Review of “PMSE Layers During Solar Maximum and Solar Minimum” by Jozwicki, Sharma, Huyghebaert and Mann

This a quite interesting study of PMSE layers and should be published in AnGeo after suitable corrections are made.  I have very little questions concerning the observations.  This was done quite well.  However I do have some comments/questions about the conclusions and implied physical causes.

Comments

Title.  It would be best to spell out PMSE for the readers of AnGeo.  I suspect that many of the readership will not know what that is. Also since you have a strong focus on multilayers, perhaps a better title would be “Polar Mesospheric Summer Echo (PMSE) Multilayer Properties During Solar Maximum and Solar Minimum”? Something like that?

Since you have invoked particle precipitation as a possible source for the radar wave scattering, you should mention that the magnetic latitude of Tromso.  I believe it is 67 degrees.  The geographic latitude is 69 deg, but for particle precipitation, the magnetic latitude is more important.  The “auroral zone” is 60 to 70 deg magnetic latitude, so your observations have been taken in the center of it.  During relatively quiet intervals the auroral precipitation occurs in the auroral zone.  During solar maximum and during storms the precipitation is at slightly lower magnetic latitudes, as low as 55 deg.  The greatest auroral precipitation is now known to occur in the declining phase of the solar cycle associated with high speed solar wind streams. A reference to this can be found in JGR, 100, A11, 21717, 1995. This precipitation occurs in the auroral zone.  I think what you have stated about precipitation during solar max and solar min is okay as is.  However I point out some of the subtleties for your information.

During solar maximum there is more EUV radiation.  Depending on season, perhaps that can be a cause for greater electron densities during solar maximum? Also could solar heating expand the atmosphere slightly giving you your height difference during solar maximum?  Please discuss in the body of the text.

Abstract, lines 12-13.  I don’t see much evidence in your paper for gravity wave generation of the multilayers.  In Figures 1 and 2, I can see evidence for a two multilayer form evolving into a monolayer.  How can that be explained by gravity wavelength separation?  From that description one would expect these multilayers to be separated equidistantly.  My suggest is to point this out to the readership and stay with the observations.  If you do have cases of say 3 or 4 layers, please add a figure and  show an example of these for the readership.  Are they separated equidistantly?  Some discussion would help.   But from what I have seen so far, I would say that gravity waves are not the answer.

Introduction.  I have not seen any mention of metallic ions at ~92 km altitude which I believe come from the decomposition of micrometeoroids?  Could these be the nuclei of water condensation? Can you add this to the discussion?

Line 19, your first sentence in the Introduction.  A review paper on PMSE should be referenced here for the readership.  Or if one does not exist, perhaps the discovery paper.

Line 46-47.  I think perhaps you should mention solar EUV being stronger here during solar maximum?  Of course in your data selection I do not know if you have chosen times when the ionosphere over Tromso was sunlit or not.  Perhaps a mixture?

Line 86.  Gravity wave breaking.  If gravity waves break near the altitude of your study, that will go against the idea of the separation of the multilayers of PMSE by gravity wave wavelengths.  So you will need to chose one route or the other.  Gravity wave breaking needs a reference.  I have talked to an expert on this topic fairly recently and this person speculated that there was no such thing as gravity wave breaking.  That the gravity waves evolved into multiple waves instead.  So please be careful.

Line 102. Please give the location of the Magridal website.  I do not see it in the acknowledgement section?

Line 104.  What is a “manda code”? Please explain to the readership.

General Comment for the beginning of the paper.  Much of this background material could be put into the Discussion and Conclusion Section.  It would be good to shorten the front end of the paper and get to your analyses and results first.  Then afterwards compare your results to theory and modeling.

Lines 148. Please give references to Pearson and Spearman.

Figure 2.  I have already made comments about Figure 2 (and Figure 1).

Figures 3, 4 and 5.  The height distributions in all 3 of these histograms are quite broad.  Compared to the small distances between the mean locations, I would draw the conclusion that there are no differences.  Please discuss. You still can say that there are small differences but they may not be statistically significant. I doubt that the mean differences are statistically significant.

Page 5, 2 lines from the bottom.  “Random forests”?  Please describe for the readership.

Lines 179-181.  It is not known if the auroral particles precipitating are more energetic during solar maximum than during solar minimum at your magnetic latitude, so this argument is not valid.  During solar maximum there are more magnetic storms than during solar minimum.  But as mentioned previously these storm particles generally will precipitate at altitudes below that of Tromso.  And the storm particles form the ring current and have energies of 10 to 300 keV. These particles will deposit their energy well below 92 km.

Line 209. Is the electron density decrease with number of layers statistically significant?   This seems weak.

Figure 12 and elsewhere (many other figures), the differences seem to be small and not statistically significant?

Conclusions, first paragraph.  Yes you are correct, but I think the differences are not statistically significant.

As stated earlier, it is best to stay with the observations. If they fit theory/code results, fine.  If they do not, then fine again. 

General comments

The article investigates properties of PMSE layers, more precisely of multiple layers, during periods of enhanced and minimum solar activity. The article is clearly structured. The literature review listed in the introduction covers the necessary background information on this topic. The data selection and analysis methods are presented and referenced in a very understandable way in section 2. The analysis of the data and discussion of the results is summarised in a structured way in one chapter. Overall, I see the article as a valuable contribution to the exploration of one aspect of the long-known but still complex phenomenon of PMSE. I would like to make a suggestion that I think could improve the readability of the paper and have listed some specific comments.

The study highlights the characteristics of PMSE in terms of their organisation into multi-layered structures, but the actual multi-layered structures are somewhat lost, at least in some illustrations. For example, the mean value of a distribution of parameters obtained from signals organised in multi-structures, as shown in Figures 4, 5, 10, 11, 13 and 14, says not much about the properties of the parameter with respect to the multi-structure. Rather, it represents the properties of a virtual layer that is organised into sub-layers. With the width of the layers considered further on, it then becomes complicated, as here the widths of the layers that occur simultaneously at different heights are combined. Therefore, I would recommend the authors to separate the distributions of the parameters in these figures for the multistructures found and to colour-code them, for example, and also to treat them separately in the analysis. Then, for example, in Fig.4b two distributions in two colours around two mean values would be shown in Fig.4c three distributions in three colours around three mean values and so on. With these separated parameters, detailed statements can be made about peak height, thickness, signal strength with regard to the  occurrence in multilayer-structures and also in relation to the periods of solar maximum and minimum. This becomes particularly interesting and meaningful when, for example, the comparison to the NLC and the underlying mechanisms is made in chapter 3.4. Implementing this recommendation would, in my

point of view, improve the readability of the article with regard to the multiple layers, because one would then see their distribution in combination with an improved bin resolution (see below) in the above-mentioned figures.

Below are some specific suggestions to the authors that I think could be included in the article:

Specific comments

• P1, L18: I would not say that the waveform is characteristic of PMSE, even though it occurs occasionally if not frequently, especially in the thin layers.
• P2, L27: Latteck et al. (2021) deals with PMSE and should not be used as a reference for NLC.
• P3, L49ff: I would suggest to move the sentence starting with ”The mesopause ...” further up in this section e.g. after the references in L23.
• Section 2.1 : I suggest to include this section into the introduction section and rewrite the introduction section since some parts as e.g. gravitiy wave breaking and turbulence is already mentioned there.
• P4, L104: I suggest writing ”manda”-experiment instead of manda code and either giving a reference to a publication describing this experiment configuration in detail or summarising the most important experiment parameters here in the text.
• Fig.3, 4 and 5: Why are the height or altitude distributions of PMSE detections shown in bins of 1km in these figures, when the experimental height resolution is 0.36m?
• Fig.4 and 5: What is the average altitude of a multilayer and what can be deduced from this value? If you observe the PMSE over many years, you will notice that the distance between e.g. double layers can cover a very large range, whereas the actual layers can be very narrow.
• P9, L179ff: Here, the lower altitude of the PMSE and especially of the NLC should be discussed, which, as far as I know, is hardly subject to annual fluctuations. The increased energy input during the solar maximum at lower altitudes might therefore have no influence on the formation of PMSE at lower altitudes, as the other necessary conditions such as ice are no longer present above a certain altitude.
• Section 3.2 : There is still some discussion missing here. What does the distribution of the electron density as well as its maximum and standard deviation say about the organisation of the PMSE in mono or multilayer?
• Fig.12, 13 and 14: Why are the distributions of PMSE thickness shown in bins of 1km in these figures, when the experimental height resolution is 0.36m?

Technical corrections

• P1, L21: Remove (km).
• P1, L23: The correct use of references should be checked throughout the text, e.g. the references here should be placed in brackets. See also at [P2, L26], [P2, L27], [P4, L83], [P4, L90], [P4, L103]
• P2, L30: Remove (PMSE).
• P2, L39: Remove (NLC).
• P8, L167: I would not write layers here but detections, e.g. ”average peak altitude of PMSE heigt distribution”, as the plots in Fig.3 are probably not a distribution of predetermined layers.
• P11, L119ff : Replace echo power by average echo power.
• Fig. 12, 13, 14: I would suggest to use the correct thickness in m or km at the x-axes as well as in the text instead of altitude intervals.
• P13, L256: Remove nanometers and the brackets.