Optimizing Wheat Yield Prediction Integrating Data from Sentinel-1 and Sentinel-2 with CatBoost Algorithm
Round 1
Reviewer 1 Report
General comments:
The paper presented a study on predicting wheat yield with a combination of S1 and S2 data and investigating the performance of different machine learning approaches. Obviously, the concept and methods follow existing literature. Given the amount of studies on that topic now, the originality of the paper in a methodological point of view is very limited. In many places the authors trying to highlight things as novelty, but in reality are quite well-known to the community. There were several major weaknesses that should be addressed.
(1) Why do the authors only choose the machine learning models (CatBoost, SVM and RF) in the article? In those days, the deep learning model (e.g., CNN, LSTM) is a powerful tool for crop yield prediction. So, I think this manuscript is out of SOTA.
(2) Although VIs has been widely used as a proxy for monitoring crop biomass growth, it fails to accurately estimate yields due to an inconsistent correlation between growth and final yield (Mkhabela et al., 2011). Climatic variables, rather than VIs, can directly capture environmental stresses on crop growth (Cai et al., 2019). In addition, soil properties contribute to improving the accuracy of crop yield prediction because of their large spatial heterogeneity (Folberth et al., 2016). However, the authors appear to neglect these important variables for yield prediction.
Mkhabela, M. S., Bullock, P., Raj, S., Wang, S., Yang, Y, 2011. Crop yield forecasting on the Canadian Prairies using MODIS NDVI data. Agric. Forest Meteorol. 151, 385–393.
Cai, Y. P., Guan, K. Y., Lobell, D., Potgieter, A. B., Wang, S. W., Peng, J., Xu, T. F., Asseng, S., Zhang, Y. G., You, L. Z., Peng, B., 2019. Integrating satellite and climate data to predict wheat yield in Australia using machine learning approaches. Agric. Forest Meteorol. 274, 144–159.
Folberth, C., Skalsky, R., Moltchanova, E., Balkovic, J., Azevedo, L. B., Obersteiner, M., van der Velde, M., 2016. Uncertainty in soil data can outweigh climate impact signals in global crop yield simulations. Nat. Commun. 7, 11872.
(3) Why do the authors only choose to optimize one hyperparameter (the number of trees) of RF? Other hyperparameters such as the maximum depth of the tree and the number of features is also important for predictive performance.
(4) The authors do not clearly describe whether they mask out low-quality pixels of S2 data due to cloud contamination and reconstruct the time series by filtering methods such as harmonic regression. It is vitally important because further analysis will be affected by the low-quality pixels.
Specific comments:
L14: insert ‘.’ before ‘The’.
L18-19: please spell out abbreviations of ‘CatBoost’, as it first appears in the abstract.
L22: please added the unit of RMSE.
L80: typo error. ‘modess’ should be ‘modes’.
L96-102: This paragraph has mentioned deep learning algorithms, however, relevant literature review is lacking.
L130: insert ‘of’ before ‘yield’.
L153: ‘production’ -> ‘yield’. Please do not confuse these two words. Same as L525.
L181: delete ‘GS39’ after the parentheses.
L245: the abbreviations of these four algorithms has been spelt out before, please delete them.
L294: the sentence ‘…to handle datasets with many’ is incomplete.
L315: please spell out abbreviations of R2, RMSE and %MAE as it first appears in the main text.
L386: the abbreviations of ‘MLR’ has been spelt out before, please delete it.
L387: Figure 3 is hard to see clearly.
L486: ‘highest’ -> ‘lowest’
L559: ‘The’ -> ‘the’.
L580: ‘Vis’ -> ‘VIs’.
L680: insert ‘for’ before ‘PA’.
L697: the abbreviations of ‘PA’ has been spelt out before, please delete it.
Comments for author File: 
Comments.pdf
Author Response
We would like to express our sincere appreciation to the reviewers for taking the time to read and provide valuable feedback on our document. We highly value their comments and have carefully addressed each one in the revised manuscript. As a result of their insightful observations, the paper has been significantly improved. Once again, we thank the reviewers for their contributions to our work.
Author Response File: 
Author Response.docx
Reviewer 2 Report
The experimental work presented in the manuscript, entitled „ Optimizing Wheat Yield Prediction Integrating Data from Sentinel-1 and Sentinel-2 with CatBoost Algorithm". The article reports that, the performance of the CatBoost gradient boosting algorithm was compared to that of multiple linear regressions (MLR), support vector machine (SVM), and random forest (RF) algorithms in crop yield estimation. The results showed that the combination of S1 and S2 data with the CatBoost algorithm produced a yield prediction with a root mean squared error (RMSE) of 0.108 and R2 of 0.952. There are several shortcomings and modifications that should be included in order to enhance the manuscript for the readers.
Abstract
1- First sentence in abstract is not interested since many studies were used to estimate the wheat yield. Place replace this sentence by important one?.
2- Line 14. The error is found please add (.).
3- Line 14. This sentence (The study was conducted on 39 wheat plots, which had high-resolution yield) should be improved in writing.
4- Line 18. Please add the full name of VV and VH?
5- Line 23. Please add the full name of EI?
6- The important results of the best model should be presented in the abstract.
7- Last sentence was not written by English language.
8- Please add the conclusion sentence at the end of abstract?.
Introduction
9- Line 42. Please add the references at the end of the sentence?.
10- Line 44. by Zambon et al., in [7], please correct the reference according to journal style?.
11- The introduction is lack of the previous studies which related to the performance of spectral indices to asses wheat yield. Comparing studies should be added in introduction
12- The introduction is lack of the previous studies which related to the performance of different machine learning to asses wheat yield. Comparing studies should be added in introduction.
13- Please add the hint about the basic of remote sensing for Estimating grain yield?
Materials and Methods
14- The figure 1 should be improved.
15- Line. Please how to estimate the wheat grain yield in the field?
16- Please remove the sentence from line 187 to line 190. The table is enough.
17- Figure 2 should be improved.
18- 2.5. Machine learning algorithm. Please elaborate this section. You need to give more information about the model used (optimization technique used, kernel type, …) algorithm:
Results
19- Remove from line 367 to line 374 or provide a flow chart to illustrate the analysis performed in this experiment in M&M?.
20- Figure 3 is very low resolution and I can’t follow the results in the text. Please improved it?. Or you can present them in table, better than in figure.
21- Figure 4 should be improved.
22- Figure 5 should be improved.
23- Figure 6 should be improved.
24- Line 490. R2 value was 0.952. Please present the R2 value as two digital number like 0.95?.
Discussion
25- Discussion is well written.
26- Please, write the practical applications of your work in a separate section, before the conclusions and provide your good perspectives.
Conclusion
27- Please write about the limitations of this work in details in conclusion section.
Author Response
We would like to express our sincere appreciation to the reviewers for taking the time to read and provide valuable feedback on our document. We highly value their comments and have carefully addressed each one in the revised manuscript. As a result of their insightful observations, the paper has been significantly improved. Once again, we thank the reviewers for their contributions to our work.
Author Response File: Author Response.docx
Reviewer 3 Report
Please find attached.
Comments for author File: Comments.pdf
Author Response
We would like to express our sincere gratitude for the time and effort you have invested in reviewing my work. Your insightful comments and constructive feedback have been instrumental in improving the quality of my writing.
We are particularly grateful for the positive feedback you have given us, and we are delighted to hear that you found our work significant and original. Your encouragement means a great deal to us, and we feel happy to have received such positive comments from a respected reviewer. We are going to try to improve the manuscript following your comments.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
The authors clearly responded to my comments. The manuscript is remarkably improved. I have no further remarks.
Author Response
We would like to express our sincere gratitude to the reviewers for their invaluable contributions toward improving the quality of our work. Their constructive feedback and insightful suggestions have played a significant role in enhancing the overall coherence and clarity of our manuscript. We appreciate the time and effort that the reviewers have invested in reviewing our work, and we are truly grateful for their critical evaluation, which has undoubtedly helped us to refine our ideas and arguments. Additionally, we are humbled and honored by the favorable evaluation received, which has further motivated us to pursue excellence in our research endeavors. Once again, we extend our heartfelt appreciation to the reviewers for their exceptional service to the academic community.
Author Response File: Author Response.docx
Reviewer 2 Report
The authors did significant improvement. It can be accepted for publication.
Author Response
We would like to express our sincere gratitude to the reviewers for their invaluable contributions toward improving the quality of our work. Their constructive feedback and insightful suggestions have played a significant role in enhancing the overall coherence and clarity of our manuscript. We appreciate the time and effort that the reviewers have invested in reviewing our work, and we are truly grateful for their critical evaluation, which has undoubtedly helped us to refine our ideas and arguments. Additionally, we are humbled and honored by the favorable evaluation received, which has further motivated us to pursue excellence in our research endeavors. Once again, we extend our heartfelt appreciation to the reviewers for their exceptional service to the academic community.
Author Response File: Author Response.docx
Reviewer 3 Report
Please find attached.
Comments for author File: Comments.pdf
Author Response
Thank you for the favorable comment and for acknowledging the effort made by the authors in implementing the proposed improvements by the reviewers. The authors have taken into consideration the comments made by the reviewer and have accordingly addressed them in the subsequent sections.
Author Response File: Author Response.docx
