Make vertical levels for FHS94 consistent for all dycores.

Christiane Jablonowski has made us aware that the vertical levels used for the FHS94 compset (30 levels by default) are different for EUL than for the FV and SE dycores. This difference leads to different climates which can be clearly seen in the zonal average temperature field. The cause of the difference appears to be that the held_suarez_1994 use case file contains initial files for the EUL dycore which contain an equally spaced pure sigma coordinate rather than the default 30 level hybrid coordiate used by the cam5 physics. The initial files for FV and SE are coming from namelist_defaults_cam.xml and contain the standard 30 level hybrid coordinates.

I think the best fix is to remove the 30 level Eulerian IC files from held_suarez_1994 and replace them by the 30 level vertical coordinate file which is dycore independent. I.e., replace

atm/cam/inic/gaus/HS1994.128x256.L30_c062216.nc
atm/cam/inic/gaus/HS1994.64x128.L30_c061616.nc

by

atm/cam/inic/gaus/cam_vcoords_L30_c180105.nc