[MITgcm-support] Possible OBCS bug

[Martin Losch]

Hi Jeff,

thanks for reporting this bug. It's now fixed. (use cvs -q update to  
"feel" the fix)

Also you describe the implications correctly, if one undefines  
ALLOW_OBCS_SOUTH (which is, however not necessary even for E-W  
channels). Also this problem only appears, when one is turning on  
(and using) the non-hydrostatic code.

Martin

On 26 Nov 2008, at 15:39, Blundell J.R. wrote:

> As rather an MITgcm newbie, I'm not sure if this is the correct  
> mechanism for bug reporting.
> However, here goes:
>
> I think there is an error on line 86 of obcs_apply_w.F in /pkg/obcs
> (the first line of the following block):
>
> #ifdef ALLOW_OBCS_SOUTH
>       DO J=1-Oly,sNy+Oly
> C Western boundary
>        I_obc=OB_Iw(J,bi,bj)
>        IF (I_obc.NE.0) THEN
>         obc_mask = _maskW(I_obc+1,J,K,bi,bj)
>         wFld(I_obc,J,K,bi,bj)=OBWw(J,K,bi,bj) * obc_mask
>        ENDIF
>       ENDDO
> #endif
>
> Surely ALLOW_OBCS_SOUTH should be ALLOW_OBCS_WEST ?
> This may have some bearing on recent struggles with channel models  
> (Dave Hebert?)
> in that if you are setting up a W-E channel with solid walls N & S,  
> you will presumably
> have ALLOW_OBCS_SOUTH and ALLOW_OBCS_NORTH undefined, so as to get
> solid walls there. This will have the unfortunate side effect of  
> turning off the western BC.
> Any thoughts?
>
>                                                         Jeff Blundell
>
> ======================================================================
> |                    e-mail:  jeff at noc.soton.ac.uk                   |
> |   Jeff Blundell,  Room 256/09  |  Ocean Modelling & Forecasting,   |
> |    phone: +44 [0]23 8059 6201  |  National Oceanography Centre,    |
> |     fax : +44 [0]23 8059 6204  |  Southampton, Empress Dock,       |
> |                                |  SOUTHAMPTON SO14 3ZH, UK.        |
> |               WWW:  http://www.noc.soton.ac.uk/omf/                |
> ======================================================================
>
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support