[MITgcm-support] How to glue mnc output files with the cube-sphere exch2 grid ?
Martin Losch
Martin.Losch at awi.de
Thu Nov 21 05:53:49 EST 2013
Hi Alexandre,
this is exactly the expected result. This experiment uses the cubed sphere that does not have the simple connectivity of a regluar lon-lat-grid, so you cannot simply glue the faces together in the same way in that you cannot create a 2D cube.
There are some matlab or python functions in the utils directory that you can use to visualize the output of scalar fields. E.g. in matlab:
t = rdmds('THETA'); % gives you a field 192x32x15 field, right?
xg=rdmds('XG'); % 192x32
yg=rdmds('YG'); % 192x32
then
cs_pcol(xg,yg,theta(:,:,1)) % I am attaching this function, as it does not seem to in the repository
should give you the surface temperature on a simple cyclindrical map projection.
There is more in the utils/matlab/cs_grid directory that maybe interesting for you.
in python something like (I am using output of global_ocean.cs32x15 here):
import MITgcmutils as mit
g=mit.rdmnc('grid.*')
s=mit.rdmnc('state.*','Temp')
mit.cs.pcol(g["XG"],g["YG"],s["Temp"][0,0,:,:])
Martin
On Nov 21, 2013, at 11:32 AM, Alexandre Pohl <Alexandre.Pohl at lsce.ipsl.fr> wrote:
> Hi,
>
> I ran the verification/cpl_aim+ocn experience and I produced NetCdf files.It uses a cube-sphere Exch2 grid and I got one file per face (6 files).
>
> I previously used gluemnc to glue the cartesian output files together, but it does not work with the exch2.
>
> I also tested the rdmnc.m script. It extracts the variables from the 6 tiles but the result is not correct. The faces are glued to each other in no particular order. I join the result. We can see Africa on the left and Antarctica on the right.... not the expected result!
>
> Is there a way to do that, like gluemnc does for a cartesian grid?
> I think you probably encountered this problem before?
>
> Thanks a lot and with kind regards,
> Alexandre<Assemblage-cube_sphere_avec_rdmnc.eps>_______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cs_pcol.m
Type: application/octet-stream
Size: 6069 bytes
Desc: not available
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20131121/6efd659f/attachment.obj>
More information about the MITgcm-support
mailing list