Mapping of submarine permafrost remains an important scientific issue despite the increasing number of works published in this field. Undersea permafrost is much more difficult to access empirically and requires the use of a new, original methodological approach.  Systematic research covers only the last few decades, so the recognition of undersea permafrost and especially the active layer is still imperfect and requires a lot of scientific effort. The presented article addresses both of these issues: substantive and methodological, contributing to this important topic with new, original research.

The authors try to present a number of scientific and methodological findings that may certainly be of great cognitive value to the growing number of undersea permafrost researchers. With this very general sentence I want to express my rather positive opinion about this work. I think it will be more interesting to present some comments and suggestions that would help improve this article. I will present them generally in this review, which is also significantly supplemented by detailed comments attached to the text of the article in the PDF version, in the form of notes

1. In the title, the authors rightly bring to the fore the substantive issue, i.e. determining the extent of permafrost in the studied area. I think it would be good to maintain this order in the structure of the entire work. However, in some chapters methodological issues come to the fore, such as limitations in the application of the method, difficulties in field work, uncertainty of some results, etc. This is an understandable and honest approach to the issue, but it slightly weakens the scientific result of the article. I would rather reverse this order.
2. An extremely important issue is to clearly define what is actually the subject of the authors' research. Just saying it's about permafrost degradation is not enough and not very original. Although the definition of permafrost applies to its terrestrial and undersea occurrence to the same extent, in the undersea environment the situation becomes much more complicated, for example because permafrost is not a purely climatic geological formation (subaerial). Therefore, in my opinion, a chapter is needed that characterizes the differences in the occurrence of terrestrial and submarine permafrost. This is even more important because a clear definition of the subject of research has important methodological consequences, i.e. in the interpretation of research results. This is not a difficult issue. In the published articles of the team of prof. Angelopoulos it is clearly presented.
3. Ice-bearing permafrost (IBP) is not the same as permafrost (PF) in general. IBP is to PF as a part is to the whole. In this context, it is important to comment on the issue of cryotic permafrost in your work. This would be a very interesting and necessary statement.
4. Regarding the method used. Of course, I agree that, just like on land, also in the case of undersea permafrost, its detection is associated with a sudden increase in resistance. It is worth noting, however, that this applies only to IBP but not to PF in general. A big problem arises when it comes to the issue of the cryotic state, especially present in sedimentary material saturated with salt water. This requires clarification and a broader discussion. In this respect, the conclusions chapter should also be clarified.

I also have a few minor comments:

1. I think it is more important to present at what depth the IBPT is located, not only under the water surface, but also under the seabed.
2. It seems to me necessary to present and comment on many more profiles analogous to the one shown in Fig. 4. The reader should be able to compare. It doesn't have to be an extensive description.
3. It would be very interesting for substantive and methodological reasons to see several ERT profiles with DOI isolines shown on its interpretation.
4. I think it would be very interesting to speak about the active layer that usually accompanies the occurrence of permafrost. This is a relatively difficult issue and there are few publications in this field, but that is why it is worth mentioning at least briefly on this occasion.

As I mentioned, the text also contains comments and some (not all) highlights that are the result of quite significant controversies. I hope they will be helpful in improving this interesting work.

The manuscript introduces marine Electrical Resistivity Tomography (ERT) as a technique for locating undersea permafrost. Using a relatively straightforward setup comprising 13 floating electrodes and an echosounder for determining water depths, multiple survey profiles around Tuktoyaktuk Island were measured. To mitigate corrupted ERT data, the cable was equipped with two GPS devices to monitor its curvature during the measurements. Various inversion methods were then tested and compared.

The manuscript demonstrates at which cable bending the geometry factor deviates significantly. These values are then filtered out. Additionally, a synthetic model was employed to investigate the impact of varying lateral and vertical smoothing constraints during the inversion process. Using the 'best-suited' inversion approach, the results indicate a slowly dipping permafrost table on the sea-facing northern side of Tuktoyaktuk Island and a more steeply dipping permafrost table on its southern side.

Overall, the manuscript is well written and easy to follow. However, to clarify the authors' approach further, a few structural adjustments could be beneficial (see general comments). When explaining the inversion process and its results, the authors could provide a more detailed introduction, analysis of the sensitivity and Depth of Investigation (DOI), as the results are primarily close to the resolution limits. Additionally, providing a brief overview of complementary methods and potential improvements to the inversion process at the end would help to properly contextualize the manuscript. These issues, along with several specific comments and technical corrections listed below, should be straightforward to address.

General comments

Structure

Beginning the “Results” chapter, Chapter 3, with uncertainties might come across as somewhat pessimistic, and readers typically expect to encounter results immediately in this section. Perhaps Section 3.1 could be relocated to the analysis chapter, Section 2.3, or alternatively, to the end of Chapter 3.

Additionally, Chapter 3.2, "Inversion and IBPT Depth," could be divided into two distinct parts: 1. Analysis and 'calibration' using synthetic data, and 2. Analysis of real data.

Presentation of inversion results

Please provide additional information on how the inversion parameters, such as DOI and inversion RMS (e.g., in comparison to data residual), are determined. A sensitivity analysis of e.g. the synthetic example would clarify how much the resistivity distribution at different depths can be expected to affect the measured data.

Including a figure that compares the synthetic model with the inversion result of the synthetic data would enhance the understanding of the synthetic study. Additionally, it would be beneficial to provide a second example of the real data inversion for comparison.

Outlook and other methods

Are there complementary methods/approaches to determine the depth of undersea permafrost? For an outlook the authors could also address possibilities to improve the inversion (e.g. 3D inversion or structural constraint inversion with different layers).

Specific comments

Line 2: comma after seawater: “Once inundated by seawater, permafrost usually begins to degrade.”

Line 11: Maybe more precise: “We discuss how marine ERT can be improved by accurately recording electrode positions, although choices made during data inversion are likely a greater source of uncertainty in the determination of the IBT position.”

Line 15: “permafrost degradation”

Line 39-40: Please give a reference for this prediction

Line 42: “sea floor” or “seafloor”? Both is used in the manuscript.

Line 47: Are “frozen fresh water saturated sediments” equivalent to permafrost? Is there a sharp boundary between these two layers, or is it more of a transition?

Line 48: “…a flexible cable in marine ERT survey relies on potentially varying electrode positions.” -> “…a flexible cable in marine ERT survey results in potentially varying electrode positions. “

Line 49: Fig. 2 mentioned before Fig. 1

Line 55 / Figure 1: Figure appears before referenced in text

Line 60: “ice bonded” -> “ice-bonded”

Line 60: “random massive ice bodies” -> “randomly distributed massive ice bodies”

Line 61: “On top of the cliff” -> “Above the cliff”

Line 67: Please explain “thermoerosion”

Line 68: Please introduce GSC

Line 86: Why reciprocal Wenner-Schlumberger array? Please discuss your choice concerning S/N ratio.

Line 87: Why did you just use 13 of 22 electrodes? Maybe give a short explanation.

Line 89: You could include 'DOI' in brackets here to clarify the abbreviation for future references. Additionally, how was the DOI determined?

Line 93: “The water depth was measured from the boat for every sounding using an echosounder.” -> Please specify a bit more: Did you measure during the ERT measurement directly on the boat (velocity?)? In that case, there would be an offset of around 50 m to the point of maximum sensitivity and DOI of the ERT measurement.

Line 104 / Figure 2: - What model was used to determine DOI? - Maybe add a depth scale and a label for the water body

Line 110: Figure 3a appears much later; perhaps show subpicture 3a) earlier

Line 111: How was the geometric factor k calculated? Was the full problem solved (3D electrode positions) or only the variation of electrode distances?

Line 117: Are the penalizing parameters of these four inversion types only applied to vertical resistivity contrasts or also to lateral resistivity contrasts?

Line 127: Why did you choose 10 Ohmm as the resistivity of the unfrozen sediment? Including a picture of the synthetic model could be generally helpful.

Line 141: Please introduce RMS

Line 145: Please introduce GIS.

Line 168: Here, mainly Fig. 3c)

Line 169-171: Probably only a secondary effect, should be in the range of the error linked to 2D assumption as resistivity also depends on variation of resistivity perpendicular to the profile.

Line 177: - “were removed” with d at the end

- “Soundings for which the difference varied were removed” What was the threshold here?

Line 185: When interpreting the data without additional information, whether the boundary layer lies above or below the DOI should not be an argument. Both are valid possibilities.

Line 195: Maybe not with cm accuracy here.

Line 196: What is the grid cell size and what resolution could be expected according to the used electrode array in that depth?

Line 200: How is the data residual determined, and what is the difference compared to the inversion RMS?

Line 204: Here you could split the chapter into “Synthetic model analysis” and “Real data analysis”

Line 206: Was the water body a constraint in the inversion? Was there a variation in the CTD measurements or was it constantly 6 Ohmm?

Line 235: Couldn't the movement of the cable also be determined by the history of the GPS data, allowing you to simulate the cable's movement?

Line 254: To address this issue, you could potentially constrain the water layer within a range, e.g. between 5-6 Ohmm.

Line 269: Maybe you could provide this explanation earlier (see comment on line 185).

Line 276: First 'or' in the sentence could be replaced with a comma.

Line 285: This is the resolution of the mesh. The capability to resolve resistivity contrasts of the inversion could be in a different range depending on your electrode configuration and resistivity distribution.

Line 294: A finer grid does not necessarily improve the resolution of the measurement. An analysis of the sensitivity would be interesting to modify or adapt the electrode configuration in later studies.

06/06/2024, Johannes Hoppenbrock

TU Braunschweig

Institute for Geophysics and Extraterrestrial Physics

Braunschweig, Germany