[MITgcm-support] Ross Sea simulation crash

[Jody Klymak]

Hi Enrico,

You can check the cfl criteria by looking at STDOUT.0000 (advu_cfl, advw_cfl etc). You don’t say anything about your vertical resolution, but if it is fine enough to use non-hydrostatic then you are quite possibly violating the CFL criteria in the vertical velocity.  

For testing purposes you may want to turn nonHydrostatic off (and maybe consider whether 0.5 degree runs can properly represent non-hydrostatic effects).  That may not help your cfl criteria, but the model will run faster.  

Cheers,   Jody

> On 22 Oct 2019, at 06:53, Pochini, Enrico <epochini at inogs.it> wrote:
> 
> Dear MITgcm support group,
> 
> I'm trying to develop a model of the Ross Sea for my PhD. Until now I've generated the grid and I've started testing it with constant atmospheric forcing, boundary conditions, and initial conditions, but I'm crashing a few kilo-seconds after initialization.
> 
> To generate it I followed what is written in this page:
> https://github.com/knaughten/mitgcm_python/wiki/Creating-a-new-MITgcm-domain#grid <https://github.com/knaughten/mitgcm_python/wiki/Creating-a-new-MITgcm-domain#grid>
> 
> However even minimal runs with disabled SHELFice, OBCS packages, imposing no surface forcing, with uniform initial conditions, still it develops instabilities and crashes.
> 
> I'm absolutely new to MITgcm and ocean modeling and I would need your help to understand what is causing the issues, in particular if the bathymetry has some pathologies or the (spherical) grid violates CFL somewhere .. (?) I'm using 150 seconds timestep, 0.5 degree grid.
> 
> I've attached four minimal test outputs, with and without ice shelf, with and without smoothing of ice shelf draft and bathymetry. The number with the file is the last iteration printed. 
> Thank you.
> 
> Best regards,
>  file_support.tar.gz <https://drive.google.com/a/inogs.it/file/d/1SM5mtzq3WgF61EbLh4404N5cz3czIuSI/view?usp=drive_web>
> 
> Enrico Pochini
> 
> PhD student, University of Trieste, Italy
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support

--
Jody Klymak    
http://web.uvic.ca/~jklymak/





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mitgcm.org/pipermail/mitgcm-support/attachments/20191022/8275fb5a/attachment.html>