Shallow Water Bathymetry Based on Inherent Optical Properties Using High Spatial Resolution Multispectral Imagery

Round 1

Reviewer 1 Report

This is high-quality research, well done! You provide a novel method for satellite derived bathymetry and a comprehensive error analysis. Extensive analysis and appropriate figures strongly support the research findings. Grammar should be improved throughout the manuscript as English writing detract from the research at times.

 

A few questions, comments, and suggestions:

 

-        Did you use near IR band to mask out land from analysis?

-        Why are there negative depths in figures 6 and 7?

-        How do you explain differences in accuracy between the two study sites?

o   “The error of Saipan Island in the depth of 0 - 5 m is larger than that in the Dongdao Island, which may be due to the undulating terrain and irregular topography of the Saipan Island”

• Is this also due to differences in sediment type?

-        Error analysis is computed by using kriging interpolation of the verification points as the “truth”. However, the accuracy of interpolation itself is not considered, and it should at least be mentioned. The accuracy of interpolation methods typically depends on the measurement sampling density and distance to measurements, as well as the terrain slope and curvature (Aguilar et al 2005, Amante and Eakins 2016, Erdogan 2009, Erdogan 2010, Guo et al 2010). Also, the verification points themselves also contain error (Amante 2018, Calder 2006, Jakobsson, Calder, and Mayer 2002)

-        Suggested citations and references therein:

o   Aguilar, F.J.; Agüera, F.; Aguilar, M.A., and Carvajal, F., 2005. Effects of terrain morphology, sampling density, and interpolation methods on grid DEM accuracy.

o   Amante and Eakins 2016. Accuracy of Interpolated Bathymetry in Digital Elevation Models.

o   Amante 2018. Estimating Coastal Digital Elevation Model Uncertainty

o   Calder, B., 2006. On the uncertainty of archive hydrographic data sets.

o   Erdogan, S., 2009. A comparison of interpolation methods for producing digital elevation models at the field scale.

o   Erdogan, S., 2010. Modelling the spatial distribution of DEM error with geographically weighted regression: An experimental study

o   Guo, Q.; Li, W.; Yu, H., and Alvarez, O., 2010. Effects of topographic variability and lidar sampling density on several DEM interpolation methods.

o   Jakobsson, M.; Calder, B., and Mayer, L., 2002. On the effect of random errors in gridded bathymetric compilations.

-        Why is 3m used as a cutoff between small and large errors in Figure 8?

 

Suggested citations and additional references therein:

 

Aguilar, F.J.; Agüera, F.; Aguilar, M.A., and Carvajal, F., 2005. Effects of terrain morphology, sampling density, and interpolation methods on grid DEM accuracy. Photogrammetric Engineering and Remote Sensing, 71(7), 805–816.

Amante, C.J., 2018. Estimating Coastal Digital Elevation Model Uncertainty. Journal of Coastal Research, 34(6), 1382 – 1397. Coconut Creek (Florida), ISSN 0749-0208. 10.2112/JCOASTRES-D-17-00211.1

Amante, C.J. and Eakins, B.W., 2016. Accuracy of Interpolated Bathymetry in Digital Elevation Models. In: Brock, J.C., Parrish, C.E., Gesch, D., Wright, C.W., and Rogers, J. (eds.), Advances in Topobathymetric Mapping, Models, and Applications. Journal of Coastal Research, Special Issue, No. 76. 10.2112/SI76-011.

Calder, B., 2006. On the uncertainty of archive hydrographic data sets. IEEE Journal of Oceanic Engineering, 31(2), 249–265.

Erdogan, S., 2009. A comparison of interpolation methods for producing digital elevation models at the field scale. Earth Surface Processes and Landforms, 34(3), 366–376.

Erdogan, S., 2010. Modelling the spatial distribution of DEM error with geographically weighted regression: An experimental study. Computers & Geosciences. 36(1), 34–43.

Guo, Q.; Li, W.; Yu, H., and Alvarez, O., 2010. Effects of topographic variability and lidar sampling density on several DEM interpolation methods. Photogrammetric Engineering and Remote Sensing, 76(6), 701–712.

Jakobsson, M.; Calder, B., and Mayer, L., 2002. On the effect of random errors in gridded bathymetric compilations. Journal of Geophysical Research: Solid Earth, 107(B12), 439–443.

Author Response

Dear Editor and Reviewer:

We would like to thank Remote Sensing for giving us the opportunity to revise our manuscript. Thanks for the reviewer’s comments concerning our manuscript entitled “Shallow water bathymetry based on inherent optical properties using high spatial resolution multispectral imagery” (Manuscript ID: remotesensing-896671).

Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied comments carefully and have made correction which we hope meet with approval. The main corrections in the paper and the responds to the reviewer’s comments are as follows, and all revisions are presented by using the "Track Changes" function in the revised manuscript.

Responds to the reviewer’s questions, comments and suggestions:

1. “Did you use near IR band to mask out land from analysis?”

Response:

We are very sorry for our negligence of the introduction of mask method. Since the remote sensing images include some waves along the coast of the island, we use visual interpretation to mask out lands and waves, which can make the results of mask more accurate.

And we have also added the description in section 2.2.1, please see Line 124-126 of the revised manuscript: ‘After the abovementioned processing, we also use visual interpretation methods to mask the land and sea wave areas to obtain the final result.’

2. “Why are there negative depths in figures 6 and 7?”

Response:

Considering the Reviewer’s question of the negative depths in Figure 6 and Figure 7, we explained and added the reasons in the section of 4. Discussion, compared and analyzed the same problem with other articles. The main reasons are as follows:

Optical remote sensing of bathymetry is based on the correlation between brightness or reflectance values and water depth, so the values have a considerable influence on water depth. And because of the different reflectance of the sediment, the situation of positive or negative water depth values will also be caused. This result is similar to the conclusions of previous researchers. In the very shallow offshore waters, a negative water depth may be estimated.

Please see Line 446-463 of the revised manuscript for details:

‘In addition, for the coastal shallow water areas of the two islands, the scatter plots both show some negative water depth values (see Fig .6,7; Fig. 15,16). Optical remote sensing of bathymetry…’

3. “How do you explain differences in accuracy between the two study sites? Is this also due to differences in sediment type?”

Response:

Different types of sediment also affect the accuracy of water depth retrieval. In the Saipan Island, there are many coral reefs in the shallow waters, and sediment of the medium water depths are most coral sand, while the Dongdao Island is the opposite. At the same time, the sediment types of Saipan Island intersect, while those of Dongdao Island are relatively uniform. Last, because the actual water depth of Dongdao is measured by single beam echosounder, the accuracy of it is higher than the nautical chart of Saipan. This may also be one of the reasons for the large error in Saipan. We are sorry for not mentioning the accuracy of the actual water depth, it has been added in section 2.1 The Study areas and Data sets.

Please see Line 103-109 of the revised manuscript:

‘The actual water depth data of Dongdao Island were obtained by a single beam echosounder in 2011, and the depth error and position error are 0.01 m and 1 m. The water depth data of Saipan Island is from the electronic nautical chart data measured in 2013, it meets the standard of Category of Zone of Confidence (CATZOC) proposed by the International Hydrographic Organization (IHO). The data in the nearshore areas meet the B standard, that is, between 0 m and 10 m water depth, the depth uncertainty and position uncertainty are less than 1.2 m and 20 m. The deep areas meet the A2 standard, the depth and position uncertainty are less than 1.6 m and 50 m between 10 m and 30 m water depth.’

4. “Error analysis is computed by using kriging interpolation of the verification points as the “truth”. However, the accuracy of interpolation itself is not considered, and it should at least be mentioned. The accuracy of interpolation methods typically depends on the measurement sampling density and distance to measurements, as well as the terrain slope and curvature (Aguilar et al 2005, Amante and Eakins 2016, Erdogan 2009, Erdogan 2010, Guo et al 2010). Also, the verification points themselves also contain error (Amante 2018, Calder 2006, Jakobsson, Calder, and Mayer 2002)”

Response:

It is really true as Reviewer suggested that the accuracy of interpolation itself should be mentioned. The standard deviation of the Kriging interpolation results of the two study areas is within 2m. Although there are few verification points in our study areas, there will be some certain errors. However, the spatial distribution of the selected points and their distribution in different water depth sections are uniform, so the interpolation results can be approximately regarded as the real.

We added the introduction of Kriging method and the explanation of its accuracy in the section 2.4. Accuracy evaluation methods. And we also added the citations provided by the reviewer. Please see Line 178-200 of the revised manuscript for details: ‘Besides, we use the Kriging interpolation method to interpolate the verification points to obtain the overall results of the study areas…’

5. “Why is 3m used as a cutoff between small and large errors in Figure 8?”

Response:

We have chosen other smaller values as the cutoff, but there will be some very small patches at this scale in the figures. These patches make the difference between the three models not obvious, which is not conducive to judging which model has better accuracy. After many attempts, 3m is determined as the cutoff between the small and large errors, which can clearly show the range of error. And make the information in the figures more obvious. The selection of the cutoff of the Saipan Island is the same, due to its error is greater than that of the Dongdao Island, so 4m is selected as the cutoff.

Besides, as our manuscript needed English editing, we have reviewed our manuscript and sent it for language revision by a professional manuscript editing company. For the convenience of reading, they are not marked in the manuscript.

We would like to express our great appreciation to the reviewer for comments on our paper. We appreciate for Editor’s and Reviewer’s warm work earnestly, and hope that the correction will meet with approval.

Thank you and best regards.

Yours sincerely,

Xuechun Zhang

E-mail: [email protected];

Corresponding author:

Name: Yi Ma

E-mail: [email protected];

Author Response File: Author Response.docx

Reviewer 2 Report

General comments

The manuscript entitled “Shallow water bathymetry based on inherent optical properties using high spatial resolution multispectral imagery” presents a novel “empirical” method for derived satellite bathymetry. In my opinion, the method described is a combination of previous methodologies but not described before in that way. The method,  named “Inherent Optical Parameters Linear Model (IOPLM) was tested in Worldview-2 images, and it was compared with Stumpf´s model and the log-linear model showing lower MAE, MRE and RMSE. In my opinion, the manuscript presents scientific interest and quality for being published in Remote Sensing. However, authors should modify some aspects of the paper before being published.

-Some elements of the methodology need to be explained in more detail. For example, the selection of the coefficients in section 2.31 or the kriging method in section 3.2.

-Methods and Results are mixed in the paper. Some methodological processes appear by the first time in the results or the discussion sections. Authors should explain all the methodological steps in the methodology section and their results in the results section

- Results should be summarized. I find some paragraphs redundant. For example, the comparison at 5 m interval and the 1 m interval. I think these aspects could be summarized or combine, reducing the number of tables, numbers and plots.

-The discussion section should be improved. Currently, authors show part of their results here, but this section should include the interpretation and significance of the findings presented in the manuscript and the comparison with other studies.

-English should be also reviewed

 

Specific comments

Line 36-38. These lines are vague for me. I suggest eliminating them as this information is developed in more detail in the paragraph.

Line 68-69 I suggest moving this sentence to line 72 after the dot. “… information. The blue and green…”

Line 104. Please add information about the error (uncertainty)  of these measurements (single beam and electronic nautical chart)

Line 128 Why authors refer here to the North Island as this area is not included as a study area? Please explain

Line 144 the selection of these coefficients needs more explanation. What are they representing? The value of 0.0895 comes from Gordon et al  1988 and the value of 0.1247 comes from Lee et al 1999. Please reference them correctly.

The method developed by Lee et al 2002 was developed for optically deep waters, however your study areas have rising corals (e.g. Saipan). Please explain here if your study areas meet the requirements for applying the model of Lee et al.

1. R. Gordon, O. B. Brown, R. H. Evans, J. W. Brown, R. C. Smith, K. S. Baker, and D. K. Clark, (1988). “A semi-analytic radiance model of ocean color,” J. Geophys. Res. 93, 10909–10924
2. P. Lee, K. L. Carder, C. D. Mobley, R. G. Steward, and J. S. Patch, (1999) “Hyperspectral remote sensing for shallow waters. 2. Deriving bottom depths and water properties by optimization,” Appl. Opt. 38, 3831–3843

Line 175-176 I don´t see these values in Table 1

Line 176 For Dongao island R2 of Stumpf´s model is very similar to IOPLM (0.95 vs 0.96)

Line 180 For Saipan Island again R2 of Stumpf´s model is very similar to IOPLM (0.94 vs 0.93)

Line 182 Please include R2 values in Table 1 if you are referring then in the text

Line 222 I suggest to eliminate table 2 and include these values in the scatterplots for more visual interpretation

Line 223-224 This sentence is difficult to read. Please rewrite. Possible option:  “The scatterplots of the estimated water depth and the actual water depth,  resulting from the different models for both study areas are drawn in Fig.6 and fig. 7” .

Line 251 Based on what parameters have you chosen this limit?

Line 249-263. Please explain in more detail how kriging has been calculated. Which type of kriging has been used?  Kriging, as a model also has some errors associated. For this reason,  please describe the accuracy of kriging model for validation and comparison with your satellite-derived bathymetry models. I suggest including a section for the description of the kriging model in the methodology.

Line 262-263 Taking into account that kriging is a model  I think is quite risky to state that IOPLM is superior to the Stumpf´s model and the log-linear model. I would say that IOPLM is more similar or coincident with the kriging results.

Figure 10 and 11. Please mark the beginning of the transects in the picture. Thus the reader can differentiate the beginning and the end of the transect for correct interpretation on the plot.

Line 341  In my opinion, the description of the SVM for mapping sediment types should be included together with a more exhaustive description and its main purpose in the methodology. Their results should be included in the results section.

Lines 340-402. The discussion section should be rewritten. Currently, it is quite “weak” and authors present here parts that should be included in methodology and results. In the discussion, authors should explain the significance of their results and compare these results with other studies.  All these aspects are currently missed in the discussion, and this section is often considered the most crucial part of a research paper. I give the authors here some suggestions up to their discretion:

For example, it would be interesting the comparison of the results obtained in this manuscript with the ones obtained for Sentinel-2, with 10 m spatial resolution but data free of charge for users.

The authors reported some negative predicted values in shallow waters. Why are these negative values appearing? Other studies such as  Vahtmae and Kutser, 2016 or Casal et al., 2019 also found these negative values using similar methods. Authors could compare their results with these studies and discuss if the causes are the same.

• Vahtmäe, E., Kutser, T (2016). Airborne Mapping of Shallow Water Bathymetry in the Optically Complex Waters of the Baltic Sea. Journal of Applied Remote Sensing 10 (2): 025012.
• Casal, G., Monteys, X., Hedley, J., Harris, P., Cahalane, C., McCarthy, T. (2019). Assessment of empirical algorithms for bathymetry extraction using Sentinel-2 data. International Journal of Remote Sensing, 40 (8): 2855-2879; DOI: 10.1080/01431161.2018.1533660

 

Other studies have used Worldview-2 for deriving satellite bathymetry. Are the results similar? Is IOPLM improving the outcomes in comparison with these studies?

• Cao, B., Fang, Y. Jiang, Z., Gao, L., Hu, H. (2019). Shallow water bathymetry from WorldView-2 stereo imagery using two-media photogrammetry. European Journal of Remote Sensing, 52:1, 506-521.
• Hamylton, S.M., Hedley, J Beaman, R.J. (2015). Derivation of High-Resolution Bathymetry from Multispectral Satellite Imagery: A Comparison of Empirical and Optimisation Methods through Geographical Error Analysis. Remote sensing 7:16257-16273.

How are the water column constituents affecting the satellite-derived bathymetry? Are waters clear in both study areas?

Line 404-406 Too long sentence

Author Response

Dear Editor and Reviewer:

We would like to thank Remote Sensing for giving us the opportunity to revise our manuscript. Thanks for the reviewer’s comments concerning our manuscript entitled “Shallow water bathymetry based on inherent optical properties using high spatial resolution multispectral imagery” (Manuscript ID: remotesensing-896671).

Those comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches. We have studied comments carefully and have made correction which we hope meet with approval. The main corrections in the paper and the responds to the reviewer’s comments are as follows, and all revisions are presented by using the "Track Changes" function in the revised manuscript.

Responds to the reviewer’s general comments:

1. “Some elements of the methodology need to be explained in more detail. For example, the selection of the coefficients in section 2.31 or the kriging method in section 3.2.”

Response:

We are very sorry for our negligence of some details of the methodology. In response to these problems, we have re-written these parts and make a supplement in section 2.3.1 and section 2.4 Accuracy evaluation methods. Please see No.5 and No.12 of the Responds to the reviewer’s Specific comments for details.

2. “Methods and Results are mixed in the paper. Some methodological processes appear by the first time in the results or the discussion sections. Authors should explain all the methodological steps in the methodology section and their results in the results section.”

Response:

It is really true as reviewer suggested that we should explain all the methodological steps in the methodology section and the results in the results section. We have re-written these parts according to the reviewer’s suggestion.

In section 2.4. Accuracy evaluation methods, we have added the introduction of Kriging interpolation methods. And we also added a section to introduce the principle of SVM method in 2.5 Sediment classification method. And we move the result of classification of the different sediment types in 3. Result, the name of this new section is 3.2.4. Comparison of accuracy in different sediment types.

3. “Results should be summarized. I find some paragraphs redundant. For example, the comparison at 5 m interval and the 1m interval. I think these aspects could be summarized or combine, reducing the number of tables, numbers and plots.”

Response:

Considering the reviewer’s suggestion, we have deleted the redundant paragraphs and tables. And left the brief results of the comparison at 1m interval. we have re-written the section of 3.2.3. Comparison of accuracy in different water depths, please see Line 352-381 of the revised manuscript for details.

4. “The discussion section should be improved. Currently, authors show part of their results here, but this section should include the interpretation and significance of the findings presented in the manuscript and the comparison with other studies.”

Response:

We are sorry for not understanding the purpose of the Discussion section correctly. We have re-written the 4. Discussion part according to the Reviewer’s suggestion. Please see No.16 of the Responds to the reviewer’s Specific comments for details.

5. “English should be also reviewed.”

Response:

We are very sorry for our incorrect writing of some words and sentences since English is not our first language. We have reviewed our manuscript and sent it for language revision by a professional manuscript editing company. For the convenience of reading, they are not marked in the manuscript.

 

Responds to the reviewer’s Specific comments:

1. “Line 36-38. These lines are vague for me. I suggest eliminating them as this information is developed in more detail in the paragraph.”

Response:

This problem has been modified in the revised manuscript. Line 36-38 was deleted.

2. “Line 68-69 I suggest moving this sentence to line 72 after the dot. ‘… information. The blue and green…’”

Response:

This problem has been modified in the revised manuscript. Line 68-69 were corrected as: ‘…water depth information. The blue and green bands of the multispectral image…’

3. “Line 104. Please add information about the error (uncertainty) of these measurements (single beam and electronic nautical chart).”

Response:

We are very sorry for our negligence of the introduction of the error (uncertainty) of single beam echosounder and electronic nautical chart. it has been added in section 2.1 The Study areas and Data sets. Please see Line 103-109 of the revised manuscript:

‘The actual water depth data of Dongdao Island were obtained by a single beam echosounder in 2011, and the depth error and position error are 0.01 m and 1 m. The water depth data of Saipan Island are from the electronic nautical chart data measured in 2013, they meet the standard of Category of Zone of Confidence (CATZOC) proposed by the International Hydrographic Organization (IHO). The data in the nearshore areas meet the B standard, that is, between 0 m and 10 m water depth, the depth uncertainty and position uncertainty are less than 1.2 m and 20 m. The deep areas meet the A2 standard, the depth and position uncertainty are less than 1.6 m and 50 m between 10 m and 30 m water depth.’

4. “Line 128 Why authors refer here to the North Island as this area is not included as a study area? Please explain.”

Response:

We are very sorry for our negligence of the explanation of why the North Island was not included as a study area. Since the North Island is too small, and the northeast part of the island is basically covered by waves, the bathymetric mapping result and accuracy evaluation of the North Island area are not performed in the following paper. It is only used as an example to show the IOPLM under different sediment types and water depths. We also added the explanation in 2.3.1. Inherent optical parameters linear model, please see Line 158-161 in the manuscript for details.

5. “Line 144 the selection of these coefficients needs more explanation. What are they representing? The value of 0.0895 comes from Gordon et al. 1988 and the value of 0.1247 comes from Lee et al 1999. Please reference them correctly. The method developed by Lee et al 2002 was developed for optically deep waters, however your study areas have rising corals (e.g. Saipan). Please explain here if your study areas meet the requirements for applying the model of Lee et al.”

Response:

We have made correction according to the reviewer’s comments. Please see Line 151-155 of the revised manuscript:

‘Here and are model constants that change with a variety of water optical properties, the values of them may vary with particle phase function, which is not known remotely. For oceanic case-1 water, ; 4 [22]. For coastal water with high scattering, and is more suitable [23]. In order to be applied to both coastal and open water bodies, here the averaged values taken by Lee [24] is used, and .’

The method developed by Lee et al. 2002 was developed for optically deep waters, but the value of the parameters can both be applied to the coastal and open water bodies.

Related references have also been corrected as follows:

[22] Gordon H. R.; Brown O. B.; Evans R. H. A semianalytic radiance model of ocean color. J. Geophy. Res. 1988, 10909-10924.

[23] Lee Z.; Carder K. L.; Mobley C.D.; Steward R. G.; Patch J. S. Hyperspectral remote sensing for shallow waters. 2. Deriving bottom depths and water properties by optimization. Appl. Optics 1999, 38, 3831-3843.

[24] Lee Z.; Carder K. L.; Arnone R. A. Deriving inherent optical properties from water color: a multiband quasi-analytical algorithm for optically deep waters. Appl. Optics 2002, 41, 5755-5772.

6. “Line 175-176 I don´t see these values in Table 1”

Response:

We are very sorry for our negligence of R² in Table 1. These values are added to Table 1. Please see Table 1 in the revised manuscript:

Study area	Model	Parameters			R2
Dongdao Island	IOPLM	= 31.734	= -30.729		0.963
	Log-linear	= 20.844	= -23.051	= 15.233	0.92
	Stumpf	= 83.69	= -82.869		0.95
Saipan Island	IOPLM	= 25.898	= -20.507		0.94
	Log-linear	= 16.426	= -17.478	= 12.605	0.89
	Stumpf	= 64.093	= -58.499		0.93

7. “Line 176 For Dongao island R2 of Stumpf´s model is very similar to IOPLM (0.95 vs 0.96).”

“Line 180 For Saipan Island again R2 of Stumpf´s model is very similar to IOPLM (0.94 vs 0.93)”

Response:

Due to the similar forms of the two models, they are all ratio models. Therefore, although the difference of R² values is not large, IOPLM shows good performance in the following accuracy evaluations. Moreover, the results of R² are quite excellent, it is difficult to continue to improve.

8. “Line 182 Please include R² values in Table 1 if you are referring then in the text.”

Response:

We are very sorry for our negligence of R² in Table 1. These values have already added to Table 1.

9. “Line 222 I suggest to eliminate table 2 and include these values in the scatterplots for more visual interpretation.”

Response:

Considering the reviewer’s suggestion, we have deleted Table 2, and added the results into scatterplots in Figure 6 and Figure 7. Please see Figure 6 and Figure 7 in the revised manuscript.

10. “Line 223-224 This sentence is difficult to read. Please rewrite. Possible option: ‘The scatterplots of the estimated water depth and the actual water depth, resulting from the different models for both study areas are drawn in Fig.6 and fig. 7’”.

Response:

We are very sorry for our incorrect writing. We have corrected it in to ‘The scatterplots of the estimated water depth and the actual water depth, resulting from the different models for both study areas are drawn in Figure 6 and Figure 7’. Please see Line 275-276 in the revised manuscript.

11. “Line 251 Based on what parameters have you chosen this limit?”

Response:

We have chosen other smaller values as the cutoff, but there will be some very small patches at this scale in the figures. These patches make the difference between the three models not obvious, which is not conducive to judging which model has better accuracy. After many attempts, 3m is determined as the cutoff between the small and large errors, which can clearly show the range of error. And make the information in the figures more obvious. The selection of the cutoff of Saipan Island is the same, due to its error is greater than that of Dongdao Island, so 4m is selected as the cutoff.

12. “Line 249-263. Please explain in more detail how kriging has been calculated. Which type of kriging has been used? Kriging, as a model also has some errors associated. For this reason, please describe the accuracy of kriging model for validation and comparison with your satellite-derived bathymetry models. I suggest including a section for the description of the kriging model in the methodology.”

Response:

It is really true as reviewer suggested that we should add the description of the Kriging model in the methodology and describe the accuracy of it. In section 2.4. Accuracy evaluation methods, we added some introductions:

‘Besides, we use the Kriging interpolation method to interpolate the verification points to obtain the overall results of the study areas. And the results will be used as the approximate “true” water depth results to compare with the retrieval results. Kriging is also called spatial local interpolation. It is an advanced geostatistical process…

…The accuracy of interpolation typically depends on the measurement sampling density and the distance to the measurements, as well as the terrain slope and curvature [28-32]. The standard deviation of the Kriging interpolation results of the two study areas is within 2 m. Although there are few verification points in our study areas, there will be some certain errors. However, the spatial distribution of the selected points and their distribution in different water depth sections are uniform, so the interpolation results can be approximately regarded as the actual results.’

Please see Line 177-199 in the revised manuscript for details.

13. “Line 262-263 Taking into account that kriging is a model. I think is quite risky to state that IOPLM is superior to the Stumpf´s model and the log-linear model. I would say that IOPLM is more similar or coincident with the kriging results.”

Response:

We have made correction according to the Reviewer’s comments. The sentences are corrected to: “According to the above results, the overall retrieval results in the two study areas of the IOPLM are more coincident with the Kriging interpolation result than that of the Stumpf’s model and the log-linear model.”

Please see Line 314-316 in the revised manuscript.

14. “Figure 10 and 11. Please mark the beginning of the transects in the picture. Thus the reader can differentiate the beginning and the end of the transect for correct interpretation on the plot.”

Response:

We have modified the figures and marked the start and end points of the transects, please see Figure 10. and Figure 11. in the revised manuscript.

15. “Line 341 In my opinion, the description of the SVM for mapping sediment types should be included together with a more exhaustive description and its main purpose in the methodology. Their results should be included in the results section.”

Response:

We are very sorry for our negligence of the description of the SVM method. We have added a section to introduce the principle of SVM method in 2.5 Sediment classification method. And we move the result of classification of the different sediment types in 3. Result, the name of this new section is 3.2.4. Comparison of accuracy in different sediment types.

Please see Line 200-213 and Line 382-429 in the revised manuscript for details.

16. “Lines 340-402. The discussion section should be rewritten. Currently, it is quite “weak” and authors present here parts that should be included in methodology and results. In the discussion, authors should explain the significance of their results and compare these results with other studies. All these aspects are currently missed in the discussion, and this section is often considered the most crucial part of a research paper. I give the authors here some suggestions up to their discretion:

For example, it would be interesting the comparison of the results obtained in this manuscript with the ones obtained for Sentinel-2, with 10 m spatial resolution but data free of charge for users.

The authors reported some negative predicted values in shallow waters. Why are these negative values appearing? Other studies such as Vahtmae and Kutser, 2016 or Casal et al., 2019 also found these negative values using similar methods. Authors could compare their results with these studies and discuss if the causes are the same.

Other studies have used Worldview-2 for deriving satellite bathymetry. Are the results similar? Is IOPLM improving the outcomes in comparison with these studies?”

Response:

We are sorry for not understanding the purpose of the Discussion section correctly. We have re-written the 4. Discussion part according to the Reviewer’s suggestion.

We added the comparison with the result of IOPLM obtained by Sentinel-2A and explain the difference between WorldView-2 and Sentinel-2A:

‘The results show that the accuracy of Sentinel-2A is not as good as WorldView-2 in both study areas (As the study area of Saipan Island contains a small amount of clouds, for the sake of unification, there are some differences in the selection of verification points. Therefore, the WorldView-2 results have changed.). Figure 19 shows the accuracy at different water depth sections (interval of 5 m). Similarly, the Sentinel-2A satellite has a weak retrieval capability. The reasons mainly include that (1) Sentinel-2A has a lower spatial resolution, the resolution of 10 m may not show areas with relatively drastic water depths and topographical fluctuations. And the proposed model may be more suitable for satellite with higher resolution. Also, (2) The image quality is not good; there are a few solar flares on the sea surface.’

We added the reasons for negative water depth values in the Discussion section:

‘In addition, for the coastal shallow water areas of the two islands, the scatter plots both show some negative water depth values (see Fig. 6,7; Fig. 15,16). Optical remote sensing of bathymetry is based on the correlation between brightness or reflectance values and water depths, so the values have a considerable influence on water depth. In addition, because of the different reflectance of the sediments, the situation of positive or negative water depth value situation will also occur. This result is similar to the conclusions of previous researchers [37-38]. In the very shallow offshore waters, a negative water depth may be estimated. In our study, there are few points with negative water depth estimates for Dongdao Island, and they are basically distributed within 0 - 2 m. After the classification, it was found that the sediment type is coral reef or coral sand, and the area with unclear sediment type has no negative value. There are relatively more negative water depths points in Saipan Island, roughly between the depths of 0 – 3 m. It was found that most of them are located in the unclear sediment type, that is, in the lagoon on the west side of the island. The sediment is not obvious and the bottom signal is weak, resulting in negative water depth values. There is a small amount of negative values distributed in the coral reef sediment type, while the sediment type of coral sand shows none. However, it is shown that the IOPLM produces less negative water depth values than the other two models. Therefore, for different study areas, topographic changes and differences in the distribution of sediment types will have varying degrees of impact on the estimation of water depth, which needs further study.’

We added the comparison with other studies about Worldview-2:

‘Based on the WorldView-2 multispectral images, the water depth retrieval experiments were carried out for the two areas. Similarly, many other methods were also carried out using WorldView-2, such as a two-media photogrammetry technique for WorldView-2 stereo multispectral imagery proposed by Cao et al. [35]. The research areas selected in the article are also located in the Paracel Islands (Zhaoshu Island and Ganquan Island) in the South China Sea. In the water depth ranges of 5 - 20 m, the MRE is about 17% and 14%, and the RMSE is 2.09 m and 1.76 m, respectively. The accuracy is similar to the results of Dongdao Island, but it does not include waters of 0 - 5 m depths, which relative errors in this section are fairly high. Our study uses a variety of perspectives to evaluate the model, in contrast to previous studies. Hamylton et al. [36] calculated the difference between the modeled water depth and the measured water depth. Then, they drew the geographical distributions maps of model residuals, and calculated the spatial autocorrelation. As a novel way of comparing different models, this technique has also achieved satisfactory results.’

Please see section 4. Discussion in details.

17. “How are the water column constituents affecting the satellite-derived bathymetry? Are waters clear in both study areas?”

Response:

Water quality is also an important factor affecting water depth inversion. The parameter selection of the model also takes into account the influence of water column, which can be applied to both high-scattering coastal waters and open waters. Dongdao Island is inaccessible, and the water is clear. The water quality of Saipan is more complicated than that of Dongdao Island. It is more affected by human activities, but its water quality is also good and its transparency is high.

18. “Line 404-406 Too long sentence.”

Response:

We have corrected this sentence as ‘In this study, we used the blue and green bands of WorldView-2 to develop an Inherent Optical Parameters Linear Model. Then, water depth retrieval experiments were carried out in Dongdao Island and Saipan Island.’

We would like to express our great appreciation to the reviewer for comments on our paper. We appreciate for Editor’s and Reviewer’s warm work earnestly, and hope that the correction will meet with approval.

Thank you and best regards.

Yours sincerely,

Xuechun Zhang

E-mail: [email protected];

Corresponding author:

Name: Yi Ma

E-mail: [email protected];

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Authors have satisfactorily addressed all questions and comments from my previous review

Author Response

Dear Reviewer:

We would like to express our great appreciation to the reviewer. Your comments are all valuable and very helpful for revising and improving our paper, as well as the important guiding significance to our researches.

Thank you and best regards.

Yours sincerely,

Xuechun Zhang

E-mail: [email protected];

Corresponding author:

Name: Yi Ma

E-mail: [email protected];