Optimizing Optical Coastal Remote-Sensing Products: Recommendations for Regional Algorithm Calibration

Round 1

Reviewer 1 Report

The Author evaluated different combinations of satellites, atmospheric corrections, algorithms, and bands to estimate turbidity and SPM in Patos Lagoon (Brazil). And the author used the GeoCalVal method to perform regional parameter correction on single-band and multi-band algorithms and innovatively proposed a new evaluation index (GoF). This new metric favors both the best overall performance and the most balanced performance. The conclusion is consistent with the evidence and arguments presented. this study fits within the scope of this journal and will be of interest to its readers. I recommend this work for publication after some revisions:

1.      In Section 2.2.2, The author mentioned three remote sensing satellites (L8/30m, S2/20m and S3/300m). The spatial resolution of these satellites is quite different. Is the influence of this factor considered in the in-situ verification of turbidity and SPM?.

2.      In Section 2.3.1, The author mentioned two atmospheric correction algorithms, PLOYMER and ACOLITE. Have other atmospheric correction algorithms (e.g. CR2CC) been considered and, if so, please provide details.

3.      In Section 2.3.2, the authors mentioned the matching of in-situ data and satellite data. What is the amount of matching data? Can it effectively support the research work of the manuscript?

4.      In Figure 6 (C, F), the SPM estimation errors of the N10 (865), N17 and T20 algorithms of the S3 satellite are large. Please try to analyze the impact on SPM estimation accuracy due to different atmospheric correction algorithms.

5.      In Section 4.2, The Author mentioned “adjacency effects could cause overestimation of the SPM concentration.”. The author can discuss in detail the impact of advertising effects.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

This study examines the performances of multiple correction algorithms on the turbidity and suspended particulate matter data provided by various satellites in the Patos Lagoon region in Brazil. This study is limited by observational data, and there is a lack of basis for comparing the effectiveness of different algorithms for the entire lagoon region. The available data only supports comparisons between areas in the south and north and has limited reference value for the vast sea area. Moreover, the results do not provide new knowledge in selecting appropriate correction algorithms. Therefore, I can not support the publication of this paper. The authors should reconsider the focus and contribution of this study before completely rewriting the entire manuscript. Here are my specific comments or suggestions.

(1) The in-situ observation data used in this study is limited and only concentrated in the south and north regions, with no observations conducted within the lagoon. It should be noted that the parameter selection of the calibration algorithm and the corresponding calibration effect are closely related to the characteristics of the geographical environment. Since the observation data is mainly concentrated in two areas, it is recommended that the research focus on comparing the effectiveness of the correction algorithms in these two areas. The focus now is on the applicability of different correction algorithm combinations to multiple satellites, which requires a more sufficient amount of data with a more uniform distribution in time and space. Otherwise, there will be insufficient persuasiveness. At present, the data in this work is only suitable for comparing different regions and is obviously weak for comparing different algorithms.

(2) From Figure 11, it can be seen that this study is analyzing the entire Patos Lagoon estuary, so using data from the southern and northern sea areas alone is far from enough. The caption of Figure 11 mentions the need to achieve regional recalibration, but the current data can only achieve local recalibration. The south, north, and central lagoon situations may vary significantly. These differences may be related to geopolitical factors, including water depth, ocean currents, waves, biological species, suspended material composition, etc. If there is no difference, please provide sufficient proof. If there are indeed significant differences, it is recommended that the research focus on comparing the differences between the southern and northern sea areas.

(3) At the beginning of this paper, it is pointed out that there is "no method to prepare for all satellites (Line 12)" regarding the correction of turbidity and SPM. However, from the results and conclusions, this work has not addressed this issue or provided significant progress in addressing it. The results suggest "in situ radiometric data to better evaluate atmospheric corrections (Line 24)"; however, such a suggestion does not require research to understand that it should be done. "The specificity of remote sensing estimates (Line 611 in the Conclusion)" is a well-known issue. This does also not require research to prove. This study should provide more specific and sound conclusions, such as what algorithm is best for the northern region and what algorithm is best for the southern region. Or what algorithm is best when the water depth is greater than or less than which levels. What algorithm is best regarding how far offshore it is, and so on. Such results would effectively enhance people's understanding and improve the technical level of calibration turbidity and SPM satellite remote sensing data. Clear answers to any of the abovementioned or similar questions would be sufficient progress.

(4) This study proposes an indicator (GoF) for evaluating the quality of satellite data. I have many questions: what is the specific calculation formula for this indicator? What is the mathematical and physical meaning of this indicator? How was it derived? Why is it more effective than indicators such as MAE, MAPE, RMSE, COR, etc? If it is an existing indicator, it is necessary to introduce its origin and characteristics, or if it is a newly proposed indicator, then it is essential to clearly explain how this indicator was constructed and how its formula was derived.

(5) The term "Coastal Remote Sensing Products" in the title refers explicitly to "the turbulent and suspended particulate matter products," and this information cannot be omitted from the title. Your method may not be applicable to other variables, such as waves, currents, temperature, chlorophyll, etc., in the coastal regions.

(6) The title should also include the region information, as the conclusions obtained from this study are probably not applicable to other regions.

(7) What is the GeoCalVal method in Line 18? The Abstract should avoid abbreviations or provide clear explanations

(8) In the Conclusion section, the first sentence should clearly state that "algorithms" refers to "bio-optical algorithms" unless you have only done atmospheric calibration.

(9) There are too many points of conclusions, with unclear emphasis. You should condense them into two or three points.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

The authors present in form of thorough review the performance of various algorithms employed to retrieve the turbidity and the SPM from observations collected by different sensors. In this point of view, it is an interesting contribution to the remote sensing community interested in these specific topics.

The manuscript would have benefited of a deeper physical discussion based on radiative transfer models, as reported in the conclusions. Nevertheless, the huge effort to inspect the optimization of the SPM and turbidity retrieval deserves to be published.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

The paper has made significant improvements compared to the previous version. However, I will only support the publication of the paper if the following issue is well addressed:

Considering that the “Center” part of the data (Figure A7) is quite essential, it should be treated equally to the south and north parts. It is necessary to include the data of the “Center” part and reperform all relative analyses in the paper. Also, Figure A7 should be included in the main text. All sites of the “Center” panel of Figure A7 should be indicated in Figure 1. If the conclusion changes due to adding the “Center” data, it must be revised accordingly.

Author Response

Dear Reviewer,

            We are grateful for the suggestions. We moved Figure A7 to the main text and added a map of the data distribution (Figure 12). The authors acknowledge that data from the central part of the lagoon are missing in the paper and that this data is relevant to the analysis.  However, the data sources used here do not provide any matchups in this portion of the lagoon. For turbidity, the moored buoys are restricted to the estuary and, thus, limited to the south part. For SPM, the compendium from Távora et al. (2021) gathers measurements from multiple projects in the region and has some data points in the "Center", as shown in Figure 12. However, all these measurements were made before 2013 and, therefore, before the satellites were launched (2013 for L8, 2015 for S2A, and 2016 for S3A). To clarify this, we added this information in rows 663 and 664.

Sincerely Yours,

Rafael Simão, Juliana Távora, Mhd. Suhyb Salama and Elisa Fernandes.