Review Reports - Seabed Mapping Using Shipboard Multibeam Acoustic Data for Assessing the Spatial Distribution of Ferromanganese Crusts on Seamounts in the Western Pacific

Round 1

Reviewer 1 Report

The manuscript presents a case-study in which shipboard MBES backscatter data is employed to identify the distribution of Fe-Mn crust in the deep sea. It is definitely an interesting datasets in the sense that deep-sea MBES datasets are rare. The figures are well designed and most of the data description is well done. The correlation plots between backscatter and the different seabed-types are well produced and are very informative. The strong points of the paper are the interesting study area and the amount of seamount considered. Also very interesting to see is the amount of information one can extract from the backscatter data of a shipboard multibeam in a water depth of about 2000m (i.e. seamounts) visualized by the various patterns in backscatter mosaics. However, I also have to say that overall the manuscript lacks some novelty in terms of methodology and experimental results. For example, the positive correlation between backscatter and deep-sea nodules or crust is already know for a long time (de Moustier (1985), Weydert 1985) and also shown in more recent work (de Moustier (1985), Weydert 1985) In addition, the final classification map is based on a very rudimentary approach of manually defining the boundaries for the classes instead of applying more sophisticated classification methods (see Brown et al. (2011)). I can deal with that but still a final accuracy assessment, which means how well the classes correspond to the video observations, is clearly missing.

As a case study, the manuscript present clearly the valuable use of shipboard MBES backscatter as a tool for pre-assessing the potential of Fe-Mn crust over a large area and also outlines clearly the limitations. Beside the novelty concern, I do have some minor to major comments listed below. The comments and recommendations below will help to improve the manuscript.

General recommendation

I would definitely mention more literature in the introduction dealing with MBES data for deep-sea mapping (e.g. de Moustier (1985), Weydert 1985, s Lee and Kim (2004), Ko et al. (2006), Alevizos 2008).

I would recommend to use dark colours for low backscatter and light colours for high backscatter levels. This is the most common presentation in literature and the other way around confuses the reader.

The description of the physical nature of scattering/backscattering needs some modifications (see below). A distinction between backscatter intensity, the echo level (measured value at MBES), processed backscatter level and backscatter strength is needed. The backscatter strength is a clearly defined parameter, as the ration between the incident and backscattered intensity per unit area in decibel [dB re m^2 per m^2]. The backscatter strength can be modelled with interface and volume scattering (see Jackson model or APL-model). The MBES processing aims to retrieve from the measured value (Echo level ) the backscatter strength as close as possible in order to remove effects of sound propagation, ensonified footprint or system effects. The backscatter intensity is often used in literature but it could be both the measured or the processed value. I think it need to be clearly stated somewhere what you are considering. A good location would be Line 52-56. There you can define the backscatter strength; what the sonar is measuring and how to retrieve a measure for the backscatter strength from the sonar data. And then you can make a link to Line 142 to 154 where you describe the processing of FMGT.

As already said, an accuracy test for Figure 11 is required.

You describe the backscatter patterns as normal and non-normal distributed. I would be cautios using the phrasing normal-distribution in terms of backscatter pattern. Normal-distribution has a mathematical expression and I would not use it to describe backscatter patterns. You could rather say that the backscatter histograms show two separated peaks (you referred as non-normal) and a single peak (normal distribution). I think this phrasing is also much clearer for the reader. If I read normal-distribution I get confused.

Detailed comments

Line 34 -38 Sentences need to be modified. First sentence: What sonar techniques? Second sentence: From this sentence I could understand that the side scan sonar is a state of the art tool for bathymetric measurements, which is of course not true.

Line 23-27: Very long sentence. I recommend to make two sentences. (…FE-Mn crust. This correlation…)

Line 36 and Line 39: I recommend to be consistent with the terms: Use consistently multibeam echosounder instead of multibeam sonar

Line 95 What means predicted bathymetry.? A little bit more insights are required.

Line 131 -132 It is unclear what the difference between bathymetry range and the operating depth is. For what angles is the acoustic footprint calculated?

Line 146-152 Long sentence difficult to read.

Line158 What do you mean with assume? Do you not know whether the system parameters has changed?

Line 169 What means to the nearest decimetre? Explain

Line 253 I would not mix backscatter and reflectivity. Stick to backscatter.

Line 260 I recommend to remove Table1. It does add necessary information for the reader and only takes space

Line 287 Make a new section. Something like Correspondence between MBES backscatter and video footage.

Line 305 Its actually a plus sign and not a star.

Line 305 – 306 I don’t understand “ boxes represent middle half of data”. I thought the boxes represent the 25% to 75% interquartile. The box width looks for me everywhere the same.

Line 336 -338 What’s the difference between the interface relief and micro-scale roughness? Its actually the same. I assume that you refer to specular reflection dominant at normal incident which can be modelled by facet-reflection and scattering away from normal incident which is described by bragg scattering (that how it is defined in Lurton (2010)). I would just say that backscatter is affected by the interface roughness (e.g. coarseness of the sediment, small-scale ripples) and the sediment volume heterogeneities (e.g. buried shells, bioturbation, changes of sediment acoustic parameters) The contribution of both mechanism to the total scattering depends on the hardness (i.e. acoustic impedance).

Line 339-345 I think in these lines there are some wrong statements. I would not say that the micro-scale roughness contribution is negligible using a 12 kHz MBES. The contribution of the micro-scale roughness might be lower compared to a 100 kHz signal but not negligible. The contribution of the micro-scale roughness depends on the relation between wavelength and roughness-scale expressed in the Raleigh parameter . The wavelength is 12 cm for 12 kHz and 1.5 cm for 100 kHz. This defines the sensitivity of the signal. The signal footprint only defines the instanteously ensonified footprint and the contribution of the ensonified area to the signal strength. By correction for the footprint (see Sonar equation and you also state it in Line 149) this contribution is taken out as it is defined in the backscatter strength [dB re m^2 per m^2] (see above). I also don’t understand the phrasing the seafloor geometry in that context. This paragraph needs some modification.

Line 372 -373 I don’t see a video track line in Figure 7 going towards the center of the summit proofing that the S-type is common in the center

Line 403 – 406 Why do you assume that the footprint of the MBES is larger than of the video? You know the footprints of both systems that means you clearly know that the MBES footprint is larger.

Line 455 -457 I don’t see the link between the sediment dominated summit and non-uniform BS distribution. Probably you mean the existent of sediment dominate areas and crust dominated areas resulting in two peaks in the histogram?

Figure comments

Figure 6 Is there a possibility to improve the quality of Figure 6e? It’s hard to see the sediments.

Figure 7 and 8 Length-scale is missing

Grammar Comments

See PDF document

Comments for author File: Comments.pdf

Author Response

General recommendation

I modified the introduction and added the suggested references in the revised manuscript (see lines 84-92)

I Fixed as requested (See Figure 4, 5, 7 and 8)

The description of the physical nature of scattering/backscattering needs some modifications (see below). A distinction between backscatter intensity, the echo level (measured value at MBES), processed backscatter level and backscatter strength is needed. The backscatter strength is a clearly defined parameter, as the ration between the incident and backscattered intensity per unit area in decibel [dB re m^2 per m^2]. The backscatter strength can be modelled with interface and volume scattering (see Jackson model or APL-model). The MBES processing aims to retrieve from the measured value (Echo level) the backscatter strength as close as possible in order to remove effects of sound propagation, ensonified footprint or system effects. The backscatter intensity is often used in literature but it could be both the measured or the processed value. I think it need to be clearly stated somewhere what you are considering. A good location would be Line 52-56. There you can define the backscatter strength; what the sonar is measuring and how to retrieve a measure for the backscatter strength from the sonar data. And then you can make a link to Line 142 to 154 where you describe the processing of FMGT.

I modified the introduction to reflect the reviewer’s suggestion in the revised manuscript (see lines 65-77).

As already said, an accuracy test for Figure 11 is required.

I added a new table to provide the result of accuracy test in the revised manuscript (see Table 3)

Fixed as requested (line 243, 246, 249, 254, 255, 351, 352, 378, 420, 425, 439 and 493)

Detailed comments

I modified sentences in the revised manuscript (see lines 47-51)

Line 23-27: Very long sentence. I recommend to make two sentences. (…FE-Mn crust. This correlation…)

I split the sentence as suggested (see lines 23-25 and 25-27).

Line 36 and Line 39: I recommend to be consistent with the terms: Use consistently multibeam echosounder instead of multibeam sonar

I used the same term for consistency in the revised manuscript (see line 49 and 52).

Line 95 What means predicted bathymetry.? A little bit more insights are required.

According to Global Multi-Resolution Topography Data synthesis(GMRT)’s website, GMRT bathymetry is maintained as a multi-resolution gridded global Digital Elevation Model (DEM) that includes ship-based multibeam sonar data (Source data include : Ship-based multibeam swath bathymetry data, terrestrial elevation data and several gridded seafloor depth data). Therefore, I adapt the keyword of “the predicted bathymetry dataset” about GMRT bathymetry. I added this information in the manuscript (see lines 127-128)

Line 131 -132 It is unclear what the difference between bathymetry range and the operating depth is. For what angles is the acoustic footprint calculated?

The “bathymetry range” means a depth of the whole surveyed area. The “operating depth” represents a depth extracted from obtained multibeam data about target’s seamount. I corrected sentence the “acoustic footprint as “signal resolution” (see lines 154-155).

Line 146-152 Long sentence difficult to read.

I split the sentence as requested (see lines 171-174 and 174-175).

Line158 What do you mean with assume? Do you not know whether the system parameters has changed?

I modified the sentence (see lines 181-182). I confirmed with acoustic engineer that the system parameter controlling the acoustic backscatter did not change.

Line 169 What means to the nearest decimetre? Explain

Due to distance between two lasers is 20 cm, the observer estimated the sizes of something from the video footage relatively. I used the term “materials” instead “decimeter” (see line 192).

Line 253 I would not mix backscatter and reflectivity. Stick to backscatter.

I used the term “backscatter” only in the revised manuscript as requested.

Line 260 I recommend to remove Table1. It does add necessary information for the reader and only takes space

I removed Table 1 as requested. (Table 2 -> Table 1, Table 3 -> Table 2)

Line 287 Make a new section. Something like Correspondence between MBES backscatter and video footage.

Fixed as requested (see line 309)

Line 305 Its actually a plus sign and not a star.

Fixed as requested (see line 328).

Line 305 – 306 I don’t understand “boxes represent middle half of data”. I thought the boxes represent the 25% to 75% interquartile. The box width looks for me everywhere the same.

> I removed the wrong sentences.

I modified sentences to reflect the reviewer’s comments (see lines 363-367).

Line 339-345 I think in these lines there are some wrong statements. I would not say that the micro-scale roughness contribution is negligible using a 12 kHz MBES. The contribution of the micro-scale roughness might be lower compared to a 100 kHz signal but not negligible. The contribution of the micro-scale roughness depends on the relation between wavelength and roughness-scale expressed in the Raleigh parameter. The wavelength is 12 cm for 12 kHz and 1.5 cm for 100 kHz. This defines the sensitivity of the signal. The signal footprint only defines the instanteously ensonified footprint and the contribution of the ensonified area to the signal strength. By correction for the footprint (see Sonar equation and you also state it in Line 149) this contribution is taken out as it is defined in the backscatter strength [dB re m^2 per m^2] (see above). I also don’t understand the phrasing the seafloor geometry in that context. This paragraph needs some modification.

I modified sentences to reflect the reviewer’s comments (see lines 367-374).

Line 372 -373 I don’t see a video track line in Figure 7 going towards the center of the summit proofing that the S-type is common in the center

I modified sentences (see lines 401-402). Original sentences meant that the S-type seabed which was frequently observed near the center of summits (DCS transects: DCS130201’s A’ location of DCS130201, B’ location of DCS130202 and C’ location of DCS130203).

Line 403 – 406 Why do you assume that the footprint of the MBES is larger than of the video? You know the footprints of both systems that means you clearly know that the MBES footprint is larger.

I wrote that sentence wrong. I modified sentences (see lines 432-434).

I modified sentences (see lines 491-494).

Figure comments

Figure 6 Is there a possibility to improve the quality of Figure 6e? It’s hard to see the sediments.

It is not possible to improve the quality of Figure. If it display only the picture of sediment-dominant area, it is easy to see the sediments.

Figure 7 and 8 Length-scale is missing

Fixed as requested.

Grammar Comments

See PDF document
peer-review-5729418.v2.pdf

Fixed as requested.

Reviewer 2 Report

This manuscript is well organized, well presented in terms of illustrations, and well written. No changes are recommended except for very rare spelling, font and punctuation as suggested in the pdf document.

I have one concern only, and this arises from my inexperience with Mn nodules in general. It is not clear if the backscatter character arises from the nodule habit itself or is it more by proxy? Stated another way...Is the acoustic character due to the nodules, and thus a true reflection of their presence, or is it that the crust develops on a substrate which has characteristics that produce the acoustic signature regardless if there exists a crust? Perhaps this could be made clear in the discussion.

One final comment. There is ample evidence of sediment transport. Presumably this can bury Mn crusts, or conversely, expose non-crusted substrate. In such a situation, the "mineable" content might be mis-represented some. Might the suggest multi-spectral MB system address this some. Is this worth mentioning in the manuscript?

Author Response

I modified “Discussion” section according to the reviewer’s comments in the revised manuscript (see lines 360-480).

After the initial exploration, it is necessary to estimate the spatial distribution of seabed mineral deposits for high-resolution quantitative mapping. Advances in MBES transducer technology have resulted in wider operating bandwidths, with some systems on the market now being capable of spanning several hundred kHz to offer improved range resolution for bathymetric data collection. However, to date, most of these systems only operate at a single frequency at any one time, and therefore the acquired dataset is still single resolution in nature. Therefore, he application of a broadband MBES system that can produce multispectral backscatter data may provide further details on the various seabed types.

Reviewer 3 Report

The authors conducted mapping surveys at several seamounts in the western Pacific and investigated the spatial distribution of Fe-Mn crusts. They provided sufficient data on the relationship between MBES backscatter intensities and seabed types determined by DCS observations. Although spacial resolution of shipboard MBES is not enough high to distinguish seabed types in small area, their dataset showed backscatter data are useful for early stage exploration.

Overall, these results in this study deserve publication. However, there are some issues to be address. First, the introduction should be improved. The background information about deep-sea mining and ferromanganese crusts is missing. For example, the paragraph describing Fe-Mn crusts in 2.1. Study area (lines 101-107) should be moved into the introduction. Second, the criteria that the authors showed in Table 2 are unclear. As they admitted in line 428, there are no significant difference between the backscatter intensities of C1-, C2-, and C3-type seabeds. Thus, I could not understand why the authors interpreted -30~-20dB as mixed area and >-20dB as crusts. Furthermore, I wonder why the ranges of backscatter intensities in Type 1 seamounts is broader than those in Type 2 seamounts (Figure 9). I require the authors to add the interpretation of that.

Minor comments are follows:

Lines 108-112: Add the official names of these seamounts.

Figure 6: Add seamount name for each photo.

Figure 11: Insert seamount names in the figure.

Author Response

(1) I modified the introduction (see lines 34-46). (2) I corrected Table 1 to provide clear grouping. (3) I modified some sentences to explain about the broader distribution of backscatter intensity in Type 1 seamounts (see lines 338-340 and 351-353).

Minor comments are follows:

Lines 108-112: Add the official names of these seamounts.

There are no the official name of these seamounts, as a result of checking with the International Seabed Authority.

Figure 6: Add seamount name for each photo.

Fixed as requested.

Figure 11: Insert seamount names in the figure.

Fixed as requested.

Round 2

Reviewer 1 Report

Dear authors,
As I already said during the first review, the methodology and the experimental results are not very innovative, nevertheless this manuscript represents an interesting datasets in the sense that deep-sea MBES datasets are rare The figures are well designed and most of the data description is well done.

I appreciate your effort to account for my comments. Even though I like your manuscript as a case study, I am not satisfied with the revision you have made. The sentences, which were added during the revision, are very difficult to read. I am not a native speaker but the English language is difficult to follow. In addition, I am struggling with the explanation of the main physical quantity, the backscatter strength, in the introduction (Line 67 to 79) and the conclusion you draw from the measured backscatter data based on the acoustic theory (Line 546-548). The explanation of the backscatter strength does not follow my previous comments or the given literature (Lurton 2010 or Jackson 1997). But maybe just the English language hamper a clear understanding.

In general, most of the revised parts hamper the reading flow of the paper and I have the feeling that you added some sentences without considering the meaning of it and neglecting the reading flow. I highly recommend to ask a native speaker for editing the English language.

I listed some comments below and in the pdf document to give some examples of possible improvements. However, before I can recommend publication, not only the listed comments need to be addressed but also the entire paper requires an extensive editing of the English language in order to improve the reading flow. In the current state, the paper is hard to read. Furthermore, the explanation of the backscatter strength and the parts where the theory is used to draw conclusions from the measured data needs to be highly improved as well.

Line 67 to 79. Very unclear. The theory is not explained well. This paragraph needs to be significantly improved with respect to the physics and the English grammer before I can recommend publication. Maybe reduce the complexity of the explanation. Check Lurton 2010 or some other papers explaining backscatter in the introduction.

Line 492 “…more than twice as larger as…”: bad english

Line 504-506 The added sentence is very unclear due to lacking English

Line 546-548 The sentence does not make sense in my opinion. There seems to be a fundamental misunderstand of the backscatter strength and difficulties with the English language.

Line 708-712 How can you use the acoustic data to estimate the thickness of Fe-Ms crust???
Probably it is just a wrong formulation otherwise the results do not agree with this conclusion.

see attachment

Comments for author File: Comments.pdf

Author Response

> The reviewer’s main concern and suggestion is improving English of the manuscript. We agree to the reviewer’s comments and tried to improve English of the whole manuscript. We also reorganized some part of the manuscript to remove redundancy and for better reading flow. For example, we removed the sentences for manganese nodules survey part from Introduction section instead of including a pertinent references as follows : “to classify the seafloor types for mapping distribution of seabed resources (e.g., manganese nodules) [5,6,11-17]” instead of “to classify the seafloor types for mapping distribution of seabed resources [5,6,11,12]”. The whole manuscript is checked by native English speaker. We hope the revised manuscript is readable.

As the reviewer pointed out, we have revised all the sentences related to drawing conclusions from measured data based on the acoustic scattering theory.

Line 49 to 52. (i.e., multibeam echo sounder, singlebeam echo sounder or side scan sonar)

I am not aware of a side scan-sonar measuring precisely the bathmetry, I would check that. Otherwise maybe restructure the sentence where you clearly see that MBES and SBES can be used for both, bathymetry and backscatter, and side scan only for backsactter (i.e., only sediment property)

> We have modified the sentence to reflect the reviewer’s comments (see lines 44-45).

> We have added more detailed descriptions for this paragraph to reflect the reviewer’s comments in the revised manuscript (see lines 59-65)

Line 492 “…more than twice as larger as…”: bad english

> We have modified sentence in the revised manuscript (see lines 274-275)

Line 504-506 The added sentence is very unclear due to lacking English

> We have modified the sentence to reflect the reviewer’s comments (see lines 284-286).

Line 546-548 The sentence does not make sense in my opinion. There seems to be a fundamental misunderstand of the backscatter strength and difficulties with the English language.

> We have modified the sentence to reflect the reviewer’s opinion in the revised manuscript (see lines 301-304).

Line 708-712 How can you use the acoustic data to estimate the thickness of Fe-Ms crust???
Probably it is just a wrong formulation otherwise the results do not agree with this conclusion

> We have modified sentence in the revised manuscript (see lines 380-381)

Author Response File: Author Response.pdf