In this manuscript, the authors assess the performance of the long-term Antarctic sea ice motion product derived from passive microwave (PMW) observations at NSIDC using buoy observations and high-resolution SAR-derived sea ice drift data. The authors show that the PMW-derived Antarctic sea ice drift generally agrees well with both SAR and buoy observations. However, they also identify a systematic underestimation of sea ice drift speed in the PMW product, particularly in regions where sea ice concentration (SIC) is lower than 50%. The authors further suggest that the optimal interpolation procedure is likely responsible for this behavior.

Overall, I found the objectives and conclusions of this manuscript clear and easy to follow. The analyses are generally solid, and the manuscript is well organized. I enjoyed reading this paper. I recommend minor revisions before the manuscript can be accepted for publication.

Major comment:

• After carefully reading the manuscript, I feel that the paper focuses more on the evaluation/validation of the NSIDC Antarctic sea ice drift product rather than a comprehensive uncertainty assessment. Therefore, I suggest revising the title so that it better reflects the actual content of the manuscript.
• Scientists from EUMETSAT Ocean and Sea Ice Satellite Application Facility have also developed a widely used Antarctic sea ice drift product. It would be very helpful if the authors could discuss how the NSIDC sea ice drift product compares with the OSI SAF product, including but not limited to retrieval methodology, spatial/temporal resolution, and more importantly the relative performance against independent observations. Including such discussion would substantially improve the scientific significance and broader relevance of this manuscript.

Reference: Lavergne, T. and Down, E.: A climate data record of year-round global sea-ice drift from the EUMETSAT Ocean and Sea Ice Satellite Application Facility (OSI SAF), Earth Syst. Sci. Data, 15, 5807–5834, https://doi.org/10.5194/essd-15-5807-2023, 2023

Minor comment:

• P2 Lines 42-44, this statement is somewhat new to me. I wonder whether there are existing studies or references that support this claim.
• P12 Lines 231-233, does hypothesis (2) also contribute to the underestimation of sea ice drift speed? Please clarify this point further.
• P15 Lines 253, please check the wording “all seasons”, as it appears inconsistent with the earlier statement that only results from March to November are shown (P8, Lines 193–194).
• Since the optimal interpolation may degrade the performance of the sea ice drift product, which version of the dataset is publicly distributed and recommended for scientific research: the interpolated product or the raw PMW-derived product?

Figures:

• Figure 2, it may be clearer to show only the buoy locations that were actually used for the validation analysis.
• Figure 9, I noticed that even without interpolation, the PMW_raw sea ice drift still shows an underestimation under low-SIC conditions. Does this imply that factors other than interpolation also contribute to the bias? If so, please discuss the possible mechanisms.

This paper details a methodology for use of SAR sea-ice drift vectors for validation of PMW sea-ice drift, allowing for full coverage validation of the Antarctic where in-situ data is only available in limited areas, and rigorously analyses the robustness of this methodology. In addition there is an excellent analysis of the sources of error in the NSIDC Polar Pathfinder sea-ice drift product, with recommendations for future improvement. The techniques and analysis in this paper will be a great reference to inform future development for not only the NSIDC product but for other products as well. I recommend publication of this paper with minor changes.

Major comments:

• Section 2.4. Once spatially sampled, the PMW and SAR ice drift vectors are representing the average ice drift within sub-regions of radius 25 km, whereas the buoy vectors represent point-like sub-regions. This may be worth a sentence or two of discussion, since the SAR and buoy ground truths represent different spatial scales.
• Given that most of the larger deviations of the PMW vectors from buoy vectors come from areas of low SIC, and that SAR vectors have only been compared to the buoy vectors in regions of SIC > 85% while the PMW vectors have been taken from regions of SIC > 15%, I would really appreciate seeing statistics (extra figures might inflate the paper too much) for PMW vectors where SIC > 85%, so there is a more direct comparison between the PMW-buoy and SAR-buoy validations.

Minor comments:

• A table summarising the statistical values from the different validations could be very helpful to the reader, rather than needing to find the individual values from within the text to compare.
• L50-55, Just as a point of interest, SH PMW ice drift has also been validated vs buoys in the OSI SAF validation reports for the near-real time sea-ice drift product (https://osi-saf.eumetsat.int/products/osi-405-d, doi:10.15770/EUM_SAF_OSI_NRT_2007) and sea-ice drift climate data record (https://osi-saf.eumetsat.int/products/osi-455, doi:10.15770/EUM_SAF_OSI_0012).

• L137 Please define the FT abbreviation on first use.
• L142. Is it worth explaining/justifying why you omit SAR ice vectors above 40 km  d-¹ - is this dependent on your constraint that SAR data is only used for SIC > 85%?

• Figure 1 could perhaps have larger individual panels, as some of the magenta arrows do not show up well at the current resolution.
• Probably too complicated, but it would have been nice to plot figure 9a with the data from figure 5a as a background, to be able to compare easily.
• Figure 11b label should probably specify PMW_raw rather than PMW.