3D diagnostics

[ericaligo-NOAA]

Hello, I used to be able to write the 3D tendencies from the physics and dynamics by setting ldiag3D = .true. and qdiag3d = .true. I added the the lines below to my diag_table, however, none of these fields were written out. Did this process change?

"gfs_phys", "dt3dt_phys", "dt3dt_phys", "fv3_history2d", "all", .false., "none", 2
"gfs_phys", "dt3dt_nophys","dt3dt_nophys","fv3_history2d", "all", .false., "none", 2
"gfs_phys", "dv3dt_phys", "dv3dt_phys", "fv3_history2d", "all", .false., "none", 2
"gfs_phys", "dv3dt_nophys","dv3dt_nophys","fv3_history2d", "all", .false., "none", 2
"gfs_phys", "du3dt_phys", "du3dt_phys", "fv3_history2d", "all", .false., "none", 2
"gfs_phys", "du3dt_nophys","du3dt_nophys","fv3_history2d", "all", .false., "none", 2
"gfs_phys", "dq3dt_phys", "dq3dt_phys", "fv3_history2d", "all", .false., "none", 2
"gfs_phys", "dq3dt_nophys", "dq3dt_nophys", "fv3_history2d", "all", .false., "none", 2
"gfs_phys", "dq3dt_o3phys", "dq3dt_o3phys", "fv3_history2d", "all", .false., "none", 2
"gfs_phys", "dq3dt_o3nophys","dq3dt_o3nophys", "fv3_history2d", "all", .false., "none", 2

[dustinswales]

@ericaligo-NOAA Did you check if the tendencies are in the sfcf*.nc history files, not the atmf*.nc files?

[ericaligo-NOAA]

Hi, I think I just figured it out. I was reading the old instructions. The new instructions indicate the diag_table names are different now. For example, temperature tendencies from physics and dynamics would be written as follows:

"gfs_phys", "dtend_temp_phys", "ttendphys", "fv3_history", "all", .false., "none", 2
"gfs_dyn", "dtend_temp_nophys","ttendnophys","fv3_history", "all", .false., "none", 2

What's interesting/odd is that the dtend_temp_nophys temperature tendency still gets placed in the phys file. Note my RRFS output is either phy.nc or dyn.nc. Is this correct that even though the non physics temperature tendency is specified under gfs_dyn that it would get placed in the phy.nc file?
Eric

[dustinswales]

My understanding is that the physics diagnostics get placed into the sfcf*.nc, previously the phys*.nc files, regardless if the fields are 2D or 3D. This makes sense with the old dyn/phys file naming convention, but not so much with the sfcf/atmf file naming convention.

[ericaligo-NOAA]

Thanks. I've never seen a situation in which the field is listed under the 'gfs_dyn" column and shows up in the phy.nc file. Normally in this case it would show up in the dyn*nc file. I just want to make sure this is working correctly.