[MITgcm-support] Abnormal end

Klymak Jody jklymak at uvic.ca
Sat Sep 10 23:03:20 EDT 2011


Hi Himansu,

Not sure exactly what you are trying to force, but…



> I am a using MITGCM model to find Internal waves in Bay of Bengal , Indian Ocean. Initially i have taken a coarser gird to investigate it. i have used the same "obcs_calc.F" given in the model. My queries are : 

>    SOLUTION IS HEADING OUT OF BOUNDS: tMin,tMax= -7.761E+02  1.428E+03
>    exceeds allowed range (monSolutionMaxRange=  1.000E+03)
>    MON_SOLUTION: STOPPING CALCULATION at Iter=      1957                               ]

When this happens it is almost always worth a look at *where* it is out of bounds.  i.e. the instability starts somewhere.

> 
> I have checked both CFL criteria and the stability criteria. They are well below their limits.

Did you make sure to check the vertical CFL?  You have really small vertical resolution, and quite coarse horizontal, which might cause problems for internal waves.  

>     (iii) What more parameters should i use in my data file for my Internal wave simulation. My data file is as follows:
> 
>  viscAz=1.E-3,
>  viscAh=100,
>  no_slip_sides=.FALSE.,
>  no_slip_bottom=.TRUE.,

I never use this unless I think I'm going to resolve the bottom boundary layer (for dz>20 m, you aren't).  Better to use a quadratic drag or free slip.   

>  diffKhT=,
>  diffKzT=,
>  diffKhS=,
>  diffKzS=,

I'm not sure what setting these empty does.  If it sets them to zero, that could be your problem.  I'd just set them to be the same as for viscosity for most problems.  

>  nonHydrostatic=.TRUE.,

Are you sure you need this?  Certainly not for your coarse runs - you might later.  

You may also want to play around w/ the advection scheme, particularly if you aren't setting the diffusivities!

> 
>  deltaTmom=30,
>  deltaTtracer=30,

OK< this is certainly fine for dx=25000.  But for dz=20 m, you are restricted to w<1/3 m/s or so.  Thats a high vertical velocity, but maybe you have some steep topography?

Good luck,

Cheers,  Jody

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20110910/3306951a/attachment.htm>


More information about the MITgcm-support mailing list