[Wrf-users] Problems with WRF, two nesting way, instability problem

[Lionel Renault]

Hello eveybody,
I'm a PhD student. I'm using WRF to study the mesocale impact of the
atmosphere on a regional oceanic model along the western south american
coast. I'm using  a two way nesting with 3 domians: 56km/18km/6km. The
time resolution is respectively 90/30/10 s. I encontred an instability in
the high spatial resolution domain, which occurs about 30 hours into the
forecast. The error message is:

Timing for Writing wrfout_d01_2000-10-01_06:00:00 for domain        1:
8.76400 elapsed seconds.
Timing for processing lateral boundary for domain        1:    2.76100
elapsed seconds.
Timing for Writing wrfout_d02_2000-10-01_06:00:00 for domain        2:
24.43500 elapsed seconds.
Timing for Writing wrfout_d03_2000-10-01_06:00:00 for domain        3:
41.69000 elapsed seconds.
         12045  points exceeded cfl=2 in domain             3  at time
30.00278      hours
  MAX AT i,j,k:            32            9           56  cfl,w,d(eta)=
        nan  -2.4245966E-02   0.1276000

The instability does not occur when we do the model run without the 6 km
domain. Also note that the model time-step is already quite small and I
don't want to reduce if any futher it thats at all possible.
Anybody is aware of what can cause instablities like this and do you
know of any methods that can be used to avoid them? I would be very
grateful for any advice anybody can give me.
Thanks you very much, best regards,
Lionel

--
RENAULT Lionel
First year PhD Student,
LEGOS (France)
--
Departamento de Geofisica,
Blanco Encalada2002
Santiago, Chile
--
Compañia de Jesus 2153, departamento 204-A
Santiago, Chile
--
Tel. : 56-2-6968686 / 56-0-99150060