charles H - PGPW-5" <cxross@xxxxxxx>
• Subject: RE: decoding the UKMET (fwd)
• From: Robb Kambic <rkambic@xxxxxxxxxxxxxxxx>
• Date: Wed, 13 Jun 2007 13:20:16 -0600 (MDT)

Charles,

the UKMET data can be decoded with the gribtonc decoder. Since it's a thin

grid the program gribtocdl has problems getting the cdl correct so i
created one from scratch, hopefully it's correct. it might need some minor
mods, i'll attach the cdl ukmet. This is the command line i use to decode
the data.

gribtonc -q "lin,dlat=1.5,dlon=1.5" ukmet.cdl ukmet.nc < UKMET_20070613
_1200.grib1


Robb...



---------- Forwarded message ----------
Date: Tue, 29 May 2007 07:03:12 -0700
From: "Ross,Charles H - PGPW-5" <cxross@xxxxxxx>
To: decoders@xxxxxxxxxxxxxxxx
Subject: RE: decoding the UKMET

I’m rephrasing my question…



Is anyone out there currently (successfully) decoding the UKMET model off
the Noaaport using either the gribtonc or the unidata java decoder?



Thanks for taking the time to check…



If this is the wrong forum to be asking this question, please direct me to
the proper unidata forum.  Thanks…



Charles



________________________________

From: owner-decoders@xxxxxxxxxxxxxxxx [mailto:owner-decoders@xxxxxxxxxxxxxx
du] On Behalf Of Ross,Charles H - PGPW-5
Sent: Wednesday, May 23, 2007 11:41 AM
To: decoders@xxxxxxxxxxxxxxxx
Subject: decoding the UKMET



I’m having trouble creating a valid CDL (thus netCDF) for the UKMET
.



Here’s my pqact.conf entry:



HDS     ^H..... EGRR ([0-3][0-9])([0-2][0-9])([0-5][0-9])

        FILE    /usr/local/models/(\1:yyyy)(\1:mm)\1\2_uk.grib





When I run a simple gribtocdl 2007052212_ukmet.grb > uk.cdl comman
d I continually get this as my dimensions in the file:



dimensions:

        record = UNLIMITED ;   // (reference time, forecast time)

        lat = 73 ;            // latitude

        lon = -1 ;            // longitude

        level = 12 ;           // isobaric levels

        time_len = 21 ;        // string length for datetime strings

        valtime_offset = 13 ;   // number of offset times

        nmodels = 4 ;          // number of models

        ngrids = 9 ;           // number of grids

        nav = 1 ;              // for navigation

        nav_len = 100 ;        // max string length for navigation strin
gs



Notice lon=-1.



Further down in the data section it tells me that:

 lon =-30.00 ;



I believe lon should equal 144 but obviously something is not working.



We take this same data and run it through the gempak decoders without a pro
blem.  Anyone have any ideas what is going on with the gribtocdl program on
this grib file?  I’ve run lots of  other model data through the gr
ibtocdl program and never had a problem such as this.



Also, the java grib decoder has issues with this as well…



[ldm@metbox4 ukmet]$ java -Xmx512m ucar.nc2.iosp.grib.Grib2Netcdf
2007052212_ukmet.grb uk_test.nc

Wed May 23 18:32:59 GMT 2007 ... Start of Grib2Netcdf

read grib file=2007052212_ukmet.grb write to netCDF file=uk_test.nc

Using default table:resources/grib/tables/wmo_2_v3.tab (74:-1:-1)

Exception in thread "main" java.lang.NullPointerException

        at ucar.nc2.Variable.setDimensions(Variable.java:1132)

        at ucar.nc2.NetcdfFileWriteable.addVariable(NetcdfFileWriteable.ja
va:335)

        at ucar.nc2.FileWriter.writeToFile(FileWriter.java:158)

        at ucar.nc2.FileWriter.writeToFile(FileWriter.java:65)

        at ucar.nc2.iosp.grib.Grib2Netcdf.main(Grib2Netcdf.java:87)





Thanks for any help on this one..



Charles