The authors are thanked for their response to the detailed comments of the two reviewers. There are still a few items that require further clarification before this manuscript is ready to proceed to full publication. The reviewer comment, the author response and an explanation where further clarification is required on each aspect are included below.

Reviewer #1

Reviewer Comment: P2 L28 (and other places). The 30-60 days oscillation found is similar to that of MJO, but there is no evidence in this analysis that it is actually due to MJO. It is only speculation so claiming that it is “related to’ MJO is too strong a claim.

Author Response: Thank you for your suggestion, we have made a change in the abstract. The sentence is changed as: In addition to the AO, the mesopause densities over low latitudes also clearly show an intraseasonal variation with a periodicity of 30-60 days.

Editor comment: the assertion that the intraseasonal density variations are related to the MJO is repeated on page 24, line 10 and has not yet been corrected.

Reviewer #2

Reviewer Comment: The authors determine ambipolar diffusion coefficient at each height from the radar data. Is there any sorting of data by elevation angle? Meteors recorded at low elevation angle will be at long ranges and so even a small error in elevation angle may thus correspond to a significant error in height. Conversely, errors in elevation angle will produce smaller errors in height near the zenith. The authors should explain if they used any sorting and comment on this possibility.

Author Response: Thank you for pointing this out. We apologize that we forgot to refer to data processing in this study. In the data processing, we actually did the zenith selection for the meteor echoes. The data processing is followed by Yi et al. (2017, 2018a). In order to avoid the possibility of excessive error in the height estimates of individual meteors, trail detections for this study were restricted to zenith angles of less than 60°. We have added this in the revised manuscript: In order to avoid the possibility of excessive error in the height estimates of individual meteors, trail detections for this study were restricted to zenith angles of less than 60°.

Editor comment: Please provide some estimate of the radars' elevation angle measurement error, and how this will impact the height determination in the worst case scenario of a meteor trail recorded at a zenith angle of 60 degrees.

Reviewer Comment: The authors use MLS temperatures in combination with the ambipolar diffusion to estimate density. However, at these heights the vertical resolution of MLS is poor, e.g., at z = 81 km the vertical resolution is 14 km. Given that the atmosphere at these height can have sharp temperature gradients associated with the mesopause, how does this impact the analysis? Is this not a major source of uncertainty in the determination of density given that the actual atmospheric temperature at a particular height may be rather different from the one derived from MLS measurements of low resolution?

Author Response: First, we should point out that we chose the MLS temperature measurements because...

Editor comment: I don't think the referee was implying that the MLS data are inaccurate, nor that the data coverage is poor in the polar regions. Rather that the MLS temperature measurements are an average value derived over 14 km of altitude at z=81km. As the referee correctly points out there can be large temperature gradients in the atmosphere at these altitudes. Some quantitative estimate of the uncertainties this may introduce in your relative density measurements - especially the difference between summer and winter - is required here.

Reviewer Comment: The manuscript refers to “densities”, but the measurements are actually “relative densities”. This should be corrected throughout to avoid confusion.

Author Response: Thank you for pointing this out, we have corrected throughout in the revised manuscript.

Editor comment: This has been corrected in five places (page 9, line 17; page 12, line 6) and in the captions to figures 3, 5 and 7. The rest of the manuscript continues to discuss "densities" and not the preferred "relative densities". As the referee correctly discusses, the measurements reported are relative densities (or density variations). At no point are absolute densities reported. The authors should clarify this important point throughout the manuscript (including in the paper title and the abstract).

publish as is