[MITgcm-support] CD scheme + thin walls = help!

[Christopher L. Wolfe]

Hello modelers,

I'd like to run a simulation in a simple domain with "thin" walls so that I can freely vary the resolution without varying the extent of the topography. Most of the simulations are fairly coarse (80km), so I'd like to use the CD scheme to suppress the grid noise associated with the Coriolis terms at low resolution. The intro to cd_code_scheme states that the CD scheme won't work with thin walls "as coded" and I do, in fact, see noise near the walls with CD scheme + thin walls that I don't see when I use either individually. I'd like to modify the scheme so accommodate thin walls (at least in my special case), but I can't for the life of me figure out where the scheme is going wrong on account of the thin walls. Could someone (perhaps Alistair or Jean-Michel?) give me a hint?

More generally, should I be using the CD scheme at all? From scanning conversations on the mailing list, it seems like people have a bad opinion of the CD scheme, but why? If you don't use the CD scheme, how do you control the low-res C-grid noise, aside from beating at the grid noise with explicit (biharmonic?) dissipation?

Many thanks,
Christopher

-----------------------------------------------------------
Dr. Christopher L. Wolfe              	   858-534-4560
Climate, Atmospheric Science, and Physical Oceanography
Scripps Institution of Oceanography, UCSD  clwolfe at ucsd.edu
-----------------------------------------------------------





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20101103/0d0b5f91/attachment.htm>