[Wrf-users] WPP postprocessor

[Jan Ploski]

wrf-users-bounces at ucar.edu schrieb am 09/17/2008 04:57:05 PM:

> 1./ I just can't use NCL on WPP output data to draw decent charts -- 
> apparently because
> of false units of the ELON - NLAT fields, and some other problems.

Hi,

I just checked in my installation (with the gfortran-compiled WPPV3, WRF 
run over Germany). The ELON/NLAT fields produced by WPPV3 are only 
slightly different from the XLAT/XLONG produced by WRFV3:

(0)     min elon-xlong = -0.00484753
(0)     min nlat-xlat  = -0.00458908
(0)     max elon-xlong = 0.00881386
(0)     max nlat-xlat  = 0.00773621
(0)     rmse elon-xlong = 0.00354025
(0)     rmse nlat-xlat = 0.00333095

> 2./ I would like to retrieve discrete values at fixed 
> longitude/latitude positions. I
> tried "degrib" (that I use for GFS), but the retrieved 
> longitudes/latitudes of ELON/NLAT
> are slightly shifted from the requested ones (slightly means about one 
> degree, which is
> way beyond the grid resolution).
> Retrieving the values with "wgrib" is a little more precise, but still 
> no bilinear
> interpolation of the surrounding gridpoints. Do you have an idea ?
>
> What is the best way to retrieve discrete values at given 
latitude/longitude?

I use some custom NCL scripts for that purpose, specifically the functions 
wrf_latlon_to_ij and dsgrid2. NCL can read GRIB data just like NetCDF.

> 3./ After running WPP post processor, the categorical snow/rain/ice 
> pellets are all 0 valued.

No idea about that.

Regards,
Jan Ploski

--
Dipl.-Inform. (FH) Jan Ploski
OFFIS
FuE Bereich Energie | R&D Division Energy
Escherweg 2  - 26121 Oldenburg - Germany
Phone/Fax: +49 441 9722 - 184 / 202
E-Mail: Jan.Ploski at offis.de
URL: http://www.offis.de