Isotropic and anisotropic mixing

Topics related to the dynamics/numerics of the WRF model
Post Reply
Posts: 10
Joined: Mon Nov 11, 2019 3:33 pm

Isotropic and anisotropic mixing

Post by MatzeG » Mon Apr 13, 2020 7:58 am

I'm doing idealized LES to study the initiation of deep convection by thermal circulations in mountains (with diff_opt=2 and km_opt=2). I have a question about the mix_isotropic option for the SGS mixing model.
I read that the isotropic mixing option can be used when vertical and horizontal grid spacing are of similar magnitude and that this option is recommended for LES. What's the recommendation, how isotropic the grid should be for the isotropic mixing option to be applicable? So what's the maximum grid aspect ratio? Or why can't I always use the anisotropic mixing option, since both options converge if the grid becomes isotropic? I guess this is because it contradicts the assumption of locally isotropic turbulence that the SGS model tries to model?
When using high model tops to study deep convection, I'm using a stretched vertical grid, e.g. 20 m at the ground and 100 m at the top (converted to pressure levels) with dx=100 m.
In this way, the grid is isotropic higher up, but not close to the ground. What to do in this case?

Thank you very much!
Matthias Göbel

Ming Chen
Posts: 1456
Joined: Mon Apr 23, 2018 9:42 pm

Re: Isotropic and anisotropic mixing

Post by Ming Chen » Mon Apr 13, 2020 10:18 pm

It is correct that anisotropic mixing is more general and would be
better for this case because the grid is not isotropic at the surface.
The isotropic setting is for some sensitivity studies, I believe.
WRF Help Desk

Post Reply

Return to “Dynamics and Numerics”