[Wrf-users] Dimension problems between 3DVAR and WRF

[Michael Mcatee]

Thorsten,

I don't know that I fully comprehend what the problem is but I can tell 
you there is a  bug in WRF Var version 2.1 concerning mercator 
projections.  I don't believe a bug fix  was ever released via the WRF web 
page.  I got a fix from one of  the WRF Var developers.   I noticed the 
bug when running WRF Var over Hawaii.  Many observations than were clearly 
in the domain and time window were not being used by 3DVAR.  There was a 
problem in the observation preprocessor and in WRF Var itself.  Both had 
to do with calculating grid indexes from latitude and longitude and only 
occured for mercator projections.   I don't know if your problem is 
related but it might be worth an email to wrf help.

Mike

___________________________
Michael D. McAtee
The Aerospace Corporation
106 Peacekeeper Dr. STE 2N3
Offutt, AFB, NE 68113-4039
email:  michael.mcatee at aero.org
Phone: (402) 292-1017
Fax: (402) 291-3137



Thorsten Beisiegel <tbeisiegel at unal.edu.co> 
Sent by: wrf-users-bounces at ucar.edu
04/03/2007 03:52 PM

To
wrf-users at ucar.edu
cc

Subject
[Wrf-users] Dimension problems between 3DVAR and WRF






Dear Users,

while comparing the output from 3DVAR with output from WRF we notice a 
difference in dimension between the wrf_3dvar_output file and the 
wrfout_d0* file for the same domain. The result is an error in grads using 
variables from both files in the same ecuation. 

The horizontal extensions of the 3DVAR grid seems to be reduced by almost 
one grid point distance, anyway it has the same number of points and grid 
point distance for the cartesian grid in each direction as the wrfout 
file.

As grid spacing an number of grid points are the same for both files, the 
problem seems to represent a difference in the proyection. But both files 
where generated with the same proyection parameters, 

 MAP_PROJ_NAME =  'mercator'
 MOAD_KNOWN_LAT =   10.60
 MOAD_KNOWN_LON =  -73.15
 MOAD_STAND_LATS =  10.60, 0
 MOAD_STAND_LONS = -74.15

Also the grads input files for both cases were written with the same 
wrf2grads executable.
Anyway with the "levels" option in the .ctl file  the number of points is 
the same in x and y but the values for latitude and longitude are not, 
they differ for 3DVAR output and WRF output. 

I wonder if the fact that 3dvar only uses initial conditions while WRF 
includes the boundary data could be a reason.

Many thanks for any hint

Thorsten 
_______________________________________________
Wrf-users mailing list
Wrf-users at ucar.edu
http://mailman.ucar.edu/mailman/listinfo/wrf-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/wrf-users/attachments/20070403/3d57f53f/attachment.html