justin-richling
commented
1 year ago Some comments from Dan about current TEM ADF calcs/plots:
I think we really need the seasonal averages to judge the model circulation and forcing. We now have the infrastructure in place and just need to tweak the figures to make them the most useful.
We may also want to have the capability to compare the test case to the baseline side by side.
Some comments from Rolando:
Comments/suggestions:
Basically, the output looks good. It is insufficient for a full TEM momentum budget because accelerations due to (parameterized) GW are missing. Ideally, one would add terms like UTGWORO, etc., but I am aware that, in the past, it has been judged to be inefficient to output those terms on a daily basis.
One might also want to look at the TEM thermodynamic balance, in which case zonal-mean T and heating terms can be added to the output (QRL_TOT, QRS_TOT, etc.). I have often done this, and it is a good consistency check between TEM momentum and thermodynamic budgets.
Color table scales need tweaking. In some cases, log scales would help, e.g., when showing Fy, Fz, which are strong functions of altitude (they grow as exp(z/H) unless the waves are dissipated.)
Color tables vs. line contours: For published work, I find I often need to show contours. Color tables, in particular the blue-red one used here, do not allow fine discrimination among contour levels.
It is not clear why “zonal-mean meridional wind” is shown in a set of TEM diagnostics. It doesn’t hurt, but it is not needed for TEM analyses.
The advective tendency terms (last two panels shown) are not particularly useful outside the Tropics, where the TEM momentum budget often reduces to -fv = div(F) + GWdrag. But for the QBO they are important, especially w dU/dz, assuming the particular model being run actually has a QBO. Of course, one can always compute these offline from U, v and w. In short, these are useful but not strictly necessary.
Altitude range shown should depend on the model being run, obviously. For full WACCM, I often cut things off at ~110 km (~1.5e-4 hPa), as the last few scale heights are in the sponge layer. The results linked are (apparently) for the low top model. I wouldn’t believe much of anything shown above 10 hPa for these results, if that.
islasimpson
New feature type
New plot and/or plot enhancement
What is this new feature?
The Transformed Eulerian Mean diagnostics would be a good addition to the ADF as an option similar to the QBO.
I've spoken with @dan800 and have his working code. There will need to be a little discussion about specifics like how the plots come out, single vs multiple time steps desired, etc.
If you are looking to use this set of diagnostics, please comment below on what specifics the ADF should use for the TEM calcs/plots.
Assistance required?
No, I will make a PR when the feature is ready
Extra info
I currently have a working version in the ADF, with all plots in one image: