Evaluation and Normalization of Topographic Effects on Vegetation Indices
Round 1
Reviewer 1 Report
This work sought to assess the impact of topographic effects correction on vegetation indices (NDVI EVI and NIRv) based on Sentinel 2 MSI remotely sensed data. This work is topical and relevant in motoring mountainous vegetation ecosystems. However, there a few concerns that the authors need to consider prior to the paper being recommended for publication. These are as follows;
In the abstract section the following abbreviations should be written in full SE, SCS+C and PLC.
Line 43 delete “been”
Lines 45-51 the concept of empirical topographic effects correction methods are not fully developed and clearly explained by the authors. This section is very critical in motivating why these have to be considered. What exactly are they what are their advantages? As long as such questions are still lurking on the readers mind the motivation for conducting this study is weak and inconsequential.
Line 62 -66 more information on the characteristics of the DEM are required. For instance what is the accuracy of the DEM?
Lines 170-181 can either be integrated in the method section. And a portion could be handy in justifying the methodological procedure followed in this work.
For figure 4 consider including a legend to easily interpret the figure in addition to the caption
Line 233 on the implications of the study needs to be developed in two or more sentences.
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
remotesensing-857031
The manuscript “Evaluation and Normalization of Topographic Effects on Vegetation Indices” addresses an interesting and up-to-date subject, which adhere to Remote Sensing journal policies and topics.
In this research there was evaluated the topographic effects on vegetation indices in a mountainous region. The manuscript contains original and interesting results, well fitted in the context. In addition, the work is well conceived, realized and written, so that I did not identify deficiencies or shortcomings that would require revisions or improvements.
My only complaint is that a more important case study/study area could have been chosen. But it is understandable the desire of the authors to have the corrections as central stage of the article.
Nevertheless, I would like to congratulate the authors for the effort and scope of the article. It presents an interesting method/topic and has good readability. I also appreciate the genuine and concrete work carried out, good documentation and discussion chapter, and the relative short manuscript with direct approach.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
The authors in the paper describe the influence of topography on vegetation indices. They used two or three (?) SE, SCS+S and PLC method, already existing in the literature for the evaluation of topographic effects. As interesting as the paper is, it does not present great innovative aspects. The results, obtained, refer to only one test area and are therefore not reliable.
I would ask the authors to review the article by proposing two or three test areas and possibly propose it as a review.
Many of the sentences are unexplained and have been highlighted in yellow in the file
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 4 Report
By way of confirming my interest in seeing this paper published, I can say that I look forward to citing it in my work. The subject itself is vital to many that are now using highly-available remote sensing products online, as the authors took care to mention. I would suggest adding to that mention regarding the use of vegetation indices in Google Earth Engine and other online platforms, illustrating how users usually perform those analyses now and how this study will streamline the way users can improve their results. I have some more specific notes on the manuscript here:
Line 40: what is the definition of "rugged terrain" in this context? Is it land areas with a slope greater than 2%? 5%?
Caption for Fig. 1: indicate which parts of the spectrum are represented by Sentinel-2A bands 8A (narrow NIR), 4 (red), and 3 (green). Those of us more familiar with Landsat band numbering will be grateful for the ease of translation.
Line 80: remove "delicately."
Subsection 2.4: I suggest reformating this subsection as a paragraph, rather than a numbered list. There are only two items in the list, and most of the information is already there, but full sentences will be much easier to read and retain for later recall.
Subsection 2.4: The goals of the proposed statistical analyses should have explicit statements. For example, if the "best" results of the correlation analysis are slopes and R^2 values closest to zero, state that and explain why. If the goal of the aspect analysis is a low CV, state that and explain why. These statements can then be reiterated for emphasis in their respective sections of the results (around lines 125 and 160, respectively).
Table 3 is practically unreadable because of the unintended line breaks in each cell and the absence of dividing lines between the columns. I suggest (a) reducing the number of significant digits from 4 to 3 in all cells; (b) split each cell of linear regression results to two lines, the slope and intercept, instead of an equation; (c) a smaller font, if necessary. Instead of separate columns for LR and R^2, you could perhaps make each result in the table a 3-line cell: LR slope, LR intercept, LR R^2.
Figures 3 and 4: I suggest rotating and flipping each polar plot so that it resembles a compass rose: 0 at the top (north) and the azimuth increasing in the clockwise direction (90 at east/right, 270 at west/left). Then the dot representing the sun azimuth will appear in the southeast (lower right) as expected for a Sentinel scene in the northern hemisphere, and the traces indicating azimuthal VI values will demonstrate correspondingly higher values in that direction. Your plots are readable, but these changes will make them more intuitively representative of the scene location and timing.
Section 5: as in subsection 2.4, I suggest reformating this section like a paragraph, rather than a numbered list. Again, there are only two items in the list, and most of the information is already there, but full sentences will be much easier to read and can incorporate the surrounding text. The last sentence of the section, on line 233, seems to stand out alone between the list and the Author Contributions and could be missed by the reader. Additional information regarding the applicability of these results to VI processing using Google Earth Engine would be appropriate here.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 3 Report
The necessary corrections have been made. The paper is now published
