|The manuscript has been substantially improved compared to the initial submission.|
I consider that it can be accepted for publication following small revisions. Indeed, I consider that some clarifications are still needed and that this paper can still be shortened, mostly by avoiding some unnecessary repetitions.
In particular, I suggest deleting from the main text most of the discussion on the comparison between GloPSI images obtained from data sets using 27 or 64 events and moving this discussion to supplementary materials. Also, I would suggest to better organize the discussion section by separating the discussion concerning the northern part of the profile and the one on the southern part. Currently, I find that there is too much back-and-forth between the two regions and between comparisons with previous studies.
Hereafter I give more minor comments for each part of the manuscript.
Section 1: Introduction
L18: “plate Adria” -> “Adria plate”
L19: “involved hundreds of kilometres, through […]” -> “involved hundreds of kilometres, of shortening, through […]”
L27: It would have been interesting to mention / compare (in the discussion) the results obtained along the EASI profile with the ones along the TRNASALP transect to get insights on the lateral variations.
L41-44: This part (about the Moho depth estimates along the Cel09 and ALP75 CSS profiles) should be moved to the next paragraph.
L48: Why do you indicate “the needs to be interpolated” ? + Not clear what is the subject of “is” in “is interpretated with a Moho triple junction or a Moho gap”.
L49-52: “The later interpretation is strongly supported by Spada et al. […]”: This sentence can be removed since your results do not support one interpretation or the other.
L54: Add “phases” after “[…] with both Ps and Sp” + the word “scattered” does not seems appropriate since the conversions can simply be associated to various interfaces.
L55-57: I would suggest removing “The interpretation of” since you are only summarizing the results of Hetenyi et al. (2018) and, later on, to remove “by different approaches” since this sentence is dealing with the results of the RF studies solely.
L63: Separate the 2 sentences + give more information on the results in terms of Moho depth (and lateral variations) based on ambient noise studies in the targeted area.
L67: Remove “new”
L75-78: the sentence “In other implementations […]” is badly positioned (it should either be positioned before the previous sentence or in the next section) and, actually, I suggest removing it from the introduction
Section 2 : Data and methods
L84-85: What is the influence of the “pure” P phases used? A supplementary figure showing the cross-section only with PKP and PKIKP phases would be useful to better see the impact of adding inline P-phases
L85-88: The sentence “We have used PKIKP […]” is just rephrasing the previous one and can be removed.
L88-89: The discarded events are unclear. If you don’t use events around 150°, you should provide the exact distance range rejected. Otherwise, I don’t understand the discarding of time windows in the sentence within the brackets.
L90: The term “a high station coverage” is unclear here. Do you want to mention azimuthal coverage?
L98: No moveout correction is applied to account for the varying incident angles?
L103-108: The spectral balancing probably also aims at having an amplitude spectrum closer to the one of a dirac function to mimic an impulse response. I’m wondering if that can be added here.
L109-113: You can mention here the source-side reverberation acronym (SSR) used afterwards
L116: I don’t understand the time range mentioned for the Hanning window (1 to 6 seconds) as the traces are much longer than 6s. Please clarify.
L120-123: I would suggest giving a little bit more information on the process of multiple removing following Verschuur and Berkhout (1997) and the effect of this step. Based on the various figures (inc. the ones in the supplementary materials), the effect of this step is not evident and only seems to lower all the amplitudes in the signal (including reflectors within the crust).
Section 3: Results
L145-149: I suggest moving the small discussion of the effect of removing/adding clusters of events (Fig S8) to the supplementary text.
L155-156 + Fig. 3: It is weird to express standard deviations (std) in terms of % of the maximum amplitude in the window. This does not allow to clearly see the real std on the amplitude of the phases. Moreover, this relative scaling can be very dependent on spurious phases.
Fig 3: Question 1: Based on Fig. 3 it seems that the std is higher (including in the northern part) after the multiple suppression. Is it just an eye effect due to the change in the std scale of a real observation? If yes, why such step should increase the std (to my mind it should decreased it) ? Question 2: Why the std is smaller for times higher than 15s? Does it mean that the results are more stable in this time range (which I do not understand)?
L159-167: I would suggest moving again this discussion between the GloPSI images with 27 and 64 phases to the supplementary materials.
L168-170: I am uncomfortable with the interpretation of the late spurious arrivals as P-S conversions. If you focus on the quasi-horizontal arrivals at depth higher than 60km, they correspond to phases with a lag time higher than 15-20s. If they are P->S conversions then they would occur within the mantle, which should be a less scattered medium than the crust. Alternatively, they would rather correspond to multiple crustal phases (PPS or PSS). Moreover, they are observed over a quite wide distance range, favoring horizontal structures. Why not simply considering that they can still represent SSR due to events present in both datasets (with 27 or 64 events)?
L171: You should remove the last part of the sentence “[…] to focus our interpretation on the Moho topography in the northern part of the profile” since you actually discuss extensively later the southern part of the profile …
L180 and after: This section related to the possible interpretation of the scattered aspect of the image in the southern part would better fit into the discussion section. Indeed, you start here to discuss your results in comparison with other ones using different techniques (Brückl et al., 2007; Hetenyi et al., 2018) as you do in the next section.
Section 4: Discussion
L200-202: You are paraphrasing the last part of the previous section (see comment above).
L209-L213: The statement abouts the “accuracy” of past CSS studies (Brückl et al., 2007; Spada et al., 2013) should be reworded/strengthen. You consider your study as the reference, but you do not provide any estimate on uncertainties about your Moho depths either due to the data selection or (most importantly) to the velocity model (that can be different from the one used by these 2 former studies) used in your inversion. It would be better to use a term like “agreement” between your results and the previous ones.
L221: I don’t understand what you mean by “and the results of this latter”
L240: I would remove “[…] the reliably resolved” statement.
L259: I don’t understand why you say that the study from Yan and Mechie (1989) does not suffer from any method or data set limitations (compared to others) …
L263: Why do you state that the Alpine orogen (in the targeted area) is characterized by “strongly dipping Moho interfaces”? (by the way I would remove the “s” from interfaces).
Section 5: Conclusion
L275-276: Please rephrase. The (potential) southern dip of the European plate is not the direct reason for delivering a clear image of the Moho with the GloPSI.
L277: I would replace “’inaccessible” by “unclear”
L278: Replace “this” by “the GloPSI”
- Give more information on which base the 27 events (Fig S8) were selected
- Fig S2: There is a typo in the minimal distance
- Fig S7: I don’t see the black dashed box mentioned in the legend
- Table : Modify the figures referring to the 27 events (no more figure 2, Fig 4a and S8