Remote Sensing Monitoring of Winter Wheat Stripe Rust Based on mRMR-XGBoost Algorithm

Round 1

Reviewer 1 Report

Good hypothesis with wonderful results. However, conclusion section can be improved how theses models can be integrated to nation wide disease alerts.

Author Response

Thanks for your suggestion on the manuscript " Remote sensing monitoring of winter wheat stripe rust based on mRMR-XGBoost algorithm". Your comment has brought important help to the revision of this article. The text of the manuscript has been revised according to it. Please see the attachment for detailed modification instructions. 

Author Response File: Author Response.docx

Reviewer 2 Report

This paper discusses an important and contemporary topic on remote sensing monitoring of winter wheat stripe rust. The paper is well written however, I believe the following issues should be addressed before being published.

1. Lines 239- 246 and 256-260 are methodological issues and should be discussed in the method section instead of the results section.
2. Is there any justification on why the Correlation coefficient value of SDy, R539, and R713 is lower than others shown in figure 3?
3. “The canopy structure index TVI and solar-induced chlorophyll fluorescence value SIF-A have obvious responses to stripe rust” need to be justified.
4. What do you mean by SPA in line 13?
5. Figure 7 shows that GI has the highest importance value than other features, so why not use GI instead of SIF in the disease index (DI) monitoring model combining mRMR and XGBoost algorithm? This needs to be justified.
6. The acknowledgement part needs to be rewritten.

Author Response

Thanks for your suggestions on the manuscript " Remote sensing monitoring of winter wheat stripe rust based on mRMR-XGBoost algorithm". Each of your comments has brought important help to the revision of this article. The text of the manuscript has been revised according to the comments. We have supplemented the description of the methods and reorganized the results in our revisions. A field survey data experiment is added to verify the universality and scalability of the mRMR-XGBoost algorithm. Please see the attachment for detailed modification instructions.

Author Response File: Author Response.docx

Reviewer 3 Report

This study proposed a mRMR-XGBoost algorithm for the information redundancy of high-dimensional and small sample data in wheat stripe rust detection, which performed well in terms of data dimensionality reduction and prediction accuracy improvement. Here are some suggestions,

(1) In section 3.2, it is not clear how to select 6 features based on the mRMR algorithm, please add relevant descriptions.

(2) The manuscript discussed the application of the mRMR-XGBoost algorithm to monitor wheat stripe rust on the canopy scale, but did not discuss its applicability in the field. Can you add the content of this part?

(3) The title of the last column of Table 3 is "Title 3", please add the title name.

(4) Do SIF-A and 3FLD-A appearing on Page 7 represent the same index? If not, please explain their respective meanings in the manuscript; if it is, please express it uniformly.

 

Author Response

Thanks for your suggestions on the manuscript " Remote sensing monitoring of winter wheat stripe rust based on mRMR-XGBoost algorithm". Each of your comments has brought important help to the revision of this article. The text of the manuscript has been revised according to the comments. Please see the attachment for detailed modification instructions. 

Author Response File: Author Response.docx

Reviewer 4 Report

Overall the manuscript, authors used the mRMR-XGBoost method to extract spectral features for the remote sensing data of winter wheat samples. The method is fine for the VIS-NIR determination. However, several details should be clarified.

1. the remote sensing data should be a hyperspectral data with spacial information. Authors should interpret the detail progress of extracting the spectra from the special-spectral data, especially taking consideration of using a suitable ROI.
2. the data stability should be validated in the data collection stage.
3. authors should demonstrate physical meaning of the extracted spectral features. Over focus on the chemometric algorithms is not recommended.
4. the Conclusion part should be extended with the key experimental results
5. English expressions should be improved throughout the paper.

Author Response

Thanks for your suggestions on the manuscript " Remote sensing monitoring of winter wheat stripe rust based on mRMR-XGBoost algorithm". Each of your comments has brought important help to the revision of this article. The text of the manuscript has been revised according to the comments. The contents in the introduction, experiments, methods, results, discussion, and conclusions of the manuscript have been revised, and field survey experiments have been added to verify the monitoring universality and scalability of mRMR-XGBoost. In addition, we have carefully revised the English expression of the manuscript. Please see the attachment for detailed modification instructions.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

The paper has now significantly improved and can be published.

Reviewer 4 Report

No comment

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.