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

20010404: Custom METARs for plotting


Tim,

The bridge library dcgbul() routine is looking for the FOS style
SOH \r \r \n through \r \r \n ETX characters or AFOS ZCZC to NNNN
sequences as Kevin mentions in order to find a complete bulletin in the
broadcast stream.

All metar reports do have 4 character IDs. The GEMPAK station tables are
backward compatible with 3 character SA format observations such that if
a METAR is from Kxxx or Cxxx, then if the 4 character ID is not found in
the table, then the 3 character ID will be searched for. You can use either
3 or 4 character id's in your station table for METARs from US and Canadian
stations.

Steve Chiswell
Unidata User Support




On Wed, 4 Apr 2001, Kevin R. Tyle wrote:

> (slightly revised version of message previously sent directly to Tim)
> 
> Hi Tim,
> 
> I'm not positive, but I believe the decoder wants things as close to
> "raw" WMO format as possible.  Thus, in addition to the WMO header, it is
> probably also expecting a bulletin sequence number, which is a 3-digit
> number that appears on a line above the WMO bulletin.  Check your raw
> METAR file as it is ingested by the LDM for the format.
> 
> *addition*
> Actually after doing some testing it seems the sequence # is
> not crucial.   The control character issue seems to be the
> important element.
> * *
> 
> I believe too that the decoder needs the ^A and ^C control 
> characters to distinguish bulletin start/stops.  Try adding
> a CTRL-A (ASCII Character 1, or CHAR(1) to the start of 
> your data, and a CTRL-C (ASCII Character 3, or CHAR(3) at the
> end.
> 
> Can you attach your sample input file?
> 
> --Kevin
> 
> ______________________________________________________________________________
> Kevin Tyle, Systems Administrator             ******************************
> Department of Earth and Atmospheric Sciences  Email:  address@hidden
> University at Albany, ES-235                  Phone:  518-442-4571
> 1400 Washington Avenue                                FAX:    518-442-5825
> Albany, NY 12222                                ******************************
> ______________________________________________________________________________
> 
> > Hi,
> > 
> > We are trying to use the gempak decoders to insert station data from local
> > mesonet sites into our surface gempak data file.  We have been following
> > the suggested guidelines for using dcmetr:
> > 
> > dcmetr -c YYMMDD/HHNN -b 9 -m 72 -s sfmetar_sa.tbl  -d dcmetr.log  -e
> > GEMTBL=$GEMTBL/tables  YYYYMMDD_sao.gem < input_file
> > 
> > Stations ID info has been added to the surface station files in $GEMTBL and
> > the data have been coded into correct METAR format.  When the decoder runs,
> > it does not give an indication of an error, but it does not decode the data
> > either.  We have tried sticking WMO headers into the METAR obs input file,
> > tried both 3 and 4 letter IDs, and various other combinations of things.
> > Nothing works.
> > 
> > Soooo.....  here's a few questions we are left with:
> > 
> > 1) Does the = sign at the end of the file do anything
> > 2) does the decoder actually read the WMO header line
> > 3) do we need 3 or 4 letter ids (code seems to truncate 4 letter id down to
> > 3 if K is leading character).
> > 4) has anybody had any luck doing similar decoding??
> > 
> > As always, thanks for your input.
> > 
> > -Tim
> > 
> > 
> > 
 > > ---------------------------------------------------------------------------
> > Tim Doggett, PhD
> > Assistant Professor
> > Texas Tech University
> > Atmospheric Science Group, Department of Geosciences
> > Phone: (806)742-3477
> > Fax: (806)742-1738
> > ---------------------------------------------------------------------------
>