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

19990713: ACARS programs

Steve,

Ok...I repacked the tar file:
http://www.unidata.ucar.edu/packages/gempak/contrib/acars_990810.tar.Z

This is the pqact.conf action I am currently using:
PCWS    ^FSL\.NetCDF\.ACARS\.QC
        PIPE    -close  /usr/local/ldm/decoders/dcacars -v -b 30
        -e GEMTBL=/home/gempak/NAWIPS-5.4/gempak5.4/tables
        -l data/gempak/logs/dcacars.log
        data/gempak/acars/YYYYMMDDHHNN_acars.gem


The "-b 30" flag tells the decoder to bin the observations into 30 minute
files when the NN template is detected. So, I get 1 file for each 30
minute time period. The data within the file remains at 1 minute
observations. The -b time period is configurable....but probably not
necessary or useful to create files containing less than 30 minutes of
observations ince it would result in very few observations for
the composite sounding.

Steve Chiswell
Unidata User Support

On Tue, 10 Aug 1999, Steven Danz wrote:

> Will you be posting a new tar file with the changes? -swd
> 
> Steve Chiswell wrote:
> 
> > Steve,
> >
> > I made some additional mods, to allow the decoder to create 30 minute
> > files if the NN template was used. These seem to be a good size
> > to work with, however, the downside is the current way I have the menu
> > times only from 1 file- makes plotting data across a time file boundary
> > more difficult.
> >
> > Aside from the printing and acars mods I made, most of the other changes
> > were in xwvid6, thermo, and skparams where getting dp=0 or dz=0
> > for acars observations at the same level really made things blow up.
> > Also, since most acars data didn't have any dewpoint, I put in
> > those checks for stability calculations.
> > Other changes primarily are C convention porting problems since we
> > have 12 different OS's to port to.
> >
> > Chiz
> >
> > On Mon, 2 Aug 1999, Steven Danz wrote:
> >
> > > Yeah, things are going pretty well.
> > > I've got the first pass of 'merging' done between the two versions
> > > of nsharp.  The data is coming in OK, and the decoder seems
> > > happy.  Thought I would mention, to give myself some 'room'
> > > I up'ed the value of MMHDRS to 12000 and I've noticed that
> > > some times of the day there are over 11000 reports in an hour.
> > > Thought you might be interested.
> > >
> > > -swd
> > >
> > > Steve Chiswell wrote:
> > >
> > > > Thanks for the note.
> > > >
> > > > Hope you are having otherwise success.
> > > >
> > > > Chiz
> > > >
> > > > On Fri, 23 Jul 1999, Steven Danz wrote:
> > > >
> > > > > Hi
> > > > >
> > > > > This might be an HPUX only problem.  The write_gempak
> > > > > function makes a call to dc_fcyl using the sizeof(prmlist) as the
> > > > > length indicator.  In the other cases of passing a 2
> > > > > dimension array from C to Fortran that I have been able to find,
> > > > > the value should be the length of an element of the array,
> > > > > in this case sizeof(prmlist[0]).  With that change, prmlist gets
> > > > > the expected values on the HP, without nasty memory stomping
> > > > > takes place ;-)
> > > > >
> > > > > Thought you might want to know...
> > > > >
> > > > > -swd
> > > > >
> > >
>