Suppression of Wind Ripples and Microwave Backscattering Due to Turbulence Generated by Breaking Surface Waves

Round 1

Reviewer 1 Report

The work “Suppression of wind ripples and microwave backscattering due to turbulence generated by breaking surface waves” deals with an experimental study of the behavior of the short gravity-capillary waves within a breaking event and its effect on the radar backscatter. Within the study authors are proving the assumption, that the wind ripples are suppressed by turbulence, taking the transversal velocity component behavior after and before intense wave train passage (the breaking event). Thus a correlation between attenuation ratio as a function of the wave steepness to the corresponding suppression ratio of the radar backscater intensity (also as a function of the wave steepness) was demonstrated. In literature some alternative explanations of the short wave component’s suppression might be found (see e.g. “The velocity field under breaking waves: coherent structures and turbulence” by W. K. Melville, F. Veron and C.J. White, JFM,2002). In this work the authors attribute the short wave part of the spectrum’s suppression by the fact of the blockage due to wave-current interactions. I think authors might do deeper literature investigation to find more clear proofs of their assumption of the dominant role of turbulence in the suppression of short waves’ part of the spectrum. The last part of the Discussion section 4 (lines 420-453) can be considered to be stronger argument in favor of turbulence if the approach used for damping of capillary waves due to surfactant films will be justified in its applicability to the current study. In contrast to the theoretical justification, experimental results from all the instruments obviously show the fact of the short wave suppression after breaking events, even though not clearly revealing the mechanism causing this process. Current experimental results looks convincing. I think it would strengthen the introduction and discussion part if the authors discussed the contradiction of their experimental results to the results previously reported in the paper of A. Rosenberg and M. Ritter Laboratory study of the fine structure of short surface waves due to breaking: Two-directional wave propagation, JGR, 2005. It is really interesting why the effect of the short wave suppression and its manifestation on a radar backscatter was not reported before. Probably a broader literature review is needed in the current study. I also recommend the authors to substantially revise their work regarding English language editing and plot representation. Some particular recommendations will be enlisted below. From general impression and recommendations I would like to move to particular comments, which will be enlisted in the order they appear in the text of the paper:

1. Lines 20-23. The sentence is long and not quite clear. Please try to reformulate and divide into two sentences.
2. Introduction lines 39-50. The word “peculiarities” does not fit the meaning and is repeated 3 times, try to use synonyms (features, aspects, etc.), which are more plausible,
3. Line 45: delete the extra space in parenthesis.
4. Lines 52-52: should be Normalized Radar Cross Section (NRCS).
5. Lines 64-67: long sentence, try to divide in two.
6. Lines 76-78: “… scale of turbulent eddies is comparable in size with wavelength.” Which wavelength are meant? We have whole spectrum.
7. Line 83: “The paper is organized as follows...”- new paragraph.
8. Line 98: Try to avoid using more than two ...of ….of. “A beach was mounted at the end of the working part of the tank….” might me changed to “A beach was mounted at the working part end of the tank….”. Please fix the same issue in other similar situations.
9. Figure 1: Would be great to put the pointer and write names of the instruments if they visible in the photo (radar, wire gauges, wave maker). In the scheme the wind fan is missing. Also for all figures, please, check the fact that all axes has labels and they are given in the format “Value (units)“.
10. Line 107: should be “...thus shortening the wind wave fetch...”.
11. Line 113: should be “(see below)”.
12. Figure 2: Since the wave train is depicted on the different fetch it would be more correct to give the figures as a function of the following argument: (t-x/c_g)/T_0, where c_g is the group velocity, and T_0 is the carrier wave period. If the linear frequency modulation was applied to the initial wave train if would be beneficial to also show pseudo-frequency wavelet spectra (or spectrograms) of the initial and focused wave trains. And present them as four panels 2*2.
13. Figure 3: put more ticks to the x-axis to see the peak at 60 Hz clearly. Empty space in y-axis might be deleted to compact the figure.
14. Lines 162-164:regard the change of “...which evidences...” to “...which is evident due to…” and “… in our experiment partly overlapped and this made difficult to select the spectra of...” to “… in our experiment were partly overlapping and this made a difficult to select spectra of...”.
15. Line 168: delete the extra space in parenthesis.
16. Figure 4: what is the reason for the transversal component to have a low frequency trend, deviating it’s mean value from zero? b) Regard the possibility to give the spectra in f/f_p coordinates to clearly see higher harmonics. f_p value might be given as a reference in the caption.
17. Lines 178,183 delete comma in parenthesis (see Figure 5), (see below).
18. Figure 5: a) Check the values and put the units of the x and y lengths; b) add a legend.
19. Figure 6: a) Zero frequency should be set to zero subtracting the mean values of the corresponding time series. Please, put more ticks to the x-axis. b) Please, give the definition (formula) of the “weighted frequency of dominant wind waves”. Discuss what is the trend and why it is important to show it.
20. Lines 209-210: Discuss what is the reason for such a behavior of the Doppler shifts, give references if needed. What is the nature of the lower peak?
21. Figure 7: a) remove empty space; b)the same, should be “VV and HH polarizations”. Define the “weighted Doppler shift” in the text.
22. Line 224 “...corresponds to theoretical values in the frame of two-scale Bragg theory” (reference is missing)
23. Line 234: should be “studied area”.
24. Figure 9: Check units on all the axes including color codes. Time axes can be given just ones in the bottom figure if it is the same to all of them with proper label and units. Line 242: should be “series”. Please give the legend or entitle in caption three velocity components.
25. Figure 10: Regard a possibility to unite all the three figures one under another with common x-axis label (given just for the lowest figure), add title to each case on the figure itself e.g. “LF before breaking”, etc. Legend might be given just one. Delete blanked spaces to compact the figure.
26. Figure 11. the same as for figure 10.
27. Think if it’s possible to unite somehow figures 12 and 13. As an option to give them as 4 panels 2*2.
28. Line 305 should be “(see Figure 6(b))”
29. Line 306-308: Not a clear sentence. Why the maximum should be higher?
30. Line 336: The abbreviation GCW appears the first time, please give the expansion (Gravity-Capillary Waves?).
31. Lines 325-332 and 334-335. Contradicting facts are given. Free and bound waves propagates with the same celerity, as it was stated before (ines 325-332). Please, fix the contradiction, may be something else was meant?
32. Figure 15: put the limiting line on the slope of 0.22-0.23 separating the breaking and non-breaking cases making it more clear why the abrupt change of the behavior is present. Define suppression ratio explicitly as a formula in the paper text.
33. Figure 16: the same as figure 15. y-axis should be AR (amplification ratio)?
34. In general try to justify more accurately, why the transversal component of the current is the only one associated with turbulence. In general, turbulence is a 3D process and all the velocity components can not be regarded as independent ones. In addition the wave field is not a 2D one, since wind component at least is essentially 3D. So it might influence the transversal component of the current as well as some standing waves might appear in the lateral direction.

Comments for author File: Comments.docx

Author Response

We are very thankful to anonymous reviewer 1 for very valuable comments, particularly for information and suggestions regarding papers by Melvill, Veron&White 2002 (MVW), Rosenber&Ritter 2005 (RR). We made corrections and added some text taking into account the reviewer’s comments.

Author Response File: Author Response.doc

Reviewer 2 Report

This paper reports laboratory investigations of suppression of radar backscattering in Ka-band after passage of the wave breaking wave. Experiments were performed at the low wind conditions with and without mechanically generated long waves. The breaking of the long waves was triggered by “dispersive focusing”. Wave gauge and acoustic Doppler velocimeter provided the detailed information about surface waves and subsurface turbulence generated by breaking waves.

Analysis of the radar Doppler spectra revealed that at the background conditions (no mechanical wave), radar scattering was supported by either the resonant Bragg scattering from capillary wind waves and non-polarized radar returns from non-linear feature of dm-length waves.

As found, during the passage of breaking crest through the radar footprint, radar backscatter is strongly enhanced in a wide range of the Doppler frequencies. After the passage, the radar backscatter has rapidly dropped and then gradually recovers to the pre-breaking level. After carefull inspection and analysis of the measurements of radar backscatter, wave spectra and sub-surface turbulence, the authors came to the conclusion that the governing mechanism leading to remarkable attenuation of radar backscatter behind the breaking wave is suppression of short wind waves by sub-surface turbulence. The authors provided physical arguments justifying such suggestion. It was also concluded that suppression of radar signal in turbulent wake behind the breaker, may erroneously be treated as radar backscatter suppression due to surface films or effect of wind shadowing.

All in all, the paper is clear written and contains interesting experimental material which is worthy of publication in Remote Sensing. However, I have some comments that must be properly addressed prior to the publication.

1. Reported results are obtained at very specific lab conditions, in particular – at very low wind (from 1m/s to 2 m/s), and rather short breaking waves. I would strongly advise to discuss of how the authors' results can be extrapolated to real and higher wind conditions.
2. In particular, in sec.4 the authors discuss observed suppression ratio, and suggested a simple physical model to interpret observed phenomenon. This model is based on the energy balance between wind input (Phillips mechanism) and wave dissipation due to sub-surface turbulence. This model and estimates are only valid at very low winds (mainly due to Phillips mechanism). However, this idea could be extended to higher wind conditions. To that end, one needs to modify the wind energy input (replacing Phillips' to Miles' generation), to keep the same estimates of energy dissipation due to sub-surface turbulence, and to add the non-linear dissipation, similar to what the authors did before with slicks, or similar to [4]. Such extension of the lab finding to the higher wind is very desirable, since it helps to assess importance  of the considered mechanism in real conditions.
3. In lines 231-238 the time series of radar backscatter before-during-after breaker, shown in Fig.9, are discussed. It is stated: “When the breaking waves enter the radar footprint the backscatter is strongly enhanced. After the intense waves leave the radar return decreases below the initial level ….”. However, inspection of Fig.9 reveals  neither enhancement nor decrease of the radar return. What a reader may find from this figure is a constant, in the mean, radar signal with wave-induced undulations. On the other hand, referring to the time evolution of the Doppler spectrum in the same Fig.9, one may find that described features are really existing. Please modify the figure or clarify the text around.
4. In this context, it would be interesting (and important) to check, what is the overall mean impact of individual breaking waves on radar signal, - increase or decrease. By other words, what is the net impact of radar signal enhancement by the breaker and consequent radar signal suppression behind the breaker?
5. In sec. 4 in lines 362-379 the authors, using velocity of facets derived from Doppler shift and measured PR, came to the conclusion that either resonant scattering and non-polarized return from non-linear wave structures contribute to radar signal. However, no quantitative estimates on relative contribution of these scatters to total NRCS are suggested. Yurovsky et al. (2017), did a similar analysis in field conditions, and estimated contribution of Bragg and non-polarized scattering to the total NRCS in Ka-band. It would be worthy, following [5] and Yurovsky et al. (2017), to compare lab and field measurements.

 

Reference:

Yurovsky Yu. Yu., V. N. Kudryavtsev, S. A. Grodsky, and B. Chapron, (2017). Ka-band Dual Co-Polarized Empirical Model for the Sea Surface Radar Cross-Section. IEEE Transactions on Geoscience and Remote Sensing, 55(3), 7769183, 1629-1647, DOI: 10.1109/TGRS.2016.2628640).

Author Response

We are very thankful to anonymous reviewer 2 for very valuable comments, information and suggestions We made corrections and added some text taking into account the reviewer’s comments.

Author Response File: Author Response.doc

Reviewer 3 Report

This paper can be published after minor correction as pointed out.

This paper reports an experimental study on the scattering of microwaves by water surface, and the experimental results are well described.
The point to be revised is that some of the results are similar to previous studies, but there is no comparison with them.
Please write clearly what is different from the past studies including the studies mentioned below.

Ebuchi, Naoto, Hiroshi Kawamura, and Yoshiaki Toba. "Physical processes of microwave backscattering from laboratory wind wave surfaces." Journal of Geophysical Research: Oceans 98.C8 (1993): 14669-14681.

Plant, William J., et al. "Bound waves and Bragg scattering in a wind‐wave tank." Journal of Geophysical Research: Oceans 104.C2 (1999): 3243-3263.

Rozenberg, Anatol D., Derek C. Quigley, and W. Kendall Melville. "Laboratory study of polarized microwave scattering by surface waves at grazing incidence: The influence of long waves." IEEE transactions on geoscience and remote sensing 34.6 (1996): 1331-1342.

Walker, D. T., et al. "Radar backscatter and surface roughness measurements for stationary breaking waves." Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences 452.1952 (1996): 1953-1984.

Zuo, Lei, et al. "An efficient method for detecting slow-moving weak targets in sea clutter based on time-frequency iteration decomposition." IEEE Transactions on Geoscience and Remote Sensing 51.6 (2012): 3659-3672.

Author Response

We are very thankful to anonymous reviewer 3 for very valuable comments, information and suggestions We made corrections and added some text taking into account the reviewer’s comments.

Author Response File: Author Response.doc

Reviewer 4 Report

See attached file

Comments for author File: Comments.pdf

Author Response

We are very thankful to anonymous reviewer 4 for very valuable comments. We made corrections and added some text taking into account the reviewer’s comments.

Author Response File: Author Response.doc

Round 2

Reviewer 1 Report

The work “Suppression of wind ripples and microwave backscattering due to turbulence generated by breaking surface waves” was substantially revised within the first iteration of the peer-review process. I appreciate that authors did a serious work to address all the comments, which were quite a number. Now the paper looks to be much easier to read and follow and the main idea and its explanation are more clear. I would be happy to recommend the paper for the publication after minor spell-check correction and addressing a few rested recommendations.  

1. Figure 2: The previous comment: “Since the wave train is depicted on the different fetch (x-coordinate) it would be more correct to give the figures as a function of the following argument: (t-x/c_g)/T_0, where c_g is the group velocity, and T_0 is the carrier wave period. If the linear frequency modulation was applied to the initial wave train if would be beneficial to also show pseudo-frequency wavelet spectra (or spectrograms) of the initial and focused wave trains. And present them as four panels 2*2”. I see that the idea of the time axis redefinition was not understood correctly. Please check again the formula and write it in a correct way both in the axis itself and in the caption. There should not be negative start times on the farther fetch. In the closest fetch case x (fetch) can be set to zero and the second one should be calculated as a distance between the gauges. T_0 is the carrier wave period. Think about what period to take in the nearest fetch as it is obvious that the signal is frequency modulated. Ideally recalibrated axes should start the same (zero) time and have the same length. I also recommend to put timeseries one under another and move the corresponding wavelet spectra to the right two panels.
2. Figure 16: The figure is missing now for a reason.
3. Please check the formula characters style (font sizes and types, words in indexes, brackets sizes, etc.) to be the same and appropriate within the discussion section.

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.pdf