[MITgcm-support] how to balance the flow flux during preprocessing

[David Hebert]

Hi Tongya,

In addition to the useOBCSbalance, keep in mind that you essentially 
have a control volume for your domain. SSH will be dependent on the 
amount of mass that comes in or goes out of your domain. You could check 
your boundary conditions beforehand by integrating the volume flux over 
each boundary for the entire simulation. (This is basically the 
integration of "normal velocity times area" on each boundary. 
Mathematically this is the divergence theorem). This gives you the net 
flux into your domain for the entire simulation. You could then alter 
each boundary velocity to remove this net flux. We did this and were 
surprised how low the velocity modification was needed, in most cases on 
the order of 0.1 cm or less.

The main difference is OBCSbalance (I believe) will balance the flux on 
each timestep, whereas the method above balances the flux for the entire 
simulation and might help keep any tidal signatures in your domain Of 
course, you would need to know your entire simulation boundary 
conditions and modify them beforehand.

Good luck,
David

On 01/03/15 22:18, Dustin Carroll wrote:
> Hi Tongya,
>
> Set useOBCSbalance=.TRUE. in your data.obcs file to balance the net 
> flow across all open boundaries.
>
> You can also control the behavior of individual open boundaries with 
> OBCS_balanceFacN/S/E/W.
> See section 6.3.1.5.6 in the documentation for how to set the flags 
> for this parameter.
>
> Cheers,
> Dustin
>
>
>> On Jan 3, 2015, at 7:30 PM, liutongya13 at mails.ucas.ac.cn 
>> <mailto:liutongya13 at mails.ucas.ac.cn> wrote:
>>
>> Hi all,
>> I am running a experiment wiht open boundary conditons and have read 
>> many previous emails about the obcs. In these emails I realised that 
>> I should do something to balance the flow flux during preprocessing 
>> in order to avoid large or small SSH. But how to do it? Just magnify 
>> or minish the entire U(W&E) or V(N&S) of a boundary ? Is there some 
>> feasible methods to balance the flow flux?I wonder the balance that 
>> just change a component of velocity will affect the model result.
>>
>> In my case, there are three boundaries(E,S,N). The NEC flow into the 
>> field from eastern boundary and Kuroshio flow out of the field from 
>> the north boundary. If I use the OBCS_balance, how should I set these 
>> parameter?
>>
>> Any avdice is high appreciated!
>>
>> Tongya Liu
>>
>>
>>
>>
>> _______________________________________________
>> MITgcm-support mailing list
>> MITgcm-support at mitgcm.org <mailto: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

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20150105/af464d21/attachment.htm>