# What variable to use for diffusion flux?

#### juliettelavoie

##### New Member
Does DIA_IMPVF_TEMP (TEMP Flux Across Bottom Face from Diabatic Implicit Vertical Mixing) include all vertical diffusive flux in the ocean?

If yes, what is KPP_SRC_TEMP (TEMP tendency from KPP non-local mixing term) ? The diffusive flux calculated from KPP_SRC_TEMP seems very large.

#### klindsay

##### CSEG and Liaisons
Staff member
Juliette,

DIA_IMPVF_TEMP only includes the tracer gradient dependent term of diabatic (cross-isopycnal) mixing.

KPP_SRC_TEMP arises from a term in the KPP vertical mixing scheme that is independent of tracer gradient. It is sometimes referred to as the non-local mixing term.

Please note that DIA_IMPVF_TEMP is a flux and KPP_SRC_TEMP is a tendency.

Note that eddy parameterization of the model's lateral diffusion scheme also has a vertical component, due to mixing along isopycnals and eddy induced flattening of isopycnals. The combined effect these processes is in the variable HDIFB_TEMP.

Keith

#### juliettelavoie

##### New Member
Thank you for your help Keith!

Just to confirm how the variables relate to equations.
In the pop reference manual, equation (7.66) (shown in picture) is D_v (\phi)= (\kappa(\phi_z - \gamma))_z.
where _z is a derivative in z.

DIA_IMPVF_TEMP=\kappa \phi_z
and
KPP_SRC_TEMP=(-\kappa \gamma)_z.
where \phi is potential temperature.

so D_v = (DIA_IMPVF_TEMP)_z + KPP_SRC_TEMP ?

Also, HDIFB_TEMP is called a flux (TEMP Horizontal Diffusive Flux across Bottom Face). Why does it have unit of tendency ( degC/s)?
To get the tendency, should we do (HDIFB_TEMP)_z * \Delta z?