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

20010619: grib and wgrib



>From: "Alliss, Randall J." <address@hidden>
>Organization: TASC
>Keywords: 200106121930.f5CJU9p16378 wgrib

Randy,

Sorry this has take me so long to respond to.  I have been frantically trying
to get my house ready for sale...

>ok..how about bufr files. Do you have utilities which will read ncep bufr
>files.

We did a quick scan of the NCAR site and came up with the following URLs:

 http://dss.ucar.edu/libraries/grib/index.html
 http://dss.ucar.edu/libraries/bufr/bufr.html

When you go to the grib page, you will see a link to the wgrib stuff you
asked about previously.

The second URL contains links to Fortran and C BUFR decoders.

I hope that this helps.

>My problem:
>
>We have a UW furnished Noaaport ingestor. There is a decoder called
>GribFILER and GRIB which converts bufr files to grib files as the data is
>ingested. One of the accompanying programs which is used was written by Chad
>johnson of UW who of course is no longer there. In addition, no one at UW
>will help us with it w/o of course $.

Hmm...  It may be that they can't help you without a lot of effort, and
this always translates into money for them.

>The program we think does not always properly to the decoding from bufr to
>grib therefore we are lacking the necessary grib data to run mm5. I speaking
>of the ETA model, however, grib is grib and bufr is bufr. My thought is we
>are getting in the bufr files alright however, the conversion does not
>always work. Mary Ellen has cross checked the gribs with the McIDAS GRIDS
>and says that the GRIDS contain ALL the data so i think the bufrs are coming
>in ok.

I understand the problem.  You need to convert the bufr to grib before it
can be easily used in MM5.

>any thoughts.

Hopefully, the code accessible through the URLs above will get you going.

Tom

>From address@hidden Thu Jun  7 12:47:20 2001
>To: "'address@hidden'" <address@hidden>,
>   "'address@hidden'" <address@hidden>
>Cc: "'address@hidden'" <address@hidden>,
>   "'address@hidden'" <address@hidden>
>Subject: question re. processing eta model  bufr files to grib files
>Date: Thu, 7 Jun 2001 14:45:15 -0400 

Russ & Tom,
        Hello!  At TASC, we transform the ETA model BUFR files ingested by
our noaaport receiver into GRIB format which we store on-line. We use the
GRIB files to initialize our MM5 model runs.  Not all ETA model BUFR files
are being converted to GRIB format, specifically model 84, geographic id 212
and 215. (We list the fields in the GRIB files using the wgrib utility).
However, all the ETA model BUFR files ARE successfully converted to GRID
format.  We used the GRDLIST command on the dataset RTGRIDS/ETA to verify
this.  
        The program we use to convert the BUFR files to GRIB format is
Mcgrbbfrdec.c (Mcidas - grib-bufr dec) called by dmbin.f.  Chad Johnson
(SSEC former employee) wrote this c program.  Since all ETA grids appear to
be available, it looks like there is a break down getting the BUFR files
converted to GRIB files...(i.e., bug in program?)   Another issue we noticed
today is duplicates in the ETA GRIB files.  For example, today's 12:00UTC
ETA GRIB files contained three copies of the geopotential height fields for
each forecast hour.
        Attached is an excel spread sheet which shows the GRIB data we have
on-line for the past two days at 00:00 and 06:00 UTC. Notice the data
drop-outs.  I have also attached the programs dmbin.f and Mcgrbbfrdec.c.
 
        Any help you can provide would be greatly appreciated .  Thanks!

Mary Ellen

Mary Ellen Craddock
Meteorologist
Northrop Grumman Litton TASC
4501 Stonecroft Blvd.
Chantilly, VA 20151
(703) 633-8300 x4022
address@hidden