I changed the budget analyses to remove the domain-mean energy. Instead, I analyze the budget of the squared planetary-scale anomaly from the domain mean.
This analysis is more relevant to the propagating zonally non uniform signal. See this ipython notebook for an analysis: https://atmos.washington.edu/~nbren12/reports/2017/2.0-ke-budget-analysis.html. I also changed the discussion Sec 2.2 and 2.5 to explain this difference. Basically, I decompose f=f¯+f^+f′, where f^ is the planetary scale anomaly from the zonal mean. Then, I analyze the budget of f^2, which has a very similar structure to the budget we were looking at before.
Dec 28
I further refined the budget analyses.
Keeping track of all the various terms is difficult, so I first plot the aggregate effect of all vertical divergence, horizontal divergence, and other known terms for the moisture, buoyancy, and moisture variance budgets. See this budget analysis notebook.
That notebook shows that vertical advection only plays a significant role in the moisture and zonal velocity variance budgets. So I only really need to do the multiscale decomposition of the various vertical advection terms for those two budgets. This is what I do in this vertical advection decomposition notebook.
That notebook shows that
The main source of planetary scale zonal velocity variance is the synoptic-scale eddy flux term (u′w′)PuzP.
Vertical advection of the domain mean by the planetary scale vertical velocity is the most important vertical advection term in the moisture variance budget.
For the dry static energy budget s=T+Γz, the vertical advection of the domain mean by the planetary is the most important.
I think that only exploring the complicated quadratic terms when absolutely necessary makes the results much easier to explain. Also, estimating the source terms as a residual is very noisy because it involves taking a time derivative. It is better to compute the source terms as the residual of the integrated values. This gives much less noisy results.
add vertical advection+PE budget analysis to results/2.1-… notebook
Change the ansatz in the paper to f=f¯+fP+f′
Change the discussion of budgets in the paper for the new approach with computing the total advection terms and then decomposing them only if needed.
Use new definition of ⟨⟩ in the paper
Jan 19, 2018
I talked with Qiu about the paper. We discussed a lot about the figures and the outline. the main new work I need to do is:
Add active radiation case to hovmoller diagrams
Add figures from poster to the paper. See Figs. 8 and 9, 11 below
Dec 15
Dec 28
Jan 19, 2018