[MITgcm-support] Alternate Grid Package!
Andrea Molod
molod at ocean.mit.edu
Tue Aug 28 11:56:26 EDT 2007
hi van thinh,
On Tue, 28 Aug 2007, Van Thinh Nguyen wrote:
> Hi Andrea,
>
> Thanks for your info,
> The idea is almost the same as atmosphere, however I just want to apply for
> an ocean modeling.
> I would like to run different components of the MItgcm, one with a coarse
> resolution for a subdomain far away from a topography/boundary, and one with
> a higher resilution for a subdomain near a boundary/topography (3D).
> Could you please tell me which routines, shall I work on them? I will be
> thankful if you could give me some testcases what you have worked on.
>
> Thanks a lot!
>
> Van Thinh
if you take a look at MITgcm_contrib/plumes
you will see how the routines look for an implementation that does
what you are talking about except that the 'higher resolution' is
in the horizontal only. in the routine do_plume.F there is a call
to a routine called 'plume' which is a stub for the sequence of model
routines that you would want to call at the higher resolution.
plume_wrapper is the driver routine for this implementation of gridalt
(in the checked in version the routine is fizhi_wrapper)
this routine calls the interpolation routines and do_plumes and steps
the higher resolution fields forward with the tendencies.
do_plumes basically is a little driver for the 'plume' call. (see
do_fizhi)
plume_tendency_apply.F sees the commons with the tendencies from
the high res grid after they have been interpolated to the coarse
grid and contains the routines called from the model to add the
tendencies to the total model tendency.
if what you are talking about is running the whole model at the higher
resolution, that is a whole different story and requires a lot more
effort. that requires keeping track of different 'versions' of the model
at different resolutions on different processors, probably, and keeping
track of the communications. way beyond what i can advise you about.
good luck with all this, and if i can help you more with this,
let me know.
andrea
More information about the MITgcm-support
mailing list