[MITgcm-support] to give delX=360*0.25, delY=360*0.25 in curvilinear coordinates in MITgcm

Martin Losch Martin.Losch at awi.de
Mon Oct 1 03:51:22 EDT 2018


Hi Yi-Chih,

curvilinear grids are not so trivial, once you've set usingCurvilinearGrid=.TRUE., you don’t need any delX/Y yOrigin information anymore, because all of that should be in the grid files. I suggest that you have a look at ini_curvilinear_grid to understand what the model expects and then generate your files.
I found it easier to use the “OLD_GRID_IO” option, if you are not interested in cubed sphere of LLC configuration, but that’s a matter of taste (and what you are used to).

Martin


> On 24. Sep 2018, at 15:37, Yi-Chih Huang <dscpln at gmail.com> wrote:
> 
> Hello,
> 
>     We conduct idealized simulations to study atmospheric convection near the north pole.  To avoid the singularity problem in spherical coordinates, we are interested in curvilinear orthogonal coordinates in MITgcm.  I don't need blank tile and do not include exch2 in cope/packages.conf.  I use MPI to expedite running processes.  If I didn't specify delX and delY or gave the delX and delY below in the data file, I got normal end after execution the mitgcmuv.  When giving delX=360*0.25, delY=360*0.25, I got the error message looking for tile002.mitgrid as the following.  How to explain these?  In the manual it said the grid spacing can not be set via the namelist; instead, the grid descriptors are read from data files, one for each descriptor.  The forcing data is Tinit.bin so we don't need extra external forcing as exf.package.  I wonder what is the right way to give delX=360*0.25, delY=360*0.25 in curvilinear coordinates in MITgcm?    
> 
> forrtl: severe (29): file not found, unit 9, file /project/6001267/MITgcm/MITgcm_c66k/verification/rotating_tank2/run_n_360/tile002.mitgrid
> 
> # Gridding parameters
>  &PARM04
>  usingCurvilinearGrid=.TRUE.,
>  deepAtmosphere=.TRUE.,
>  rSphere=60000.E3,
>  delR=50*200.E3,
>  delX=320*0.25,
>  delY=320*0.25,
>  ygOrigin=0.,
>  &
> 
>     Thanks much,
> 
>                          Yi-Chih
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support



More information about the MITgcm-support mailing list