Image Spectral Resolution Enhancement for Mapping Native Plant Species in a Typical Area of the Three-River Headwaters Region, China
Round 1
Reviewer 1 Report
The manuscript entitled “Image Spectral Resolution Enhancement for Mapping Native Plant Species in a Typical Area of the Three-River Headwaters Region, China” face and important frontier of vegetation classification by enhancing image processing. Although, the structure of the manuscript and some methodological flaws undermine the soundness of the study. In general, there is a lack of sound scientific hypotheses. In some parts, the manuscript reads more like a technical report than a scientific paper. The methods are too much detailed in many parts but lacks of important information about statistical analysis.
Introduction is confused; some parts are mixing local issues with real scientific questions. These parte are more appropriated in study area description. The field sampling design and reason/definition of noxious species need to be better justified.
The authors are continue fighting with abbreviations; first mentions and typos need to be fixed. I would suggest to merge material and method in one section.
Conclusions are in some parts are not sustained by the study results.
The manuscript needs also to be tweaked in language in both scientific and English. The manuscript has to be carefully proofread.
Some other detailed comments:
Line 21: add SVM and RF extended names
Line 37: What authors mean with “study grass degradation”
Line 40: Indexes = more book index, Indices=more statistical index.
Line 41: I would prefer mention “for vegetation characterization” rather than “for vegetation identification”
Line 42: Names are not allowed in references, all the references have too many brackets
Lines 43: EVI ?
Line 69: Already mentioned, use the abbreviation
Line 69 to 80: This section is a mix of introduction and study area (material of methods). Please reshape it.
Line 92: “efficient data” is an odd collocation
Line 123: revise the title, “mainly native species” is meaningless
Line 128: “the area has sufficient sunlight radiation”, what does it means?
Line 138: “A previous study”
Line 138-139: Please define what noxious weed mean.
Line 175-192: many details (e.g. correction of special references) could be omitted as not influent.
Line 298 -300: another example of a by far too detailed description of methods. It is rather obvious that preliminary clipping and spatial adjustments of the layers is needed. I suggest to revise the methods section keeping in mind that this is a scientific paper, not a technical report.
Line 332-333: obvious
Line 428-432: it would be nice to test such result and not merely describe their differences
Line: 434-444: not mentioned in the statistical methods, what kind of correlation parameter (pearson r?)
Line 445-450: these are methods
Line 474: I guess “then” is better that “thus” in this sentence.
Table7: table caption needs to be more accurate in the description of the content
Author Response
Dear reviewer,
We are very grateful to your comments for the manuscript. All the comments are all valuable and helpful for improving our manuscript. They have been carefully studied and major modifications were made to the introduction, methods and results.
The response letter is replied point by point. Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This is interesting topic covered by authors, and with in the scope of journal. Authors missed several points and failed to explain them. They are talking about something and suddenly explaining other things..My major comments are:
- need to link paragraphs.
- check the contents which should be moved to methodology sections instead of results.
- field survey require extensive write up.
- check the tables and complete them
Abstract:
Line- 20- what is synthetic HSI, are you referring to fused or integrated image. ?
Statistical results are completely missing from the abstract.
Figure 2 (B) - where is sample plot no 7. Missing from the figure,if it is not there. this means it is jumping from 6 to 8. check and implement.
What is the reason that Figure 2 (a) and (c) is different for the quadrat. I assume that same quadrat was recorded by your team. Or just to display that team is recording the quadrat...??
Line- 171-173- field collection or sample collection.? make it clear.
Figure 2 (d) your total percentage coverage of the test site comes to be 92% (what are the rest features- 8%). you did not mentioned them, does it affect your accuracy ? Please provide them too.
How many samples were collected? you are showing 9 samples in 30 m. I assume you are working with only 1 pixel of MSI size.
How is your accuracy ? how you just rely on only 9 samples collected in 30 m regions for your study.?
Ground survey needs more contents. Authors need to provide clear texts for field survey.
Line 163-164- "Ground truth data were obtained through two field surveys. The first field survey started on 10 August and finished on 19 August 2017. The second field survey was from 12 to 23 August 2019."
Does this mean- why such a long gap of two years..? any differences in the field survey must be reported?
Line- 383-384- 310 plant leaves.... this is from your research work or any previous one.. need clarity. why 310 plant species here?? correct it.
line 385-"Some biochemical indexes,...... " need re-writing.
line 385-386 --- you were talking about NDVI,nGDVI, nNDVI and suddenly you came to conclusion for other indices as well.... like nLCI, ARI2 and nPRI ..... No discussion on these indices, but you are providing a support sentence to them... needs clarification,
Table 4 - provide additional column for the reference source of the VIs.
Section 3.2. SVM/ RF- check the citation- for your clear contents and sources. https://doi.org/10.1080/10106049.2019.1629647
Please find the references which should be included in the manuscript. 1. Antonarakis A, Richards KS, Brasington J. 2008. Object-based land cover classification using airborne LiDAR. Remote Sens Environ. 112(6):2988–2998
2. https://www.tandfonline.com/doi/full/10.1080/10106049.2019.1629647
3. Fauvel M, Benediktsson JA, Chanussot J, Sveinsson JR. 2008. Spectral and spatial classification of hyperspectral data using SVMs and morphological profiles. IEEE Trans Geosci Remote Sens. 46(11):3804–3814
4. Dixon B, Candade N. 2008. Multispectral landuse classification using neural networks and support vector machines: one or the other, or both? Int J Remote Sens. 29(4):1185–1206. DOI: 10.1080/01431160701294661.
Figure 5 - check the caption for Grammatical mistakes. (a) (b) .... represents... correct it.
There are many English Grammatical mistakes in the MS, please check with native speaker for any corrections.
Results:
line 445- 446: suddenly you are saying that SAM, ERGAS and UIQI is also used in the study..please provide the information in methodology section. Do not come to sudden conclusion or add any thing abruptly.
For SAM no source or citation is provided. Check.
Line- 452-453- The SAM value ? what is the techniques, must be explained in methodology section to enable readers to follow and replicate them. understand the basic behind it.
You have obtained the SAM values using all mentioned techniques, but failed to represent them... ?
Table 5- as per my understanding it is incomplete....why only up to count of 0-10, and 10-20 %...? Even you said that authors implemented SAM, and other techniques. Why not represented here,.. to provide a comparative assessment.
Author Response
Dear reviewer,
We are very grateful to your comments for the manuscript. All the comments are all valuable and helpful for improving our manuscript. and they have been carefully studied and major modifications were made to the introduction, methods and results.
The response letter is replied point by point. Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Dear Authors,
The manuscript "Image Spectral Resolution Enhancement for Mapping Native Plant Species in a Typical Area of the Three-River Headwaters Region, China" presented a novel method (Improved spectral resolution enhancement method - ISREM) to combine multi and hyperspectral images into a new product which increases spatial and spectral resolution.
The quality of the presented method and the science discussed in the manuscript looks fine to me and needs no further improvement.
However, I would encourage the authors to work on the quality of the manuscript, particularly on the way it shows the introduction, methods, results, discussion and conclusions.
The introduction show parts which, in my opinion, should be included in the methods section. There is an "over-explanation" of several terms (like, SVM on line 396). The explanation of so many "study area" characteristics may not be necessary since the main objective of the manuscript is to show the quality of the method, comparing it to other methods. I would also include a list of symbols to help the readers.
Regardless of the improvements described above, I believe the manuscript can be published.
Author Response
Dear reviewer,
We are very grateful to your comments for the manuscript. All the comments are all valuable and helpful for improving our manuscript. They have been carefully studied and major modifications were made to the introduction, methods and results.
The response letter is replied point by point. Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The author did a great job during paper revision process. Most of the flaws have been addressed. I have no other major comments on the current form of the manuscript.
Reviewer 2 Report
Authors have revised the manuscript as per suggestions. Now I can recommend for acceptance.
