It is clear that the authors put considerable effort into the revision. While the paper remains an unsatisfying mix of qualitative comparisons and speculation regarding dynamical connections, with Sec. 5 being particularly unfocused and having too wide of a scope, there is enough interesting and new material to warrant publication. Below are suggestions for minor alterations.
Page 2, line 3: By saying ‘contrasts with’, do you mean ‘is balanced by’?
P. 2 L. 8: With regards to ice injected into the lower stratosphere; isn’t evaporation due to mixing with sub-saturated air a more likely moistening process than sublimation?
P. 2 L. 9: Change ‘emphasized to contribute’ to ‘emphasized as a contributor’ or, maybe, ‘recognized as a contributor’.
P. 6 L.6 (also discussion of Figs. 2-7): Regarding using extremes for comparing models: I would expect there to be (in general) more model-to-model variability with extremes (such as maximum velocity potential within a box) than with average quantities. To put the inter-model comparisons into perspective, it would be good to mention comparisons of average values as well.
P. 7, L. 15 and 27: The use of ‘time-mean’ to described the second term on the RHS of Eq. 1 is confusing because you also say that is describes a time-varying Walker circulation. Perhaps it would be better to separate the two eddy terms based on time-scale. The second term of Eq. 1 being a slowly varying (inter-annual) quantity and the last term being a more rapidly varying (mostly seasonal) quantity.
P. 8 L. 8-9: What is meant by ‘the two innermost tropical latitudes? Please be more specific; e.g., specify a latitude range over which values are averaged.
P. 8: Regarding Eq. 2: Please describe the performance of this regression. In particular: How independent are each of the terms? And how much of the low-frequency variance is unaccounted by the primary terms of the regression (or, equivalently how much of the low-frequency variance is accounted for by the residual)? Later in the manuscript, the authors discuss the statistical significance of their regressions – however, statistical significance does not necessarily indicate a meaningful regression relationship since one can render any relationship statistically significant simply by using more data. That is, one quantity might only explain 1% of the variance of another, yet with enough data, that 1% can be statistically significant.
P. 10 L. 13: It is not clear what is meant by ‘the onset of the divergent winds’. For example, do you mean that the location coincides with where winds become divergent (i.e., switch from convergent to divergent) or that the location coincides with the maximum values of divergence, or maybe some other meaning?
P. 10 L. 13: ‘vice-versa’ seems to be misused here. Do you mean that the negative peaks values of chi correspond to convergence? If that is the case consider rewriting to something like ‘The positive peaks … of upwelling, coincide with divergent winds while the negative peaks coincide with convergence.’
P. 10 last paragraph: You seem to be stating that the large values of humidity in the northern extra tropics seen in the multi-model-mean (Fig. 3) is due to lateral transport from the tropics. If that is true, then the fact that the values of humidity in the extra-topics is as large as values within the anti-cyclone implies that a large fraction of extra-tropical air was recently transported northward by the anti-cyclone (if not then air from the anti-cyclone would be mixing with air that is, presumably, drier, which would reduce the humidity). That implication is difficult to believe.
P. 10 L. 3: Change ‘diabatic’ to ‘radiative’.
P. 12 L. 3-6: This paragraph discussing temperature has no obvious bearing on the paper and the reasoning seems weak. Either clarify or remove.
P. 12 L. 7-8: Remove the phrase ‘and as it is not affected by dehydration’
P. 14 L. 1-4: Are you trying to say that ozone values over Asia exhibit variations that coincide with variations of the anti-cyclone on time scales less than the seasonal values you show? Please revise this paragraph to make its meaning clear or remove it since your study does not examine sub-seasonal variability.
P. 14 L. 5 – P. 15 L. 6: The paragraph that discusses Fig. 6 is very weak. I suggest removing it along with Fig. 6.
P. 19 L. 1: Vice versa is misused. Revise to something like ‘… dipole structure with warmer conditions on the southern edge of the AMA during wet phases with the oppositely signed structure evident in the dry phases’. If you want to use ‘vice versa’ you could say ‘the temperature structure has a warm anomaly in the south and a cold anomaly in the north during the wet phase and vice versa during the dry phase.’ [although I would not use vice versa this way either.]
Last paragraph of P. 19 and first paragraph of P. 20; Fig. 8: Don’t you find it interesting that the strongest wet anomaly is nearly collocated with a strong cold anomaly?
P. 6 L. 5: Remove comma after ‘CCMs’
P. 8 L. 8: Change ‘wind in 50 hPa’ to ‘wind at 50 hPa’.
P. 12 L.3: Either use ‘The temperatures on … are influenced’ or ‘The temperature on … is influenced’.
P. 12 L. 8: Change ‘ASM are the’ to ASM is the’.