vertical mixing coefficient for ana_vmix

[logvinov]

Hi all,

In my problem I use ANA_VMIX, where I define AKv and AKt to be equal to constant AKv_bak=1e-3 and AKt_bak=1e-3 respectively.

if defined EKMAN_WALL
DO k=1,N(ng)-1
DO j=JstrR,JendR
DO i=IstrR,IendR
Akv(i,j,k)=Akv_bak(ng)
END DO
END DO
END DO


if defined EKMAN_WALL
DO k=1,N(ng)-1
DO j=JstrR,JendR
DO i=IstrR,IendR
Akt(i,j,k,itemp)=Akt_bak(itemp,ng)
Akt(i,j,k,isalt)=Akt_bak(isalt,ng)
END DO
END DO
END DO

But in output his.nc file the Akv an Akt are constant in all the domain but for the very top and bottom points where they equal 0.

Which makes me wandering what kind of Akv is used to satisfy stress and hit flux boundary conditions?

Thanks a lot,
Evgeny

[jcwarner]

AKv and AKt are prescribed at vertical "w" points, so they range from k=0 to N. In the ana file, you are setting the values of AKv and AKt from 1 to N-1. That is why you get zeros at the top and bottom (see Modules/mod_mixing.F).

The mixing occurs in step3d_t for tracers, and step3d_uv for momentum.

[logvinov]

Thanks a lot John!