[Wrf-users] trouble with 2-way nest

[John Krasting]

I can't help you very much, but I encountered the problem with a
12km/4km two way nested run using GFS data.   I also played with the
time step by it didn't help.  I changed my resolution to 15km/5km and
the problem resolved.   I never figured out why WRF didn't like the
12km/4km configuration.

-John

On 10/17/05, mpborkow at ncsu.edu <mpborkow at ncsu.edu> wrote:
> I am running a two-way nest, 12km outer domain with a 4km inner domain
> using a NARR dataset.  The setup seems to be fine but I am having trouble
> with the model run itself.  I do get one ouput file for each domain for
> the initialization time only.  The SLP from these output files match up
> well with the reanalysis data in some spots, in other spots the SLP seems
> very artificial with the isobars looking erratic.  Here is a sampling of
> the errors I am getting:
>
> Timing for Writing wrfout_d01_2004-03-16_00:00:00 for domain        1:
> 21.71200 elapsed seconds.
> Timing for processing lateral boundary for domain        1:    5.28800
> elapsed seconds.
>  WRF NUMBER OF TILES =   1
>          90190  points exceeded cfl in w_damp
> (module_big_step_utilities_em.F)
>   MAX AT i,j,k:           174           60           30  cfl,w,d(eta)=
>         nan             nan   2.1000000E-02
> Timing for Writing wrfout_d02_2004-03-16_00:00:00 for domain        2:
> 18.27400 elapsed seconds.
>  WRF NUMBER OF TILES =   1
>          21257  points exceeded cfl in w_damp
> (module_big_step_utilities_em.F)
>   MAX AT i,j,k:             5           55           30  cfl,w,d(eta)=
>         nan   3.5428979E-02   2.1000000E-02
>          41992  points exceeded cfl in w_damp
> (module_big_step_utilities_em.F)
>   MAX AT i,j,k:            10           55           30  cfl,w,d(eta)=
>         nan   5.5973802E-02   2.1000000E-02
>
>  -------------- FATAL CALLED ---------------
>  FATAL CALLED FROM FILE:  module_cu_kfeta.b  LINE:    2054
>  KAIN-FRITSCH, istop=1, diags
>  -------------------------------------------
> [4] MPI Abort by user Aborting program !
> [4] Aborting program!
>
>
> My first thought was that it was a time step issue, but my time step is
> set at 60s for the 12km domain, 20s for the 4km domain.  I did try
> reducing them to 48s and 16s, but this did not solve the problem.  Given
> the final error, looks like there may be some error with the KAIN-FRITSCH
> scheme, which I only have running on the 12km domain.  Does anyone have
> any thoughts as to what the problem may be?  Thanks!
>
> -Matt
>
> _______________________________________________
> Wrf-users mailing list
> Wrf-users at ucar.edu
> http://mailman.ucar.edu/mailman/listinfo/wrf-users
>


--
****************************************************
John Krasting

Center for Environmental Prediction
Rutgers University
14 College Farm Road
New Brunswick, NJ  USA 08901-8551
Phone:  +1-732-932-3890
Fax:  +1-732-932-8644
****************************************************