Remote Sensing of Sea Surface Artificial Floating Plastic Targets with Sentinel-2 and Unmanned Aerial Systems (Plastic Litter Project 2019)

Round 1

Reviewer 1 Report

<!-- *The comments were formatted in Markdown. -->
### remotesensing-808013
The manuscript *Remote sensing of sea surface artificial floating 2 plastics with Sentinel-2 and unmanned aerial systems 3 (Plastic Litter Project 2019)* by Konstantino Topouzelis and colleagues describes ongoing efforts to advance in the development of remote sensing techniques for the detection and quantification of the presence of large patches of floating plastic debris in the sea using hyperspectral sensors. The authors detail a series of efforts to deploy large plastic targets (5 x 5 m) to test the effectiveness of an inversion algorithm. Initial results seem promising even with the moderate sample size analyzed and the multiple challenges faced in an study of these characteristics. 
The overall thematic and the approach pursued in the manuscript are of general interest and match public concerns raised about marine pollution in recent years. From a scientific and academic point of view, the study describes research at very early stages, and in some passages the manuscript fails to deliver a clear message about the actual accomplishments and future prospects. I also think the authors should frame and motivate their work in a more general context; 
1. Questions like the fate of plastics in the sea, their degradation to microplastics (and the availability of methods to quantify them), the rates at which they are accumulating would help to attract readers.  2. The potential applications of the proposed methods, and the feasibility and ultimate usefullness of monitoring and removal approaches are other aspects that readers would expect to see in a manuscript like this one.  3. Specifically, the authors should ellaborate more their case about the advantages of developing targets *vs* perfoeming or using available surveys of plastics already drifting in teh sea (*e.g* I can easily argue that as plastics age their spectral signature will be alteared, etc.).
Finally, the manuscript requires some restructuring to streamline the flow of ideas and avoid providing unnecessary details. For instance, it seems that the pilot study was usefull to characteriza the spectral signature of PET. Besides that, perhaps any other reference to that study can be omitted to avoid distracting readers. On the other hand, some passages might benefit from a less detailed narrative. The description of the field work in particular seems to provide details that would fit better in a project report rather than in a scientific article. 
In summary, and to avoid that my main message gets lost in the criticism about the current structure of the manuscript, I think this study reports important advances towards the detection of plastic debris from space and a series of valuable experiments conducted in an allways challenging environment. At the same time, I think the authors should made an extra effort to present their approach and results in a more streamlined way. For this reason, I recommend a major revision.
*Minor comments*
- L30 - please intorduce PET or provide some context for why you targeted plastics made of PET  - L60-70 - perhaps it is not necessary to provide so much detial about the pilot study, it would be better to refer to published results or just provide any important detail to understan the current work when corresponds - L107 - this level of detail is probably more appropriate for a project or cruise report, rather than for a manuscript - L 125 - be more precise, Posidonia is actually optically active ... but does not reflect too much radiation ... - L188 - this information is key, perhaps should appear earlier - L193 - UAS or UAV? - L269 - are fp and fw areal fractions insttead of abudances? - L304 - this paragraph can be easily reabsorbed in methods - L313 - the tunning of these parameters and their interpreatation should be detailed in methods - L340 - this paragraph seems to mix methods and discussion ... not many results, please consider summarizing these materials and/or moving them to other sections - L399 - again a very long paragraph that seems to be better suited for methods and the discussion ... - L507 - perhaps the issues can be presented in a more strighforward way [i.e. Ssunglint removal algorithms may introduce artifacts that compromise the detection of plastics, so it is necessary to improve .. ] - L536 - this seems an important aspect, epsecially in terms of the developmetn of practical applications ...

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

This is a very well thought-out, planned and executed study - a very enjoyable read. It is a valuable contribution to a rapidly growing, relevant and of high public interest area. The authors display the limitations of their approach which is very important. I have a few comments, mostly regarding the figures and data visualization, and a couple regarding clarity of statements.

Line 48 (& a few other locations): suggest adding the authors names before citation when you do “by [13]” e.g. “by Maximenko et al [13]”. It makes for easier reading.

 

Fig 1 and accompanying text: What was the 4th quadrant? A mixed section? Could you please specify in the text.

 

Jump from Fig1 to Fig 5 - please number and order the figures as they appear in the text. I see why you’ve done it in this way, but I’d suggest not referring to Fig 5 in Line 132.

 

2.1.1 Configurations and plastic coverage: perhaps present this information in a table? And perhaps consider naming each configuration to make it easier to refer to throughout the rest of the text? 

 

No reference to Figure 2

 

Fig 3: Suggest adding (a), (b), (c) and (d) to the four panels. The information in the table is labelled as a percentage, but the data are given as a fraction of 1. Can the authors please present the data as a percentage i.e. of 100?

 

L279/280: I’m slightly confused as to why you used pixel percentage coverages from PLP2018? Why aren’t you using the numbers calculated in the previous section (2.4.1) and that are shown for one example in Fig 3?

 

L327: What’s the spectral angle?

 

Figure 4: No reference to figure in text. Can you make the caption more descriptive? What are the different targets? (This comes back to the suggestion of naming the different targets/configurations). No units on figure.

 

L402-404: Are you talking about the four pixels identified in Fig 3 or from the 2018 image? If the 2018 image, why not use any of the 2019 images? Can you please clarify in the text?

 

L426-428: Can you please expand on this statement? What do you mean by “coverage amplitude”? This sentence and the next few (ending in Line 434) seems weirdly placed… they’re mixed in between the discussion of Figure 5, and these data aren’t shown in Figure 5.

 

Figure 5: I like this figure, but I wonder about presenting the percentages in the last column in a table alongside the actual percentages of material (which is known), to give a clearer idea of how the matched filtering is working?

Author Response

Please see the attachment

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

<!-- *The comments were formatted in Markdown. -->
### remotesensing-808013
I read the two cover letters prepared by the authors in response to comments about the previous submission and also checked  the revised version of the manuscript *Remote sensing of sea surface artificial floating 2 plastics with Sentinel-2 and unmanned aerial systems 3 (Plastic Litter Project 2019)* by Konstantino Topouzelis and colleagues. The authors solved all my concerns and criticism and all the changes introduced in the manuscript seem to clarify the text. I just found a single minor issue listed below.
This is an important contribution; accurately monitoring marine plastic debris is an urgent need. I recommend the manuscript to be accepted.
*Minor comments*

- L29 - the comment about teh sue of the acronym PET in the abstract referred to the need to provide some context; e.g. introduce it in L23 [detail that your targets were made of PET]. The current syntax forces the casual reader to guess why you start talking about PET.