[MITgcm-support] cell size variation

[Martin Losch]

Patrick,

your upper layer thickness dz(1) (or drF(1) in MITgcm-lingo) should  
be much larger than your maximum expected wave height, otherwise, the  
surface layer is going to drain. I'd go with waveheight < 20% drF(1)  
at least.

I am not sure if this works with non-hydrostatic dynamics, but maybe  
you want to try the r*-coordinate (something like a surface following  
sigma-coordinate) for your experiments?

Martin

On 9 Apr 2008, at 12:28, Patrick Rosendahl wrote:
> Thank you, I get much better results!
>
> Next thing I was wandering, if that equation has to be fulfilled  
> for zero-level (no waves) only? I have ~6cm waves in a 5cm upper  
> layer (the wave reduces it to 5-3=2cm and enlarges it to 5+3=8cm).
> Is there a rule-of-thumb for the maximum streching of the upper  
> most layer?
>
> Best,
> Patrick
>
> Martin Losch wrote:
>> the rule of thumb is:
>> abs(delz(k+1)/delz(k)) < 1.4
>> In your case: 0.03/0.005 = 6
>> Martin
>> On 9 Apr 2008, at 10:34, Patrick Rosendahl wrote:
>>> Hello all,
>>>
>>> I get unstable results with the following configuration in free-  
>>> surface and non-hydrostatic mode:
>>>
>>> delZ=1*0.03,100*0.005,
>>>
>>> The depth is 0.5m, divided into a 3cm upper layer (that will   
>>> contain the free surface), and the rest is divided into 5mm layers.
>>>
>>>
>>> In the mail "Re: [MITgcm-support] z coordinate" from Baylor Fox-  
>>> Kemper:
>>> >Hi Rima,
>>> >  The MITgcm is not accurate when you change the delZ so  
>>> rapidly.   You
>>> >are much better off using the 20 levels, if you can afford to  
>>> run it
>>> >that way.  If not, you can make a few levels of delz=30m, then
>>> >gradually increase until you reach bottom.  The grid size shouldn't
>>> >increase too rapidly from level to level, approximately 20% at  
>>> most.
>>> >This will be even more important in nonhydrostatic mode.
>>> >   -Baylor
>>>
>>>
>>> Is there a guide for the cell thickness? Is 20% really necessary?
>>>
>>> Best,
>>> Patrick
>>> _______________________________________________
>>> MITgcm-support mailing list
>>> MITgcm-support at mitgcm.org
>>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>> _______________________________________________
>> MITgcm-support mailing list
>> MITgcm-support at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support