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

19990416: access to gridded data other than 211 (fwd)


===============================================================================
Robb Kambic                                Unidata Program Center
Software Engineer III                      Univ. Corp for Atmospheric Research
address@hidden             WWW: http://www.unidata.ucar.edu/
===============================================================================

---------- Forwarded message ----------
Date: Fri, 16 Apr 1999 14:50:50 -0600
From: Unidata Support <address@hidden>
To: "Molenar, Deb" <address@hidden>
    address@hidden
Subject: 19990416: access to gridded data other than 211 


Deb,

Remember that TABS are important in the pqact.conf file. Cutting
and pasting in the mail will replace tab characters with spaces. That
is probably the cause of problems you had in using what I sent you.

If you are only receiving the 211 grids from your upstream site,
then its possible that either you, or your upstream site is filtering
the broadcast. When we transitioned from FOS to NOAAport, many
additional grids were added to the broadcast- and some sites which 
could not handle the additional bandwith chose to restrict what they receive.

Check your "request" line in ~ldm/etc/ldmd.conf and see if you
are filtering what you are requesting by using a patter other than .*.

As for the grib products in the IDD stream, all should have a /m tag
at the end which identifies the model either by name or number.
Other products which are not grib, do not have the tag applied.
All of the HRS  ^[YZ].* should have the /m tags. The NCEP models
will be from KWB. (the 4th character varies depending on the model).
The grid naming under FOS is spelled out in the grib bulletin as to
which qrid is represented by each letter, and what grid fields
are represented by the second letter, and which forecast time are
represented by the fourth letter. I will send you the references
separately (we are out of our offices for rewiring and I don't have the
reference handy currently). The LDM adds the /mXXX tag since the FOS
name is ambiguous for identifying the models (especially for the
^[YZ].* eta/ruc and ngm headers. The LDM pqact.conf actions I sent
usually try to find the easiest way to disambiguating the model.

It looks like Aurora.atmos.colostate.edu is feeding auriga.cira.colostate.edu.
I believe Aurora filters their request to the upstream- though I thought that
cira was going to switch roles and be upstream of the atmos department.
Aurora was not a suitable site to feed NSBF (balloon facility) since
they were not obtaining the entire feed. Has there been action to
moving cira to upstream to request the complete feed? I believe
NSBF would be interested as well, and will pass this on to Robb Kambic
who handles IDD routing.

Steve Chiswell
Unidata User SUpport


>From: "Molenar, Deb" <address@hidden>
>Organization: .
>Keywords: 199904161842.MAA15323

>In response to my question on how to access gridded data other than the 211
>grids on April 14, Steve Chiswell provided a sample pqact.conf file.  I
>could not get ldm to start using the sample file==first error was HRS not
>recognized.  Changed to WMO, then I just received a "child process exited
>error" with no other indication of what the problem was.  So I modified the
>pqact.conf to try and just bring the 212 ETA grids using the format I've
>been using for the 211 grids.  Here is the line from pqact.conf:
>
>WMO    (^[YZ].[R].*/mETA) (....) ([0-3][0-9][0-2][0-9][0-5][0-9])      PIPE
>       -close  /usr/local/ldm/addccb.pl /data/ldm/\1\2.\3
>
>The ldm started fine, but has been running for over 1/2 hour and has not
>received any data.  Is this normal for this data set, or is there a problem
>with my pattern action code?
>
>While the file Steve sent was helpful, it still does not clear up many of my
>questions about how the raw data is actually named on the feed.  Is there
>any hard documentation about the naming conventions?  If not, my most recent
>questions include:
>
>1)  It looks sometimes /*mXXX is used, where XXX is the model name, and
>sometimes it isn't.  What is the convention for this?
>
>2)  It looks like sometimes a site header (KWB.) is used and sometime not.
>Conventions?
>
>3)  It looks like sometime a time field is used and sometimes not.  Again,
>conventions?
>
>4)  And the big question:  according to the sample file sent,  my code of 
>(^[YZ].[A-U].*) should be getting ALL of the different grids.  However, only
>the 211 grids come in.  What is wrong with what I'm using?
>
>Thanks much.
>
>
>___________________________
>Debra A. Molenar          address@hidden
>Computer Specialist     NOAA/NESDIS/RAMM  E-RA2
>phone:  970-491-8447    fax:  970-491-8241
>
>