[MITgcm-support] adding uniform background flow to all water grid points smoothly.

[Jody Klymak]

Hi Jianing Li,

Its actually a little unclear what you are trying to do and what you are trying to study which will definitely affect how you force the model. 

In general, the obcs radiation conditions are not very robust.  They can be tuned to remove mode-1 reflections, but are unlikely to radiate higher modes, or, in 3D internal waves that impact the boundary layer from an angle.  

That leaves you using a sponge, which can work quite well if you exercise some care in setting up the sponge.  In general the sponge should physically be as wide as a mode-1 wavelength, and the sponge time scale should not increase too suddenly or you will get reflections.  You can make the sponge cheaper by telescoping the values of dx in the sponge so you don’t need as many grid cells.  You may want to look at Carter and Merrifield 2007, Ocean Modelling.  

If you also want the velocity to be nudged everywhere in the domain, RBCS can also do that for you. Or you can add a body force to represent an along-channel pressure gradient. 

To my mind, the only reason to use OBCS versus RCBS is if you want something cheaper, since RBCS requires sponge and background values to be calculated for the whole model domain.  However RBCS is far more flexible, allowing you to also specify in-coming internal waves (i.e. Klymak et al 2015 JPO).  

Cheers,   Jody


> On 21 Nov 2020, at 04:12, Li Jianing <li.jia.ning at outlook.com> wrote:
> 
> Hi Martin,
> 
> With your suggestion, I altered my configuration by enabling rbcs and it works fine, but a new tricky thing occured when I try enable rbcs and obcs both.
> The rbcs derived the flow field as I expected, however, the boundary reflections was too noisy, so I enabled steven obcs to remove the reflections as you mentioned. But it seems like the obcs suppressed the rbcs and the U field kept quite small (~1e-18 to ~1e-5 depending on different obcs configs) after 5 days in model time. Since the rbcs worked well by itself, I supposed the problem is on obcs and separately altered some obcs parameters which may be responsible for it in many model runs, like defining&undefining N/S OBCs, defining&undefining OBCS prescribe, using Olanski, seting rbcs masks to 0 at OBCs... But all these makes no improvment.
> By far my rbcs is applied to all wet points with varying masks depending on their distances to the topography. And the attachments is some configs maybe usful. So what should I do to fix this? 
> Many thanks for your help.
> 
> Jianing Li
> <CPP_OPTIONS.h><data><data.obcs><data.rbcs><OBCS_OPTIONS.h>_______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org <mailto:MITgcm-support at mitgcm.org>
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support <http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support>
--
Jody Klymak    
http://ocean-physics.seos.uvic.ca/~jklymak/






-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mitgcm.org/pipermail/mitgcm-support/attachments/20201121/cec4c8f5/attachment.html>