[MITgcm-support] partial cells and model stability

[Matthew Mazloff]

Actually the model is more stable with partial cells on when I use a 
very smooth bathymetry file, but much less stable with partial cells on 
when I use a high resolution rough bathymetry file.  This is surprising 
to me as it seems partial cells should smooth the bathymetry.


Matthew Mazloff wrote:

> Hello support line,
>
> I am looking for information regarding partial cells and model 
> stability.  I have run the model with smooth bathymetry and coarse 
> bathymetry and there is little difference with model stability. The 
> majority of my abyssal depth levels have delZ=250.0 and when I first 
> turn on partial cells and leave hFacMindz = 250.0 the model is 
> stable.  When I turn hFacMindz=225.0 the model becomes significantly 
> less stable.  My vertical stability parameters are well within a safe 
> range so that I do not believe the change in dz from 250 to 225 is 
> causing a problem.  The model is, however, very sensitive to viscAh.  
> For my quarter degree model I currently have viscAh set to 6.65E3 and 
> making it any larger or smaller requires me to shorten the time step 
> (current timestep=1140sec).  I would greatly appreciate any comments.
>
> Thanks for your advice,
> Matt
>
> PS> I am running with open boundaries but I am running on a C-grid and 
> I have made sure that the open boundary and the adjacent cell has the 
> same bathymetry level.  Are there any issues regarding partial cells 
> and open boundary conditions that I should be aware of?
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support