Next Article in Journal
Development of Sustainable Radiation-Shielding Blend Using Natural Rubber/NBR, and Bismuth Filler
Previous Article in Journal
Assessing the Effects of Natural Resource Extraction on Carbon Emissions and Energy Consumption in Sub-Saharan Africa: A STIRPAT Model Approach
 
 
Article
Peer-Review Record

TP Concentration Inversion and Pollution Sources in Nanyi Lake Based on Landsat 8 Data and InVEST Model

Sustainability 2023, 15(12), 9678; https://doi.org/10.3390/su15129678
by Lei Ding 1,*, Cuicui Qi 2, Geng Li 1 and Weiqing Zhang 1,3
Reviewer 1: Anonymous
Reviewer 3:
Sustainability 2023, 15(12), 9678; https://doi.org/10.3390/su15129678
Submission received: 21 May 2023 / Revised: 13 June 2023 / Accepted: 14 June 2023 / Published: 16 June 2023

Round 1

Reviewer 1 Report

The paper titled "Discussion on TP concentration inversion and pollution sources

in Nanyi Lake based on Landsat 8 data and InVEST model" presents a comprehensive and well-executed study on the estimation of total phosphorus (TP) concentration in freshwater ecosystems. The authors propose two main frameworks, an empirical method and a machine learning algorithm, to develop the Nanyi Lake algorithm for TP estimation. I believe the authors have done an excellent job in addressing the significance of remote sensing technology and its application in estimating TP concentrations, as well as identifying the contribution of TP inflow load in the surrounding areas of the lake body.

 

The comparative analysis between the empirical method and machine learning algorithm demonstrates the superior performance of the latter in TP inversion prediction for Nanyi Lake. I find the results particularly compelling, highlighting the BST algorithm as the most accurate and capable of capturing variations in TP concentration.

 

Furthermore, the integration of the InVEST model to simulate the TP inflow load and identify key pollution source areas in the Nanyi Lake Basin adds another layer of depth to the study. The identification of the southern and northeastern parts of the basin as significant contributors to the pollution load of the lake area provides vital guidance for effective prevention and control measures. The authors' efforts in correlating simulation results with water quality monitoring and management reflect their commitment to improving the overall water quality of Nanyi Lake.

 

Overall, I think with the authors' have done a meticulous job. The paper is well-structured and effectively communicates the significance of remote sensing and machine learning in TP estimation. The findings contribute meaningfully to the field of water quality monitoring and provide valuable insights for the management and preservation of Nanyi Lake. 

Author Response

Thank you very much for your positive comments and we will continue to work hard in the future.

Reviewer 2 Report

Overall, a well-written paper. The authors have maintained quality and logic throughout each section. There are a few minor edits I suggest:

Line 11: Please expand TP - Total Phosphorus (TP) 

Line 17: Please expand inVest

Line 20-21: Please expand ML models - SVM, BP, etc

Line 65:  Remote sensing technology is a new technology for lake water quality monitoring…” is not true. It is not a “new” technology. Consider modifying the sentence.

 

Line 67: “It not only makes up for the shortcomings and shortcomings of traditional..”  please modify this sentence..

 

Line 104:  Considering adding the full form of inVEST

 

Line 112: “Landsat8 OLI” -> “Landsat 8 Operational Land Imager (OLI)”

 

Line 157: “Landsat8” -> “Landsat 8” - Please correct this across the paper. There should be a space between Landsat and 8 or you can even write it as “L8”.

 

Line 157: “TRS” -> “Thermal Infrared Sensor (TIRS)”

 

Lines 228-230: It is mentioned that from 78 sampling points, 57 were used for training and 21 for validation. How did you split 57/21 for training and testing? Were these random splits? 

 

Lines 230-232: “Coefficient of determination (R2), RMSE, and Bias was..are these the all used to compare the 4 ML models? 

Line 174: Expand NDR

Author Response

Dear reviewer:
Thank you for your decision and constructive comments on my manuscript. I have carefully considered the suggestion of the Reviewer and made some changes. I have tried my best to improve and made some changes to the manuscript(Modifications have been highlighted in red in the revised version). Revision notes, point-to-point, are given as follows:

Line 11: Please expand TP - Total Phosphorus (TP)

I have improved it in the article(line 11).

Line 17: Please expand inVest

I have improved it in the article(line 18).

Line 20-21: Please expand ML models - SVM, BP, etc

I have improved it in the article(line 22).

Line 65: “Remote sensing technology is a new technology for lake water quality monitoring…” is not true. It is not a “new” technology. Consider modifying the sentence.

I have corrected it in the article(line 65).

Line 67: “It not only makes up for the shortcomings and shortcomings of traditional..”  please modify this sentence.

I have corrected it in the article(line 67).

Line 104: Considering adding the full form of inVEST

I have corrected it in the article(line 99).

Line 112: “Landsat8 OLI” -> “Landsat 8 Operational Land Imager (OLI)”

I have improved it in the article(line 107).

Line 157: “Landsat8” -> “Landsat 8” - Please correct this across the paper. There should be a space between Landsat and 8 or you can even write it as “L8”.

I have improved it in the article(line 151).

Line 157: “TRS” -> “Thermal Infrared Sensor (TIRS)”

I have improved it in the article(line 151).

Lines 228-230: It is mentioned that from 78 sampling points, 57 were used for training and 21 for validation. How did you split 57/21 for training and testing? Were these random splits?

Random selection is made and the selected points are evenly distributed in the lake.

Lines 230-232: “Coefficient of determination (R2), RMSE, and Bias was..” are these the all used to compare the 4 ML models?

Yes, using multiple indicators makes the results more convincing.

Line 174: Expand NDR

I have improved it in the article(line 169).

 

Yours Sincerely,
Lei Ding

 

Author Response File: Author Response.docx

Reviewer 3 Report

Please refer to the attachment for detailed comments

Comments for author File: Comments.pdf

The English language quality of the paper is satisfactory, but the authors are advised to further improve it.

Author Response

Dear reviewer:
Thank you for your decision and constructive comments on my manuscript. I have carefully considered the suggestion of the Reviewer and made some changes. I have tried my best to improve and made some changes to the manuscript(Modifications have been highlighted in red in the revised version). Revision notes, point-to-point, are given as follows:

  1. The title of the dissertation is suggested to be ‘TP concentration inversion and pollution sources in Nanyi Lake based on Landsat 8 data and InVEST model’.

Thank you for the title suggested. The precedent version of the title has been replaced, becoming “TP concentration inversion and pollution sources in Nanyi Lake based on Landsat 8 data and InVEST model”(line 2).

  1. Among the keywords: the term remote sensing is too broad.

I have refined this and modified it to remote sensing inversion( line 31).

  1. Introduction is long and recommended to be compressed and streamlined.

I have streamlined it appropriately as suggested.

  1. Line 67 is incorrect: It not only makes up for the shortcomings and shortcomings of traditional moni…

I have perfected this sentence(line 67).

  1. Is the abbreviation extreme gradient boosting algorithm (BST) correct? Should it be abbreviated as XGBoost?

Thank you for pointing out this abbreviation error, which I have revised in the text.

  1. Which period of time does the InVEST model simulate the distribution? Whether it is an average spatial distribution of data over 15 to 21 years or something else, should be specified in the text.

The InVEST model is simulated with 2021 as the base year, which is consistent with the base year for constructing the inversion and is indicated in the text(line 170).

  1. Most of the references in the article are in Chinese literature, and it is recommended that the English literature is the main one.

As recommended, some Chinese literature has been replaced with English literature wherever possible(References:1, 2, 4, 5, 6, 7, 8, 12, 31, 33, 35, 68, 71, 72, 74, 75, 76, 77 ).

  1. 5 (b) is without unit labeling.

Thank you for pointing out this vulnerability, which is now indicated in the figure.

  1. The data R2 on the graph in Figure 5(c) could not be matched with the original analysis data (Line 334)

I have fixed this deficiency in the article(line 326).

 

Yours Sincerely,
Lei Ding

Author Response File: Author Response.docx

Round 2

Reviewer 3 Report

There are problems with the format of some references, including numbering. Please check and modify them carefully. Good luck to you!

Author Response

The references have been improved, thank you very much.

Back to TopTop