Difference in h and q points in MOM.res.in

[abhisekc]

Hi all,

I have recently noticed that there is an extra grid point for the "q" location than the "h" location in the restart files. For example, the dimensions of a restart file from a global MOM6 test simulation are shown as below:

netcdf MOM.res {
dimensions:
lath = 210 ;
lonh = 360 ;
latq = 211 ;
lonq = 361 ;
Layer = 63 ;

Note that there is an extra point for lonq and latq compared to it's tracer points.

However, in a different HPC, I didn't find any difference between lonh and lonq (and in lath and latq).

I believe this difference has something to do with how the MOM6 is built in a specific HPC. Can someone please suggest how to build the MOM6 so that q points and h points remain identical?

Thanks,
Abhisek

 

[kshedstrom@alaska_edu]

You have somehow switched from non-symmetric mode to symmetric mode. Yes, it is in how MOM6 is
compiled, or rather linked. How are you compiling? Search for "dynamic_symmetric".
For global simulations it really doesn't matter, but all the regional domains have to be in symmetric mode.

You want to link to files under $(MOM6)/config_src/memory/dynamic_nonsymmetric
not those under $(MOM6)/config_src/memory/dynamic_symmetric

 

[abhisekc]

Dear kate,

Thanks for the response. Yes, switching to nonsymmetric worked.

Best,
Abhisek