Radar Remote Sensing of Precipitation in High Mountains: Detection and Characterization of Melting Layer in the Grenoble Valley, French Alps

Round 1

Reviewer 1 Report

My comments are well addressed in the revised paper and I think this paper is acceptable.

Even though a lot of studies on the melting layer have been done, observations in a high mountain región

are valuable.

Reviewer 2 Report

The motivation, background, focus and especially the summary abstract are improved. The issue of potentially paraphrasing a cloud physics textbook review of melting processes without proper attribution has been sufficiently resolved. The final outcome is a nice summary of known physical processes and their relations to radar observations. The presented methods of melting layer identification and documentation provide only iterative improvements upon the many existing techniques in the literature using similar data. Some of the results regarding the melting layer are largely confirmatory in nature. Nonetheless, the study does demonstrate significant analysis capability to carefully characterize the melting layer properties of stratiform precipitation systems in a challenging and unique geographical location. These outcomes will therefore have useful practical application toward improving quantitative precipitation estimation (QPE) in the local region and other similar regions with terrain challenges, which is a useful contribution to a number of fields including hydrometeorology. The study does present some tantalizing results about the information content of polarimetric radar for melting layer characterization. The practical importance of these interesting results is still somewhat unclear at this point, although they may be useful to motivate future studies in polarimetric radar data assimilation into cloud models used for quantitative precipitation forecasting (QPF) as the authors now point out.  In summary, the study provides a good review of the state-of-the art of melting layer detection and characterization along with some minor improvements to methods, demonstration of capability in a challenging mountainous regime with obvious local applications and potential for broader applicability and some analysis that points to future research. With the revisions applied, publication is recommended.  I have no further suggestions other than a couple of minor suggested edits by line number.

152: "artefact" should be "artifact"

212:  "dual-polarimetric" should be just "polarimetric"  and "observations" should be "radar observations"

Reviewer 3 Report

This study is carried out over large alpine valley in French Alps, in which authors utilized various types of RADAR data for analyzing hydrometeorological characteristics and detection of Melting Layers. Dataset used in this study is from  IGE XPORT research radar and MOUC (Mt. Moucherotte) Radar; which is unique. Besides carrying out usual postprocessing of Radar observations, statistical analysis is also carried out to analyze melting layer characteristics. This study shows some promising results and scope for further improvement of the algorithm used for ML analysis. Work is promising and deserves publication for wider publicity and reading.

 

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

This paper describes results of melting layer observation using multi-parameter radar and other radars.

The results are well documented, and the relationships among several radar signatures are well presented. The melting layer structure is complicated, and it is difficult to interpret the radar signatures. However, the authors challenged

this problem and good statistical results. So, I think this paper is worthy of publishing.

 

Minor comments.

L123: A more description is desired for the homogeneity of the precipitation system. The elevation angle of the radar is low, and the criterion of the homogeneity may be severe. For example, what is a typical distance to the melting layer? This distance may limits the horizontal extent of the homogeneity.

 

Section started with L232: Breakup processes should be more described. The break up could contribute much to the radar signatures, particularly when rain is strong.

 

Paragraph started with L427: Why is “ML top” used as a parameter? The pressure effect has only minor effect. ML top may be an indicator of the precipitation type, such as, convective/stratiform as described in L447. If so, more discussion on the precipitation type should be added.

 

L436: What is “points”?

 

L448: Is “altitude of ML top” “ML top”?

 

L556: Adding several definitions such as “Zh.val.t2p”, “Zh.alt.peak”, “ML top” may help readers to grasp the contents.

Reviewer 2 Report

Unfortunately, it was difficult for me to understand the findings of this research. The manuscript spent its most pages for describing the facts found by previous studies, but it did not provide its findings with a well-organized explanation. In some parts, the paper seems too redundant. In particular, section 3 is the "Results" section but a number of sentences in section 3 did not give the "results". Rather, the vast majority of the sentences in section 3.1 were a review.  I would strongly recommend the authors to re-organize the structure of the presentation and make it much simpler than the current version.

 

In the manuscript, no sonde observations were used. It would be better to use an independent temperature profile observation for validating the ML estimation by radars. Otherwise, it would be impossible to validate the accuracy of the estimation. At least, NWP model outputs (or meso-scale analysis) should be available and it may give a reliable reference for the ML estimation. Further verifications would dramatically increase the value of the study.

Reviewer 3 Report

Review of Atmosphere 570849

Recommendation:  Reject

Summary:  Although the authors have demonstrated some capability in analyses and interpretation, this manuscript has no notable science outcomes.  The majority of the results are not original and the few that are unique do not come to any clear conclusions.  In addition, the paper has serious organizational flaws as it presents background review in the results while it still misses much of the key radar literature for its chosen subject area of identifying melting layers, their attributes and associated physical processes. Unfortunately, some of the background review presented in the results section is dangerously close to text from a well-known textbook in cloud physics.  The authors should be careful not to paraphrase other sources, especially without attribution.  In summary, the authors have shown some promise in their analysis capability for identifying melting layers and associated physical processes but the paper has serious structural flaws and the large majority of the material presented is not sufficiently original to warrant publication.   The rest of the results (e.g., application of various data analytic techniques like PCA and regression) do not come to any firm results and conclusions.  Given the capabilities, my suggestion to the authors is to reorganize and shorten the paper, refocus the science and writing around clear and novel objectives that are well motivated and follow through with another manuscript at a later date.  Specific comments and suggestions are provided below.

Specific Comments by line number:

17-18 (abstract): The review provided should have been of past literature in the background section near the introduction, which should have motivated specific scientific objectives that are followed by clear outcomes.  What is new here?   21 (abstract): Please make your specific, and if relevant, quantitative, outcomes and conclusions clear in the abstract rather than reviewing broad study goals as you might in an introduction. 32-33: should be “most critical application concerns is” 79-80: NUBF needs reference(s) 92: ML detection algorithm:  What about past efforts on this subject?  This is by no means a new area of research.  A proper review of past capabilities and results were never presented (even in Section 3). 93-95: Why?  What's the point of this basic review?  Please justify and motivate or reorganize and rewrite.  After reading all of the paper, my final strong suggestion is to rewrite. 98 (Section 1 general): There is an over-emphasis on reviewing the authors' parochial efforts related to the their specific past projects.  While relevant, the authors have omitted a number of important journal papers on topics relevant to this paper's stated goals.  The authors are suggesting that they review additional relevant background in Section 3 on the microphysics of melting but there has already been considerable work in the general radar and specific dual-polarization radar communities on these topics.  So far, I'm not understanding the need for this approach.  At a minimum this approach needs to be better justified or possibly a reorganization of the paper with more complete background referencing in radar methods will be warranted. 106: Please see the MDPI template for how to annotate figures and provide descriptions in figure captions. The authors did not follow MDPI or really standard community practices. 110: Web pages should be in references.  See MDPI guidance to authors. 112 and 116:  The 3-dB beamwidth is 1.37 degrees and yet the angular resolution (azimuthal spacing for PPI?) is set to 0.5 degrees?  That is significant over-sampling and is rather unusual and should therefore be explained. 125: 25 deg elevation angle:  Have you accounted for non-alignment of the polarization planes with the local horizontal and vertical?  25 degree elevation angle is beyond the angle after which some effects will be noticeable such that the measured values are not the same as (e.g.,, Zdr is less than) the intrinsic radar parameters.  See Ryzhkov et al. 2005 or 2016 for mathematical expressions. 127: Explain this approach.    Why use rhoHV to differentiate homogeneous vs. heterogeneous precipitation rain echo?  How was this specifically done? 126-127: Stratiform bright band seems to be a rather narrow study choice given the motivation of the overall project to look at variable precipitation types.  I thought that was one of the key points of all the measurements reviewed in the introduction?  This should be justified more clearly given how the background emphasized wider use.  The background should motivate what was actually done here.  As noted elsewhere, the introduction needs to be rewritten accordingly. 143-145: Herzegh and Jameson? This is an early paper with basic ideas regarding melting layer identification and processes using dual-polarization radar variables.  As reader and reviewer, at this point in the manuscript, I'm still waiting for a careful review of this topic which has been studied in detail in the 25+ years since this paper in 1992. 143-146: OK, specific algorithm seems fine as it relies on well-known ideas of the melting layer (ML) and polarimetric variables. But there are a number of other approaches for polarimetric radar ML detection, including much simpler techniques that work robustly. Again, the non-expert reader would be unaware of this prior published work so far in the manuscript and mistakenly think that these are new ideas, while the expert radar meteorologist aware of the last 25 years of polarimetric radar research would be losing her patience at this point with the paper. Background above needs to be refocused and completely rewritten to address these manuscript problems. 148, 150 (and elsewhere in body text): What do you mean by "gradient" here, including 1st gradient and 2nd gradient? Do you mean a first order and second order vertical derivative of the radar variables? Explain in words more clearly and provide an equation that describes the mathematical operations being done. Your terminology is non-standard English to describe the mathematical procedure.  Shorthand would be "first derivative" and "second derivative" in the vertical.  Gradient is a 3-D vector quantity.  Since this operation is directional (i.e., in the vertical), it is a (vertical) derivative. 157: Yes, agreed but past work on this should have already been reviewed to motivate what is new in this manuscript.  You should not review past work but showing it with your own data in a peer-review manuscript. 159 (and throughout text): I found this use of pseudo variables for ML properties to be awkward and less clear than plain words throughout manuscript. 164-171: This all reads like pseudo-code. Some may find it helpful if trying to reproduce the procedure in code but for general reading, it is quite awkward and should probably be supplemented with plain words when possible. 168-171: How were these distances above and below determined? Physical considerations?  Need to note dependency on range, which is specific to your study, since beam broadening with range would influence choices. 176-177: Please explain and justify the definition of RHOhv.alt.top.   Explain general procedure for rhoHV.alt.* 186-187: The literature results on these issues were not reviewed. How is this new?  How will you put your outcomes in proper context of past work? 187-188: The statement regarding “ML 1D models” is unclear. Please clarify. 204: Minor formatting point but Authors seem to be mixing referencing style of MDPI and other journals.  Didn't think you were supposed to mention author names in body text as you have the reference #.  Both are redundant? 207: should be “…receive latent heat of fusion…” 210 (Figure 3 and general methods): Zh, Zdr and rhoHV all respond to different properties of aggregate melting so how do you use all of them to define the ACTUAL melting layer?   You have separate definitions for each radar variable but it is unclear how you combine them into a final answer of the physical definition of the ML. Unclear here and in method section. 212 and other locations: Why are some text bolded for emphases?  This seems to be an organizational artifice making for lack of clear organization.  How about sub-sections? 211-231: This is textbook level review. In fact, it is a bit too close to the summary in textbook by Puppacher and Klett (1997, Sec. 16.3.2, p. 697) not to be referenced.  First, I don't understand the point of a textbook summary of literature regarding melting aggregate.  Please justify.  Why isn't it in the background instead of in results?  Second, if it is needed, please reference Pruppacher and Klett.  This text is dangerously close to a paraphrase of much of this textbook section, although it does appear to integrate additional information obtained from these other references provided, most of which were pulled as secondary references from the textbook.  My suggestion is to get rid of this review and reference the textbook and make whatever succinct key points needed for your science in a rewritten introduction and background. 232-237: Why is this information on dual-polarization radar variables in results? If needed, this information should be in data/method and/or background.  These are not results.  So far result section contents are lots of literature review, background and 1 example profile of ML.  Nothing new so far. 238-254: Although a bit too lengthy, it is a good summary and demonstration of well understood processes and signatures of Zh in ML.  But what is new?  This text should not be in results.  It is background 255-274: Again, entirely a review of prior knowledge, including how the microphysical processes related to polarimetric radar. Should not be in results. Should be shortened and placed in background section in or near introduction to the extent that it motivates new science. 264: Aggregates also have very low bulk density and low dielectric such that their shapes are "effectively isotropic" even if they are physically anisotropic. In other words, dielectric plays a strong role in addition to physical shape.  This is well known.  See the paper you referenced earlier, Herzegh and Jameson, for an example intended for a general meteorology (non-expert) community back in early 1990’s. 280: As note earlier, all polarimetric variables are sensitive to elevation angle especially as it goes above about 15-20 degrees.  Variations from intrinsic exceed typical bias and random errors. 275-290: Again, this is all review.  Not results. 291-313: More review.  Really the entire section 3.1 presented as results is really background review.  Much of it well known (e.g., although well summarized, still textbook level and probably unnecessary for a peer-review journal article). 332-336: This text answers my question in #26 above but this material should be in methods, not results. It is not a science result but a method decision based on well understood behavior of radar variables in the ML. 338 (Figure 4): Usually if a height is relative to (or with respect to) a reference, one computes Relative height = height - reference height where here reference height is rhohv.alt.peak.  Your figure seems to be reversed. To clarify, especially if non-standard, please define in figure caption what relative height is shown. 347: “some part” - maybe most part in it? In other words, to be expected. 349-350: ok good that your results are consistent....but not new...where are the new results? I still do not understand the point of this paper and it is half way through the manuscript. 364: Why?  I think you can speculate based on your knowledge shown in review above. 366-370: It should be clear at the outset of results section where you are going with science and why. There needs to be some rewriting here, including proper focusing, movement of method and background to their proper sections and reduction in bloated background.  The paper meanders at this point forward. 378: very difficult to read this text and follow with the shorthand notation. It seems unnecessary.  You should strive not to make your results difficult to follow for the reader, which is why I recommend removing the shorthand notation. 426 and eqn (1): This equation is method and should probably not be in results. 438-440: So statistics with no easy physical interpretation. I can sympathize but as a reviewer I have to ask “so what is the point then?”  What do we do with this information?  It's new but the practical relevance is unclear.  On the start of the section you suggested that the analysis would tie radar signatures to physical processes.  At the end, you noted that it is complex and unclear.  Sure, but how does that help move community knowledge forward?  Put another way, why should the reader care? 451: OK so other factors are important to ML width and 0C isotherm height is often not the dominant control.  If correct, please say so. 453-457: Again much of this should be in background to motivate new work. What is new here? Are you confirming past results by Zawadski et al.?  Refocus your science and present your nice analysis capabilities more clearly. 457: is riming and snow density the only influence on wsnow? How much influence relative to other factors?  Again, this should be sorted out before using the idea to explore new results. 459: Why 26 dBZ?  Please explain and justify. 469: “investigate more deeply the information content” - what do you mean?  As presented, it seems like fishing with data analysis without further justification.  Again PCA should be in methodology.  I think this can be cleaned up.  The authors start to get to some new directions with these last results in later parts of Section 3.2 and 3.3 but they are not well motivated, clear or conclusive. 484-486: This seems like the beginning of something.  But why is that important? Significance? 491: “exploratory analysis” – Unfortunately, much of it seems duplicative with existing literature, which was not properly reviewed at the beginning of the paper such that the paper was not well motivated.  The science directions in last portions of the results section were more novel but not clear or conclusive.  502 (paragraph starting here): There are few if any new science outcomes here. 512-513: all of these results are well known and in the literature.  Not new.  522-523:  statistical analysis is consistent with similar observations in similar weather/climate context.  So not new.    The rest of the results from correlation and PCA do not seem to come to any final outcomes summarized here.  What are the new science results and outcomes? 558: Although I don’t feel strongly about it, I don’t think Appendix A.1 Formula List is necessary as it is all routine textbook information and equations about radar.

Reviewer 4 Report

Authors have done hard work and presented their research nicely. This study clearly indicates usefulness and importance of Radar installation in mountainous regions. Another important uniqueness about this work is observations of rain using X-band, X-band XPORT and K-band Radar.  Analysing data from all three provides unique solution for many problems of high mountain physical meteorology.