Transferability of Models for Predicting Rice Grain Yield from Unmanned Aerial Vehicle (UAV) Multispectral Imagery across Years, Cultivars and Sensors

Round 1

Reviewer 1 Report

The manuscript is well written and enough well structured anyway some parts have to be added before going ahead in the pubblication phase.

Firstly I suggest you to move table 1 in a section of material and methods comparing the sensors becuause in introduction section can mislead the reader. Then I sugget you to discuss in the introduction a little bit more about the role of UAV in agriculture application compared to satellite. In particular I advice you to include these work in the introduction section:

- https://doi.org/10.3390/ani12081049

- https://doi.org/10.3390/su10010051

 

Moreover, I strongly suggest to include flight parameters (sensor view angle, speed, fligth quote etc) these information are crucial to scale the research and are absolutely necessary. Then I advice you to include a section concerning the sensor calibration and the results validation.

Finally consider to describe which software you adopted to perform the whole workflow.

 

 

 

Author Response

Please check the file that I repond to all the reviewers' comments.

Author Response File: Author Response.docx

Reviewer 2 Report

drones-2067126-peer-review-v1

Transferability of models for predicting rice grain yield from UAV multispectral imagery across years, cultivars and sensors.

 

Comments and Suggestions for Authors

The present paper is based on a novel algorithm and its applications. The article can be considered only if it is revised.

1. The abstract must be written completely again. I found nothing attractive in the abstract. The abbreviations are not defined. 

2. Do not use abbreviations in the title.

3. The authors discussed several times what they are going to propose but did not discuss the research gap or what are the sole reasons for this research. 

4. The literature review is not sufficient enough. Please provide more details about the literature review by considering the latest articles.

5. A comparative study if included would have an impact.

6. The authors wrote the conclusion in a rush. Please describe your achievements in detail. 

7. The quality of the figures is too low. I did not understand the results in Figure 5. Please explain and them and redraw it.

8. Add a transition paragraph and describe every section as the last paragraph of introduction.

9. Develop the conclusions section to include the unique contributions of the paper, theoretical and managerial implications, limitations of the research and future research directions. This section must be well detailed.

10. There are some English problems throughout the paper. So, it should be amended accordingly and rechecked by a qualified English proof-reader.

11. Do not use so many abbreviations in the title, abstract and elsewhere. In the absence of stringent space constraints, the use of abbreviation is not a good idea because it decreases ease of reading if a person has to remember all the abbreviations. The paper is at places practically unreadable due to the excessive use of abbreviations.

12. And an abbreviation is used only if the term appears at least five times in the main text. (The Abstract, conclusion, figures, and tables don’t count.) If the term or phrase is used only two, three or four times it should not be abbreviated (The Chicago manual of style).

***

Author Response

Please check the file that I repond to all the reviewers' comments.

Author Response File: Author Response.docx

Reviewer 3 Report

The presented manuscript studies the potential of spectral and textural information extracted from multispectral images to predict rice crop yield. More importantly, the feasibility of model transferability between different years, cultivars, and sensors in predicting rice grain yield has been investigated.

It seems to be solid research, with a good organization of the essential materials, and clear presentation of the results. Also, a strong literature review as well as appropriate discussion of the results are conducted. The limitation of the research as well as perspective for future works is also provided in the manuscript which can be helpful to preparation of future studies.

 

The present work is good and can be published after minor revision.

However, an area which the authors can further improve the quality of the work is to adopt novel and state-of-the-art transfer learning algorithms to calibrate models between different variables. For example, the knowledge acquired from learning to predict grain yield in one year can be used to predict the yield in another year with more accuracy. To this end, some modifications need to be made on the existing algorithms and some variables which are effective on grain yield can also be taken into account.

 

The minor issues that need to be taken care of before publication are as follows:

In the Results section (section 3.1), the authors are presenting results regarding influence of N fertilization on grain yield, while no previous information has been provided about this subject. Also, the relationship between fertilization and crop yield was not the objective of the study, nor was it covered in the Material and Methods’ section.

Also, some selected Figures related to the multispectral images of the rice field should also be added to the manuscript.

Author Response

Please check the file that I repond to all the reviewers' comments.

Author Response File: Author Response.docx

Reviewer 4 Report

General comments:

1.      The uncertainty of the predicted models must be quantified in order to confirm these results. However, I suggest adding a Models uncertainty analysis. Pls, See the section Models uncertainty analysis: https://doi.org/10.3390/rs14164080

 

2.      Pls, Add the Variable (predictor) importance measures. Pls, See the section relative attribute/predictor importance: https://doi.org/10.1007/s11356-022-21890-8

Author Response

Please check the file that I repond to all the reviewers' comments.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors well aimed the suggestion given. Therefore the manuscript can be published on Drones.

I adivice only to consider to add in supplementary material the filght parameters of Drones

Reviewer 2 Report

No comment

Reviewer 4 Report

Based on the author's previous revisions, the manuscript is recommended to the public.