[Wrf-users] RE: GRIB2 updates for ungrib.exe

[Kemp, Eric M.]

I agree with the additional changes for SOILHGT.  I am cc:ing wrf-users and Kip Smith.

-Eric

-----Original Message-----
From: wrfhelp [mailto:wrfhelp at ucar.edu]
Sent: Thu 9/27/2007 7:10 PM
To: Kemp, Eric M.
Subject: Re: GRIB2 updates for ungrib.exe
 
Thank you for your report. We have made the necessary change in our  
repository, and will
soon update the WPS tar file.

We noticed that GRIB code for SOILHGT is not quiet correct in your  
Vtable. Here is a correct line:

    7 |   1  |   0  |      | SOILHGT  | m       | Terrain field of  
source analysis        |  2  |  0  |  7  |   1 |

See if you agree with it.

wrfhelp

On Sep 26, 2007, at 2:43 PM, Kemp, Eric M. wrote:

>
> Dear wrf-users/wrfhelp:
>
> Following up on my e-mail last week:  I have attempted to process NAM
> GRIB2 files with ungrib.exe to initialize WRF-ARW.  Originally I was
> unable to process the SKINTEMP variable using the WPS VTABLE.NAM file.
> Kip Smith of NOAA/ARL was kind enough to send me a revised VTABLE
> (see attached) that changes the GRIB2 attributes for SEAICE, LANDSEA,
> SOILHGT, SKINTEMP, SNOW, CANWAT, SOIL_CAT, and VEGCAT.  Kip's new
> settings agree with Section 4 of the NCEP GRIB2 documentation:
>
> http://www.nco.ncep.noaa.gov/pmb/docs/grib2/grib2_doc.shtml
>
> However, there is still one problem with the ungrib.exe program.   
> The NAM 218
> GRIB2 files from NCEP include soil temperature valid at 300 cm, and
> ungrib.exe incorrectly uses this data in place of 0-10 cm soil  
> temperature.
>
> To fix this, I've added a cycle statement in subroutine rd_grib2,
> WPS/ungrib/src/rd_grib2.F90, near line 507:
>
>                 if (j .gt. maxvar ) then
>                   write(6,'(a,i6,a,i6,a)') 'Subsoil level ',
>      &               gfld%ipdtmpl(12),' to ',gfld%ipdtmpl(15),
>      &           ' in the GRIB2 file, was not found in the Vtable'
>
>                   cycle ! EMK bug fix, avoid saving data that is  
> not in Vtable
>                 endif
>
> What happens is that ungrib.exe identifies the 300 cm soil  
> temperature as
> 0-10 cm soil temperature based on the GRIB2 discipline, category,  
> parameter,
> and elevation numbers.  Subroutine rd_grib2 then compares the soil  
> depths
> with those in VTABLE, and prints out a warning message when the depths
> fail to match.  However, without a cycle statement, the program
> continues on and ungribs the 300 cm soil temperature instead of
> skipping it.
>
> I believe that other "mix-ups" can happen without this cycle  
> statement,
> and I recommend adding it to the official ungrib program.
>
> -Eric
>
> Eric M. Kemp
> Meteorologist
> Northrop Grumman Information Technology
> Intelligence Group (TASC)
> 4801 Stonecroft Boulevard
> Chantilly, VA 20151
> (703) 633-8300 x7078 (lab)
> (703) 633-8300 x8278 (office)
> (703) 449-3400       (fax)
> eric.kemp at ngc.com
>
>
>
> <Vtable.NAM218_test>


-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/wrf-users/attachments/20070928/e7a3c432/attachment.html