I noticed that for `exp2_w4`, I did not get a perfect PV balance between the component across PV contours and the dissipative term when I use two cycles (Fig. 1). I re-did the calculation using up to 13 different cycles (all 100 day long but separated by 15 days to explore different phase). With 13 cycles, the time rate of change is much weaker and we obtain a good balance between the component across PV contours and the dissipative term (Fig. 2).

Notice something important: The dissipative term is nearly the same in both calculations. This suggests that if we have access to this term, we could compute this term only over a couple of cycles to know the pattern and amplitude of dissipation.

computed with `theory_test_several_cycles_script.m` which produces the Matlab file `diag_VC_thirteen_cycles_100day_long_every15days_exp2_w4.mat`, all in `RESEARCH/MODELISATION/HIM/studies/diss_train_of_eddies/exp2/exp2_w4/analysis_1d` on the main `ipu1` disk.