[MITgcm-support] noise in high resolution run

[Dimitris Menemenlis]

> My experience with KPP is that it tends to amplify noise, but does not generate it (?).

Yes, that's right.  Without horizontal filters, KPP is 1D so it cannot generate 
horizontal noise but it will amplify it.  What happens is that (i) smaller 
vertical stratification leads to (ii) higher vertical mixing coefficient
which leads to (iii) yet smaller vertical stratification and vice versa.

... but obviously this is not your problem here.

It is strange because I have not seen this noise in our 1/8 or 1/16-degree 
global integrations.  If it's any help, the configuration that we used is 
documented in MITgcm_contrib/quarter_degree_global

Dimitris