BoatNet: automated small boat composition detection using deep learning on satellite imagery

The authors adequatly addressed or responded to all given comments and suggestions.

I have no further recommendations and agree to release this version for publication.

The author of this manuscript reveals that in the design of the energy emission model it is also necessary to consider small boat fleets, in this case in the form of fishing and recreational boats.

In this paper, the author shows that BoatNet small boat detection model, can detect, measure, and classify small boats
even under low-resolution and blurry satellite images.

However, there are some shortcomings in this paper. Authors should explore YOLOv5l which was chosen as the model to train the dataset. Authors should present the YOLOv5l architecture and cite previous research using YOLO5vl.
In addition, it is necessary to include GPU resources and the employed framework along with computational cost analysis.
To enrich the analysis, the author should add a comparison of research results with other methods.

The article discusses maritime energy-emissions with a focus on emissions caused by ships. The authors reveal that existing energy emission models usually do not consider small boat fleets, which becomes the main focus in this paper. A model titled BoatNet for quantifying small boat fleet emission inventories is introduced, which can detect, measure and classify small boats in order to improve the accuracy of carbon inventory of small boats globally. BoatNet applies CNN on RGB very high-resolution satellite imagery.

The paper is well written, and the research is innovative, relevant and important. The main drawback is the documentation of the proposed methodology. The following, mostly Satellite-Remote Sensing issues should be considered before publication:

1. The authors missed to provide details on the imagery source and did not discuss resolution details on the available and used satellite imagery extracted that was from Google Earth Pro. For the applied eye altitude (200m) Google Earth Pro usually uses mosaiced true color composites derived from Digital Globe's WorldView-1/2/3 series, GeoEye-1, and Airbus' Pleiades, all of which provide data at around 0.5m spatial resolution. How does this refer to the resolution you extracted at 200m eye altitude?
2. Google Earth Pro does display the satellite imagery source, as for instance: “Image @ 2022 CNES / Airbus”, which refers to Airbus' Pleiades imagery. However, the authors did not mention the satellite imagery source(s) of the 694 high-resolution imagery being used.
3. The paper also leaves a few further open questions: What would be the benefit of additional spectral bands in the IR part of the EMS, if available? Worldview3 for example comes with 29 bands. PlanetScope with 5 bands. What would be the minimum required spatial resolution? Which multispectral (high-res) satellite imagery is available for free, and which one is not? A discussion on satellite imagery access, availability, costs and in particular resolution (spatial, spectral, temporal) would be beneficial for this research.
4. Di you account for the count of duplicates?
5. Details on how satellite images were extracted from Google Earth Pro are missing.
6. The issue of boats located on 2 adjacent images could have been addressed by applying tiling with spatial overlap (for instance of 50%, depending on the set tile size) …page 7 second paragraph: “…some large vessels …do not appear fully in an image”
7. The authors could also consider masking out land areas
8. Page 6: 2 ^nd paragraph: The authors mentioned shadows and clouds but did discuss in the next sentence the removal of haze. This leaves the reader with open questions about shadows and clouds (although Google Earth Pro imagery is as good as everywhere cloud-free mosaics).

Other, mostly minor comments:

1. Abstract: 1 ^st sentence: improve wording
2. Abstract: I suggest to replace the term “Techno-activity” with “Technology…”
3. Abstract: Replace GPS with “Global navigation satellite system (GNSS)”
4. Abstract: “…The work produced a methodology named BoatNet that can detect, measure and classify small boats…” – I suggest to also inform about target classes (shipping/leisure)
5. Page 1: Unit ‘Mt’ should be written in full when using the first time: “Megaton (Mt)”
6. Same accounts for CO2e: Carbon dioxide equivalent (page 2 last line)
7. Page 2: first 3 paragraphs: I suggest adding the respective literature references to back up your statements
8. Page 2 – section C.: First sentence: “Bringing deep….is essential.” Why is it essential – what for?
9. Page 2 – section C.: Third sentence: I suggest using the term ‘resolution’ instead of ‘quality’
10. Page 2 – section C.: Forth sentence: I suggest rewording into something like: “Machine learning is widely used for satellite imagery analysis.
11. Whole text: You may consider replacing “Satellite image” with “Satellite imagery”
12. Page 3 – line 8: “…and fuel used data” Did you mean fuel-used data? The sentence wasn’t clear to me.
13. Page 3 – line 10: “CO2e” Does “e” refer to estimate? Please write full form when using an abbreviation for the first time in the text.
14. Page 3 – section B – paragraph 2: “…each number is neither zero nor new, but…” Are you sure that is correct? What is a “new” number? Please adjust to improve clarity.
15. Page 5 – end of chapter II: I recommend adding a summary of the literature review including argumentation for why Yolo CNN was selected for this work (just before the last paragraph)
16. Page 5 – end of chapter II- last paragraph: “…aims at detecting small boats…” I recommend adding the fact that the proposed model intends to detect specific boat types (fishing, recreational)
17. Page 6 - Fig. 5 caption: reference should be replaced by ref 51 (Dwivedi …Yolov5), or at least ref 51 should be added.
18. Page 6 - the last paragraph: “ To validate… was done beteen BoatNet…” à correct between
19. Page 7: what is the loss of prediction (detection) accuracy due to image resizing? Depending on research goals (and other factors), it is sometimes worth running a multiple day training.
20. Page 8: false counts caused by nearby located boats: couldn’t the boat type classification allow to at least distinguish between different (adjacent) boat types?
21. Page 8: last paragraph: “Nevertheless, as Figures 11, 12, 13 demonstrate, the model still detects most small boats in poorly detailed satellite images,…” What exactly did you refer to with “poorly detailed” ?
22. “precision of training can be up to 93.9%,”  … why isn’t this result not explained in the result section? How did you derive 93
23. Page 10, fourth paragraph: “Due to the low data quality of the selected regions, the images are less suitable as training datasets.” What exactly did you mean with “ low data quality”