Dronar—Geoforensic Search Sonar from a Drone
Round 1
Reviewer 1 Report
I enjoyed reading this manuscript! It is interesting work, albeit fairly straight-forward. But the examples are well-presented and relevant.
Check on the frequency of the WPR. In the text it is 100 MHz and in the figure it is 200 MHz.
I really only have two suggestions:
1) Your use of commas is sometimes confusing. Perhaps ask another person to read it through again and cut down on the number of commas. As it stands it is comprehensible, but just a little difficult to read in places.
2) The arrows on your figures are really faint. Perhaps these would be better in white? Of course, the picture quality may be much better in the published version. Similarly, the text in the figures is small and a bit difficult to read.
Author Response
Reviewer 1.
Check on the frequency of the WPR. In the text it is 100 MHz and in the figure it is 200 MHz. – Now corrected.
I really only have two suggestions:
1) Your use of commas is sometimes confusing. Perhaps ask another person to read it through again and cut down on the number of commas. As it stands it is comprehensible, but just a little difficult to read in places. Checked and hopefully clarified throughout.
2) The arrows on your figures are really faint. Perhaps these would be better in white? Of course, the picture quality may be much better in the published version. Similarly, the text in the figures is small and a bit difficult to read. All figures now have larger fonts and thicker/different colour arrows etc.
Reviewer 2 Report
I am not an expert in forensic science, but I do research and publish on a variety of remote sensing methods, including collecting and processing shallow bathymetric data. For these reasons, I cannot comment on whether the audience of this journal will find this study contains new or novel information. However, assuming this paper represents progress in this field, I can recommend it for publication following moderate revision.
At present, the authors make a clear case for incorporating UAV-mounted sonar to improve search efforts by (a) rapidly focusing into the best search area and (b) gathering data at increased rates and accuracies. However, the paper never explicitly states this or any other hypothesis, research question, or aim. This is one of the many (currently missing) elements that must be added to the paper to meet the standards of a scientific peer-reviewed publication. I do not think the paper needs additional data collection or analysis. The authors allude to robust data collection and processing throughout, but they must increase their descriptions of these, and reorganize, so that this paper meets the standards of a peer-reviewed scientific journal. There are also minor grammatical corrections needed throughout
The attached PDF contains itemized comments. Many of these are minor corrections, such as the misuse of semicolons or small typos. However, some are more intense corrections that are needed. I have summarized those comments below. If the authors can make these changes, I believe the paper is publishable.
Summary of major comments in attached PDF:
1) A clearly stated hypothesis, research question, or primary aim of the study should be added to the introduction, repeated at the top of results, and repeated in the conclusion.
2) Of note for this entire paper: The term "drone" is well known, colloquial, and acceptable in most commercial circles. However, in academic research, more specific language is used. The DJI machines described in section 2 are "unmanned aerial vehicles" or UAVs. I would suggest using the term UAV throughout this paper. However, I can also imagine the authors want to ensure that non-academics can easily find and understand their paper. It would also be an acceptable edit to add in the introduction a definition of the academic term UAV, and explain that the commercial term "drone" used throughout the paper refers to such a UAV.
3) On most figures: The text labels are often too small. I suspect the figures were designed for a larger interface (perhaps a presentation) and have not been properly scaled for the sizes used in this paper. Please ensure that all text and symbols are clearly legible when the PDF is viewed at 100%. Most journals require a minimum text size of 11 (but I am not certain of this journal’s formatting requirement).
4) In multiple places throughout: Language is too conversational or casual. This might be how we talk to each other at happy hour, but it is not appropriate for a publication. Be specific with descriptions. For example: instead of saying the system is “superb” say "All sites yielded surveyed surfaces within +/-X m elevation (where validated) and captured target objects." Be quantitative wherever possible.
5) Reorganization needed throughout. If you want to describe the validation process as its own section, make it a subsection of methods or move it before results. It's not appropriate to talk about final results and then describe validation. Please rearrange information to the scientific format, where these sections are used and contain the following:
Methods:
- Specific descriptions of every instrument (including how each instrument was mounted to specific vehicles)
- specific descriptions of the signal frequencies used and resolutions captured - how different data sources were combined into a single bathymetric surface
- what data was used for validation
- how validation was performed (DEM differencing?)
Results:
- Including results of validation
--- Ex: "RMSE for each survey was ______ when compared by [what method of analysis] to the [validation data]."
- other quantitative measures of how the system being tested was proved successful
Discussion*:
- how the system and results compare to previous publications
- how the work can move forward
Conclusion:
- one short paragraph reminding the reader:
--- the goal of the project
--- the key finding(s)
--- what this means for the field
*There is not currently a discussion section, but the material that should be there is scattered throughout results, validation, and conclusions. Here, you should describe specifically how the methods presented in this paper build on or combine the two cited (Bandini et al. and Giambastiani et al.).
6) Increased detail required for many of the statements currently in the paper. For example, rather than saying only “we followed Name et al.” A complete sentence in the methods might say “we followed Name et al. by using [this] to do [this].” Increased detail is also required regarding equipment. Which models of which instruments were used? What was the frequency range of each? What were the margins of error when each survey was validated against locations with known elevations? Be specific, and use measured values as often as possible. In many places the details of methods and results are poor or missing entirely.
Comments for author File: 
Comments.pdf
Author Response
Reviewer 2
At present, the authors make a clear case for incorporating UAV-mounted sonar to improve search efforts by (a) rapidly focusing into the best search area and (b) gathering data at increased rates and accuracies. However, the paper never explicitly states this or any other hypothesis, research question, or aim. Now done in the new first paragraph
This is one of the many (currently missing) elements that must be added to the paper to meet the standards of a scientific peer-reviewed publication. I do not think the paper needs additional data collection or analysis. The authors allude to robust data collection and processing throughout, but they must increase their descriptions of these, and reorganize, so that this paper meets the standards of a peer-reviewed scientific journal. We have not repeated the excellent statistical tests of Bandini et al., and Giambastiani et al., so these are quoted more frequently throughout the revised text, such as ‘validation of the method by Bandinia/Giambastiani’. In addition, we have expanded on the validation section.
The attached PDF contains itemized comments. Many of these are minor corrections, such as the misuse of semicolons or small typos. However, some are more intense corrections that are needed. I have summarized those comments below. If the authors can make these changes, I believe the paper is publishable.
Summary of major comments in attached PDF:
1) A clearly stated hypothesis, research question, or primary aim of the study should be added to the introduction, repeated at the top of results, and repeated in the conclusion. This is added to the Introduction, and repeated as asked for.
2) Of note for this entire paper: The term "drone" is well known, colloquial, and acceptable in most commercial circles. However, in academic research, more specific language is used. The DJI machines described in section 2 are "unmanned aerial vehicles" or UAVs. I would suggest using the term UAV throughout this paper. However, I can also imagine the authors want to ensure that non-academics can easily find and understand their paper. It would also be an acceptable edit to add in the introduction a definition of the academic term UAV, and explain that the commercial term "drone" used throughout the paper refers to such a UAV. The reviewer is exactly correct, if now out of date as there are many more UAV platforms than drones (e.g. fixed wing), so we have done as asked and added this clarification, but used drones (as that is what we mean, plus communication as Rev. 1 states).
3) On most figures: The text labels are often too small. I suspect the figures were designed for a larger interface (perhaps a presentation) and have not been properly scaled for the sizes used in this paper. Please ensure that all text and symbols are clearly legible when the PDF is viewed at 100%. Most journals require a minimum text size of 11 (but I am not certain of this journal’s formatting requirement). Many thanks, Reviewer 1 makes this point as well, so we have made the images clearer. Unfortunately, this work has never been presented (as in a Powerpoint), due to Lockdown, more it is our fault for seeing the images on a good clear retina screen, apologies if others have less visible computer screens.
4) In multiple places throughout: Language is too conversational or casual. This might be how we talk to each other at happy hour, but it is not appropriate for a publication. Be specific with descriptions. For example: instead of saying the system is “superb” say "All sites yielded surveyed surfaces within +/-X m elevation (where validated) and captured target objects." Be quantitative wherever possible. Many thanks – we have been less casual in our happy hour language and made the descriptions more staid, as suggested.
5) Reorganization needed throughout. If you want to describe the validation process as its own section, make it a subsection of methods or move it before results. It's not appropriate to talk about final results and then describe validation. Good point, perhaps we have over-stated the work, which is not an actual search, where this comment is absolutely correct, but more feasibility – and thus validation of the experiments (‘results’) is needed. We altered the text to reflect this, and to make it clear, made the section heading ‘Initial Results’ – which naturally require post-collection validation.
Please rearrange information to the scientific format, where these sections are used and contain the following:
Methods:
- Specific descriptions of every instrument (including how each instrument was mounted to specific vehicles) – we do this, same for each device and drone.
- specific descriptions of the signal frequencies used and resolutions captured: now in the new Table
- how different data sources were combined into a single bathymetric surface Again perhaps our description is poor and the reviewer has misunderstood (apologies) – the frequencies of the sonar and outputs are the same throughout, just tethered to a different drone. Table added to section 2.
- what data was used for validation added plumb bob and WPR
- how validation was performed (DEM differencing?) as above - direct measurement
Results:
- Including results of validation – WPR and plumb bob added
--- Ex: "RMSE for each survey was ______ when compared by [what method of analysis] to the [validation data]."
- other quantitative measures of how the system being tested was proved successful – as above, WPR and plumb-bob
Discussion*:
- how the system and results compare to previous publications: the abundant reference to the two main previous works is now repeated in Conclusions
- how the work can move forward. We added a section on this.
Conclusion:
- one short paragraph reminding the reader:
--- the goal of the project
--- the key finding(s)
--- what this means for the field
Now added
*There is not currently a discussion section, but the material that should be there is scattered throughout results, validation, and conclusions. Here, you should describe specifically how the methods presented in this paper build on or combine the two cited (Bandini et al. and Giambastiani et al.). As above, going back to the previous two key works.
6) Increased detail required for many of the statements currently in the paper. For example, rather than saying only “we followed Name et al.” A complete sentence in the methods might say “we followed Name et al. by using [this] to do [this].” Updated section 2
Increased detail is also required regarding equipment. Which models of which instruments were used? What was the frequency range of each? Table 1 added
What were the margins of error when each survey was validated against locations with known elevations? Be specific, and use measured values as often as possible. In many places the details of methods and results are poor or missing entirely. We add some margins of error, but as stated in the text, 'known elevations' (measured bathymetry) is impossible, as plumb-bob/probe nearly always meets a layer of thixotropic sediment. Even the radar return is inaccurate, as the dielectric shift between water and sediment, will be at a wavelength-dependant threshold that varies for each antenna or Sonar.
