[MITgcm-support] MITgcmutil - cubed sphere - creating a broken line file

[Jean-Michel Campin]

Hi Flynn,

There has been some updates to this broken-line script in Oct 2023 (PR #729) plus some
fixes applied in Jun 2024 (PR #792) with a simpler way of generating the brokenr-lines
but also using a different convention to store it in the broken-line file (I think the
default file name has changed too, to reflect the fact that it's in a different format).

So if you use an old broken-line file (generated from scripts in 
 MITgcm/utils/matkab/cs_grid/bk_line/ )
you can continue to use this file the same way as before ; 
But if you use a new broken-line file (generated from script MITgcm/utils/matkab/cs_grid/mk_isoLat_bkl.m ), you will need to adjust the way you use it (when computing the stream function) 
as done in the example use_isoLat_bkl.m (you may have to apply the same "reshape" as done
in "use_isoLat_bkl.m", lines 92-96).

Let us know if this is your case and if you need more details to fix it.

Cheers,
Jean-Michel

On Wed, May 14, 2025 at 03:31:18PM +0000, Flynn Ames wrote:
> Dear MITgcm people,
> 
> Hello - I've run the MITgcm in a cubed sphere configuration for two different resolutions: cs32 and cs96. I am in the process of analysing the output, specifically computing the meridional overturning streamfunction, and have been using the MITgcmutils cubed-sphere MATLAB functions for this.
> 
> To compute the overturning streamfunction, a 'broken line' file is required. I have one of these for cs32 (given to me by my supervisor) and have used this successfully to compute the overturning streamfunction (see attached figure). I don't have a broken line file for the cs96 configuration, so I created an isoline for this configuration using 'mk_isolat_bkl.m'.
> 
> My problem: I get nonsense output when trying to compute the overturning stream function from my cs96 configuration, using the broken line file I created.
> Furthermore, as a test, I tried recreating the broken line file for the cs32 configuration. When I use this file (instead of the file I had been given which works), I get nonsense output here too (see attached figure). This suggests there is something wrong with how I'm creating the file.
> 
> My question: Does anyone have any ideas for what could be going wrong?
> 
> Inspecting the broken line files, the only major difference I can see is that the parameter bkl_IJuv (which lists the indexes of velocity points to assign to a broken line file) looks quite different when I plot it (see attached figures). Not sure if this would be causing the issue without delving in deeper.
> 
> I use the same number of points in latitude for creating the cs32 broken line file (59) as the working file I was given, so would have expected the two files to be practically identical.
> 
> 
> Thanks in advance!
> Flynn





> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support