Next Article in Journal
PolSAR Image Classification via Learned Superpixels and QCNN Integrating Color Features
Next Article in Special Issue
Dual Frequency Orbiter-Radar System for the Observation of Seas and Tides on Titan: Extraterrestrial Oceanography from Satellite
Previous Article in Journal
Random Noise Suppression of Magnetic Resonance Sounding Data with Intensive Sampling Sparse Reconstruction and Kernel Regression Estimation
Previous Article in Special Issue
Satellite Derived Bathymetry Using Machine Learning and Multi-Temporal Satellite Images
Open AccessTechnical Note
Peer-Review Record

Leveraging Commercial High-Resolution Multispectral Satellite and Multibeam Sonar Data to Estimate Bathymetry: The Case Study of the Caribbean Sea

Remote Sens. 2019, 11(15), 1830;
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3: Anonymous
Reviewer 4: Anonymous
Remote Sens. 2019, 11(15), 1830;
Received: 27 June 2019 / Revised: 26 July 2019 / Accepted: 2 August 2019 / Published: 6 August 2019
(This article belongs to the Special Issue Satellite Derived Bathymetry)

Round 1

Reviewer 1 Report

This paper describes how a new methodological framework featuring a combination of an empirical linear transformation, cloud masking, sunglint correction, and pseudo-invariant features allows spatially independent calibration and test of a satellite-derived bathymetry approach.


Unfortunately, I do have a major concern linked to this manuscript for publication in Remote Sensing as it is. My concerns are not with the authors' methods, the paper is arguably at its best in the technical details of data processing. I know the methods represent a quite big amount of work, and I applaud the authors for their thoroughness.

The problems I see in this manuscript are more fundamental:

The authors' claims and analysis need to be tested not only in very clear water like the one they are using in the Caribbean areas. The innovation missing from this paper—and ultimately preventing it from being a suitable contribution to Remote sensing—is perhaps the great challenge of remote sensing and specifically the innovation and contribution of the satellite approach for very clear water. The author’s have to make a more comprehensive analysis of several other attempts like:

 Cappucci, S.; Valentini, E.; Del Monte, M.; Paci, M.; Filipponi, F., and Taramelli, A., 2017. Detection of natural and anthropic features on small islands. Journal of Coastal Research, Special Issue No. 77, 73-87.

Quadros N. D., Collier P.A., Fraser C.S. (2008) - Integration of Bathymetric and Topographic LiDAR: a Preliminary Investigation. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol. XXXVII, Part B8 (Beijing 2008): 1299-1304.


Guenther G.C. (2007) - Airborne LiDAR bathymetry, Digital Elevation Model Technologies and Applications: The DEM Users Manual, 2nd Edn. American Society for Photogrammetry and Remote Sensing, Bethesda, Maryland: 253-320.


Stewart, C., Renga, A., Gaffney, V., & Schiavon, G. (2016). Sentinel-1 bathymetry for North Sea palaeolandscape analysis. International Journal of Remote Sensing, 37(3), 471-491.


Hamylton, S. M., Hedley, J. D., & 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(12), 16257-16273.


Taramelli, A., Valentini, E., Innocenti, C., & Cappucci, S. (2013, July). FHYL: Field spectral libraries, airborne hyperspectral images and topographic and bathymetric LiDAR data for complex coastal mapping. In 2013 IEEE International Geoscience and Remote Sensing Symposium-IGARSS (pp. 2270-2273). IEEE.


Discussing the paper output considering turbidity and sediment suspension is a tricky part that the author’s are completely missing. I do see the innovative issue specifically in this context.

Wherever it gets published, any subsequent version of this work needs to clarify explicitly the authors' research question, motivation, hypothesis, and outcomes in relation not only to a specific test area but more in the context of bathymetric research. I do strongly support the author’s to improve the manuscript and submit a new version because I do see the innovation point of the arguments they are proposing in connections with the new satellite bathymetry detection.


Best regards.

Author Response

Dear Reviewer,

Thank you for your time identifying your concerns with our submission, and pointing us to relevant research for the future.

Please find attached our responses (in red). Please be aware that all mentions of figures and line numbering are taken from the updated manuscript.

All the best,


Author Response File: Author Response.pdf

Reviewer 2 Report

Hello. First of all, I would like to thank you for your work and for preparing this article. It is a pertinent and current topic, with relevance for both the advancement of scientific method but, perhaps most importantly, for their adoption by other activity sectors.

I would recommend to review, the language and style, and particularly the structure of the sentence, aiming at improving the clarity of the message. I believe you will find some sentences too long, as well as some typos or words lost in the last revision/edition process.

I would also suggest revising the figures. Due to the multiplicity of images (6 per figure), it would be very helpful for the reader if you could include more visual cues and information such as titles. Please also make sure that the colour palette provides appropriate contrasts, and that all data included is featured in the legend (eg land and cloud mask).

Please refer to the attached documents for specific comments for different sections of your work, including several considerations regarding the methodological choices.

Comments for author File: Comments.pdf

Author Response

Dear Reviewer,

Thank you for your time, your suggestions, and comments, for our manuscript submission to the satellite-derived bathymetry special issue of Remote Sensing.

Please find attached our responses (in red).  Please be aware that all mentions of figure and line numbering are taken from the updated manuscript.

All the best,


Author Response File: Author Response.pdf

Reviewer 3 Report

This paper presents a case study based on the published algorithms about the satellite data preprocessing and SDB. The presented results extend the results published by authors in previous works.

Even if this paper presents a case study [only for two studied sites] based on the improved algorithms, the data analysed in this paper present a good correlation and interpretation, so I consider that this can be recognised and this paper can be accepted for publication.


Suggestions for the future studies:

About the modelling of atmospheric column conditions and parametrization constraints, in addition to those already exposed in this paper, I suggest to the authors for furthers studies to consider the variations of refractive index vs wavelength of the atmosphere [e.g. due to the pollution events] that can be obtained, as example, from AERONET [sun/lunar photometer network data]. Moreover, the accuracy will be surely improved by considering the atmospheric turbulence also, calculated from proved models based on the active remote sensing techniques [e.g atmospheric lidar, lidar networks]. As example, the dust storm events can affect the sunlight correction for a few days….and so on.

Author Response

Dear Reviewer,


We would like to thank you for the ideas you provide that can support future work (many ideas on that – issues with funding and data availability towards automatization of the workflow).

The optimal for atmospheric column correction (as to that you refer to AERONET) is to collect data using a spectroradiometer at the same time of satellite overpass, which cannot be operational. On the other hand, AERONET stations and data can support such activities.

And over the coastal zone neither the land-based AERONET data nor a generic algorithm can provide a full correction; as such, more dedicated algorithms (e.g. ACOLITE) are developed, supporting for now only Landsat/Sentinel data.

It is expected that in the next 6 months, versions for commercial satellites will become available (we are in collaboration with ACOLITE persons for that).

All the best,


Reviewer 4 Report

This research sets out to modify the previously developed empirical SDB methodology outlined in Traganos et al. 2018 for use with Pleiades datasets, and test it on two study areas within the Leeward Islands. Methodological differences between this and the previous study are clearly outlined, differences mostly due to the use of Pleiades instead of Sentinel, and QGIS instead of a proprietary software. Results show depth RMSE values of between 1.76 and 6.93 m depending on depth interval and study site.

Broad Comments:

The vertical accuracy results seen in this study don't seem to line up with other similar research, which tends to have around 1 m to sub-metre RMSE values (which the authors themselves identified in section 4.2). This is not in itself a problem, since the authors are using sensors not used in these other studies, however they fail to adequately explain the source of the discrepancy.

In particular, the slope of the linear fit on the validation plots for BVI is significantly different from x=y, indicating a methodological problem. The stated conclusion in line 276 "suggesting an increase in vertical error with depth" doesn't fully describe what appears to be happening. If that were the case, the plots would appear to lie on the x=y, and have a 'horn' shape towards deeper points (such as that seen somewhat in Fig 4B).

The authors need to add x=y (one-to-one) lines onto the validation plots, and explain why linear fit of the data lies so far off x=y, or else modify the methodology to fix this.

Specific Comments:

81-82: "in-situ data availability" -- photogrammteric and physics-based models are specifically developed to not require in-situ data availability.

167: For method point 2, masking of land -- it is my understanding that in the original Traganos18 methodology, a CART was manually trained on several Sentinel-2 bands to delineate the water/land boundary, however since those specific bands are not available with Pleiades, the existing OSM coastline was used. Did the authors try a near-infrared threshold, or an NDWI with green and NIR bands? For research purposes, OSM would be fine, but I would be concerned with using it for delineating coastline in navigational applications.

Figure 4: Must add one-to-one lines on validation plots. Also, the axis values here are odd. Depth is shown as negative values on the axes, though when mentioning depth intervals in the figure description and throughout the paper, depth is denoted as positive (e.x. 10-30 m) values. Standard nomenclature here would be to have positive values indicating depth below chart datum. For clarity, it might also be a good idea to label each sub-plot directly with BVI/Anguilla and the depth interval, something like "(a) Anguilla 0-10 m" instead of only "(a)".

The paper is well structured and written, with just a few unclear sentences or typos:

51: Sentence is a bit hard to understand.

62: "emerged" -> emerging

127: " the, for the Darwin..." -- typo/repetition?

420: "we applied", "was applied" -- repetition

Author Response

Dear Reviewer,

Many thanks for your time, the broad and specific comments, and the suggestions you have made for our submission to this special issue of Remote Sensing.

Please find attached our responses (in red).  Please be aware that all mentions of figures and line numbering are taken from the updated manuscript.

All the best,


Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The paper is now ready for publication

Reviewer 4 Report

Revisions are sufficient for publication.

Back to TopTop