[MITgcm-support] Water balance problem

[Martin Losch]

Hi Terry,

with non-hydrostatic dynamics, the non-linear free surface is not implemented. You only need to specify eta with a non-linear free surface.

If you don’t want to use OBCSbalance (which really is only a hack that is not generally useful as you have found out yourself), then you need to balance your inflow/outflow “offline”, i.e. before you apply it to the model. This is a little tricky, because you have to take into account the exact geometry of your boundaries (including the hfac’s); also you should specify your obcs fields as double precision (exf_iprec_obcs = 64, or readBinaryPrec = 64).

There are many posts about this and related questions, please search the support list for details.

Martin


> On 11. Dec 2017, at 09:09, Terry Chen <505883357 at qq.com> wrote:
> 
> Hello everyone,
>   I am trying to apply this model with NON-hydrostatic mode in a coastal region, but I meet great water loss near the offshore boundary. Thus, the model cannot represent ideal Eta result now. 
>   Since in the NON-hydrostatic mode, the elevation at the boundary cannot be prescribed, so I need some way to balance the transport now. But the OBCSbalance module in the model just balances the mass transport at each time step, which actually remove some signal of tide and wind.  
>   Does anyone have any idea about a better way to keep both the water balance and the signal of tide or wind? 
>   Thanks a lot.
> Best,
> Terry
> ------------------
> 
>  
>  
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support