Modeling Phenols, Anthocyanins and Color Intensity of Wine Using Pre-Harvest Sentinel-2 Images
Round 1
Reviewer 1 Report
I enjoyed reading this paper. The English was slightly off but I think it’s good enough. I had a couple of minor concerns, listed below. I think the only real concern is my L483 comment, I don’t think it is backed up by the facts presented.
L214 need citation for NDVI
L215 need citation for GNDVI
L217 need citation for chlorophyll if [44] is more of a general citation (?)
L236 here you say 8 dates were used, fig 5 shows 5 dates, L319 you say 4 dates, which is it?
Table 3, it would be better for the reader if the caption listed wavelengths instead of band number
L483 but I thought correlations in 2018 were best ~1 month earlier than in 2017, did the harvest occur 1 month earlier in 2018?
Comments for author File: 
Comments.pdf
Author Response
Please,
See the attachement.
Author Response File: 
Author Response.docx
Reviewer 2 Report
It's an interesting topic on estimating wine quality variables using the pre-harvest satellite images. However, I am afraid that the modeling method and the quality of the figures and tables should be improved. Please refer to the following specific comments.
Line 179, please show the producer and version of this software.
Lines 212–213, please add the references for the spectral indices.
I suggest moving Figure 1 to an appendix. I suggest using DOY (days of the year) or days of the growing season to replace the dates in the figures and tables.
Some of the values in Figure 6 are inconsistent with those in Table 1 and please check that. Actually, I think the information in Figure 6 and Table 1 is partly repeated with that in Table 2. I suggest moving some of them into the appendix and showing Table 2 as a graph.
Lines 273-289, please clarify if the four winemakers use any quantitative indices to evaluate the four types of wine factors. Please also explain why the values of the total anthocyanins of 2018 are more than twice that of 2017.
For optimal model selection, I suggest using statistical variables to show the degree of collinearity and then selecting the variables. Besides, the physical meanings of the spectral bands and vegetation indices that were selected should also be discussed. What's more, overfitting of the models should also be avoided. I also suggest using different samples for modeling and evaluation, respectively. In terms of the comparison of different models with different variables, I think only R2 is not adequate.
Author Response
Please,
See the attachement.
Author Response File: Author Response.docx
Reviewer 3 Report
Τhis study modeled key wine quality variables, such as phenols (TP), anthocyanins (TA) and color intensity (CI) based on Sentinel-2 image bands and spectral indices to provide different sensory characteristics of wine. This is an interesting study that demostrated how remote sensing based tools can be used to define a differentiated vintage and estimate the accumulation of phenolic compounds and the intensity of wine color, key elements in the final quality of the wine. Consequently, I recommend the publication of this study once the few issues noted in the attached manuscipt will be taken care of.
Comments for author File: Comments.pdf
Author Response
Please,
See the attachement.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
I think the authors have revised or explained according to my comments, and this manuscript could be of interest to a wide readership. I am happy to recommend acceptance of this manuscript.
