[MITgcm-support] Question about the grid of LLC90 (ECCO V4)

[Martin Losch]

Hi Fei,

when distribute the llc90 domain onto 30x30 tiles, you get 117 (=3*3*13) tiles. Some of these tiles are empty, i.e. they do not contain any ocean. These tiles can be listed in data.exch2: blanklist, and in SIZE.h you can then reduce the number of “tiles” nSx*nPx. In you configuration, you obviously have 21 blank tiles.

I found this helpful:
https://wiki.math.uwaterloo.ca/fluidswiki/index.php?title=MITgcm_tips#EXCH2

with exch2, you run your configruation with the full number of tile, ie:

sNx=sNy=30
nSx=nSy=1
nPx = 117
nPy = 1

then find the empty tiles like this:

grep Empty STDOUT.*

this gives you a list of empty tiles and more, which you need to remove before you can use the list in data.exch2, blanklist	

Martin


> On 30. Nov 2019, at 03:11, 彭启华 <979583702 at qq.com> wrote:
> 
> Hi everyone,
> I am running the LLC90 configuration (from ECCO V4). I have a question about the LLC90 grid :  
> 
> The size of LLC90 should be 90*(90*13=1170), but in the SIZE.h, the  Nx  = sNx*nSx*nPx=30*96, Ny=30.The total grid number of Nx*Ny is not equal to 90*1170 (the total grid of LLC90).
> 
> I don't know how the Nx*Ny comes out. I'd like to modify the sNx, nSx, nPx, but I don't know what principle (rule) I should follow.
> 
> Thank you so much.
> 
> Best,
> Fei
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support