Performance Evaluation of Inherent Optical Property Algorithms and Identification of Potential Water Quality Indicators Using GCOM-C Data in Eutrophic Lake Kasumigaura, Japan

Round 1

Reviewer 1 Report (New Reviewer)

Comments and Suggestions for Authors

Comments/Suggestions:

(1)Table 1 could be re-organized in a concise format

(2)Number of measurements or matchups should be given together with statistics, e.g., Table 1, Table 2, Figure 3, etc.

(3)Unit of Rrs is sr^-1, not str^-1

(4)Since you have field measurements of IOPs, water quality and Rrs together, why not make correlation analysis with field data in Section 3.3? In other words, what is the advantage of using GCOM-C/SGLI-estimated IOPs?

(5)In section 3.2, it is required to describe the spatial-temporal match-up between OC data and in-situ measurement. Furthermore, what is the purpose of illustrating Rrs comparison here?

(6)detailed information should be provided with some references, e.g., website or ISSN

(7)Sequential number of Figures and Tables have to be checked, and they should be properly referenced in the context

(8) Highlighted context at the beginning of Page 3: This statement should be treated carefully. To my knowledge, machine learning methods is only one of the available IOP inversion models, and empirical/semi-analytic models are those frequently preferred. ML method emerges in recent years and highlighted.

(9) The last passage on Page 9: Here, two different inversion algorithms are used for different IOPs. Attention should be paid to their consistency when it comes to the closure of total absorption and constituents' absorption.

(10) The passage just below Figure 3: If you used the same dataset in Table 3 and Figure 3, no need to re-state the performance in this passage after final selection of best models in your study as described in previous paragraph.

Or two can be merged properly.

(11) The passage just before Section 3.3: To quantify the error across different wavelengths, it would be better to use RMSE relative to its mean, because Rrs may vary significantly across wavelengths.

Also, it would benefit readers if the Rrs spectra is shown somewhere in the text.

(12) Section 4.1: To quantify the error across different wavelengths, it would be better to use RMSE relative to its mean, because Rrs may vary significantly across wavelengths.

Also, it would benefit readers if the Rrs spectra is shown somewhere in the text.

(13)more can be found in the attached document (highlighted areas and their note)

Comments for author File: Comments.pdf

Comments on the Quality of English Language

The manuscript could be more compact when the authors are explaining and comparing performances of different IOP models.

Author Response

The details of the modifications are provided in the attached word document. Please check it.

Author Response File: Author Response.docx

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

• General comments

This paper investigates the performance evaluation of inherent optical properties (IOPs) algorithms based on GCOM-C satellite data for eutrophic waters in Lake Kasumigaura, Japan, and explores the potential of using IOPs as water quality indicators. The study compares the performance of five IOP algorithms (QAA, GSM, GIOP, PML, and LMI), and analyzes the relationship between IOPs and water quality parameters (such as chlorophyll-a, suspended solids, etc.) by integrating field measurements and satellite remote sensing data. The research concludes that the QAA and GSM algorithms perform excellently in estimating specific IOPs, and that IOPs are significantly correlated with water quality parameters, making them effective indicators for water quality monitoring.

Overall, the topic of the paper has strong scientific significance, as remote sensing of water quality is a critical and practical issue in the management of eutrophic lakes. The research methodology is systematic, the data sources are reliable, and the analysis results are convincing. However, the paper has some shortcomings in terms of innovation, methodological details, depth of result interpretation, and clarity of presentation. Further improvements are needed to enhance its academic value and readability. While the study could potentially contribute to existing knowledge by refining its overall approach and clarifying its data interpretation, the current version of the manuscript leads me to make the decision to recommend major revision.

To wit:

First of all, although the paper evaluates the performance of five IOP algorithms and analyzes their relationship with water quality parameters, the overall research framework and methodology are rather conventional, failing to present significant innovations based on existing literature. For example, the superiority of the QAA and GSM algorithms in complex water bodies has been reported in several studies (e.g., references [10][91]), and the findings of this paper are highly consistent with those conclusions. Additionally, the use of IOPs as water quality indicators is not a novel topic.

Secondly, the paper does not sufficiently describe the preprocessing steps of the GCOM-C/SGLI data (such as radiometric calibration, cloud detection), as well as the specific atmospheric correction methods. These steps are crucial for the accuracy of remote sensing reflectance (R_rs), especially in eutrophic waters where atmospheric correction is particularly challenging.

Thirdly, although five algorithms are listed, the paper does not provide detailed information on their parameter settings, assumptions, or adjustment processes. For instance, the specific input parameters for QAA or the nonlinear optimization process for GSM are not mentioned, which could affect the reproducibility of the results.

Fourthly, some of the manuscript's expression is not appropriate, and some of the statements in these parts have no corresponding literature support. Literature research on this statement is needed to support the statement.  I will give a detailed explanation in the "specific comments" below, please refer to it carefully and modify it.

Lastly, the writing of the manuscript requires some English editing, and there are many errors in grammar, spelling, and sentence structure.  I have provided some suggested changes in the "specific comments" below to improve this situation.

2) Specific comments

Throughout the manuscript:

1. All abbreviations in each part of the manuscript including the abstract and contents, as well as in the Table and Figure, should be introduced for the first time despite how common or not the abbreviation is.

 

Abstract

1. It is recommended to add a sentence outlining the background and scientific issues of the study before "This study examines...".
2. Which version of the quasi-analytical algorithm (QAA) is used? The QAA family includes QAA_V5, QAA_V6, QAA710, QAA716, QAA750E, and the TC algorithm. Please specify the exact version used in this study to ensure clarity and reproducibility of the results.
3. The QAA demonstrated the best performance in... What are the evaluation criteria? The explanation here is too qualitative; it is recommended to use quantitative accuracy metrics for a more precise evaluation. Similarly, the performance of the GSM algorithm should be assessed using specific quantitative metrics.
4. With Pearson correlation coefficients... What are the specific values? It cannot remain too qualitative. The exact numerical values should be presented for clarity and precision.
5. Strong correlation... Which specific band does this refer to, or is it referring to all bands? Please clarify and revise accordingly.
6. Moderate to strong correlations. What is this being compared to in order to define 'moderate'? Please clarify and revise accordingly.
7. It is recommended to add the significance of this study and its implications for other research in the last sentence of the abstract. Please elaborate and revise accordingly.

 

Introduction

1. It is recommended to change 'GEOKOMPSAT-2B GOCI-II/Geostationary Ocean Color Imager-II (GOCI-II)' to 'GEOKOMPSAT-2B/Geostationary Ocean Color Imager-II (GOCI-II)'. Please revise accordingly.
2. The effectiveness of estimating Inherent Optical Properties (IOPs) has been extensively studied, with various algorithms developed for this purpose. Currently, the algorithms for IOP estimation are mainly categorized into four types: empirical models, semi-analytical models, quasi-analytical models, and machine learning models. Among these, the quasi-analytical models are particularly diverse, with several versions available, including QAA_V5, QAA_V6, QAA710, QAA716, QAA750E, and the TC algorithm. The following references provide key insights into the development and application of these algorithms:

Lee Z, Lubac B, Werdell J, et al. An update of the quasi-analytical algorithm (QAA_v5) [J]. International Ocean Color Group Software Report, 2009, 1: 1-9.

Lee Z, Shang S, Du K, et al. Enhance field water-color measurements with a Secchi disk and its implication for fusion of active and passive ocean-color remote sensing [J]. Applied Optics, 2018, 57(13): 3463-3473.

Huang J, Chen L, Chen X, et al. Modification and validation of a quasi-analytical algorithm for inherent optical properties in the turbid waters of Poyang Lake, China [J]. Journal of Applied Remote Sensing, 2014, 8(1): 083643-083643.

Li J, Zheng Z, Li Y, et al. A hybrid algorithm for estimating total nitrogen from a large eutrophic plateau lake using Orbita hyperspectral (OHS) satellite imagery [J]. International Journal of Applied Earth Observation and Geoinformation, 2024, 131: 103971.

Zheng Z, Huang C, Li Y, et al. A semi-analytical model to estimate Chlorophyll-a spatial-temporal patterns from Orbita Hyperspectral image in inland eutrophic waters [J]. Science of the Total Environment, 2023, 904: 166785.

Huang C, Zheng Z, Li Y, et al. Enhanced algorithm for water transparency estimation in turbid plateau waters using Orbita Hyperspectral (OHS) Imagery [J]. IEEE Transactions on Geoscience and Remote Sensing, 2025.

1. Xue, R. Ma, H. Duan, M. Shen, E. Boss, and Z. Cao, “Inversion of inherent optical properties in optically complex waters using Sentinel-3A/OLCI images: a case study using China’s three largest freshwater lakes,” Remote Sens. Environ., 225, 328–346 (2019).

Liu G, Li L, Song K, et al. An OLCI-based algorithm for semi-empirically partitioning absorption coefficient and estimating chlorophyll a concentration in various turbid case-2 waters [J]. Remote Sensing of Environment, 2020, 239: 111648.

3. "Using measured Rrs" should be modified to "Using measured R_rs,"This correction clarifies the need for a comma after "Rrs" for better sentence structure and readability.

 

Materials and methods

1. Missing data processing details: The paper does not adequately describe the preprocessing steps for GCOM-C/SGLI data, such as radiometric calibration, cloud detection, and the specific atmospheric correction methods. These steps are crucial for the accuracy of remote sensing reflectance (Rrs), especially in eutrophic waters, where atmospheric correction is particularly challenging.
2. Unclear implementation details of IOP algorithms: While five algorithms are listed, the paper does not provide sufficient details on the parameter settings, assumptions, or adjustments made during the implementation. For example, the specific input parameters for QAA or the nonlinear optimization process of GSM are not mentioned, which may affect the reproducibility of the results.
3. Field measurement representativeness: The field data from 2017-2018 only cover 10 sites and have a limited temporal range. The paper does not adequately discuss how the spatiotemporal representativeness of these measurements supports the validation of satellite data from 2018-2022.
4. It is recommended to change '2.1. Study area description.' to '2.1. Study area.
5. It is recommended to redraw Figure 1. For the left image, please add a scale bar, latitude and longitude grid, and the boundary of Japan, along with subplot numbers. For the right image, please add the directions 'N' and 'E'.
6. For the optical... it is recommended to include the formula for calculating remote sensing reflectance to provide clarity for the readers. Please elaborate and make the necessary revisions.
7. Please include the specific steps for data preprocessing and algorithm implementation in the '2.2 Data Collection' and '2.4 Semi-analytical IOP Algorithms' sections, such as the atmospheric correction model used (e.g., 6SV, MODTRAN) and its parameters.
8. Please include an explanation of the criteria for selecting field measurement sites and the sampling frequency, discussing their potential impact on the results (e.g., seasonal variations).
9. Provide supplementary materials for the algorithm implementation (such as code or parameter tables) to enhance the reproducibility of the study.
10. What are the full forms of aph, adg, and bbp? Regardless of how common the abbreviations are, their full forms should be provided the first time they appear. Please elaborate and revise accordingly.
11. The statement 'The Redfield ratio is...' lacks theoretical support. It is recommended to cite relevant literature to substantiate this claim. Please elaborate and revise accordingly.
12. It is recommended to change 'with a resolution of 250 m' to 'with a spatial resolution of 250 m.' Please revise accordingly.
13. Is the secondary product of GCOM-C/SGLI an atmospheric correction product? What atmospheric correction method was used? Please elaborate and revise accordingly.
14. The discussions by Gordon et al. [79] and Smyth et al. [39] should be part of the introduction, not placed in the methods section. Please elaborate and revise accordingly.

 

Results

1. Insufficient depth in statistical analysis: Although MAE, RMSE, and r were used as indicators, no statistical significance tests (such as t-tests or ANOVA) were performed to assess the differences in algorithm performance, which diminishes the rigor of the comparison results.
2. Inadequate interpretation of the Taylor diagram: Figure 4 (Taylor diagram) presents the algorithm performance, but the discussion is somewhat general, lacking a detailed analysis of the physical reasons behind the algorithm's performance across different wavelengths or IOPs.
3. Limitations in correlation analysis: Table 3 and Figure 7 show strong correlations (e.g., aph and Chl-a with r=0.84), but the underlying bio-optical mechanisms or potential sources of error (such as spectral interference) are not discussed.
4. What steps are involved in processing the secondary products of GCOM-C/SGLI? Does it yield surface remote sensing reflectance or remote sensing reflectance above the water surface?
5. It is recommended to include the MAPE (Mean Absolute Percentage Error) metric in Figure 5.
6. It is recommended to change 'ranging from 0.0005 str^-1' to 'ranging from 0.0005 sr^-1.
7. Lines 393-398: Why are the correlation coefficients between the inherent optical properties derived from the QAA algorithm and water quality parameters used in Figure 6, rather than the correlation between the measured inherent optical properties and water quality parameters? Please elaborate and make the necessary revisions.

 

Discussion

1. Insufficient comparison with other studies: The discussion section does not adequately compare the results with existing studies on Xiapu Lake (e.g., [16][58][59]) or similar water bodies (e.g., [26][113]), failing to highlight the unique contributions or limitations of this study.
2. Weak analysis of limitations: While the impact of atmospheric correction on Rrs is mentioned, other potential sources of error (e.g., sensor noise, lake hydrodynamics) and their possible effects on the results are not discussed.
3. Lines 480-574: This section seems more like an extension of the results. It is recommended to deepen this section into an explanation of the mechanisms behind why the QAA algorithm outperforms other algorithms. Please elaborate on why the QAA algorithm performs better than other semi-analytical models, and provide a detailed analysis of the factors contributing to its superior performance.
4. Lines 578-580: Why was the OCx algorithm chosen for comparison, rather than using the more commonly applied three-band or four-band algorithms for inland lakes? Please provide a detailed explanation and revision.
5. Why was suspended solids (SS) differentiated into organic suspended solids (OSS) and inorganic suspended solids (ISS) rather than directly conducting a correlation analysis with total suspended solids (TSS)? Please provide a detailed explanation and revise accordingly.

Comments on the Quality of English Language

The writing of the manuscript requires some English editing, and there are many errors in grammar, spelling, and sentence structure.

Author Response

The details of the modifications are provided in the attached word document. Please check it.

Author Response File: Author Response.docx

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

This manuscript conducted a comparative evaluation of the accuracy of five classical inherent optical property (IOP) algorithms including QAA, GSM, PML etc. algorithms, in calculating remote sensing reflectance (Rrs) of water bodies. Additionally, it analyzed the relationship between IOPs and water quality parameters. Although this work represents fundamental algorithm validation without so much innovative methods, it holds significant reference value. The experimental data are comprehensive and well-documented, offering important insights for the selection of IOP algorithms and the inversion of water quality parameters in related research fields.

Author Response

The details of the modifications are provided in the attached word document. Please check it.

Author Response File: Author Response.docx

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.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Suggest publishing this manuscript after minor revisions. Please refer to the attachment for detailed suggestions.

Comments for author File: Comments.docx

Author Response

Reviewer 1

Thank you for your valuable comments. We have thoroughly reviewed each one and addressed all the points you raised. We believe that your suggestions have significantly enhanced the quality of the paper. We kindly request your review and confirmation.

Comment 1: The first paragraph lacks a description of the ecological functions of lakes. It is suggested to supplement the writing style and content related to lake functions.

Response 1: Thank you for pointing this out. We agree with this comment. Therefore, we have supplemented the first paragraph by including a description of the ecological functions of lakes. These revisions can be found on page 2, paragraph 1. To enhance the introduction section, two additional paragraphs were added to emphasize key aspects of the research content: the significance of IOP validation for accuracy, discussed in Page 2, Paragraph 3, and an overview of research efforts in Lake Kasumigaura, presented in Page 3, Paragraph 5.

Introduction

Paragraph 1

The uneven distribution of lakes and marshes, as inland water bodies, across the globe leads to notable regional variations in water availability, biodiversity, aquatic ecosystems, and biogeochemical properties [1, 2]. These water bodies play a vital role by contributing to nutrient cycling, environmental stability, supplying drinking water, supporting aquaculture, enabling recreational activities, and offering other vital benefits that contribute to human well-being and the health of global ecosystems [3, 4]. However, in environments where primary production is high and water exchange is limited, problems such as organic pollution and eutrophication are increasing occasionally, emphasizing the need for sustainable monitoring and management of water systems to preserve ecological balance [5]. Additionally, numerous studies have highlighted the effects of recent climate change on lake ecosystems [6, 7]. Therefore, it is crucial to consistently monitor ecosystem trends and changes in water quality [8–10].

Paragraph 3

The effectiveness of inherent optical properties (IOPs) depends on using well-validated algorithms, particularly in environments like highly eutrophic and optically complex waters [22]. These algorithms come in various forms and are designed to generate a wide range of optical data products, each serving different purposes [10, 23]. Ensuring these IOP algorithms are validated is crucial for generating accurate, reliable, and actionable data, minimizing uncertainty, which supports water quality monitoring, ecosystem management, and the assessment of biogeochemical characteristics [21, 23, 24]. Accurate in situ IOP measurements are essential for ensuring reliable remote sensing validation, which enhances the ability to estimate critical water quality parameters such as chlorophyll-a, suspended solids, and dissolved organic matter. However, optically complex waters often rich in phytoplankton, organic material, and non-algal particles introduce unique challenges for optical measurements[25, 26]. By developing and refining strong algorithms, we can significantly improve remote sensing capabilities. This includes tracking harmful algal blooms, monitoring ecosystem health, and adapting tools to fit a wide range of aquatic environments. Ultimately, these developments offer dependable, high-quality data that facilitate informed decision-making and promote sustainable water resource management [27, 28].

Paragraph 5

Remote sensing has been used to explore the optical properties of water across various regions and a wide range of aquatic environments [5, 26, 32]. This study examines the case of Lake Kasumigaura, the second-largest lake in Japan, which is a vital component of the region's ecosystem, economy, and water resource management system [55]. However, like many freshwater bodies worldwide, it faces significant environmental challenges, including eutrophication, pollution, and water quality deterioration, driven by both anthropogenic activities and natural processes [56]. Furthermore, climatic variations and seasonal hydrodynamics exacerbate water quality issues, complicating management and restoration efforts. Previous research has developed various methods to assess water quality in Lake Kasumigaura, focusing on the application of machine learning and algorithms to predict Chl-a and suspended solid using remote sensing data [16, 57–59]. However, addressing these intricate challenges requires extensive water quality parameters to understand the drivers of water quality changes and to develop effective mitigation strategies for sustainable lake management [60, 61].

 

Comment 2: The references in the introduction lack research from the latest year. Currently, the latest literature in the article was published in 2023. It is recommended to supplement relevant literature published in 2024.

Response 2: Thank you for this observation. we have incorporated recent and relevant research articles published in 2024 to strengthen the introduction and provide up-to-date context. The introduction section incorporates references [1, 19, 42, 43, and 49] from 2024.

Comment 3: Table 3: Compared to tables, scatter plots often present richer and more intuitive information. Tables can only provide numerical information, but cannot reflect whether sample points are evenly distributed around the 1:1 line. Therefore, it is suggested that the author modify Table 3 to a scatter plot.

Response 3: Thank you for this suggestion. We agree that scatter plots offer a more intuitive way to visualize the data distribution and relationships. Therefore, We have modified Table 3 to highlight a scatter plot illustrating the distribution of sample points around the 1:1 line. This change is reflected on page 9, Table 3.

 

3. Results

Table 3. Accuracy assessment results of QAA, PML, LMI, GSM, and GIOP algorithms using in-situ measured a(λ), a_ph(λ), a_NAP(λ), and a_CDOM(λ).

IOP Algorithm	Error metrics	QAA	PML	LMI	GSM	GIOP
a(λ)	MAE	0.23	0.23	0.29	0.42	0.21
	RMSE	0.29	0.31	0.42	0.57	0.33
	r	0.98	0.96	0.95	0.94	0.95
	% Between ±30 to 1:1	91.81	85.83	79.31	75.04	85.97
aph(λ)	MAE	0.05	0.15	0.11	0.15	0.09
	RMSE	0.08	0.24	0.17	0.21	0.14
	r	0.97	0.69	0.83	0.75	0.88
	% Between ±30 to 1:1	89.17	56.11	63.89	65.83	75.17
aNAP(λ)	MAE	0.23	0.25	0.27	0.22	0.29
	RMSE	0.31	0.35	0.34	0.30	0.38
	r	0.85	0.74	0.73	0.84	0.73
	% Between ±30 to 1:1	74.19	69.58	61.25	75.53	62.03
aCDOM(λ)	MAE	0.19	0.27	0.20	0.17	0.21
	RMSE	0.23	0.31	0.25	0.21	0.27
	r	0.85	0.65	0.83	0.87	0.79
	% Between ±30 to 1:1	71.25	52.50	66.67	73.33	59.56

Comment 4: The distribution of content between different paragraphs should be roughly uniform. The writing style of this study is that some paragraphs are extremely short, while others have a lot of content. It is recommended that the author handle them in a balanced manner.

Response 4: Thank you for highlighting this. We have revised the manuscript to ensure that the content distribution is more balanced. Short paragraphs have been expanded with relevant details, and longer paragraphs have been broken down for clarity. These revisions are reflected throughout the manuscript, particularly on introduction section pages 2, 3 and on discussion section page 15, 16 and 17.

Thank you again for your constructive feedback, which has greatly improved the clarity and quality of our manuscript.

 

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

   This study provides validation results of IOP algorithms and comparison result between IOPs and water quality parameters. The whole structure is quite simple, which limited the novelty of this work. The main object of this research should be clarified. Methods should be specified to do match-up comparison between in situ and remote sensing data. More discussions about the performance of different algorithms could be added. I  think it can't meet the requirement of acceptance for the journal at this stage. 

Comments on the Quality of English Language

Could be improved. 

Author Response

Reviewer 2

Thank you very much for your important comments. We have carefully reviewed each one and have addressed all the points raised in your comments. We believe that, thanks to your suggestions, the paper has been significantly improved. We kindly ask for your review and confirmation.

Comment 1: The whole structure is quite simple, which limited the novelty of this work.

Response 1: Thank you for your insightful comment regarding the simplicity of the structure and its impact on the novelty of this article. We acknowledge the importance of addressing this concern and have taken significant steps to clarify the scientific rigor and originality of our research work.

To better reflect the depth of our research, we have revised the title from “Validation of Inherent Optical Property Algorithms Using GCOM-C/SGLI and Relationships with Multiple Water Quality Parameters in Lake Kasumigaura, Japan” to “Performance Evaluation of Inherent Optical Property Algorithms and Identification of Potential Water Quality Indicators Using GCOM-C Data in Eutrophic Lake Kasumigaura, Japan.” This change emphasizes the dual focus of our work: (1) the systematic evaluation of IOP algorithms and (2) the identification of water quality indicators, which collectively provide insights for IOP-based water quality monitoring in eutrophic lakes.

Furthermore, we have added a detailed study flow chart to visually articulate the methodological framework, illustrating how the integration of IOP validation and remote sensing data analysis enables the identification of water quality indicators. This flowchart not only clarifies the research process but also emphasizes our approach, including the utilization of satellite data for identifying water quality indicators.

To enhance the introduction section, two additional paragraphs were added to emphasize key aspects of the research content: the significance of IOP validation for accuracy, discussed in Page 2, Paragraph 3, and an overview of research efforts in Lake Kasumigaura, presented in Page 3, Paragraph 5.

Introduction

Paragraph 3

The effectiveness of inherent optical properties (IOPs) depends on using well-validated algorithms, particularly in environments like highly eutrophic and optically complex waters [22]. These algorithms come in various forms and are designed to generate a wide range of optical data products, each serving different purposes [10, 23]. Ensuring these IOP algorithms are validated is crucial for generating accurate, reliable, and actionable data, minimizing uncertainty, which supports water quality monitoring, ecosystem management, and the assessment of biogeochemical characteristics [21, 23, 24]. Accurate in situ IOP measurements are essential for ensuring reliable remote sensing validation, which enhances the ability to estimate critical water quality parameters such as chlorophyll-a, suspended solids, and dissolved organic matter. However, optically complex waters often rich in phytoplankton, organic material, and non-algal particles introduce unique challenges for optical measurements[25, 26]. By developing and refining strong algorithms, we can significantly improve remote sensing capabilities. This includes tracking harmful algal blooms, monitoring ecosystem health, and adapting tools to fit a wide range of aquatic environments. Ultimately, these developments offer dependable, high-quality data that facilitate informed decision-making and promote sustainable water resource management [27, 28].

Paragraph 5

Remote sensing has been used to explore the optical properties of water across various regions and a wide range of aquatic environments [5, 26, 32]. This study examines the case of Lake Kasumigaura, the second-largest lake in Japan, which is a vital component of the region's ecosystem, economy, and water resource management system [55]. However, like many freshwater bodies worldwide, it faces significant environmental challenges, including eutrophication, pollution, and water quality deterioration, driven by both anthropogenic activities and natural processes [56]. Furthermore, climatic variations and seasonal hydrodynamics exacerbate water quality issues, complicating management and restoration efforts. Previous research has developed various methods to assess water quality in Lake Kasumigaura, focusing on the application of machine learning and algorithms to predict Chl-a and suspended solid using remote sensing data [16, 57–59]. However, addressing these intricate challenges requires extensive water quality parameters to understand the drivers of water quality changes and to develop effective mitigation strategies for sustainable lake management [60, 61].

We believe these revisions significantly enhance the novelty and scientific contribution of our work, addressing the concerns raised and demonstrating the broader implications of our findings for water quality monitoring and remote sensing applications.

Comment 2: The main object of this research should be clarified.

Response 2: Thank you for your feedback. We have clarified the main objective of the research and included specific objectives to guide the study’s direction. These revisions are presented on page 3, paragraph 6.

The aim of this study is to identify suitable IOP algorithms and key IOP indicators for enhanced water quality monitoring in eutrophic aquatic environments, assuring higher accuracy and minimal errors. Specifically, the study evaluates the performance of five IOP algorithms, each developed with distinct characteristics of optical properties, and to explore the relationship between IOPs and water quality parameters. The research emphasized understanding the light environment in aquatic systems through IOPs and evaluating their potential for estimating water quality parameters. Using measured Rrs and IOP data from Lake Kasumigaura, the accuracy of the five IOP algorithms was validated. The study also examined the precision of atmospheric correction and IOP estimation using GCOM-C satellite data. Lastly, the relationship between the estimated IOPs and water quality metrics was analyzed to assess the viability of IOP-based monitoring for water quality evaluation.

Comment 3: Methods should be specified to do match-up comparison between in situ and remote sensing data.

Response 3: Thank you for your valuable comment regarding the need for clearer specification of methods for match-up comparisons between in situ and remote sensing data. We fully agree that a transparent methodology is critical for ensuring the reliability and reproducibility of our findings.

To address this, we have significantly enhanced the methodological description in the revised manuscript. Specifically, we have introduced a Section 2.3 study design and comprehensive flowchart (Figure 2, located on pages 6) that systematically outlines the research design and procedural steps. This flowchart makes it easy to see the main steps of our study, such as collecting data, evaluating the performance of the IOP algorithms, validating the GCOM-C/SGLI Rrs, and identifying indicators of water quality.

2.3 Study design

The study is structured into two main phases to achieve its objectives. In the first phase, the performance of various IOP algorithms is evaluated to select the best-performing algorithm for each IOP. The second phase focuses on identifying key water quality indicators by analyzing the correlation matrix, with an emphasis on parameters showing strong correlations.

Figure 2 Study flowchart

We believe these revisions significantly strengthen the methodological transparency and scientific rigor of our study, ensuring that readers can fully understand the match-up comparison process.

 

Comment 4: More discussions about the performance of different algorithms could be added.

Response 4: Thank you for your suggestion. We have expanded the discussion section by including a detailed analysis of the performance of five IOP algorithms by comparing IOP algorithms them with each other  and with findings from other researchers. This expanded discussion can be found on pages 15–17, in five paragraphs, section 4.1.

4.1. Performance evaluation of IOP algorithms and R_rs

The performance of the IOP algorithms across IOPs and wavelengths highlights significant differences in accuracy and reliability. The QAA algorithm consistently outperformed others, demonstrating the lowest error metrics and highest correlation coefficients for most parameters. For a(λ), and a_ph(λ)the QAA algorithm achieved the highest accuracy with a low MAE and RMSE, alongside a correlation coefficient (r) of 0.98 and 0.97. This performance is corroborated by previous studies [10, 90, 91] that emphasized QAA’s ability to provide precise absorption estimates. The percentage of retrievals within ±30% of the 1:1 line was 91.81%, further confirming its reliability. By contrast, the GSM algorithm, while achieving a reasonable r value of 0.94, exhibited higher MAE (0.42) and RMSE (0.57), indicating greater variability in its estimations [92]. The PML and LMI algorithms showed weaker performance, with lower r values and higher error metrics. The GIOP algorithm followed with an r value of 0.88 and reasonable error metrics (MAE: 0.09, RMSE: 0.14), supporting its competitive accuracy as noted by Jorge et al. (2021) and Werdell et al. (2013) [38, 91]. The PML and GSM algorithms, with r values of 0.69 and 0.75, respectively, and high MAE and RMSE values, showed limited accuracy for a_ph(λ).

The GSM algorithm exhibited slightly better performance for a_NAP(λ), next to QAA, with the lowest MAE of 0.22 and RMSE of 0.30, achieving an r value of 0.84 and 75.53% of retrievals within ±30% of the 1:1 line. However, the QAA best performance slightly higher error metrics and a comparable r value of 0.85. For a_CDOM(λ), the GSM algorithm achieved the highest accuracy, with an MAE of 0.17, an RMSE of 0.21, and an r value of 0.87, aligning with Betancur-Turizo et al. (2018) and Lewis & Arrigo (2020) [11, 92], who reported GSM’s precision in estimating CDOM absorption. The QAA algorithm also performed well, with an r value of 0.85 and slightly higher error metrics (MAE: 0.19, RMSE: 0.23). The GIOP algorithm had moderate accuracy, with an r value of 0.79, while the PML algorithm showed the weakest performance, with the lowest r value (0.65) and the highest MAE and RMSE values.

The moderate performance exhibited for some IOPs, GIOP algorithm demonstrated competitive performance for a_ph(λ) and a_NAP(λ), while not as consistently accurate as the QAA or GSM algorithms [10, 90, 91]. With an r-value of 0.88 for a_ph(λ), GIOP outperformed both the PML (r = 0.69) and LMI (r = 0.83) algorithms, achieving moderate error metrics (MAE: 0.09, RMSE: 0.14) that indicate its potential for use with fine-tuning [10]. For a_NAP(λ), GIOP’s performance was like LMI and PML, with slightly higher error metrics (MAE: 0.29, RMSE: 0.38) and an r-value of 0.73. For a_CDOM(λ), GIOP achieved moderate results (r = 0.79) but fell short of the GSM algorithm (r = 0.87), which excelled for this parameter. Despite its variability, GIOP showed greater versatility compared to PML and LMI, which generally had weaker correlation coefficients and higher error metrics across most IOPs. These findings suggest that while GIOP is not the most robust algorithm overall, its competitive performance in specific scenarios positions it as a viable alternative, particularly when optimized for specific optical environments [10, 106].

The performance of IOP algorithms is dependent on the underlying equations and IOPs used in their development, which can significantly influence their accuracy and effectiveness [10, 90]. PML and LMI algorithms exhibited weaker performance, as evidenced by their higher error metrics and lower correlation coefficients across most IOPs. For a_CDOM(λ), the PML algorithm showed the highest MAE (0.27) and RMSE (0.31) among all algorithms, with a significantly lower r-value of 0.65, indicating its limited accuracy in retrieving a_CDOM absorption. Similarly, for aph(λ), PML's performance was poor, with an MAE of 0.15, an RMSE of 0.24, and an r-value of 0.69, reflecting its inability to handle phytoplankton absorption accurately. The LMI algorithm demonstrated similarly inconsistent performance, with moderate error metrics for a_CDOM(λ) (MAE: 0.20, RMSE: 0.25) and a_ph(λ) (MAE: 0.11, RMSE: 0.17), but lower r-values of 0.83 and 0.73 showed moderate and consistent [106]. These weaknesses underline the limitations of PML and LMI in accurately retrieving IOPs compared to the more robust QAA and GSM algorithms.

For shorter wavelengths, the error metrics were higher for algorithms such as the GSM algorithm, indicating potential issues with signal attenuation or noise [90]. Despite the variability observed for the GSM algorithm at lower wavelengths, there are notable similarities between the results of the GIOP and GSM algorithms, both of which utilize the Levenberg-Marquardt algorithm for solving unconstrained nonlinear least-squares problems [11]. The QAA and the GSM algorithm have the best accuracy, with consistent low error metrics and high correlation coefficients across various conditions[91]. The GIOP algorithm has potential but requires further tuning and optimization to improve its accuracy and reduce its variability[10]. Overall, the QAA is the most reliable for a(λ), a_ph(λ), and a_NAP(λ), while the GSM algorithm offers strong performance in specific cases, particularly for a_CDOM(λ).

Thank you again for your constructive feedback, which has greatly improved the clarity and quality of our manuscript.

Author Response File: Author Response.docx