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

19990926: AVN Grids and McIDAS (cont.)

>From: weather <address@hidden>
>Organization: NMSU/NSBF
>Keywords: 199909242206.QAA23500 McIDAS-XCD

Robert,

On Friday, you noted:

>>I seem to be missing a lot of AVN grids lately on McIDAS
>>(data outside North America) that is present under GEMPAK.

and I responded:

>I have noticed this on our system as well.

I have been looking into the problem on our ingest machine and note
that the cause for me was an old gbtbpds001.av1 file in the working
directory for XCD decoders.  This was a strange one for me since it was
correct a few weeks ago.  I just found that my gbtbpds001.av1 file
still had the aviation grids labeled as MRF.  This caused, among other
things, thinned grids to be filed in the wrong (read as unexpected)
GRID files on adde.unidata.  I just switched the file for the correct
one, so I should start seeing AVN grids filed in AVN GRID files with
this evenings run.  I'll let you know if this was not the problem for
me.

Your problem seems to be different.  I logged onto psnldm and note
that your gbtbpds001.av1 file is the correct one and you are decoding
at least some of the thinned grids (I didn't do an inventory, so I
don't know if/how many you are missing).

Can you describe which GRIDS are not getting processed by McIDAS but
are available through GEMPAK?  Is the decoding failure consistent?

If there is some sort of sporadic problem, the only conjecture I can
make is that the data are coming in faster than dmgrid.k can read the
grib products out of HRS.SPL.  If the data wrapped in HRS.SPL (i.e. if
the filling of the file went past the point where dmgrid.k was
reading), then grids would be lost.  If this really is the problem, you
will need to use a larger spool file.  This is controlled in the file
~mcidas/workdata/HRS.CFG:

SPOOLENG=16          # size of the spool file to use in megabytes

If 16 MB is not big enough, I would start by doubling SPOOLENG.
Remember that you make modifications to HRS.SPL, you will have to stop
ingebin.k and let it get restarted so that it will reread the
configuration options.

Please let me know if the above helps.

Tom