[MITgcm-support] OBCS：balance the flow on the open boundaries

[Matthew Mazloff]

If you are using partial cells you need to use the hFacs. 
Dump the model grid and use:

V*DRF*DYG*hFacS
U*DRF*DXG*hFacW

And watch the corners -- best to just set U and V = 0 at the corners. Or make it a land point. And for the west and south you should step one index in, e.g.:
V*DRF*DYG(:,2)*hFacS(:,2,:)

This will work is you are very careful with the gridding. Any inconsistencies between your method and what the model does will make this not work.

Matt


On Oct 6, 2014, at 12:22 AM, 刘成彦 <killy at nuist.edu.cn> wrote:

> 
> Dear all:
> 
> I was still trying to design an experiment with proper open boundary conditions. In my experiment, I set the Western boundary, the Eastern boundary, and the
> Southern boundary as open boundaries. In order to keep the volume conservation, the inflow should be equal to the outer flow, by integrating the orthogonal
> velocity along the open boundaries.
> 
> After reading the former emails in the The MITgcm-support Archives, I learned that formers suggested that the velocity in open boundaries could be balanced
> before we started run our case for convenient. In my case, the OBNv*S1+OBWu*S2+OBEu*S3, S1 S2 S3 denotes the areas of the southern western and
> eastern boundaries respectively, should be zero.
> 
> Now the velocity of open boundaries has already been prepared from a global simulation experiment. I have to compute each area of the open boundaries,
> and I faced the problem: How to compute the areas! In the manual, I learn that the MITGCM is on the C grid. Sorry for my poor knowledge of model grid design,
> and I can’t make sure whether my method of compute boundary areas is right or wrong.
> 
> For example, my grid is designed in data as follows:
> &PARM04
> usingSphericalPolarGrid=.TRUE.,
> delX=640*0.0376,
> delY=640*0.02191,
> ygOrigin=-73,
> xgOrigin=66,
> delR = 10.0, 10.0, 10.0, 11.0, 12.5, 14.1, 15.5, 17.5, 18.5,
> 20.5, 22.2, 23.7, 25.1, 26.6, 27.7, 28.5, 29.5, 29.5, 29.2,
> 29.2, 28.5, 27.5, 26.5, 25.5, 24.7, 23.8, 23.0, 22.3, 21.5,
> 21.1, 20.5, 20.0, 19.5, 19.0, 19.8, 20.8, 21.1, 21.3, 21.7,
> 22.3, 23.0, 24.1, 25.5, 27.5, 29.5, 32.5, 35.5, 40.5, 45.5,
> 50.5, 55.5, 60.0, 70.5, 80.5, 90.5, 100.5, 110.5, 120.5, 130.5,
> 140.5, 155.0, 170.0, 185.5, 200.5, 215.0, 230.0,250.0, 250.0, 250.0,250.0,
> &
>  
> in the northern boundary, the inout-flow is calculated by integrating v*dz*dy. V is given in my OBNv* file. My dz is derived from the data, the same value as delR.
> dy is derived from DYG.data, and in my case, I set dy equal to the DYG(:,1:640). The west boundary and east boundary are computed as the similar method.
> 
> Finally, I want to ask whether is my method right or wrong. I think there should be something wrong with my method, because when I balance the flow through
> my method, the volume does not keep conservative.
> 
> Thank you for your considerations!
> 
> Cheers
> Liu
> 
> 
> 
> ----------------
> *********************************************
> Dr. Chengyan Liu
> Nanjing University of Information
> Science & Technology
> No.219, Ningliu Road, Nanjing, Jiangsu,
> China(Postcode: 210044)
> Phone: 18251854735
> E-Mail: killy at nuist.edu.cn
> 
> *********************************************
>  
> 
> _______________________________________________
> 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/20141006/0896d770/attachment.htm>