Estimation of Total Phosphorus Concentration in Lakes in the Yangtze-Huaihe Region Based on Sentinel-3/OLCI Images

Round 1

Reviewer 1 Report

General comments:

In this manuscript, the authors present a machine learning algorithm for remote sensing of Total Phosphorus (TP) in inland lakes within the Yangtze-Huaihe region. This is based on in-situ measurements coupled with concurrent remote sensing data. Developing accurate algorithms for TP is challenging given that TP is not optically active, which underscores the significance of this study. However, I question the universality of the proposed algorithm across the various lakes in the region, as each may exhibit unique TP sources and characteristics. Moreover, the accuracy of the developed algorithm seems less than optimal. I would urge the authors to delve into these concerns.

 

Minor issues:

1.      In line 23, consider revising "The spatial distribution analysis revealed that..." for better fluency.

2.       In line 30, remove the unnecessary punctuation after "enclosed fish farming nets.".

3.       In the introduction, there are redundant and repetitive phrases that could be streamlined. For instance, line 75's "It has found wide application in the retrieval of water quality parameters." can either be removed or integrated more succinctly with surrounding sentences.

4.      Lines 80 to 87 should provide a more detailed discussion on the limitations of machine learning in predicting total phosphorus. The present content is too vague.

5.      There might be a typographical error in line 143, specifically "within the spectral range of 21-400 nm". This should be reviewed and corrected as necessary.

6.       Line 263's claim, "surpassing the threshold for algal bloom occurrence (0.02 mg/L)", requires a reference for verification.

General comments:

In this manuscript, the authors present a machine learning algorithm for remote sensing of Total Phosphorus (TP) in inland lakes within the Yangtze-Huaihe region. This is based on in-situ measurements coupled with concurrent remote sensing data. Developing accurate algorithms for TP is challenging given that TP is not optically active, which underscores the significance of this study. However, I question the universality of the proposed algorithm across the various lakes in the region, as each may exhibit unique TP sources and characteristics. Moreover, the accuracy of the developed algorithm seems less than optimal. I would urge the authors to delve into these concerns.

 

Minor issues:

1.      In line 23, consider revising "The spatial distribution analysis revealed that..." for better fluency.

2.       In line 30, remove the unnecessary punctuation after "enclosed fish farming nets.".

3.       In the introduction, there are redundant and repetitive phrases that could be streamlined. For instance, line 75's "It has found wide application in the retrieval of water quality parameters." can either be removed or integrated more succinctly with surrounding sentences.

4.      Lines 80 to 87 should provide a more detailed discussion on the limitations of machine learning in predicting total phosphorus. The present content is too vague.

5.      There might be a typographical error in line 143, specifically "within the spectral range of 21-400 nm". This should be reviewed and corrected as necessary.

6.       Line 263's claim, "surpassing the threshold for algal bloom occurrence (0.02 mg/L)", requires a reference for verification.

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Thank you for submitting a well written scientific study. I read your paper with great interest and found it very informative and a nice contribution for the remote sensing community. 

I have 2 comments: 

1) Page 5, Lines 192-197. Please broaden your discussion of the XGBoost algorithm package to include more background information (i.e., how it can be obtained, computing requirements, and how its been previously applied in terms of water quality studies.

2) Please make your discussion of the comparison of XGBoost-derived TP concentrations with measured values from other studies more detailed and quantitative, if possible.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

In this manuscript, the authors estimated the total phosphorus (TP) concentration in lakes in the Jianghuai region based on OLCI images using the XGBoost model. This study contributes to the understanding of TP concentration in a large-scale region, but there are areas that need further improvement and clarification.

The authors should provide a clearer explanation of the contributions and innovations of their study.  In line 324, it is suggested to change "SPM concentrations" to "TSM concentrations" for accuracy. Please provide a detailed explanation of how the authors determined the most appropriate input variables for their model. Please explain the process of hyperparameter tuning for the XGBoost model. Which hyperparameters were adjusted, and how were the optimal values determined? Was cross-validation used during the model training phase? Could the authors please provide a detailed explanation of the methodology used for model validation? Did the selection of the samples have an impact on the results? Did the authors tune the hyperparameters of the XGBoost model? If yes, please explain the tuning process and methodology, and state the final chosen hyperparameter values. It is suggested to consider merging paragraphs 4 and 5 of the introduction for better cohesion. In lines 88-91, the repetition of the preceding information should be removed or integrated into a single sentence. In line 375, when mentioning "Xiong et al. analyzed the TP concentration in Taihu Lake...", please provide the appropriate citation for the reference. Overall, addressing these points would enhance the clarity and robustness of the manuscript.

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The manuscript can be accepted. 

Reviewer 3 Report

The authors have fully responded to my comments and there are no more comments available.