[Wrf-users] Re: wrfprep.pl and NNRP

[Ricardo C. Almeida]

Dear Don,

I found out that the problem is related to the 6-hours time offset between 
the pgb and grb2d files. In wrfsi, if all files are located in the same 
directory and  the grd2d file corresponding to 6 hours before the initial 
time is not found, the LSM information is read from the pgb file and wrfprep 
runs without error. However, the files produced miss some data, and the 
problem  I reported happens when running real. exe.  All I had to do was to 
separate the pgb anb the grb2d files (as suggested by you) in different 
directories and to add the grb2d file corresponding to 6 hours before the 
initial time. Then everything worked fine.

Thanks for your help,

Ricardo.

----- Original Message ----- 
From: "Don Morton" <Don.Morton at umontana.edu>
To: "Ricardo C. Almeida" <rcalmeida at terra.com.br>
Cc: <wrf-users at ucar.edu>
Sent: Thursday, October 26, 2006 3:11 PM
Subject: Re: [Wrf-users] Re: wrfprep.pl and NNRP


> Hi Ricardo,
>
> I'm afraid I can't add much, and what I can add comes from my
> somewhat feeble memory (I sent an email to someone in
> Colorado a couple of weeks ago with more specifics, but can't
> find the email now).  The major thing that I do remember is that
> Oriol had suggested putting both the surface and analysis NNRP
> files in the same directory when degribbing, but that caused problems
> for me that didn't really show up until I was running real.exe.  The
> problems seemed related to the LSM, and I kind of wonder if
> those are the same sorts of problems you're running into.
>
> At any rate, the one major thing I did different (with respect to Oriol's
> advice) is that I put my NNRP surface files in one directory, my
> NNRP analysis (3D) files in a separate directory, then degribbed them
> separately.  I'm guessing that when I failed to do this, the degribbing
> routines were taking the wrong fields out of the wrong files.  I've 
> learned
> that grib_prep.pl is VERY sensitive to having exactly the needed files in
> a directory all by themselves - even putting a subdirectory in amongst
> your needed files can sometimes cause issues.
>
> I can send you copies of my grib_prep.nl, Vtables, wrfrsi.nl, etc. later 
> on
> if you want, but I'm hoping your problem is as simple as the one I 
> describe
> above.
>
> Cheers,
>
> Don
>
> On 10/25/06, Ricardo C. Almeida <rcalmeida at terra.com.br> wrote:
>>
>>
>> Dear Don,
>>
>> I'm trying to run WRF (ARW) using NNRP data as input data. I was having 
>> the
>> same problem you had with the surface data and made the changes 
>> recommended
>> by Oriol, and prepared the input data (using wrfsi GUI).
>> When I ran real. exe I had the following error message:
>> .
>> .
>> .
>>
>>
>> DYNAMICS OPTION: Eulerian Mass Coordinate
>>
>> Time period # 1 to process = 2005-11-01_00:00:00.
>>
>> Time period # 2 to process = 2005-11-01_06:00:00.
>>
>> Total analysis times to input = 2.
>>
>> -----------------------------------------------------------------------------
>>
>> Current date being processed: 2005-11-01_00:00:00.0000, which is loop # 1
>> out of 2
>>
>> configflags%julyr, %julday, %gmt: 2005 305 0.0000000E+00
>>
>> error in the TSK
>>
>> i,j= 1 1
>>
>> landmask= 1.000000
>>
>> tsk, sst, tmn= 0.0000000E+00 0.0000000E+00 285.2493
>>
>> error in the TSK
>>
>> i,j= 2 1
>>
>> landmask= 0.0000000E+00
>>
>> tsk, sst, tmn= 0.0000000E+00 0.0000000E+00 9.9999999E+36
>>
>> -------------- FATAL CALLED ---------------
>>
>> FATAL CALLED FROM FILE: module_initialize_real.b LINE: 420
>>
>> TSK unreasonable
>>
>> -------------------------------------------
>>
>> ====================================================================
>>  In wrf namelist I set sf_surface_physics = 0. (is that correct ?)
>>
>> Since you have some experience on running WRF with NNRP input data I 
>> would
>> like to know if you (or any other WRF user) could help me with this 
>> problem.
>>
>> Thanks,
>>
>>
>> Ricardo C. de Almeida
>> Brazilian Navy Hydrography Center
>> _______________________________________________
>> Wrf-users mailing list
>> Wrf-users at ucar.edu
>> http://mailman.ucar.edu/mailman/listinfo/wrf-users
>>
>>
>>
>
>
> -- 
> Don Morton
> Department of Computer Science
> The University of Montana - Missoula
> http://www.cs.umt.edu/~morton/
>
>