Next Article in Journal
Pile Installation Assessment of Offshore Wind Jacket Foundation in Completely Weathered Rock: A Case Study of the South China Sea
Next Article in Special Issue
Numerical Study of Circulation and Seasonal Variability in the Southwestern Yellow Sea
Previous Article in Journal
Environmental Compatibility of the Parc Tramuntana Offshore Wind Project in Relation to Marine Ecosystems
Previous Article in Special Issue
Investigation of Vortex Structure Modulation by Spume Droplets in the Marine Atmospheric Boundary Layer by Numerical Simulation
 
 
Article
Peer-Review Record

The Gradient-Boosting Method for Tackling High Computing Demand in Underwater Acoustic Propagation Modeling

J. Mar. Sci. Eng. 2022, 10(7), 899; https://doi.org/10.3390/jmse10070899
by Dominic Lagrois 1,*, Tyler R. Bonnell 1,2, Ankita Shukla 1,3 and Clément Chion 1
Reviewer 1: Anonymous
Reviewer 2:
J. Mar. Sci. Eng. 2022, 10(7), 899; https://doi.org/10.3390/jmse10070899
Submission received: 13 May 2022 / Revised: 21 June 2022 / Accepted: 23 June 2022 / Published: 29 June 2022
(This article belongs to the Special Issue Numerical Modelling of Atmospheres and Oceans)

Round 1

Reviewer 1 Report

This paper presents a machine learning algorithm to model the broadband noise received by marine mammals.  Generally, the objective of the work is interesting and the methodology seems sound.  

My major comment is on the quality of the presentation to improve the flow, in addition to reviewing the English for grammatical and syntax errors. 

Specifically:

1 - I may, at the end of section 1, summarize the structure of the paper for the benefit of the reader. 

2 - In section 4, the authors refer to details on the stochastic process (line 122).  However, we have no idea what will be stochastic at that point in the model.   It makes it hard to understand why the authors are discarding the first day. 

3 - On line 162, the authors refer to equation (3) in a citation.  I think it would be preferable to have the equation repeated in this paper for convenience.  

4- In the same paragraph, the authors refer to their analytical model as a range-independent one.  In that case, it is not surprising at all that the output doesn't match the full RAM simulation.  I am surprised that the authors even try to compare those.   An explanation would be required on the methodology, and rationale. 

5 - The authors tend to explain their methodology in great detail, but the objective is not always clear.   For example, I would, at the beginning of Section 4 explain the objective to compare the model using RAM, and Gaussmann. 

6 - It would be useful that the conclusion reads easily on its own.  The way it is written it is sowewhat vague.  I understand that the first sentence summarizes the problem.  The second sentences has syntax errors, so that it is not clear.

There are minor typos or syntax errors throughout the paper (the paper may not have been written by a native English speaking author, but it still needs to read better), and we should improve the English throughout the paper: lines 222 ("retrieved", should be retrieve), line 170 (I would not use the word "put"), line 64, (I would not use the word "made"), line 45 (works), line 32 ("the impact in time", I would use "the impact as a function of time").  These are only a few examples.  

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper used the Gradient boosting XGBoost under R (?) in the social-ecological agent-based model 3MTSim to interpolate, in the time domain, sound pressure levels received at the agents’ positions occupied by the endangered St. Lawrence Estuary and Saguenay Fjord belugas and caused by anthropomorphic noise of nearby transiting merchant vessels.

Main comments:

1.     What is R in the abstract?

2.     Is 3MTSim open and everyone can test it? The paper did not mention it or provide a link.

3.     Are there other similar applications besides 3MTSimPlease refer them and compare among them and 3MTSim. Then modify Part 1 of the paper.

4.     Words and figures in Figure 1 are not clear enough.

5.     In Table 1, what are the References? And use reference papers [16,17] may not appropriate. We think you only use part, even small part of them.

6.     In Part 3, the GBM is clearly stated. Please add a figure to illustrated it and state why this method chosen.

7.     Any other machine learning method can be use in this paper? Why not chosen?

8.     Table 2: The contents of the table could be centered.

9.     Line 133-144: the steps are too long to read. Please rearrange them.

10. Are your equations well descripted? For example, in Eq. (3), unit Hz should not appear.

11. Table 3: if this table necessary? Or if the presentation appropriate?

12. In my view, Figure 2 and 3 are similar to each other. Please clarify why two figures needed.

13. your paper and reduce the number of parts.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Back to TopTop