[MITgcm-support] Help me get rid of my noise!

[Dimitris Menemenlis]

Ryan, it's very long ago so I don't remember the details nor
am I sure whether relevant to your problem, but we did have
some issues with grid-scale noise and KPP early on.
In case it is still useful, here is an "old" description of this problem:
https://dl.dropboxusercontent.com/u/40292774/ECCO/kpp.pdf
One thing to note is that Gradient-Richardson-based schemes
tend to maintain grid scale noise patterns because K is increased
in unstratified grid columns and decreased in stratified grid columns,
a positive feedback loop.  This KPP grid scale noise led to
development of following KPP_OPTIONS:

C o When set, smooth shear horizontally with 121 filters
#define KPP_SMOOTH_SHSQ
#undef KPP_SMOOTH_DVSQ

C o When set, smooth dbloc KPP variable horizontally
#define KPP_SMOOTH_DBLOC

C o When set, smooth all KPP density variables horizontally
#undef KPP_SMOOTH_DENS
#ifdef KPP_SMOOTH_DENS
#  define KPP_SMOOTH_DBLOC
#endif

C o When set, smooth vertical viscosity horizontally
#undef KPP_SMOOTH_VISC

C o When set, smooth vertical diffusivity horizontally
#undef KPP_SMOOTH_DIFF

C o Vertically smooth Ri (for interior shear mixing)
#undef ALLOW_KPP_VERTICALLY_SMOOTH

On May 8, 2014, at 9:06 AM, Ryan Abernathey <ryan.abernathey at gmail.com> wrote:

> Hi folks,
> 
> Thanks to everyone for your feedback. I am going to resurrect this thread, because I still have some weirdness happening in my mixed layer. 
> 
> After integrating longer, the stuff over the slope settled down. However, there are still apparently some strong, shallow convection cells happening in my mixed layer in the northern part of the domain, just as Jody suggested. (See new attached figure.) There are little convective chimneys with the width of a single grid point, so perhaps this is the hydrostatic convection that Christopher mentioned. These cells lead to lots of grid scale weirdness at the base of the mixed layer. In particular, there are sharp grid-scale meridional gradients in the vertical diffusive heat flux (the step-like patten visible in the third panel), accompanied by strange looking patterns of GGL TKE. These cells are completely steady and persistent in time. (The GMREDI parametrization is not the culprit--its fields are relatively smooth. And almost the exact same thing happens with KPP instead of GGL.) 
> 
> I just don't understand why I should be having resolved "convection" happening in my mixed layer. Isn't the job of the boundary layer parameterization? This is problematic for me because the meridional heat transport associated with these mixed layer cells results in the surface heat flux being unevenly communicated to the interior. For my particular application, I would really prefer for these fields to be more smooth.
> 
> So is this also physically realistic? My feeling is no. And if not, how can I eliminate them!?
> 
> Thanks again.
> 
> -Ryan

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20140508/fd54b888/attachment.htm>