Prediction of Areal Soybean Lodging Using a Main Stem Elongation Model and a Soil-Adjusted Vegetation Index That Accounts for the Ratio of Vegetation Cover

Round 1

Reviewer 1 Report

Dear authors,

As per my understanding of the subject and written content presented in the paper. I am agreeing that I should be published in its current form. 

Good luck

Author Response

Thank you very much for your kind comment.

Reviewer 2 Report

Please see the attachment.

Comments for author File: Comments.pdf

Author Response

Dear authors,

Prediction of areal soybean lodging is an important topic in agriculture remote sensing. The authors proposed to use a main stem elongation model and a new soil-adjusted vegetation index to predict soybean lodging. The manuscript is interesting and has the potential to be published on condition that the following problems are solved.

Thank you for your valuable comments. We revised the manuscript based on the comments, surely enhancing the information to the readers.

• Page 1 Line 33 Please reorder the keywords in alphabetical order by the first letter

We revised them in alphabetical order (P1L33).

• Page 2 Line 67 Add the full name of "SPAD". Abbreviations should be given their full name when they first appear. Check out this problem throughout the manuscript.

The full name and explanation of SPAD was added (P2L68).

• Page 2 Line 94 Add a reference to Nakano et al.

The reference was added (P3L100).

• Page 3 Line 128 Missing first line indent.

The first line indent was added (P4L134).

• Page 3 Line 131 It is recommended to change all variables throughout the manuscript to italics, such as "n". Besides, it is recommended to add an explanation of each variable the first time they appear. Check this problem throughout the manuscript.

All variables were presented in italics and their explanations were added (ex. P4L163).

• Page 4 Line 140 The same title “Main stem elongation model from emergence date to blooming stage (R1)” for section 2.1.1. and 2.1.2.

Revised the title of 2.1.2 to "Main stem elongation model from blooming stage (R1) to peak main stem length" (P4L147).

• Page 4 Line 138 Table 1 The font of ℃. Check this problem throughout the manuscript.

The font has been unified to "Times New Roman" (Table 1 and 2).

• Page 5 Line 191 modify "2.1.3" as "Section 2.1.3".

We corrected it (P6L199).

• Page 7 Line 237 Square brackets are used first, followed by brackets. Check this problem throughout the manuscript.

We corrected them throughout the manuscript (P7L245).

• Page 7 Line 239 It seems that the value range of Lvc is from 0 to 1, corresponding to the value range expressed by the reference. However, recent studies show that the optimal L value can be negative. How to consider the negative value?

We consider that the negative value for L is acceptable in some cases, especially in sparse vegetation. We add a following discussions:

Regarding L, Ren et al. reported that the highest accuracy was obtained to estimate aboveground living biomass in arid grasslands at a negative L [49]. They suggested that the negative value reduced the effect of soil noise ascribable to the width of a global soil line in sparse vegetation conditions. However, the condition in this study was soybean with uniform vegetation. Accordingly, we applied Lvc, ranging from 0 to 1 (P21L578).

• Page 9 Line 324 "1.5 m2" Please write 2 as the top corner.

It was corrected (P9L332).

• Page 12 Line 365 The x-axis label is "DAS" rather than "DAE".

It was corrected (Figure 4).

• Page 15 Line 427 Figure 9 Better to write the x-axis label as "Vegetation indices" because the current label may mislead the reader to the multiplication between NDVI and SAVIvc. Besides, better to add the figure of SAVI and MSAVI into the figure 9.

Figure was corrected as your comment (Figure 9).

• Page 18 Line 483 Please write "R2" in the correct way.

It was corrected (P20L492).

• Page 19 Line 552 MSAVI uses an induction function to potentially consider the relationship between soil adjustment factor L with foliage cover. Besides, this induction function is also used for transformed soil-adjusted vegetation index to modify the constant soil adjustment factor to a function. Although the authors have shown in the results section that the use of foliage cover is significantly better than the use of the inductive function, it is recommended that the differences between the two methods (foliage cover and inductive function) be fully discussed in the discussion section.

We added following sentences to the discussion.

MSAVI is a modified version of SAVI: L of MSAVI is optimized by iteratively calculation [20]. In the optimization, L was started from 1 -SAVI and the present equation was obtained after repeated calculation. Consequently, L in MSAVI is not obtained from actual measurements of vegetation coverage. On the other hand, SAVI_vc in this study requires s the quantification of L_vc by binarizing the NIR images each observation.  Since the binarization clearly separates vegetation and soil, SAVI_vc is a better index to estimate the growth rate of soybean (P21L563).

• Page 19 Line 556 Please modify "NDAVI" as "NDVI"

It was corrected (P21L589).

• Page 20 Line 622 Page information missing: 2020, 23(3): 247-259 (https://www.tandfonline.com/action/showCitFormats?doi=10.1080%2F1343943X.2019.1702468)

It was corrected (P23L674).

Thanks!

Reviewer from Remote Sensing 

Reviewer 3 Report

Dear authors,

I am pleased for the opportunity to review your paper. The methodological section, results, and discussions are nicely presented and well explained. It shows the authors have a good understanding about the aerial imagery and the challenges that come with imagery analysis.

Overall, the manuscript is really well written. No major flaws are observed. An extra sentence in the conclusion speaking in more details about future applications or approaches would be beneficial.

I would suggest the following very minor corrections, before the MS could be published:

·         Ln 68-70: consider replacing one of the two “Furthermore” to avoid repetition in consecutive sentences.

·         Ln 72: “[15;16]” or “[15&16]”

·         Ln 106: same as above

·         Ln 181: nth to n^th

·         Ln 183: n+1th to n+1^st

·         Ln 238: instead of “in SAVI, in formula (7)”, you can write “in SAVI (formula 7)”

·         Table 4: upper and lower limits are reversed.

·         Conclusion: you mention that the method “is expected to be used in the future”. I t would be nice to add another sentence elaborating more in future applications, challenges, or opportunities for improvement.

·         Article: citations are not presented in a same way throughout the text. Some of them are presented with a corresponding number (e.g. [1]) and some of them using a parenthesis with the publishing year (e.g. “Yamamoto et al. (2019)” instead of “Yamamoto et al. [x]”). I would suggest sticking with one way throughout the text.

Comments for author File: Comments.docx

Very minor comments. Please find them attached.

Author Response

Comment to the Author

I am pleased for the opportunity to review your paper. The methodological section, results, and discussions are nicely presented and well explained. It shows the authors have a good understanding about the aerial imagery and the challenges that come with imagery analysis.

Overall, the manuscript is really well written. No major flaws are observed. An extra sentence in the conclusion speaking in more details about future applications or approaches would be beneficial.

I would suggest the following very minor corrections, before the MS could be published:

Thank you for your valuable comments. We have made the following corrections to the points you raised. In particular, we have described future applications and issues in the conclusion section.

• Ln 68-70: consider replacing one of the two “Furthermore” to avoid repetition in consecutive sentences.

Deleted "Therefore" in Ln 68 and modified it to "This suggests that it is ~" (Ln72).

• Ln 72: “[15;16]” or “[15&16]”

It was corrected to [16-17] (Ln77).

• Ln 106: same as above

It was corrected (L112).

• Ln 181: nth to n^th

It was corrected (Ln189).

• Ln 183: n+1th to n+1^st

It was corrected (Ln191).

• Ln 238: instead of “in SAVI, in formula (7)”, you can write “in SAVI (formula 7)”

It was commented (Ln246).

• Table 4: upper and lower limits are reversed.

It was corrected (Table 4).

• Conclusion: you mention that the method “is expected to be used in the future”. I t would be nice to add another sentence elaborating more in future applications, challenges, or opportunities for improvement.

The following issues has been added.

• Prediction of lodging would be possible.
• The prediction enables farmers to take counter measure of pinching against the lodging.
• The approach combining UAV remote sensing and machine learning is also worth considering.
• Article: citations are not presented in a same way throughout the text. Some of them are presented with a corresponding number (e.g. [1]) and some of them using a parenthesis with the publishing year (e.g. “Yamamoto et al. (2019)” instead of “Yamamoto et al. [x]”). I would suggest sticking with one way throughout the text.

The citations were checked and corrected throughout the text (e.x. Ln74).

Round 2

Reviewer 2 Report

Thank you for your consideration of the comments. The quality of the article has been greatly improved and I recommend its immediate publication.