|I thank the authors for their time spent in carefully addressing the concerns raised in the first review. The revised manuscript is much improved and quite informative. Please find comments below that I hope will be found useful. |
Page 1, line 24. I suggest changing "affects" to "may affect" given that it is a conceptual model.
Page 1, line 26. I suggest changing "of thick Antarctic snow" to "of thick snow on perennial Antarctic sea ice" since these are the limiting criteria for applying the conceptual model, which is important to convey.
Page 2, line 16. Please consider adding "positive" before albedo-feedback here.
Page 4, lines 11-13. It would be important to note that snow salinity is not negligible during the spring period. Observations by Eicken et al., 1994 and Worby and Massom (1995) show larger ranges in snow salinity during spring. Also, Massom et al., 1997 observed values of 8 psu in summer. Is this negligible for passive microwave, or just low?
Page 3, line 25. To clarify my earlier concern, please include a statement in the manuscript that these locations may not fully represent multiyear or seasonal sea ice regimes since there may be inter-annual variability in the coverage of ice types at any of these locations. Before, I presumed you had used an ice type product to distinguish multiyear from seasonal ice regimes at these locations.
Page 5, line 5. Do you mean 8 rather than 9 nearest neighbors? What is the 9th neighbor?
Page 7, lines 30-33. My concern here is that flooding is not limited to areas with 30% sea ice concentration, and this should be noted in the text. It would be reasonable to state that the effects of flooded ice on the passive microwave signature cannot be fully accounted for in the analysis given the limitations of space-borne observations.
Page 8, figure 4. It would be helpful to show the corresponding sea ice concentrations as additional figures in supplementary information due to their large influence on backscatter, rather than only showing when concentrations are less than 70%.
Page 9, lines 9-10. "...or the appearance of liquid water in the pendular regime in the middle or lower snowpack." My concern here is that this doesn't appear to be an observationally-based statement, but rather a speculation that liquid water occurs in the middle or lower snowpack first. How do we know that liquid water first occurs in these particular levels of the snowpack? Please state that this is speculation or give references to support this interpretation.
Page 10, line 13. I'm sorry, but it's still not clear what you mean by regionally adaptive here. The paragraph describes an approach for each specific location, but how is this translated to a regional approach?
Page 12, line 14. It would be helpful to state here that the values differ that those in Haas (2001) due to the different number of analysed pixels between studies, in case a person would like to reproduce the analyses.
Page 19, Section 4.3. There are errors that need to be addressed in this section, which are listed below. My concern is that parts of the argument are not supported by the literature and that previous works are referred to inappropriately. While the authors present an interesting concept, more transparency is needed on the knowledge gaps and uncertainties on this topic.
- The Brandt and Warren, 1993 study did not observe sub-surface maximum temperatures or melt in snow. Other works refer to this analysis as a theoretical modeling study. Rather, the study demonstrates how, once the radiative transfer model's spectral resolution is improved, the sub-surface maximum temperature does not materialize in a snowpack. Rather, they show this phenomenon to be limited in blue ice only. It is erroneous to refer to this work as an observational study of sub-surface temperature maximum and melt.
- Page 20 lines 24-29. The argument that blue ice and melting snow are similar in their inherent optical properties is not supported by the literature and should be noted. The optical properties of melting snow drastically differ from those of blue ice. Blue glacial ice is the most transparent natural medium on Earth. Snow is one of the most reflective mediums on Earth. There is a wealth of literature on this topic, with one of the clearest examples shown in figure 16 of the following paper, which compares the extinction coefficients of blue sea ice (which is relatively less transparent than glacial blue ice) to melting snow: https://apps.dtic.mil/dtic/tr/fulltext/u2/a310586.pdf
- The Cheng et al., 2003 states that their the model results were difficult to verify experimentally due to the problems encountered in measuring the temperature profile using temperature sensors embedded in the snow and ice. Too much weight is placed on this study without considering the caveats of radiative heating of sensors, which Cheng describes.
Page 20, line 19. Massom et al., 1997 found summertime values of 8 psu, which seems relatively high for snow. I suggest either changing negligible to "low" or explaining what a negligible amount of snow salinity is in relation to, e.g. negligible for detecting a change in the passive microwave signature.
Page 22, lines 10-22. Somewhere within this section, it would be important to state that this conceptual model applies to thick snow on perennial sea ice in the Antarctic. That's a helpful finding to relay to the community.