[MITgcm-devel] smoothing package for dimensional controls

[Matthew Mazloff]

Hi Gael,

Can we revert the change:
- bug fix : in this case we want to do the full
   diffusion operator (as opposed to its square
   root in smooth_correl2d.F).
where the factor of 1/2 was dropped:

revision 1.1 by gforget, Fri May 20 00:52:47 2011 UTC	revision 1.2 by  
gforget, Fri May 20 18:41:08 2011 UTC
48	       nbt_in=smooth2Dnbt(smoothOpNb)/2	        
nbt_in=smooth2Dnbt(smoothOpNb)

in smooth2d.F and smooth3d.F.  I would still like to take the square  
root of the operator, so at least the length scales are consist with  
ALLOW_SMOOTH_CORREL.

Do you have any objections to reverting?  I think it makes sense as we  
are still operating on the way out (adxx) and the way in (xx).

Thanks
Matt

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-devel/attachments/20110628/8eafa7f6/attachment-0001.htm>