UAV and Structure-From-Motion Photogrammetry Enhance River Restoration Monitoring: A Dam Removal Study
Round 1
Reviewer 1 Report
I think this manuscript has potential to be of value, with high interest to a wide variety of readers. That said, I think it needs to be modified to ensure it can be repeated by other scientists or practitioners and for clarity to readers.
I hope the authors find my review constructive and helpful, as I think most of the substance for this to be a very interesting paper exists. The topic is interesting and important, but the clarity of the manuscript needs improvement. My main concerns are with the methods section, as the way it reads now makes me find it hard to believe it would be repeatable by other researchers/scientists/practitioners, and I have a hard time following some of the results and discussion based on what I took away from the methods. I also think this paper needs a bit of reorganization, as it is hard to tell whether or not this is focused on restoration 'or' a new technique. I think it'd be best framed as a restoration paper, where you are using UAV and interesting techniques to improve restoration monitoring -- rather than a paper about SfM/Remote Sensing with restoration being a small component. Perhaps I missed it, but my reason for that is I don't see anything entirely novel about your techniques (i.e., others have used UAVs, SfM, etc. to monitor other systems before), but I do find the paper novel if it can be spun as restoration monitoring is an issue and techniques exist to improve it - with the way you went about it being one potential option. I'll give broad comments on each section and then some specific corrections below.
Title - I think the title is OK, though it isn't very clear that the manuscript will be about restoration.
Abstract - Overall, I think this is pretty good. A couple comments: 1) I wouldn't have known this is related to river restoration other than a very brief mention in the last sentence. I think that could be more clearly stated earlier in the abstract. 2) I think it would help to specify what information you are able to get from drone that you cannot get with conventional methods in the abstract.
Introduction -- Overall, this section reads well. A concern of mine is that you go into extreme detail about restoration in this section, but the abstract/methods/results/discussion don't highlight restoration near as much. Based on my comments above, I think the authors must decide if this is a 'restoration monitoring' paper which utilizes UAV/SfM or is this a UAV/SfM paper which happens to take place in a restoration setting and balance the paper accordingly (not saying there is a right answer, though unless there is something entirely novel with the SfM algorithms, I think it'd be best to be a restoration paper utilizing new technology rather than the other way around).
Methods: Paragraph from lines 258-271 - VCP use is somewhat obscure and confusing, please elaborate on it so that other researchers would be able to reproduce what you did.
Was there any sort of scheme/design to how you set up GCP & VCPs? I don't see it here -- were they randomly placed within the study area? At set intervals? Based on previous literature? Again, please keep in mind that the methods should be written so that others could repeat this in the same study area or use your method in a new study area.
Paragraph Lines 279-287 -- How were CPs determined? How did you try to capture heterogeniety within the landscape? Was a random sample design implemented? Did you follow some sort of guidelines (later you mention the SBRMG)? Did you monitor vegetation height on all species or did you only pick certain species -- if so, how did you pick these species - random, technician choice, at set intervals, spatially balanced design, etc..?
Paragraph 287-293 -- what was the 'set number of steps'? How did you determine when to collect 'finer' measurements? How did transects provide useful for geomorphic measures/was that their intent (in the results it seems like not all did)? If I were to try to repeat this study, I'd be very lost by this point.
Next paragraph - you set vegetation plots where change was expected to occur following dam removal? How did you know where this change was expected? Were plots placed at every square inch of where change was expected or did you randomly/haphazardly/systematically place plots within a given area of study?
Section 2.4 -- To someone not familiar with AgiSoft or DSMs, I could see this section almost reading like a foreign language as a list with abstract/obscure concepts with little to no guidance as to how you did things (e.g., "building the dense point cloud, building the DSM, building the mesh," etc. etc.... how did you do these things? Did you have to do them in that order? A brief explanation of what a 'mesh', 'dense point cloud', etc. are would be helpful).
Results -- Overall, I think the results read pretty well, though I don't think they are matched up to the methods -- I would suggest revising the methods to make them clearer and so that an independent researcher would have no issue implementing them and then tailoring the results to them.
Discussion -- Again, I have a hard time understanding the results (and therefore the discussion) because I am still a bit lost in the methods (i.e., how did you actually set up your study to get results?). Otherwise, it reads pretty well... though, I think making the technology and restoration 'mesh' together a bit better would be helpful as it still seems like they are treated somewhat separately.
Line 664 you mention later in section 5.6 (I assume you mean 4.6?).
Other specific points:
Line 79 you have citations mixed up, it'd read better and save space at [10, 14-16]
Line 89 - seems like a weird place for citation 23 (why not at end of sentence?)
Line 120 - sentence seems incomplete ending in citation, this happens other times later in the paper with sentences ending in citations. It seems like it'd be useful to paraphrase the specific citation rather than expect the reader to scan to that citation in the reference list and read it in detail... most of the time this happens, it is at the end of the sentence, though Line 836 begins like this.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
paper is very intresting.
comments/suggestions:
- it is necessary to expand the purpose of the work.
- must be compared with analogs.
- it is necessary to highlight the scientific component and adjust the title of the work and the abstract for it.
- part of the study has gaps in the narrative - you need to make transitions and build an article according to generally accepted standards.
- Fig.6 / Table.3 - it is recommended to expand the concepts - the specified classes can intersect, as well as add images.
- when describing experiments and calculations, information on the representativeness of the sample and the validity of its sufficiency should be added
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Thank you for your revisions. I think you did a thorough job on them and greatly improved the manuscript.
I have no further revisions at this time. I would recommend future studies reduce the use of haphazard sampling, though I do not think in this instance it takes away from the merit of the paper. You did a nice job explaining the handiness of drones in unique situations (e.g., global pandemic) compared to relying on teams of individual people to meet.
