Initialization file being read incorrectly

[bperfect]

Has anyone encountered this issue in the past?

I'm using a butchered version of pyroms to write my own grid and init files. Upon inspection, the netcdf files look like everything is in place, but when ROMS reads them in, it seems to skip some fields, most notably the zeta field. So instead of a geostrophically balanced free surface, I'm getting a flat surface.

[kate]

You can look at the output of GET_STATE to see when it reads each field (or not). You can look at get_state.F to see when it thinks it should read each field.

If it's just zeta that it's skipping, does it really matter?