Please see the attached pdf

Please find comments as attached in supplement pdf

Review of the manuscript:

Sea ice melt pond bathymetry reconstructed from aerial photographs using photogrammetry: A new method applied to MOSAiC data

By:

Niels Fuchs, Luisa von Albedyll, Gerit Birnbaum, Felix Linhardt, Natascha Oppelt, and Christian Haas

This is a fantastic high-impact paper. The methodology is well described, the figures clear, and the results compelling. I suggest a copy editor or an intensive read-through to catch awkward/incorrect phrasing. I attempted to point these out as I came upon them but I may have missed some. The other main suggestions are including and clarifying a few studies in the background introduction section, and adding to the figure captions.

General:

The Figure captions are pretty short, and sometimes say “described in the text” I recommend adding text to the figure captions so that they can be stand-alone figures. I imagine this will be a useful reference text for future studies and the description of diagrams in figure captions will make the methodology clearer.

Specific:

Line 26: awkward phrasing- maybe you mean “rather simply”

Line 29: I disagree that ‘most melt pond depth obs. for models were published in Morassutti and Ledrew’. Melt pond depth measurements from SHEBA (Perovich et al., 2003, Figure 11) were used in CCSM4 parameterizations (Holland et al., 2012). You actually mention this in line 40. These thoughts could be combined

Line 34. I would say “Here we define pond bathymetry as…” because some studies will refer to bathymetry as a two-dimensional sample instead of the whole bathymetric floor.

Line 42: I don’t know if this is based on Luthje 2006. See Holland et al 2012- directly references Perovich et al. 2003 and the SHEBA measurements. Maybe this is true for the Pedersen scheme but certainly no all the links between pond fraction and depth

Line 50: ICESat-2 (correct capitalization)

Line 50: Please also include the larger study by Buckley et al., 2023 (Follow on to Farrell et al., 2020) which involves two algorithms (also include Herzfeld et al., 2023 that described the DDA algorithm) to automatically retrieve melt pond depth applied to thousands of ponds in the 2020 melt season. Still not a comprehensive database but showcases the ability to retrieve pond depths at large scales. I suggest you also include this in the discussion section about pond depth and coverage (fraction) evolution.

Line 51. Consider “Orbital path” instead of “flight track lines”

Line 54. Chiroptera flew over sea ice for the ICESat-2 summer validation campaign in 2022 and that is a ALB system. Not sure of any publications that include that information right now though.

Figure 1 caption. “Know” to “known”

Figure 1. can you make sure the arrows are contained within the image – it is hard to tell what they are.

Line 87. I’m confused about the use of “we.” You are not the Macke and Flores authors- do you mean they did that? Or the authors on this paper also happened to be on the Polarstern cruise. The second half of this paragraph is in third person. Consider clarifying or putting the whole methods section in third person.

Line 101. Replace “most probably” with “most likely” or “likely”

Line 104: why can’t these ponds be designated strictly as melt ponds?

Line 111: Reference for cloudy days being more common in Arctic summer?

Line 123: change “reached” to “ranged” and “pond depth” to “pond depth measurements”

Line 126: Can you quantify pond coverage increase?

Line 134: either leave out “pandemic related” or include COVID-19 (hopefully people in 100 years will be reading this paper and may not know what this is referring to)

Line 135: I think you said this somewhere else but can you remind us the age of the new ice floe?

Line 141: 2 km x 2 km you mean?

Line 146: how much error does this introduce?

Line 209: can you make Snell’s law a proper numbered equation in the text, or refer to Eqn. 1 here.

Line 255: (e.g., Hutter et al., 2023) – im sure there are others so add the e.g.

Line 267: (Jordahl et al., 2020; Perry, 2015; Gillies, 2013). Either but these in chronological order or if these refer to the python libraries in order add “,respectively”

Line 276: is the algo description in Fuchs 2023a or Fuchs 2023c- both are included here- what is the difference

Line 279: perhaps here can you list the main classes and not just refer to Table 3

Line 301: if you include the QGIS version here, you should include it everywhere

Figure 11: Consider a gray or dark background so the light points stand out.

Line 370: Again the pandemic comment

Line 376: Sentence doesn’t make sense – especially the ending “with partially more than 2 m)

Line 381: relatively high underestimation is a confusing statement

Line 385: does vertically elevated mean above sea level?

Line 398: this is confusing. Pond volume across the floe was constant (in space or time?) And what do you mean “has not changed much either”… has not changed since when? And then the next clause you say the ponds do deepen?

Line 404: what do you mean both Webster et al., 2022? If you are talking about the ossp classification in webster, cite webster in the first set of brackets with Wright and Polashenski

Line 410: can you describe what you see in Fig 15a and quantify how well they match.

Figure 15: can you explain the recalculation, has this been fixed in the Webster manuscript? If so can you reference the correction?

Line 476: all other instances of Mystery Lake do not have lake capitalized. Be consistent.

Line 479: too few

Line 525: sentence starting with “most interestingly…” does not make sense.

Line 545: “the here” doesn’t make sense