davbyr
commented
3 years ago From Nico:
I did follow Maria's experiments and at the end went to the less diffusive scheme. Cannot throw you the numbers like this though... Especially as I don't have access to my emails and all exchanges I had with Maria. It was just a couple of changes in namzdf_gls
Hope what is below could help a bit. Jeff might have more details as he co-authors Maria's work on mixing.
From a quick glance (if you check the AMM7 experiment here; folders name should be enough explicit; /projectsa/NEMO/nibrun/SIMULATIONS/AMM7/v3.6_early_ran_test you should see differences from namelist):
I would say something like this is more the stuff recommended by Maria
!----------------------------------------------------------------------- &namzdf_gls ! GLS vertical diffusion ("key_zdfgls") !----------------------------------------------------------------------- ln_length_lim = .true. ln_sigpsi = .true. nn_bc_bot = 1 nn_bc_surf = 0 nn_clos = 1 nn_stab_func = 1 nn_z0_met = 1 rn_charn = 100000.0 rn_clim_galp = 0.267 rn_crban = 100.0 rn_emin = 1.0e-6 rn_epsmin = 1.0e-9 rn_frac_hs = 1.3 rn_hsro = 0.003 /
this is more what used to be in the model before
!----------------------------------------------------------------------- &namzdf_gls ! GLS vertical diffusion ("key_zdfgls") !----------------------------------------------------------------------- ln_length_lim = .true. ln_sigpsi = .true. nn_bc_bot = 1 nn_bc_surf = 1 nn_clos = 1 nn_stab_func = 2 nn_z0_met = 1 rn_charn = 100000.0 rn_clim_galp = 0.267 rn_crban = 100.0 rn_emin = 1.0e-6 rn_epsmin = 1.0e-12 rn_frac_hs = 1.3 rn_hsro = 0.003 /
Regarding a different mixing scheme to CO8_AMM15 (NEMOv3.6).
From NOWMAPS 2019 report: "Over the next cycle of improvements the NEMO codebase was upgraded to the stable release v4.0.1; Wetting and Drying (V1.1.5 above) was implemented, and a physically motivated improvement to shelf sea vertical mixing scheme was implemented (V1.1.6)"
@davbyr have you found trace of Nico's choices/settings for the vertical mixing in AMM15? This needs a separate test and therefore ticket to the other things.
AC: