[MITgcm-support] gluemnc truncates Xp1 and Yp1 to the length of X and Y

[Martin Losch]

Hi Roland,

I haven’t used gluemnc in a while because the python script utils/python/MITgcmutils/scripts/gluemncbig is so much faster. As far as I can see, gluemncbig preserves the “mnc-service” of including the n+1st velocity point.

M.

> On 18 Jun 2015, at 11:32, Roland Young <Roland.Young at physics.ox.ac.uk> wrote:
> 
> Hi all,
> 
> I am using gluemnc to combine 16 tiled files into a global file. This works generally fine, except for the following:
> 
> My input files (one per tile, as standard) include 
> 
> U (Xp1,Y,Z)
> V (X,Yp1,Z)
> 
> with nx = 256 and ny=128 globally in my case (so 257 X-points in U and 129 Y-points in V). However, the .glob.nc output truncates Xp1 and Yp1 to 256 and 128 grid points respectively, so the easternmost grid point in U and the northernmost grid point in V are cut off.
> 
> Is gluemnc supposed to do this?  
> 
> In my case the domain is global so U(nx+1,y,z)=U(1,y,z) and V(x,1,z)=V(x,ny+1,z)=0, so in practice I can get around this, but I have a load of scripts that use rdmnc to read in the tiled output, which preserves Xp1 and Yp1 correctly, and I want to avoid having to go through and correct them for single-file input.
> 
> Thanks,
> 
> Roland
> 
> 
> 
> 
> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support