[RegCNET] instant CFL violation via CanESM ICBC

Константин Макушев makushevkm at mail.ru
Tue Oct 25 10:54:15 CEST 2016


Hello, dear RegCM users!

I had tried to run RegCM4.4.5.7 via CanESM global forcing, but the model instantly gave me this error message:

*************************************************************
MAXVAL ATEN T : 0.83140449859510546 
II : 12 , JJ : 69 , KK : 2
MAXVAL ATEN U : 2.3331410432147703 
II : 14 , JJ : 51 , KK : 1
MAXVAL ATEN T : 3.7800806779265788 
II : 36 , JJ : 46 , KK : 17
MAXVAL ATEN U : 2.6793612500123634 
II : 34 , JJ : 71 , KK : 1
MAXVAL ATEN V : 2.3787916083209342 
II : 35 , JJ : 54 , KK : 1
MAXVAL ATEN V : 1.9001860543930813 
II : 18 , JJ : 44 , KK : 1
-------------- FATAL CALLED ---------------
Fatal in file: mod_tendency.F90 at line: 1574
CFL VIOLATION
-------------------------------------------
-------------- FATAL CALLED ---------------
Fatal in file: mod_tendency.F90 at line: 1574
CFL VIOLATION
**************************************************************

Spatial and temporal resolution looks accepable:
************************************************************
ds = 40.0

&timeparam
dt = 120., ! time step in seconds
dtrad = 30., ! time interval solar radiation calculated (minutes)
dtabem = 18., ! time interval absorption-emission calculated (hours)
dtsrf = 600., ! time interval at which land model is called (seconds)
*************************************************************

Earlier, I had run the model with HadGEM forcing in exactly the same configuration and had no problems.

I even tried to set dt = 60, but the error still appears.

Did anyone have an idea why this problem arises and how to solve it? Maybe someone else faced the similar?

Thanks for any response!

-- 
Konstantin Makushev
Altai State University 
Barnaul, Russian Federation
makushevkm at mail.ru
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ictp.it/pipermail/regcnet/attachments/20161025/d287c9c5/attachment.html>


More information about the RegCNET mailing list