[MITgcm-support] Re: Question on kpp_calc_diff_s.F

[Jean-Michel Campin]

Hi Galen,

> I've been browsing the KPP code, trying to get my head around it. I find in
> kpp_calc_diff_S.F on the 12th to last line, it uses KPPdiffKzT for the
> k=kArg loop. Should this not be KPPdiffKzS?

Yes, it's wrong, and should be KPPdiffKzS instead.

I've just fixed it, but this should not affect the results 
since this 2nd part of the routine is used when called from
CALC_DIFFUSIVITY, for explicit vertical diffusion.
And it's never the case with KPP, which requires
implicitDiffusion=.TRUE., 
(when I try without, it stops in KKP_CHECK).

Anyway, thanks for catching this.

Cheers,
Jean-Michel