Hello,
I’m using a fork of 4.2.2 (here) to set up a “hybrid-resolution” simulation using AGRIF, where an eORCA025 domain is the “child” and the “parent” is coarsened by a factor of 3 (i.e. 0.75 degree), as in described here .
After running theDOMAINcfgtool, we found that some of the bottom layers in our coarsened/parent domain are extremely thin, as low as 0.44mm, which we’re trying to rule out or fix as the cause of a blowup in the subsequent run.
I attempted modifying the values of rn_e3zps_min and rn_e3zps_rat in the namelist, and can see in the output that the values have been updated, but it appears these values are ignored, as the output is bitwise identical regardless of their value.
Is this feature still supported?
Thanks!
Dafydd