jdha
commented
3 years ago From Romain B-B:
"same than in TKE (mixing lenght depending on ice thick) but in GLS we can not impose a mixing length in the first level as TKE, so we impose an unrealistic roughness wich will generate the wanted mixing length... As in TKE, this "ugly trick" solves the problem in artic (because of tidal mixing now the background is very low in artic ocean and fresh water is trapped in the first layers) and it has solved also my polinia problem in antarctic !... I didn't understood really why, I suspect because with the very hudge friction there is a slowdown of ocean/ice and a decrease of to strong shear in antartica...."
chris-O-wilson
jasontempestholt
Romain B-B namzdf_gls:
!----------------------------------------------------------------------- &namzdf_gls ! GLS vertical diffusion (ln_zdfgls =T) !----------------------------------------------------------------------- rn_emin = 1.e-7 ! minimum value of e [m2/s2] rn_epsmin = 1.e-12 ! minimum value of eps [m2/s3] ln_length_lim = .true. ! limit on the dissipation rate under stable stratification (Galperin et al., 1988) rn_clim_galp = 0.267 ! galperin limit ln_sigpsi = .true. ! Activate or not Burchard 2001 mods on psi schmidt number in the wb case rn_crban = 100. ! Craig and Banner 1994 constant for wb tke flux rn_charn = 70000. ! Charnock constant for wb induced roughness length rn_hsro = 0.02 ! Minimum surface roughness rn_hsro_i = 0.02 ! Minimum surface roughness under ice rn_frac_hs = 1.3 ! Fraction of wave height as roughness (if nn_z0_met>1) nn_z0_met = 2 ! Method for surface roughness computation (0/1/2/3) ! ! =3 requires ln_wave=T nn_z0_met_i = 3 ! Method for surface roughness computation under ice (1/2/3) nn_bc_surf = 1 ! surface condition (0/1=Dir/Neum) nn_bc_bot = 1 ! bottom condition (0/1=Dir/Neum) nn_stab_func = 2 ! stability function (0=Galp, 1= KC94, 2=CanutoA, 3=CanutoB) nn_clos = 1 ! predefined closure type (0=MY82, 1=k-eps, 2=k-w, 3=Gen) /