Enhanced Water Quality Inversion in the Ningxia Yellow River Basin Using a Hybrid PCWA-ResCNN Model: Insights from Landsat-8 Data
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe topic under study is interesting. The paper's objective is evident, it adheres to journal guidelines and is well-organized, although with misbalances in sections' lengths.
The research methodology is accurate, however not explained thoroughly. Although generally well written, the manuscript requires grammar interventions and especially could benefit from improving style.
Too many acronyms would be avoided by replacing them with descriptive wording.
Below are suggestions on how to improve the manuscript:
1. Title: Based on the contents of the manuscript, I suggest the following title of the paper: ‘Advanced Water Quality Monitoring in the Ningxia Yellow River Basin Using a Hybrid Convolutional Neural Network Model: Insights from Landsat-8 Satellite Data’. If it is that important to refer to hybrid modeling, an alternative title could be: ‘Enhanced Water Quality Monitoring in the Ningxia Yellow River Basin Using a Hybrid PCWA-ResCNN Model: Insights from Landsat-8 Data’.
2. Abstract. The abstract should be rewritten to clearly say what was the aim, what methodology was used and what are the main results. Authors should ‘condense’ writing and avoid detailed percentages (especially with two decimals!). Avoiding acronyms as much as possible is strongly recommended. It is sufficient in the abstract to say ‘hybrid attention mechanism’ and avoid the acronym PCWA. Also in other cases, such as LSTM, remove acronyms unless necessary. Mention that the study area is the Yellow River in China. Not necessarily all readers know where is the Yellow River.
3. Introduction: Restructure text in a way to focus references on ‘topics’ such as ‘water quality monitoring’, ‘neural networks modeling’, etc. The last paragraph should be rewritten to make it easier to read. Delete all acronyms and rather describe them briefly. Add a paragraph on what was achieved and how much it is important.
4. Materials and Methods: In 2.1. Research area, avoid GPS coordinates, and describe its position in China (mention China). In the caption of Fig. 1 add the Yellow River basin and China. Mathematics is correctly presented. Title in 2.5. is imprecise; should be ‘Statistical indicators’, or ‘Statistical evaluation of indicators’; mention which indicators (samples), that is, be specific about ‘y’ in relations (6)-(9), i.e. which samples are used in computations
5. Results and Discussion: The last paragraph from 4.2. Limitations, after adapted, should be moved into Conclusions
6. Conclusions: Expand this section and indicate future research agenda.
Other suggestions:
- Complete text formatting must strictly follow journal rules (line spacing, referencing sources, etc.)
- Check for typos and grammar errors. Do not use ‘we’, ‘our’ etc., and write impersonal (3rd person) style --The final proofread is strongly recommended after rewriting parts of the manuscript aimed at improving the style of presentation.
Comments on the Quality of English LanguageAlthough generally well written, the manuscript requires grammar interventions and especially could benefit from improving style.
Author Response
Comments 1: [ Title: Based on the contents of the manuscript, I suggest the following title of the paper: ‘Advanced Water Quality Monitoring in the Ningxia Yellow River Basin Using a Hybrid Convolutional Neural Network Model: Insights from Landsat-8 Satellite Data’. If it is that important to refer to hybrid modeling, an alternative title could be: ‘Enhanced Water Quality Monitoring in the Ningxia Yellow River Basin Using a Hybrid PCWA-ResCNN Model: Insights from Landsat-8 Data’.]
Response 1: Thank you for pointing this out. I agree with this comment.[ Therefore, I chose the second title you gave, but I changed "Monitoring" to "Inversion", because this paper is to calculate the actual water quality parameter content in the water body by combining the remote sensing image data with the measured water quality data, which is the "Inversion" content.]- page number 1
Comments 2: [Abstract. The abstract should be rewritten to clearly say what was the aim, what methodology was used and what are the main results. Authors should ‘condense’ writing and avoid detailed percentages (especially with two decimals!). Avoiding acronyms as much as possible is strongly recommended. It is sufficient in the abstract to say ‘hybrid attention mechanism’ and avoid the acronym PCWA. Also in other cases, such as LSTM, remove acronyms unless necessary. Mention that the study area is the Yellow River in China. Not necessarily all readers know where is the Yellow River.]
Response 2: Thank you for pointing this out. I agree with this comment.[ As a result, I rewrote the abstract to clarify the content of each section, and condensed the writing to avoid detailed percentages. I avoided acronyms unless necessary, and there was an acronym about the core content structure of the article, echoing the title. In addition, I modified the title and content of the research area to make it clear that this area is in China.]-page number 1 and 2
Comments 3: [ Introduction: Restructure text in a way to focus references on ‘topics’ such as ‘water quality monitoring’, ‘neural networks modeling’, etc. The last paragraph should be rewritten to make it easier to read. Delete all acronyms and rather describe them briefly. Add a paragraph on what was achieved and how much it is important.]
Response 3: Thank you for pointing this out. I agree with this comment.[ Therefore, I have rewritten the last paragraph of the introduction, removing all the acronyms and describing them briefly.]-page number 2.
Comments 4: [ Materials and Methods: In 2.1. Research area, avoid GPS coordinates, and describe its position in China (mention China). In the caption of Fig. 1 add the Yellow River basin and China. Mathematics is correctly presented. Title in 2.5. is imprecise; should be ‘Statistical indicators’, or ‘Statistical evaluation of indicators’; mention which indicators (samples), that is, be specific about ‘y’ in relations (6)-(9), i.e. which samples are used in computations]
Response 4: Thank you for pointing this out. I agree with this comment.[ Therefore, I have changed the title of Figure 1 to clarify its location in China. For the Title in 2.5, I chose the "'Statistical evaluation of indicators" you gave, and the "y" in formula (6) - (9) did not give the formula in italic form due to the problems I wrote before, and I have modified it.] -page number 7
Comments 5: [Results and Discussion: The last paragraph from 4.2. Limitations, after adapted, should be moved into Conclusions]
Response 5: Thank you for pointing this out. I agree with this comment.[ Therefore, I have adjusted the last paragraph of 4.2 to the conclusion.] -page number 12
Comments 6: [ Conclusions: Expand this section and indicate future research agenda.]
Response 6: Thank you for pointing this out. I agree with this comment.[ Therefore, I have revised the conclusion based on the fifth comment you gave, and expanded the future research agenda.] -page number 13
Other suggestions:
Thank you for pointing this out. I agree with this comment.[ This manuscript has been submitted to the MDPI official body for polishing and adjustment of the grammatical structure.]
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors proposed a new deep-learning PCWA-ResCNN model with multiple indicators to test its performance in water quality inversion using data from the Yellow River Basin in Ningxia. They concluded that the proposed model can be used as a low-cost and effective method to estimate the concentration of conventional pollutants and water quality parameters in the Yellow River Basin of Ningxia. The manuscript is well written, and the authors specifically expanded on the advantages and limitations of the proposed model; however, I have the following comments for the authors to consider:
General Comments
· Page 4, Lines 133 and 137, please remove the hyperlinks and replace them with proper intext references as other references. The hyperlinks can be included with the relevant reference in the reference section. Also, consider removing the hyperlink in line 159.
· Some of the texts in Figure 5 are difficult to read.
· In section 2.3.1, the PCWA was combined with CNN for improvement. Can LSTM also be combined with PCWA to increase its effectiveness?
· Please expand and provide more background on the class 1-IV water standard mentioned in section 3.2.
· Line 322 talked about inversion maps of the four-water quality, can this map be provided and included for reference or include a reference to it if possible?
· Lines 326-327, please add reference.
· Lines 329-321, are these from the findings of this study or from reference 30, please clarify.
· Acknowledgement: Please remove the word “Remote” before the sentence.
Comments on the Quality of English LanguageMinor editing required.
Author Response
Comments 1: [ Page 4, Lines 133 and 137, please remove the hyperlinks and replace them with proper intext references as other references. The hyperlinks can be included with the relevant reference in the reference section. Also, consider removing the hyperlink in line 159.]
Response 1: Thank you for pointing this out. I agree with this comment.[ Therefore, I have removed the original hyperlinks on lines 133, 137, and 159.]- page number 4
Comments 2: [Some of the texts in Figure 5 are difficult to read.]
Response 2: Thank you for pointing this out. I agree with this comment.[ Before Figure 5, I introduced each water quality standard and the meaning represented by each water quality parameter in the paper in detail, and then observed the water quality changes in Figure 5, it was obvious that the water quality in the middle reaches of the basin was poor, while the water quality in the upper and lower reaches was relatively good. Combined with relevant references, it was discussed that the water quality in the basin was developing in a good direction.]-page number 10
Comments 3: [ In section 2.3.1, the PCWA was combined with CNN for improvement. Can LSTM also be combined with PCWA to increase its effectiveness?]
Response 3: Thank you for pointing this out.[ PCWA is combined with CNN for improvement. In theory, LSTM can be combined with PCWA to improve its effectiveness. However, experimental results show that LSTM has poor performance in dissolved oxygen parameters and can not show good generalization ability in all water quality parameters. Therefore, the model effect is not obvious after adding this attention mechanism to improve the model.]
Comments 4: [Please expand and provide more background on the class 1-IV water standard mentioned in section 3.2.]
Response 4: Thank you for pointing this out. I agree with this comment.[ Accordingly, in Section 3.2 I have provided detailed background information on the levels 1-IV water standards mentioned.] -page number 10
Comments 5: [Line 322 talked about inversion maps of the four-water quality, can this map be provided and included for reference or include a reference to it if possible?]
Response 5: Thank you for pointing this out. [ Therefore, the inversion map of four kinds of water quality in the original line 322 can provide a map of this region, but due to the large sampling area, the Yellow River Basin looks subjectively narrow. The addition of the map and the addition of multiple lines will make the figure look chaotic. I think this can also accurately point out the specific location of this area, I hope you can understand.]
Comments 6: [ Lines 326-327, please add reference.]
Response 6: Thank you for pointing this out. I agree with this comment.[ I have modified the original lines 326-327. Since the content of this sentence comes from the analysis and summary of the content of the document published by the local government, the corresponding references cannot be found, I hope you can understand.] -page number 11.
Comments 7: [ Lines 329-321, are these from the findings of this study or from reference 30, please clarify.]
Response 7: Thank you for pointing this out. I agree with this comment.[ Original lines 329-321, these are from reference 30, which I have revised.] -page number 11.
Comments 8: [ Acknowledgement: Please remove the word “Remote” before the sentence.]
Response 8: Thank you for pointing this out. I agree with this comment.[ Due to this error in my written question, I have deleted the word 'Remote' from the acknowledgements.] -page number 13.
Reviewer 3 Report
Comments and Suggestions for AuthorsDear author,
The manuscript " "Inversion of Water Quality Parameters in the Ningxia Yellow River Based on the PCWA-ResCNN Model" presents the prediction of water quality parameters such as of turbidity, Dissolved Oxygen (DO), Chemical Oxygen Demand (COD) and ammonia nitrogen using PCWA-ResCNN model, on the example of monitoring Ningxia Yellow River Basin. The introduction provides adequate background information on the subject and research area and is well-supported with references. The aim of the manuscript is clearly defined.
Although the work is both interesting and systematically presented, there are minor omissions in the manuscript that need to be addressed. Therefore, the manuscript should be reviewed after minor revisions due to the concerns listed below:
1. Introduction: Cite the reference immediately after mentioning the author, rather than placing it at the end of the sentence. Liu et al. [7]...
2. Can the PCWA-ResCNN model be used to predict other parameters such pH and conductivity?
3. How effectively does the PCWA-ResCNN model capture and interpret the contextual relationships and interactions between various water quality parameters, such as turbidity and pH?
4. Adapt the abbreviations to match those already used in the text.
5. All the typos and grammar need to be checked thoroughly in the manuscript.
Author Response
Comments 1: [Introduction: Cite the reference immediately after mentioning the author, rather than placing it at the end of the sentence. Liu et al. [7]...]
Response 1: Thank you for pointing this out. I agree with this comment.[ Therefore, I have modified the position of the reference.]- page number 2
Comments 2: [Can the PCWA-ResCNN model be used to predict other parameters such pH and conductivity?]
Response 2: Thank you for pointing this out. [ PCWA-ResCNN model can be used to predict pH value, electrical conductivity and other parameters. This study mainly focuses on the prediction of the concentration of four kinds of water quality. If there are enough sample points in the study area, the model can be used to predict any parameter.]
Comments 3: [ How effectively does the PCWA-ResCNN model capture and interpret the contextual relationships and interactions between various water quality parameters, such as turbidity and pH?]
Response 3: Thank you for pointing this out. [ PCWA module uses local and global context information to assign different weights to different water quality parameters to enhance the sensitivity of the model to water quality characteristics. This mechanism helps capture complex spatial and temporal changes. ResCNN improves the performance of the model through residual learning. Residual blocks allow the model to maintain good information transmission while capturing deep features, thereby improving the modeling ability of the interaction between water quality parameters. Combining these two modules, the PCWA-ResCNN model is able to parse water quality data at multiple scales and levels to more accurately estimate the complex relationships of water quality parameters such as turbidity and pH.]
Comments 4: [Adapt the abbreviations to match those already used in the text.]
Response 4: Thank you for pointing this out. I agree with this comment.[ Abbreviations have been adjusted to match those already used in the text.]
Comments 5: [FAll the typos and grammar need to be checked thoroughly in the manuscript.]
Response 5: Thank you for pointing this out. I agree with this comment.[ I have thoroughly checked all the spelling errors and grammar in the manuscript, and the manuscript has been submitted to the MDPI official agency for polishing.]
Reviewer 4 Report
Comments and Suggestions for AuthorsL276-278. Please explain further why LSTM is not suitable for predicting DO. Do you have any recommendations on how to address this limitation?
L 301 – L304 As stated, turbidity is not included in the WQ evaluation but is used as a reference indicator to determine its effect on sediments, salinity, and DO. I am searching for where the comparisons were made in the paper. If these are already in the paper, kindly make a brief explanation/clarification.
L 330-342: Please clarify to what extent the improvement of WQ is in terms of NH3-N and CODMn.
Comments for author File: 
Comments.pdf
The quality of English can be improved to increase the paper's clarity.
Author Response
Comments 1: [ L276-278. Please explain further why LSTM is not suitable for predicting DO. Do you have any recommendations on how to address this limitation?]
Response 1: Thank you for pointing this out. I agree with this comment.[ The LSTM model is not suitable for predicting DO parameters in the Yellow River Basin of Ningxia mainly because the water quality changes in the Yellow River Basin involve multi-scale and nonlinear factors. The LSTM model mainly focuses on the long-term and short-term dependence of time series data, and its processing capacity for these multi-scale and nonlinear factors is limited. Consider the multi-dimensional characteristics and complexity of the data to improve the prediction effect.]- page number 8
Comments 2: [L 301 – L304 As stated, turbidity is not included in the WQ evaluation but is used as a reference indicator to determine its effect on sediments, salinity, and DO. I am searching for where the comparisons were made in the paper. If these are already in the paper, kindly make a brief explanation/clarification.]
Response 2: Thank you for pointing this out. I agree with this comment.[ In the original L 301-L304, turbidity has no corresponding standard limit value, so it does not participate in water quality assessment. However, the Yellow River contains a lot of sediment, so in this study area, turbidity inversion is particularly important, which can be used to determine the influence of sediment and sediment on the basin, and determine whether it should be strengthened.]-page number 10
Comments 3: [ L 330-342: Please clarify to what extent the improvement of WQ is in terms of NH3-N and CODMn.]
Response 3: Thank you for pointing this out. I agree with this comment.[ In original L 330-342, through searching relevant literature, this paper concluded that through continuous management and manual intervention, CODMn and NH3-N content in the Yellow River Basin were reduced by about 40% and 50% compared with ten years ago, indicating that the management effect was very good.]-page number 11
