[MITgcm-support] tidal forcing on 4-OB region (Jody Klymak)

[guopu at scsio.ac.cn]

Hello Jody,

Thanks for your help. I use DeltaT=10 because at first I used various dz and the min(dz)=10m near the surface, later I think 10-m resolution is too high then I change dz to unfied 100m but did not change DeltaT.

Now I adjust the horizontal and vertical resolutions to 1km and 100m, then use K1 tidal currents from TPXO7.2 to force all the 4 boundaries. In fact, as a MITgcm beginer I hope to finally reproduce the "near-field" model in Alford et al, 2015, Nature Letter. I tried DeltaT=120, 60 and 30 in turn. However the temperature and vertical velocity on the boundary always increase very fast. The output values change to NAN after several minutes. Is it hinting that I should use sponge layer to relax the temperature and velocity fields at boundaries? By the way, you mentioned that "use rcbs to propagate a uni-directional wave through the sponge". What is the rcbs? Do you mean the RBCS package?

Sorry but what is the meaning of "astronomically force it"? Is the explanation in that paper you recommend? I will read that paper carefully.

Best,

Pu





> 
> Message: 2
> Date: Tue, 8 Sep 2015 09:13:24 -0700
> From: Jody Klymak <jklymak at uvic.ca>
> To: mitgcm-support at mitgcm.org
> Subject: Re: [MITgcm-support] tidal forcing on 4-OB region
> Message-ID: <05E2D688-7A47-43C7-9B2F-4EF71B77ABF4 at uvic.ca>
> Content-Type: text/plain; charset="utf-8"
> 
> 
> > On 8 Sep 2015, at  8:59 AM, guopu at scsio.ac.cn wrote:
> > 
> > Hello everyone,
> > 
> > Now I am trying to force a 3D region by tidal current. All 4 boundary are open and I prescribe the tidal current on these 4 boundary by OBCS, EXF (obcsEperiod=1200) and CAL packages. There are above-sea mountains and under-sea mountains in this region. Then I expect for the generation and propagation of internal waves by tidal forcing. Horizontal and vertical resolutions are 5km and 100m, DeltaT=10.
> > 
> 
> If you need DeltaT=10 with those resolutions, something is very wrong!  You should be able to get by with deltaT>100.  
> 
> > Now I am wondering about 3 questions:
> > 
> > 1. Should I give this region the initial tidal current field which is correspond with that on boundaries.
> > 
> This can?t hurt, but nor will it particularly help.  Each run will have start up transients, so you need to let those settle down first.  ?Settle down? depends on what you are interested it, but certainly time for mode-1 to propagate across the domain a couple of times.  
> > 2. Could I use Orlanski or Sponge-layer on the boundaries to let waves go out or get weak.
> > 
> You should definitely use a sponge layer.  If your internal waves are coming from a prescribed direction, you can even use rcbs to propagate a uni-directional wave through the sponge.  If you just have velocities at the boundaries, then that is probably too hard, but you should be aware that your internal waves will not be perfect analogues of the real ocean because the real ?boundary? will be inside the sponge somewhere.  
> 
> > 3. Are there other better ways to force this region and then generate internal waves? 
> > 
> If the region is huge, you could astronomically force it.  But that will only be accurate if you resolve the whole basin.  If you have a smaller region inside a big basin, forcing at the edges works relatively well. I recommend Carter and Merrifield 2007, Ocean. Modell. as a good study showing how to implement BCs for internal tide models.  
> 
> Cheers,   Jody
> 
> 
> > Thanks for your help.
> > 
> > Best,
> > 
> > Pu
> > 
> > 
> > 
> > 
> > _______________________________________________
> > MITgcm-support mailing list
> > MITgcm-support at mitgcm.org
> > http://mitgcm.org/mailman/listinfo/mitgcm-support
> 
> --
> Jody Klymak    
> http://web.uvic.ca/~jklymak/
> 
> 



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20150916/6db5afe2/attachment.htm>