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

20030317: is the 1x1 degree AVN GRIB decoder available? (cont.)



>From: Owen Cooper <address@hidden>
>Organization: Aeronomy Laboratory/NOAA
>Keywords: 200202181908.g1IJ8Dx01947 McIDAS gribdec.tar.Z

Hi Owen,

>Thank you for agreeing to test the GRIBs for me.
>I placed 3 of the FNL GRIBs on the aeronomy lab ftp site. One for each
>year: 2000 2001 2002

OK.

>I thought it would be a good idea to test a file from each year as some
>changes have been made to the contents of the GRIBs over the past couple
>of years.

Good thinking.

re: FTP the data files
>get the 3 files in the directory:
>fnl_000520_12_00  --> 2000 May 20, 12UTC
>fnl_010520_12_00  --> 2001 May 20, 12 UTC
>fnl_020520_12_00  --> 2002 May 20, 12 UTC
>
>Okay, please give these a shot and let me know how it works.

I grabbed the files and put them in a separate directory as I recommended
you do in a previous email.  I then decoded the files with no errors/warnings/
comments of any kind and did some plots which I put into:

http://www.unidata.ucar.edu/staff/tom/gifs/owen

Take a look and see what you think.

To be more specific (it might help get you going) I did the following:

1) logged on as 'mcidas'.  This is not mandatory, but it make make things
   simplier since you will need to link against the McIDAS library
2) create a directory /home/mcidas/gribdec
3) unpack gribdec.tar.Z in that directory
4) set my MCPATH to include /home/mcidas/gribdec as the first directory
   in my MCPATH (C shell syntax):

setenv MCPATH 
/home/mcidas/gribdec:/home/mcidas/workdata:/home/mcidas/data:/home/mcidas/help

5) built the gribdec executables:

make

   Again, for this to be so simple, I had to know how McIDAS was built
   in that I knew the set of compilers used.  In my case, I was on
   a Sun using Sun compilers.

6) FTPed your data into a different directory:

mkdir /home/mcidas/gribdata
ftp ...

7) created output GRID files:

cd /home/mcidas/gribdec

/home/mcidas/gribdec% ./gribdec.k ../gribdata/fnl_000520_12_00 6000 \"2000 May 
20 12 UTC
 Input GRIB file: ../gribdata/fnl_000520_12_00
Output GRID file: GRID6001
GRIB messages:: Found: 246 Decoded: 246 Written: 246
GRIBDEC: Done

/home/mcidas/gribdec% ./gribdec.k ../gribdata/fnl_010520_12_00 6000 \"2001 May 
20 12 UTC
 Input GRIB file: ../gribdata/fnl_010520_12_00
Output GRID file: GRID6010
GRIB messages:: Found: 272 Decoded: 272 Written: 272
GRIBDEC: Done
mv GRID6010 GRID6002

/home/mcidas/gribdec% ./gribdec.k ../gribdata/fnl_020520_12_00 6000 \"2002 May 
20 12 UTC
 Input GRIB file: ../gribdata/fnl_020520_12_00
Output GRID file: GRID6010
GRIB messages:: Found: 272 Decoded: 272 Written: 272
GRIBDEC: Done
mv GRID6010 GRID6003

   Notes:

   - The name of the output GRID file is created from the base GRID file
     number you provide (6000 in the examples above) and the Julian day
     that the data represents.  For the fnl_010520_12_00 and
     fnl_020520_12_00 data files, the output GRID file name ended up
     being the same: GRID6010.  In order to not decode the third day's
     data into the file created for the second, I simply renamed the
     second file.  I did the same for the third just to be tidy.  I ended
     up with three files, one per day: GRID6001, GRID6002, and GRID6003

   - gribdec.k tells the user (you) how many GRIB messages it found in
     the input data file AND how many of those messages it was able
     to decode.

8) After decoding all of the data, I created three 500 mb geopotential
   plots as a quick check.  I then used the McIDAS function FRMSAVE
   to create GIFs (tm) so that you could look at them on the web
   page I listed above.

Again, you have to decide if the decoding is correct (looks good to me!).

>Thanks!

No worries.

Tom

>From address@hidden Mon Mar 17 15:20:46 2003
>Subject: Re: 20030317: is the 1x1 degree AVN GRIB decoder available? (cont.)

Tom,

Excellent!!!
I'll give it a try and will let you know how it goes...........

Owen

PS I checked the plots you made against some other 500 hPa plots I have and
the decoding is right on the mark.