[MITgcm-devel] grid rotation

[Jean-Michel Campin]

Hi Martin,

I don't know if grid-rotation will often be used with curvilinear 
grids (although, I needed to turn the CS-grid and ended up reading 
coriolis map from files, which can be done but is not very 
convenient, plus fCoriG has those missing corners issue).
but as long as xc,xg and yc,yg are longitude and latitude,
which is what we assume (e.g., when Coriolis is computed),
it could also be applied to curvilinear grids.

And I agree with you on CALC_ANGLE, call this S/R if the
angles are not provided in the grid files.

Thanks,
Jean-Michel

On Fri, Mar 06, 2009 at 08:41:27AM +0100, Martin Losch wrote:
> Hi Jean-Michel,
>
>> Just a comment: regarding the the grid-rotation calls, them
>> could have been put in ini_grid.F (and then used also with
>> curvilinear grids, when x & y are lon & lat), no ?
>
> Is the grid rotation relevant to any other but spherical grids? If so,  
> then I agree that it should be moved to ini_grid.F.
> Another related issue: we could call calcAngles (angleCS, angleSN)  
> online also for curvilinear grids, if they are not read ( if 
> .not.anglesAreSet at the end of ini_curvilinear_grid ); currently they 
> are not require except for non-hydrostatic code on a curvilinear grid, 
> but it's nice to have them have the correct values as they are always 
> part of the write_grid.F output (and consequently they are wrong, i.e. 1 
> and 0, in most curvilinear grid cases).
>
> Martin
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel