[McIDAS #JEY-161022]: GOES ingest during Rapid Scan Operations

Hi Fred,

re:
> During yesterday's (Nov. 17) Rapid Scan Operations, the GOES Northern 
> Hemisphere
> (EAST/NH) data were missing from the GOES data base (EAST/NH).  In looking 
> into the
> problem, I noticed that NESDIS has changed the Rapid scanning strategy to 
> include two
> South America sectors (South America Central [at XX:04] and South America 
> South [at
> XX:34]) to allow South American coverage every hour rather than every 3 hours 
> as
> previously during Rapid Scan.  These new South American sectors were captured 
> in the
> EAST/OTHER data set.  However, the northern hemisphere scans were also in the 
> EAST/OTHER
> data sets rather than the EAST/NH as would be expected.

re:
> What determines which which ingests go into which ADDE descriptor file?

The (old and now unsupported) UW/SSEC SDI package we are using to do
our GOES GVAR ingest and data serving classifies sectors into ADDE
descriptors based on coverage definitions, so the question boils down
to whether or not the file defining the coverages for the various ADDE
dataset descriptors is correct on goeseast.unidata.ucar.edu.

A quick comparison of the SDI installation on the machine that was
acting as goeseast.unidata.ucar.edu until Monday afternoon and the
machine that is now functioning as goeseast.unidata.ucar.edu and the
machine that acts as goeswest.unidata.ucar.edu shows that the coverage
definition file on the current 'goeseast' is different than those on
the other two machines.  This strongly suggests that we are using an an
incorrect definition file on the current 'goeseast', and this
is the cause of the problem.

I just updated the coverage file on 'goeseast' to match the setup on our
previous machine.  We will not know if this fixes the problem until
we go through another RSO.

re: 
> Does NESDIS have a data flag at the
> beginning of the ingest which tells which data set is being transmitted.
> What I am driving at is who is to blame for this problem?
> Is it NESDIS not putting in the proper
> documentation into the rapid scan image headers?  Is it you for not having 
> the ingestor
> file the data properly? Or is it me for not having the program logic to look 
> into the
> EAST/OTHER files for northern hemisphere data?

My guess is that the cause of the problem was an out of date coverage definition
file on goeseast.unidata.ucar.edu.

re:
> I will have to change my logic for the
> global mosaics to include these new South American sectors, but having the 
> northern
> hemisphere data in the /OTHER makes the program logic more difficult. I 
> currently
> generate the global mosaics on a 30 minute schedule.  The OTHER had some new 
> data every
> 15 minutes when the northern hemisphere in in /OTHER. If the northern 
> hemisphere data
> could be in /NH it would make my life a lot easier.

Hopefully, use of the same coverage definition file that is on our other two SDI
machines will correct the mistaken classifications that occurred yesterday.
  
Sorry for the problems!

Cheers,

Tom
--
****************************************************************************
Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
----------------------------------------------------------------------------
Unidata HomePage                       http://www.unidata.ucar.edu
****************************************************************************


Ticket Details
===================
Ticket ID: JEY-161022
Department: Support McIDAS
Priority: Normal
Status: Closed