periodic boundary condition

Discussion of how to use ROMS on different regional and basin scale applications.

Moderators: arango, robertson

Post Reply
Message
Author
simion1232006
Posts: 60
Joined: Tue Sep 29, 2009 3:50 pm
Location: School of Environment System Engineering,UWA

periodic boundary condition

#1 Post by simion1232006 » Tue Oct 04, 2011 7:57 am

Hi everyone.

I have some questions related to the periodic boundary conditions. Does anyone have the experience in specifying a periodic boundary condition in idealized upwelling case with a pressure gradient (barotropic or baroclinic). So basically only the velocity fields are periodic, and both the tracer fields and sea surface elevations are not periodic.

Is it possible to do this in ROMS? Any suggestions will help. Thanks

User avatar
arango
Site Admin
Posts: 1092
Joined: Wed Feb 26, 2003 4:41 pm
Location: IMCS, Rutgers University
Contact:

Re: periodic boundary condition

#2 Post by arango » Tue Oct 04, 2011 3:28 pm

No, this is not possible in ROMS because of the numerical kernel design. Once that you selected periodic boundary conditions, all the model variables have to be periodic. Otherwise, you cannot achieve the correct periodicity.

simion1232006
Posts: 60
Joined: Tue Sep 29, 2009 3:50 pm
Location: School of Environment System Engineering,UWA

Re: periodic boundary condition

#3 Post by simion1232006 » Wed Oct 05, 2011 2:02 am

Hi Prof Arango,

Thank you very much. However , is there a way to apply a negative pressure gradient (opposite to wind )in idealized up-welling cases.
I did some hind cast simulations fully driven by global model,but it is too complicated to analyse. I would like to simplify it by using a idealized model with a constant pressure gradient and wind stress, to see their interaction.So what kinds of boundary conditions are most applicable to this case?

User avatar
kate
Posts: 3754
Joined: Wed Jul 02, 2003 5:29 pm
Location: IMS/UAF, USA

Re: periodic boundary condition

#4 Post by kate » Wed Oct 05, 2011 5:41 pm

This was easier to do in the days of the rigid lid, where you could impose a streamfunction gradient across the domain. Perhaps you could set up an idealized channel and force a mean flow through it via boundary conditions. Specify a ubar on the incoming end, then let it radiate out the other end.

rduran
Posts: 139
Joined: Fri Jan 08, 2010 7:22 pm
Location: Theiss Research

Re: periodic boundary condition

#5 Post by rduran » Thu Oct 06, 2011 1:10 am

Im just curious how the solution of imposing a pressure gradient via barotropic velocities would compare with imposing a pressure gradient via sloping isopycnals in the boundary conditions or even how about both. Any thoughts?

simion1232006
Posts: 60
Joined: Tue Sep 29, 2009 3:50 pm
Location: School of Environment System Engineering,UWA

Re: periodic boundary condition

#6 Post by simion1232006 » Thu Oct 06, 2011 2:56 am

Thank you. Kate. I think your methods will work. (flather with specified boundary at north, radiation at south, solid both east and west).
Also it is a problem to decide the ubar corresponding to certain pressure gradients. Actually my case is more about pressure gradient via sloping isopycnals,so I am not sure whether ubar is enough to control the incoming boundary(maybe also including baroclinic velocity).
I am thinking possibly i could run a simulation just with pressure gradient(like a channel flow), and then get the ubar field. I am not sure whether it makes sense. Any suggestions?

tony1230
Posts: 87
Joined: Wed Mar 31, 2010 3:29 pm
Location: SKLEC,ECNU,Shanghai,China

Re: periodic boundary condition

#7 Post by tony1230 » Thu Oct 06, 2011 6:49 am

hi prof A and friends
I have some questions:
1> what is the
periodic boundary conditions
exactly mean
and 2> the meaning for
negative pressure gradient
.
I can not understand very well.


thanks for reply

simion1232006
Posts: 60
Joined: Tue Sep 29, 2009 3:50 pm
Location: School of Environment System Engineering,UWA

Re: periodic boundary condition

#8 Post by simion1232006 » Thu Oct 06, 2011 8:03 am

Periodic boundary conditions are a set of boundary conditions that are often used to simulate a large system by modelling a small part that is far from its edge. Basically on these boundaries the dynamic properties are exactly the same. It is suitable for simulating the system that is infinite in one direction.

The negative sign of pressure gradient is dependent on the predefined coordinate system used in your model. In my case, to north is positive while to south is negative.

tony1230
Posts: 87
Joined: Wed Mar 31, 2010 3:29 pm
Location: SKLEC,ECNU,Shanghai,China

Re: periodic boundary condition

#9 Post by tony1230 » Fri Oct 07, 2011 3:17 am

Hi simion
Thank you very much.Now basically i can follow what you said.But:?::why you have to do
So basically only the velocity fields are periodic, and both the tracer fields and sea surface elevations are not periodic.
I have no idea and can not imagine on the situation(velocity fields periodic, tracer and elevation not periodic) you described.I never use boundary condition related to periodic,but i have the curiosity about what you posted.


Any reply or discusstion will appreciated.
Tks.

simion1232006
Posts: 60
Joined: Tue Sep 29, 2009 3:50 pm
Location: School of Environment System Engineering,UWA

Re: periodic boundary condition

#10 Post by simion1232006 » Fri Oct 07, 2011 7:33 am

Oh ,this is tricky. I simply want a channel flow driven by both pressure gradient and wind. For simplicity , it could be infinite in long-shore direction. However , seems it is very hard to put into practice

User avatar
kate
Posts: 3754
Joined: Wed Jul 02, 2003 5:29 pm
Location: IMS/UAF, USA

Re: periodic boundary condition

#11 Post by kate » Fri Oct 07, 2011 4:18 pm

I was hoping someone more expert would chime in, but I think you are out of luck. ROMS is by nature sensitive to incoming velocity, not so much to pressure gradients from boundary conditions. There are other models with a pressure solve in which pressure is a primary variable - they likely could do the job easily.

rduran
Posts: 139
Joined: Fri Jan 08, 2010 7:22 pm
Location: Theiss Research

Re: periodic boundary condition

#12 Post by rduran » Sat Oct 08, 2011 3:50 am

Good to know, thanks!

simion1232006
Posts: 60
Joined: Tue Sep 29, 2009 3:50 pm
Location: School of Environment System Engineering,UWA

Re: periodic boundary condition

#13 Post by simion1232006 » Sun Oct 09, 2011 6:23 am

Thank you very much. Kate.
I will reconsider it.

User avatar
wilkin
Posts: 511
Joined: Mon Apr 28, 2003 5:44 pm
Location: Rutgers University
Contact:

Re: periodic boundary condition

#14 Post by wilkin » Sun Oct 09, 2011 8:29 pm

Yes, there is a way to apply a net pressure gradient but to do this but you need to make a couple of small changes to the code. By default ROMS will make the pressure field periodic and you don't want that. So you need to disable that.

The along-channel pressure field forcing can then be introduced by including it in the atmospheric pressure forcing data (which is already being read my the model if you are using BULK FLUXES) and activating the use of this pressure in the pressure gradient calculation (#define ATM_PRESS).

Gordon Zhang used this approach to impose an along-shelf pressure gradient in a "2-D" (i.e. a very short periodic channel) idealization of the Mid Atlantic Bight circulation published here:
http://journals.ametsoc.org/doi/pdf/10. ... 10-05009.1

Perhaps Gordon will comment here in the Forum on the precise details of which routines needed changing.
John Wilkin: DMCS Rutgers University
71 Dudley Rd, New Brunswick, NJ 08901-8521, USA. ph: 609-630-0559 jwilkin@rutgers.edu

simion1232006
Posts: 60
Joined: Tue Sep 29, 2009 3:50 pm
Location: School of Environment System Engineering,UWA

Re: periodic boundary condition

#15 Post by simion1232006 » Tue Oct 18, 2011 4:07 am

Thank you.Wilkin.
Also did you post the right link? I did not see any things related to periodic boundary with along-shelf pressure gradient in this paper. Also it is in Hawaiian , not Mid Atlantic Bight.
I really appreciate if you can send me a good link. Currently , I am working on the instability of upwelling front,and i want to have a look at the influences of the flow shear. So if i can apply a pressure gradient on periodic boundary, I might be able to get the solution.(The pure periodic boundary condition for upwelling cases are too typical and people have already looked at it)

I actually looked at the code and found the Pair variable has not been used within any periodic boundary loops. Does any one happens to know which scripts contain the boundary exchanging scheme (exchange2d.f maybe)? I still will work on it and hope to get some suggestions.

User avatar
wilkin
Posts: 511
Joined: Mon Apr 28, 2003 5:44 pm
Location: Rutgers University
Contact:

Re: periodic boundary condition

#16 Post by wilkin » Tue Oct 18, 2011 5:31 am

John Wilkin: DMCS Rutgers University
71 Dudley Rd, New Brunswick, NJ 08901-8521, USA. ph: 609-630-0559 jwilkin@rutgers.edu

simion1232006
Posts: 60
Joined: Tue Sep 29, 2009 3:50 pm
Location: School of Environment System Engineering,UWA

Re: periodic boundary condition

#17 Post by simion1232006 » Tue Oct 18, 2011 7:47 am

Sincerely appreciate your favor. Yes, this is the right one. I will have a look.

--------------------------------------------------------------------------------
I get a reply from Dr zhang. It works for me now.I can see the atmosphere pressure driving flow now even with periodic boundary condition.
Thanks to everyone in the forum.

Post Reply