MOM6 1deg parameterization 2 - lateral friction

[minghangli-uni]

In the 1deg setup of MOM6, Leith nonlinear eddy viscosity was employed instead of the Smagorinsky eddy viscosity used in MOM5. Any reasons behind why this decision was made? The relevant Leith parameters are listed below,

 631 AH = 1000000000000.0 (not sure)
 632                                 ! "[m4 s-1] default = 0.0
 633                                 ! The background biharmonic horizontal viscosity."
 634
 635 AH_VEL_SCALE = 0.0 (not sure)
 636                                 ! "[m s-1] default = 0.0
 637                                 ! The velocity scale which is multiplied by the cube of
 638                                 ! the grid spacing to calculate the biharmonic viscosity.
 639                                 ! The final viscosity is the largest of this scaled
 640                                 ! viscosity, the Smagorinsky and Leith viscosities, and AH."
 641
 642 LEITH_AH = True (not sure)
 643                                 ! "[Boolean] default = False
 644                                 ! If true, use a biharmonic Leith nonlinear eddy viscosity."
 645
 646 LEITH_BI_CONST = 128.0  (not sure)
 647                                 ! "[nondim] default = 0.0
 648                                 ! The nondimensional biharmonic Leith constant, typical values are thus far
 649                                 ! undetermined."
 650

Parameters associated with Smagorinsky eddy viscosity are provided for reference, with the reference given in paper The GFDL Global Ocean and Sea Ice Model OM4.0: Model Description and Simulation Features,

SMAG_BI_CONST = 0.06
                                ! "[nondim] default = 0.0
                                ! The nondimensional biharmonic Smagorinsky constant,
                                ! typically 0.015 - 0.06."
SMAGORINSKY_KH = True           
			        !   [Boolean] default = False
                                ! If true, use a Smagorinsky nonlinear eddy viscosity.
SMAG_LAP_CONST = 0.15           
			        !   [nondim] default = 0.0
                                ! The nondimensional Laplacian Smagorinsky constant, often 0.15.
AH_VEL_SCALE = 0.01 (both 1deg and 0.25deg)             
			        !   [m s-1] default = 0.0 
			        ! The velocity scale which is multiplied by the cube of the grid spacing to calculate the biharmonic viscosity. The final viscosity is the largest of this scaled viscosity, the Smagorinsky and Leith viscosities, and AH.
KH_VEL_SCALE = 0.01 - 1deg / 0.0 - 0.25deg           
			        !   [m s-1] default = 0.0
                                ! The velocity scale which is multiplied by the grid spacing to calculate the
                                ! Laplacian viscosity. The final viscosity is the largest of this scaled
                                ! viscosity, the Smagorinsky and Leith viscosities, and KH.

[aekiss]

All the parameters were inherited from the CESM GMOM_JRA configuration this was based on, so I don't know the reasons for these choices. We should feel free to modify these.

Our other MOM6 configurations use Smagorinsky biharmonic (smagorinsky_ah) rather than Leith biharmonic (leith_ah) - see this table. ACCESS-OM2 used isotropic biharmonic Smagorinsky - see section 3.2.5 of the tech report.

For details see the mom_hor_visc module reference.

According the GFDL OM4 paper only their 0.5° config used Laplacian Leith (SMAGORINSKY_KH) in addition to biharmonic Leith.

[dougiesquire]

Note, this was discussed in a meeting with notes here: https://forum.access-hive.org.au/t/namelist-configuration-discussion-meeting/1917