[MITgcm-support] SEAICE - OBCS

[Martin Losch]

Hi Erick,

I think that your viscosity is way too large for this configuration. Very recently, I answered a similar question on this list:
<http://mitgcm.org/pipermail/mitgcm-support/2014-March/008963.html>

There are some very high resolution configurations with sea ice online at <http://mitgcm.org/viewvc/MITgcm/MITgcm_contrib/>, for example this one might help you: <http://mitgcm.org/viewvc/MITgcm/MITgcm_contrib/arctic/cs_4km/> with your parameter choices.

Martin

PS. In my experience the model’s scaling starts breaking down with about 30x30 grid points per tile. Your tile size of 16x16 appears very small. You might spend your CPU time more efficiently by running two experiments with 32x32 tiles (so only 8x8 processors instead of 16x16)
On Mar 14, 2014, at 10:34 PM, Edwards, Erick (LCDR) <eledward at nps.edu> wrote:

> Developers,
>  
> I’m trying to run a model using both the OBCS and SEAICE packages together, but so far I haven’t had much luck.  I’ve compared the relevant files in the “code” and “input” folders to those in the seaice_obcs tutorial in an attempt to isolate the problem, but I haven’t been able to get a model to run properly yet.
>  
> After I made obvious fixes that appeared as fatal errors in previous STDERR files, the model ran for a little over a dozen iterations before succumbing to “extreme potential temperature” (XPT).
>  
> The particular set-up that I’ve attached is the closest thing to a “null model” that I want to run before adding more interesting forcing.  My gut is telling me that there’s something simple that I’m missing in my set-up, like an additional “define” or “undef” or another parameter specified in a “data” file…. 
>  
> I did lower the time step to 30 seconds, to see if this would have an effect, but it did not (it just ran for more iterations before stopping).  I could, of course, continue lowering the time step in an attempt to get rid of the XPT shut-down, but in the past I’ve always found errors not related to the delta-t and eliminated any XPT problems.
>  
> I’m out of ideas on what to try.  If y’all are so inclined, could you take a look at my set-up?  Any pointers would be helpful, I’m sure.
>  
> Very respectfully,
>  
> Erick Edwards
>  
>  
>  
>  
>  
> <code.zip><input.zip>_______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support