| 27 Apr 2026
Lake Ice Thickness retrieval using TanDEM-X immediate interferometry
Abstract. Knowledge of Lake Ice Thickness (LIT) is essential for understanding the cryosphere and monitoring current climate change impacts. However, accurately retrieving LIT at the desired spatial-temporal scale remains a challenge, as many lakes are in remote regions and LIT is a logistically expensive parameter to measure. Interferometric synthetic aperture radar (InSAR) provides a novel approach to estimating ice thickness by measuring surface deformations at high resolution. This study used TanDEM-X pursuit mode that offers minimal temporal correlation to maintain high coherence for accurate LIT retrieval in thermokarst lakes in Northern Alaska during the 2014–2015 winter season. The InSAR-derived LIT was validated against simulations from the Canadian Lake Ice Model (CLIMo), supported by in-situ snow and ice measurements. Results show consistent ice growth patterns and an RMSE of 0.08–0.26 m, demonstrating that the proposed method captures LIT evolution with reasonable agreement with CLIMo estimates. By employing an immediate interferometric approach, the present study maintains sufficient coherence to isolate and highlight the influence of volume scattering, which shifts the phase center away from the ice-water interface, and is the main factor limiting the accuracy of the LIT retrieval. These findings provide new insights into the technology of InSAR-derived LIT and suggest that SAR missions operating at longer wavelengths, such as NISAR and TanDEM-L, hold significant potential for improving retrieval accuracy by enhancing penetration and reducing sensitivity to internal scattering within the ice volume.
The manuscript “Lake Ice Thickness retrieval using TanDEM-X immediate interferometry” provides an overview of using TanDEM-X immediate interferometry to retrieve ice thickness for lakes located on the North Slope of Alaska. The research is relevant and it is good to see this methodology explored. The results presented appear promising. There is a good discussion on the possible limitations and physical processes within the ice that may introduce error when retrieving ice thickness using these data.
However, the manuscript should undergo major revisions before it is ready to publish. There are several points within the introduction that are no clear or need additional support from the literature. Additionally, there are parts of the methodology that are not cited and it is unclear whether the authors are the first to use the proposed methodology. Statistical assumptions are not considered in the results and the plots displaying the results could be improved for clarity. Finally, the discussion is strong regarding the physical basis but there is a lack of discussion on how these results compare to past results and the retrievals performed.
Please see further comments below.
Line 27-28: The reason why this is the case should be clarified to the reader. How does a decline in manual meteorological observations pose a challenge to LIT retrieval?
Line 28-30: This is not necessarily true. LIT is difficult to retrieve using remote sensing technologies and only recently through altimetry (Beckers et al, 2017; Mangilli et al., 2022; 2024) have methods developed closer to operational algorithms. Arguably early work on bedfast ice using SAR reveals information on approximate thickness but not true quantitative retrievals. It would be truer to state that there is an interest in the use of remote sensing technologies as a solution to the current issues with monitoring LIT. To amend this a note stating the large body of work using InSAR would be good. The decision to cite Duguay et al., 2015, instead of a list of key papers from the use of InSAR (Lines 65-82) could reinforce this statement.
Line 31-32: This should be more specific, and SAR is a poor example. It would be better to use radar altimeters as an example here.
Line 45-47: Why? What is it about radar systems that make this interaction useful for determining LIT? As of this point in the manuscript the main conclusion is that roughness is what drives the backscatter response from lake ice. How does this interaction translate into the retrieval of LIT?
Line 48-59: The discussion on radar altimeter methods is quite short despite the promising recent advances in algorithms. The discussion on the method used here should be expanded.
Line 60-62: It is worth stating here that fully-focuses SAR altimetry has the capacity to have a resolution of 0.5m which exceeds that of imaging SAR. One aspect missing from the discussion here is that imaging SAR can provided gridded data compared to altimetry which is restricted to tracks. I note only one mention of SWOT which is the exception via the KaRIn sensor.
Line 180-188: Because of the lack of citations, to confirm this method is developed by the authors and has not been previously used in the literature?
Line 298: The selection of the 99.7th percentile for ice thickness is unclear. There is expected spatial variability in ice thickness, as authors have mentioned, why use the 99.7th percentile for validation when the maximum retrieved value could not be used instead? This may not need to be clarified in text and just a further explanation in response to the comment here.
Table 3: This chart is oddly set up. Why are observation data separated by CLIMo simulations? It would make more sense to show observation data versus CLIMo results. I believe it is to display it as increasing values, however, it did not hold for ice thickness so it does not make sense to do it just for snow depths.
Figure 7: The boxplots are very confusing. The results presented in the text are difficult to interpret with the box plots. Primarily because when initially looking at the box plots it appears that the method has done a poor job retrieving ice thickness with mean values being >20 cm less than the CLIMo values. However, re-reading the text does help to clarify a bit the meaning and what should be examined (the red triangles). The recommendation would be to improve the text describing these figures as well as the figures themselves.
Figure 8: Similar comment to Figure 7. Does not clarify if this is all values or just the 99.7th percentile as specified in the methodology. Why use R2? You are not performing a linear regression. Additionally, the data does not appear normal, was a normality test performed? Likely a non-parametric correlation coefficient (Spearman’s Rho or Kendall’s Tau-b) would be better suited.
Discussion: The results should be placed in the context of existing literature. How does it compare to the results of other studies? Other methods (e.g., altimetry)?