|The manuscript has improved during revision, especially regarding the recognition of PMSE physics and explanation of the examples shown. Corrections or extensions were applied to the manuscript where necessary. Although the authors answered to all questions, I have subsequent questions or comments to their extensions, as explicated below. The study of NLC and MSE lower and upper edges certainly is of scientific significance (even if no substantial new concepts, ideas, methods or datasets are introduced in the manuscript), especially in comparison to the respective results at polar latitudes, as this might give indications for different formation processes. Yet this remains a difficult task and the authors tried their best to supply with temperature and wind data, but the results are inconclusive. The question is still open whether the dataset is large enough to derive significant results and allow for interpretation. At some places, analysis that could aid interpretation is lacking, or results are not properly acknowledged. I also want to make aware that the manuscript does not conform to the journals data policy, requesting the used datasets to be publically available in an online repository and to be cited accordingly. |
Layer thickness: There are several references to layer thickness throughout the text, however it was not analyzed. It is expected to be thin, but e.g. Fig 2d shows a rather thick layer. As it is just a combination of two analyzed variables, z_up - z_low, the effort to calculate at least mean values for NLC and MSE should not be unreasonable. It would aid interpretation to know if the layers are thinner than at polar latitudes. The authors suspect that, but can easily derive it from data.
I think another flaw of the previous type was introduced during revision. In p. 12, l. 16: Kaifler et al. (2011) list z_low as 82.1 km which is not 0.5 km above the mid-latitude value of 82.6 km, but below. Thus it can not be explained by the general increase of NLC altitude with latitude. With about 50 m per degree of latitude the z_low from high latitude of 82.1 km would shift down to 81.4 km, and not up to 82.6 km. Furthermore, this cited increase with latitude refers to the centroid altitude, and it is not clear at all what applies for lower edges. Again, some more intelligence on this topic might be revealed by the analysis of the layer thickness.
Related to this topic is another missing quantity which is z_up for the complete MSE dataset. It is very important for interpretation and should not be omitted. The authors hint that it could be significantly larger than 84.5 km which is z_up of NLC during MSE. This would mean that high-altitude MSE are suppressed during NLC conditions, which would be a major result. I don’t think the authors interpretation is correct, they state that the small particles at high altitudes have grown and sedimented and are thus removed from high altitude during NLC conditions. But this process would not suppress subsequent reformation of local (!) high-altitude MSE, as is obviously observed during non-NLC conditions.
The fact that most NLC had to be discarded due to missing MSE is another major result, which I think is different from polar latitudes. I understand that the authors refrain from analyzing occurrence frequencies for good reason (limited statistics), but the total hours of NLC during daylight (ionization) discarded due to missing MSE could be robust, and it would be an important result.
Representativeness for all NLC: I still wonder about the numbers, and the reason for the discrepancy between the NLC statistics of centroid height of 82.6 km and the MSE-selected NLC centroid height of 83.3 km. Looking at the distribution in Fig. 4a this seems to be a significant discrepancy not related to geophysical variance. The authors mention the removal of NLC during nighttime. This cannot be the reason, as Gerding et al. (2013) show rather slightly higher NLC centroid altitude during night, so this should not be an effect of diurnal variation. They mention the removal of weak NLC profiles during the beginning and end of the measurement, but this assumes they are at low altitude, which was not shown and which I doubt, as at polar latitudes NLC are brightest at low altitude. Then they mention the removal of very low NLC due to absence of MSE. Now this speaks for a strong coupling of NLC and MSE in line with the observations of coincident edges, and for a true (regarding physics) difference between NLC with and without MSE. Then the selection is not representative, but it does not have to be, this altitude-dependence of NLC regarding MSE conditions would rather be a major result of this study, if robust. Because Kaifler et al. (2011) found no such difference in NLC altitude due to PMSE selection, their table lists 82.1 km for both. When the authors claim that 600 or 700 m is within the accuracy or variability or due to the limited size of the dataset, then this also applies to the upper edges, and it would be even more difficult to draw conclusions from these numbers given the large uncertainty. Then I would conclude that the dataset is yet too small to derive results that can aid understanding of the formation processes.
The authors main conclusion is that mature NLC particles are advected to the observation site. While this may be true, and was claimed before, it is not totally clear if this conclusion is confirmed by the absence of MSE above the NLC layer (coincidence of upper edges) and if it can thus be strengthened by this study. For all we know, the particles could nucleate and grow within the NLC layer, or they could nucleate above the NLC layer but MSE could be suppressed by other mechanisms. So I am not yet convinced that the presented analysis allows for conclusive interpretation on this topic.
I am looking forward to the authors thoughts and data on the points I mentioned. In case a second revision should be prepared, I give the following advice for improvements of the text. Especially in the conclusion a number of formulations are imprecise or incorrect:
p. 13, l. 20 and p. 15, l. 5: This should be expressed more carefully. There are no temperature measurements available at the lower edge at mid-latitudes. So the fact that the lower edges coincide hints at a large temperature gradient. This is not the same than expecting coincident edges on the assumption of a large temperature gradient.
p. 15, l. 8: “the layer is thinner”: thickness was not evaluated, and no numbers were given for MSE-only. So we don’t really know if it is thinner.
p. 15, l. 10: “thin layer above” does this mean between the NLC top and the MSE top? But the smaller particles reach down to the NLC bottom. “no such layer at all” does that mean NLC without MSE, or coinciding upper edges?
p. 15, l. 12: “southward or northward or weak”. There is no preference from the analysis, I’d say.
p. 15, l. 13: if they don’t grow to optically visible sizes, local formation of NLC is not possible
The language could be improved regarding expressions as “provides some rough information”, “scattering happens only on structures” , “layers stretch much higher than”, “revealed from PMSE observations”, “particles have been grown to sizes”, “we partly find very large agreement”
p. 1, l. 19: For sure, PMSE have not been observed for several decades by human eye. I see that this happened during revision, but nevertheless strongly advise to pay attention to correctness of sentences.
p. 2, l. 14: Thomas (2003) formulated a question, and Russell et al. (2014) make no reference to climate change at all. This topic might be too complex to be covered in half a sentence, so I suggest to remove this statement.
p. 6, l. 3: the lower edges do not agree, they are plus or minus 1 km.
p. 8, l. 9: indeed I can imagine this (Fig. 2d, 3:30 UT) to be a FOV effect
p. 8, l. 11: one could estimate the drift time given typical wind and FOVs, should be few minutes only.
p. 11, l. 20: my interpretation would be that there is no preference for any wind direction above the NLC layer.
p. 1, l. 1: suggesting: “We combined ground-based lidar observations of noctilucent clouds (NLC) with co-located, simultaneous radar observations of mesospheric summer echoes (MSE) at a mid-latitude site in order to compare ice layer altitudes. While larger ice particles (> 10 nm) are directly observed by lidar, the echoes recorded by radar are created by a complex interplay of ice particles, ionization and turbulence. The combined lidar and radar dataset thus includes some information on the size distribution and history of the clouds. ..”
p. 1, l. 3: “first comparative study”: “first” could be removed, as it is not clear if the authors claim to do this for the first time, or if they plan a second study
p. 1, l. 6: “rough information”-> “Thus, some information on the size distribution and history of the cloud is included in the combined lidar and radar dataset.” Just a suggestion
p. 1, l. 10: “We find no difference of the lower edges”, as the accuracy of the radar is 300 m.
p. 1, l. 13: “stretch higher” -> reach higher?
p. 1, l. 17: I would turn this around: “High-altitude MSE, usually indicating nucleation of ice particles, are rarely observed in conjunction with lidar observations of NLC at Kühlungsborn.”
p. 2, l. 6: “Later on” this might refer to the 1996 citation, but the 2017 citation is in between, so this is unclear.
p. 2, l. 9: “comprehensive interpretation” -> review of PMSE physics?
p. 2, l. 9: “Though, NLC..” Thus, NLC and PMSE are both indicators for..
p. 2, l. 10: “indirect information on temperature … atmosphere where other data is sparse.”
p. 3, l. 11: were grown, have grown?
p. 3, l. 26: delete “daylight-capable”
p. 5, l. 25: “to allow for a radar backscatter signal”