[Wrf-users] Cycling in WRFDA - Updating BCs

[Sean Crowell]

I'm running the most recent version of WRFDA, and I had a few  
questions.  I'm cycling hourly with conventional observations, and  
after several hours of assimilating, I'm getting artifacts around the  
boundaries, and so I figured I must be doing the boundary conditions  
wrong.  Right now, the way that I have my scripts set up, I first copy  
my wrfbdy_d01 and wrfinput_d01 from the last hour's run to my WRFDA  
run folder, and link the most recent wrfout_d01 to the file "./fg",  
and the proper obs*.3DVAR to ob.ascii.

Then I copy a file I called "parame.in.lower" to "parame.in" and run  
da_update_bc.exe.  These are the contents of parame.in.lower:

&control_param
da_file		   = './fg'
wrf_input          = './wrfinput_d01'
wrf_bdy_file			= './wrfbdy_d01'
domain_id          = 1
debug   	   = .true.
cycling = .true.
update_lateral_bdy = .false.
update_low_bdy 	   = .true.
low_bdy_only = .true.
iswater 	   = 16
/

After that I run da_wrfvar.exe, and then copy another file  
parame.in.lateral to parame.in.  These are the contents of  
parame.in.lateral:

&control_param
wrfvar_output_file = './wrfvar_output'
wrf_bdy_file       = './wrfbdy_d01'
wrf_input          = './wrfinput_d01'
domain_id          = 1
cycling = .true.
debug   = .true.
update_lateral_bdy = .true.
update_low_bdy = .false.
low_bdy_only = .false.
update_lsm = .false.
/

Is there something I'm missing that would account for the boundary  
effects I'm seeing?  The "obs" are really just taken from another set  
of WRF runs, so noise is not really an issue.



Sean Crowell
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/wrf-users/attachments/20120203/d9a90ea9/attachment.html