[MITgcm-support] diagnose gmredi contribution to vertical tracer flux

[Daiwei (David) Wang]

Hi,

I wonder if anyone has a good way to diagnose the gmredi contribution to 
the vertical tracer (e.g., temperature) flux. To my best knowledge, 
gmredi (in its skew flux form) contributes to both the explicit part 
(DFrE_*) and the implicit part (DFrI_*). The explicit part, if 
implicitDiffusion is true, is purely gmredi's off-diagonal contribution 
(see: gad_calc_rhs.F), and is trivial to determine. The implicit part 
(gmredi's diagonal contribution), however, is lumped together with other 
contributions from the background diffusion and the kpp diffusion by the 
array KappaRX (see: gad_implicit_r.F and calc_diffusivity.F), thus hard 
to separate. The best approach that I could think of is to scale the 
implicit tracer flux output (DFrI_*) with diffusivity ratios (gmredi's 
Kwz over the total). Before I proceed with that, I'd like to check if 
anyone has done this in a better way.

Thanks,
David