Implement TEM circulation diagnostics in physics

[fvitt]

What is the feature/what would you like to discuss?

Implement TEM circulation diagnostics in physics independent of dycore. Zonal means of wind components and potential temperature are needed in the calculations of the TEM diagnostics. This would use the Zonal_Mean code proposed in PR #629 which computes zonal mean values for arbitrary grids based on m=0 spherical harmonics on arbitrary grids. The ability to output zonal mean diagnostics to history would be needed (see issue #652).

This feature would be useful to evaluate WACCM simulations unstructured grids such as the cube sphere spectral element grid.

Is there anyone in particular you want to be part of this conversation?

@nadavis-ncar @dan800

Will this change (regression test) answers?

No

Will you be implementing this enhancement yourself?

Yes, but I will need some help

[dan800]

@fvitt Can you add Peter, Julio and Isla to the conversation?

[dan800]

So I looked at the test cases: /glade/scratch/fvitt/FWsc2000climo_f09_TEM_test04/ /glade/scratch/fvitt/FWsc2000climo_ne30pg3_TEM_test04/

In f09 the fields calculated each way look very similar, e.g., UVzm and UVzaphys However, UVzaphys from the SE run looks very different I realize these are different runs (FV vs SE) and so shouldn't be identical, the magnitude of the flux terms is much reduced, which would imply the circulation is very different. Sort of looks like a bug since I imagine the FV and SE circulations are similar. @PeterHjortLauritzen @nadavis-ncar ?

[nadavis-ncar]

It's hard to tell from the contour interval, but they look maybe similar in the troposphere/lower stratosphere? Can you reduce the contouring to emphasize those low values? Otherwise I agree, I would expect much larger values in the upper half of the domain.

Nicholas A. Davis https://staff.ucar.edu/users/nadavis Atmospheric Chemistry Observations and Modeling National Center for Atmospheric Research

On Fri, Nov 18, 2022 at 1:08 PM Dan Marsh @.***> wrote:

So I looked at the test cases: /glade/scratch/fvitt/FWsc2000climo_f09_TEM_test04/ /glade/scratch/fvitt/FWsc2000climo_ne30pg3_TEM_test04/

In f09 the fields calculated each way look very similar, e.g., UVzm [image: image] https://user-images.githubusercontent.com/37961632/202792807-35b80246-0861-49b1-b59c-3a8ee87ee1c2.png and UVzaphys [image: image] https://user-images.githubusercontent.com/37961632/202792891-43ac1471-711f-4b39-ae88-6f22c5f4da29.png However, UVzaphys from the SE run looks very different [image: image] https://user-images.githubusercontent.com/37961632/202793095-947959c4-3a18-41e3-a15e-d5d4945d8312.png I realize these are different runs (FV vs SE) and so shouldn't be identical, the magnitude of the flux terms is much reduced, which would imply the circulation is very different. Sort of looks like a bug since I imagine the FV and SE circulations are similar.

— Reply to this email directly, view it on GitHub https://github.com/ESCOMP/CAM/issues/653#issuecomment-1320474831, or unsubscribe https://github.com/notifications/unsubscribe-auth/APBQOAWEON67Y6TGNM552H3WI7O5PANCNFSM6AAAAAAQDRPIA4 . You are receiving this because you were mentioned.Message ID: @.***>

[fvitt]

@dan800 Further into the run, away from the initial conditions, UVzaphys fields are more comparable. Following are the fields 2 months into the run.

FV: /glade/scratch/fvitt/FWsc2000climo_f09_TEM_test05

SE: /glade/scratch/fvitt/FWsc2000climo_ne30pg3_TEM_test05