[MITgcm-devel] please send latest MLI3

[Menemenlis, Dimitris (3248)]

Baylor, thank you for taking a look and for suggestion.
We will compare the heat flux with the ECCO heat flux.

GM-Redi is turned off in the CS510 configuration,
except for the submeso parameterization.
So there is no GM tapering issues.

Dimitris

Dimitris Menemenlis <menemenlis at jpl.nasa.gov<mailto:menemenlis at jpl.nasa.gov>>
Jet Propulsion Lab, California Institute of Technology
MS 300-323, 4800 Oak Grove Dr, Pasadena CA 91109-8099, USA
tel: 818-354-1656;  cell: 818-625-6498;  fax: 818-393-6720

On Apr 7, 2010, at 5:59 PM, Baylor Fox-Kemper wrote:

Hi Dimitris,
  I thought you said the mixed layer was too deep in your model vs. argo, not the opposite!  I do think the changes are what's expected--increased near surface stratification and a shallower ML with submeso on.
  However, I'm puzzled by the issues you have--I wonder if the heat fluxes and/or wind weren't readjusted by ECCO to prevent overly deep mixed layers?  Can you check the raw heat flux vs the ECCO heat flux in this region?
  Otherwise, we found that introducing submeso into MOM & POP is not a clear winner, but it reveals other errors that were previously not an issue.  Conceivably, GM tapering, etc, need to be revisited after submeso is implemented.  (Indeed, in the paper, we spend a lot of time discussing the AMOC issues after the GM tapering in MOM was fiddled with).
  Cheers,
   -Baylor

On Wed, Apr 7, 2010 at 6:47 PM, Menemenlis, Dimitris (3248) <Dimitris.Menemenlis at jpl.nasa.gov<mailto:Dimitris.Menemenlis at jpl.nasa.gov>> wrote:
Baylor, could you take a look at following slides:
hzhang_submeso_tests.ppt<http://ecco2.jpl.nasa.gov/data1/meeting/2010/04_07/hzhang_submeso_tests.ppt>

They compare results on cube sphere with and without your submesoscale parameterization
for mixed layer depth and for temperature profiles in the CLIMODE region.

We modified your MITgcm code to add the cutoffs you discuss in appendix A of your paper.
(Without it, the model crashed.)

We want your opinion on

a) is the mixed layer depth pattern of change consistent with what you expect
    (seems to be of right order of magnitude compared to results in your paper) and

(b) unfortunately in the CLIMODE region, things are getting worse rather than
     better. So something else is missing. Any suggestions?
    (Redi mixing below the mixed layer?).

I cc the devel list if anyone else has experience with the submeso parameterization.

D.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-devel/attachments/20100407/02e09a20/attachment.htm>