Prediction and Analysis of Spatiotemporal Evolution Trends of Water Quality in Lake Chaohu Based on the WOA-Informer Model
Abdellah Elhmaidi
Round 1
Reviewer 1 Report
Comments and Suggestions for Authors- Do you really need data from all three years, from 2021 to 2024, to predict values ​​over time intervals of only 4, 12, 24, 48, and 72 hours?
- Why didn't you predict over time intervals of one, two, and three years ahead to truly see the evolution of future water quality trends?
- Why predict the values ​​of these four parameters separately? You would have been better off using empirical formulas to calculate the overall Water Quality Index (WQI) based on the four parameters.
- In this case, you will have the WQI as the model output and the four parameters as the model input. Thematic maps of observed and predicted values ​​will be presented according to quality classes, from excellent to very poor.
- Still in this same case, the thematic maps obtained by GIS deserve to be reclassified to highlight the quality classes excellent, good, average, poor and very poor for the four parameters studied.
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe manuscript addresses an important and timely topic, offering valuable contributions to water quality prediction and environmental management through the application of advanced machine learning techniques. The approach has strong potential for ecosystem monitoring and contributes to knowledge in environmental modelling and aquatic resource management. Several aspects of the study require clarification and improvement before the manuscript can be considered for publication.
The final paragraph of the Introduction would benefit from a clearer structure. It should begin with a brief description of the methods, followed by the study’s objectives, the research questions or hypotheses, and end with the significance. Rewriting it in this order will create a logical flow and strengthen the transition to the Methods section.
All data, statistics, or factual statements included in the Materials and Methods section should be supported by appropriate references or sources. Citing these sources enhances the reliability, transparency, and reproducibility of the study and ensures that readers can verify the information presented.
The discussion currently employs informal and non-analytical expressions such as “In response…” and “In summary…”, which should be replaced with more precise and academically appropriate language. In addition, the section would be significantly strengthened by expanding on three essential elements: (1) a clearer explanation of the practical or theoretical implications of the findings, (2) acknowledgment of any limitations or uncertainties within the study, and (3) explicit priorities or recommendations for future research. Addressing these aspects would greatly enhance the coherence, depth, and scientific contribution of the discussion, bringing it in line with international standards for academic publications.
L95 It would be clearer to title this section “Materials and Methods” to align with standard scientific manuscript conventions.
L100–L102: Currently, some units are written in full (e.g., 6 meters; 2 billion cubic meters) while others are abbreviated (e.g., km²). Units should follow a consistent format, either fully written or properly abbreviated according to standard conventions. The authors should carefully review the manuscript to ensure that all units are standardized throughout the text.
L103, 109, 114, and 122 There is a broken reference citation (“Error! Reference source not found. 19”). The authors should correct this by ensuring that the reference is properly linked and appears in the reference list.
L114‒L115 The statement “According to recent monitoring data...” does not clarify the source of the data. The authors should indicate whether the data are included in the Supporting Materials, available in a public database, or accessible from another source to ensure transparency and reproducibility.
L125 “...monitoring section” should be “...monitoring stations”.
L110‒L112 The flow of this paragraph is difficult to follow. The ideas appear fragmented and lack clear connections. I recommend rewriting these sentences to improve logical development: start with the main point, then present supporting details in a coherent order.
Comments on the Quality of English LanguageThe English could be improved to more clearly express the research.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsThe paper presents a methodology for forecasting and spatiotemporal analysis of water quality in Chaohu Lake adopting the WOA-Informer Model, analyzing data from a four-year period (2021 a 2024). The paper follows a logical structure, presenting a logical sequence of introduction, methodology, results/discussion, and conclusion. However, I believe the abstract could begin by explaining the importance of lakes as ecological barriers and then provide a specific introduction to Lake Chaohu.
The introduction is well-written and describes the importance of water quality, specifically referring to the case of eutrophication in Lake Chaohu, one of China's largest. The text also presents the methodological innovation proposed in the paper, which is the use of the Informer model (an advancement over Transformer for long time series) combined with WOA, which reduces errors and increases robustness. However, the text is somewhat lengthy and repetitive, citing many similar works but not providing a critical analysis or synthesis. A comparative table of previous models (authors, methods, performance metrics) would be interesting. The introduction focuses too much on methods and not enough on the reality of Lake Chaohu (ecological importance, eutrophication level, social impacts). Furthermore, it would be important to incorporate statistical data or global number about eutrophication and water quality degradation in large lakes to better illustrate and reinforce the relevance of the topic. Several adjustments are also needed regarding the references to ensure consistency and compliance with the journal's formatting requirements.
The "Research Areas" and "Data Analysis and Processing" sections need to present more details about the region. The description of the study area could include demographic data, such as population size, the most important cities near the lake basin, as well as information on the predominant geology and climate classification. These details would help to contextualize the environmental pressures and anthropogenic activities that influence water quality.
The methodology explains in detail how Informer and WOA work, demonstrating the recurring problems of traditional methods and the consequent use of the WOA-Informer combination as a solution. However, formulas (such as K-L divergence and WOA update functions) are presented that do not directly relate to the practical problem (water quality).
Another important issue is explaining the choice of the period and the reason for collecting data for only four years. Adjustments to the references are also necessary, as we have several sentences without references and references without numbers and years.
The "Results" chapter presents a robust dataset, which lends robustness to the forecasts and increases the reliability of the results. The chapter is well organized and divided into sections that facilitate comprehension. The comparison between Informer and WOA-Informer is systematic, with clear improvements in RMSE and R².
However, several aspects of the presentation and discussion of the results could be improved. Many numerical results are presented in long paragraphs, which makes them difficult to read. These data could be presented in more concise tables and graphs, with more interpretive analyses in the text. Several repetitions of information are also observed. Sections about performance degradation with long forecasts appear two or three times in very similar texts.
The Discussion section directly highlights that WOA-Informer improves accuracy (RMSE, R²), reduces error accumulation, and has advantages in long-term forecasting, connecting the results well with the study's initial objectives. It mentions that the short data period (2021–2024) and the difficulties in generalizing to other lakes are limitations to the research. This is an interesting aspect of the paper and contributes to its credibility.
However, while these limitations are acknowledged, the discussion is superficial in its scientific interpretation. Furthermore, it is reported that the model helps predict eutrophication and provides management support, but does not quantify the impact. In other words, could a percentage improvement in R² imply an earlier forecast? Presenting this relationship between the numbers and actual effects would contribute to improving the discussion. All of these issues are described in the attached PDF.
Regarding the conclusions, this section begins with a repetition of points already presented in the methodology and results (lines 738 to 747), and therefore this part of the text could be reworded. Otherwise, I consider the conclusions to be well-written and consistent with the results presented.
Comments for author File: Comments.pdf
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsAll comments have been addressed, and now the manuscript is in good shape. I recommend acceptance for publication.
