[MITgcm-support] cubed sphere grids

[Maura Brunetti]

Hello,

I have now a working version of matlab scripts for constructing the cubed sphere grids.
I have compared what I obtain with the grids in
MITgcm/verification/tutorial_held_suarez_cs/input and I have found that in order to
reproduce the tutorial grids I need to use the `conf' method and nratio = 16.
But still there are differences (for example, in LATG, LONG, DYG and DXF I have
relative errors (LONG_MIT - LONG)/LONG of the order of 10ˆ(-5) and all the other quantities
have relative errors of the order of 10ˆ(-2), 10ˆ(-3), showing differences at
the corners of the cube faces).

If I put the forth coefficient in the
matlab script map_xy2xyz.m to the value -0.00895884801823 instead of
-0.01895884801823 (as suggested in
http://mitgcm.org/pipermail/mitgcm-support/2012-May/007759.html )
I obtain a much better agreement:
relative errors for LATG, LONG, DYG, DXF become of the order of the
machine precision, 10ˆ(-15), while all the other
quantities have relative errors as before (of the order of 10ˆ(-2), 10ˆ(-3) with
larger differences at the corners of the cube faces).

Since I need to construct other grids for my paleoclimate simulations
(I am studying theJurassic period), I would like to know which is the typical configuration
that you use to construct your grids:
do you use the `conf' method?
which is the best value for nratio? Does it depend on the grid resolution?
which is the correct coefficient in map_xy2xyz.m?

Thank you for your help,

Maura



--
Dr. Maura Brunetti
Institute for Environmental Sciences (ISE)
University of Geneva
Site de Battelle / D, 7 route de Drize
CH-1227 Carouge / GE -- Switzerland

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20131220/a7596afc/attachment.htm>