Review of "First Arctic-wide assessment of SWOT swath altimetry with ICESat-2 over sea ice" by Müller et al.

Summary:

This study evaluates the performance of the SWOT mission’s swath-based sea surface height observations in the Arctic by comparing one year of data (March 2023–April 2024) with ICESat-2 measurements at ~550 crossover locations. Results show good overall coherence, with mean standard deviations of ~8 cm (6 cm over leads), but reveal systematic offsets between left- and right-swaths (up to 50 cm) and higher deviations during the early melt season and in complex regions such as the Canadian Archipelago.

General Comments:

This paper will be helpful for future studies using SWOT data over sea ice, and helps to better understand the complexity of the SWOT data. However, in its current form, it sometimes lacks clarity. Some figures are difficult to interpret and require close attention. I therefore recommend major revisions, although the scientific content appears sound. The main issues concern presentation and clarity.”

I have some general comments that hopefully help to improve the paper:

1. The introduction lacks a clear statement of the study’s main goals. It should also emphasize that SWOT is a predecessor of Sentinel-3 NG, which will carry a similar sensor, making the results valuable for the upcoming mission.
2. I find the usage of SLA for heights in this paper slightly misleading. My understanding of SLA refers to open water only, not including sea ice. I would suggest to refer to surface heights/elevations instead.
3. Figures and text can be improved in clarity. For example, Figure 4 lacks labels, and the link between the different panels demands some attention. I made suggestions and comments below.
4. Can you make a suggestion how to correct for the SWOT cross-track errors? Have you tried?

Specific Comments: 

L9: Throughout the paper, it is referred to “Sentinel-1 grey-scale values”. Is it related to sigma0? It might be clearer to refer directly to the backscatter coefficient.

L34-42: The ICESat-2 description is maybe a bit lengthy and part of it could be moved to the Data section, especially technical details like footprint etc.

L38: Different values for the ICESat-2 footprint can be found in the literature; a commonly cited estimate is 11 m, as reported by Magruder et al. (2020).

L41: Leads are sometimes not so easy to track with ICESat-2, in contrast to radar altimetry where we receive specular waveforms. However, I would add here that ICESat-2 ice well suited to detect ridges and surface roughness (e.g., Farrell et al (2020)., Ricker et al. (2023)) 

L95: “ICESat-2 latest SLA segments from ATL07 Release”. I believe “surface heights” are meant here? The usage of SLA is slightly confusing.

L111: Here it is referred to “SSH”, SLA and SSH are not the same, see also comment above. Please clarify.

L102-104: “Tests have shown that no significant differences exist between the two laser beam types.” Does it mean tests that have been done in the framework of this study? Or is it referring to earlier studies? If the latter, a reference is needed.

L146: I find it hard to see the “good agreement” here. Instead of the zoom-in area, an along-track line-plot with SWOT and ICESat-2 heights might be better suited (like Figure 5). In any case, is the figure relevant? See my comments on the figures below.

L163-164: “In some regions, sea ice surfaces appear not to be represented in the SWOT heights or show no particular elevation structure compared to other sea ice surfaces” This sentence is unclear, what is meant with “sea-ice surfaces are either absent” ?

164-165: Why are the stars not in the upper left figure? This would make the comparison easier.

172-177: This part is not entirely clear to me. What is meant here? What are the line-like height differences? Linear kinematic features? Perhaps it can be marked in the figure.

177-179: this cross-track error looks like it could be at least reduced with a relatively simple gradient correction? Have you tried to correct for it?

L186-189: Have you checked the behaviour using the individual ICESat-2 beams? There might be also small differences/biases between the three beams.  

L194-195: “i.e. it is less sensitive to changes in height than ICESat-2” … which makes sense given the small footprint (~11 m) of ICESat-2

L245: “The precision reduces to 6 cm”. Is it not an improvement in precision? Do you mean the standard deviation?

L251: “Compared to SWOT, this indicates that the LR dataset…” … but LR is SWOT, no? Please clarify.

Figure 3 & 4: I find Figure 4 (in combination with Fig. 5) very informative, but using the same colormap for both SWOT and ICESat-2 makes it sometimes difficult to separate. May be use different colormaps like in Figure 3? On the other hand, I wonder if Figure 3 is actually needed. Is there something in Figure 3 that cannot be explained by Figure 4 + 5? However, in Figure 4, I suggest to use “a)”, “b)”, etc, and moreover, it would be good to use arrows starting at the black boxes, pointing to the respective bottom figures.

Figure 5: The right axis needs a label, even if these are relative units.

Figure 10: Colorbar label missing.

References:

Farrell, S. L., Duncan, K., Buckley, E. M., Richter-Menge, J., and Li, R.: Mapping Sea Ice Surface Topography in High Fidelity With ICESat-2, Geophys. Res. Lett., 47, e2020GL090708, https://doi.org/10.1029/2020GL090708, 2020. a, b, c

Magruder, L. A., Brunt, K. M., and Alonzo, M.: Early ICESat-2 on-orbit Geolocation Validation Using Ground-Based Corner Cube Retro-Reflectors, Remote Sensing, 12, 3653, https://doi.org/10.3390/rs12213653, 2020.

Ricker, R., Fons, S., Jutila, A., Hutter, N., Duncan, K., Farrell, S. L., Kurtz, N. T., and Fredensborg Hansen, R. M.: Linking scales of sea ice surface topography: evaluation of ICESat-2 measurements with coincident helicopter laser scanning during MOSAiC, The Cryosphere, 17, 1411–1429, https://doi.org/10.5194/tc-17-1411-2023, 2023.

Review comments for “First Arctic-wide assessment of SWOT swath altimetry with ICESat-2 over sea ice” by Felix L. Müller, Florian Seitz, and Denise Dettmering

The manuscript presents a comparison of elevations measured by SWOT and ICESat-2 in the Arctic ocean (below the latitude limit set by inclination of SWOT) - most importantly in sea ice covered areas. The paper is an important step in understanding the novel altimeter measurement of SWOT over ice infested waters. The potential of SWOT is large, but we need studies like this to be able to utilise the data. 

Overall, the paper is well written, interesting and important and most certainly deserves to be published. However the manuscript would benefit from a thorough, minor revision, adding more rigour and sharpening the overall presentation. The main results - overall agreement of ICESat-2 and SWOT elevations as well as the up to 50 cm differences when comparing left- and right-hand swaths are sound and scientifically the paper is of good quality. Thus, I would recommend minor revisions concentrating on presentation as well as adding more information on the SAR data used - including SAR / sea ice interaction as well as discussion on how snow on sea ice affects the difference between ICESat-2 and SWOT.

More specific comments:

Reviewer #1 has already raised a good point of using “SLA” for both elevations of sea ice and water. I’d recommend the authors to use another term, such as surface height or elevation as suggested by reviewer 1.

Also, already noted by reviewer 1, “sentinel-1 gray scale value” is a vague term. I’d recommend authors add a subsection in the data section explaining what it actually is and where does the data come from. Throughout the paper, S1 images are discussed quite a lot, but what lacks is the background on how SAR and sea ice actually interact. That is to say - there are many ways sea ice can produce high (or low) backscatter and not all of them affect the actual surface elevation of the sea ice surface. 

Another thing I'd like to see in the paper that is now lacking is the effect of snow on sea ice. Currently, snow is not really mentioned at all. Part of the mismatch between SWOT and ICESat-2 is due to penetration depth into snow (which is very shortly mentioned in L251). This should be elaborated.  

Extremely specific comments:

L140-143: The driving phenomena behind higher backscatter values is surface roughness. This should be explained and also connected to the fact that surface roughness of sea ice is very much connected to surface height. This is touched upon at ~ L170, but in my opinion quite lightly.

L150-160 (Figures 4 and 5): The discussion here includes few wobbly statements like “grey-scale values show a strong coherence with SWOT”, “can be monitored very well” and “captured quite well”. These should be re-written to something preferably quantitative. In the big picture of Figure 4, one can see areas where all three datasets agree, but also areas where the correlation is not apparent. 

Figure 5: X-axes should be made equal for all subplots, even if the northernmost beam lacks data in the Western end.

Figure 9: There is an interesting linear growth in STD over winter 23-24. This coincides with thickening ice, increasing surface roughness as well as thickening snow pack.

L190: I do not disagree, but would like to comment that pan-Arctic sea ice surface elevation distributions often have two maxima - one for first year and other for multi year ice. Would it be possible that the STD here would be a combination of two gaussian distributions, one for FYI and other for MYI? 

L245: Not precision (as also pointed out by reviewer 1)

L260: Would it be possible to quantify the difference caused by time difference in acquisitions?