Revisions for “Contribution of blowing snow sublimation to the surface mass balance of Antarctica”

The authors introduce a new version of the blowing snow scheme in RACMO. The new model is described and a first validation comparing the model to local as well as satellite blowing snow measurements is provided. Furthermore, the authors also discuss their results in comparison to an older version of RACMO and another continent-wide simulation over Antarctica. The paper overall fits the scope of the journal.



Below a list of comments if provided, which should be taken into account before considering the paper for publication.

General / major comments:

• For the validation against measurements as well as compared to the other model, the term “significant” is used regularly in the text. It is unclear if significance tests have been performed. Are the changes actually significant based on significance tests? And if yes, which changes are and which aren’t? If this is not feasible at least the wording should be adapted accordingly (see detailed comments below).
• The authors state, that there are differences between the satellite based estimate of blowing snow frequency and the model based results due to the fact that on one hand the satellite based results only start at 30 m above ground, while the results for RACMO include all the layers close to the surface. Furthermore, CALIPSO can only see blowing snow during cloud-free conditions or for optically thin clouds. Did the authors consider excluding cloudy days and the model levels below 30 m for the analysis? What is the reason for not applying these corrections to the displayed result? This would give the comparison more weight, as then the results could more directly be compared (relates to P16, L380ff).

Minor comments:

P1, L7: In ”a non-uniformly discretized ice particle radii to …” radii -> radius

P4, L85: TrasnfEr -> TransfEr

P5, L113: empirircal -> empirical

P5, L114: “, is” does not seem to fit the sentence. Remove “, is” or reformulate.

P9, L223: Is wind speed measured at 2m above ground? Based on Amory, 2020 it seems to be at 2.8 m originally for station D-47. Is 2m an average over the measurement period given the elevation changes due to snow? (if a change is required take it into account accordingly for P10, L261).

P11: In the comparison between the modelled and the measured data, it is shown that the results improve when simulating with blowing snow and even more when using the new version. Especially in lines 270ff the authors talk about significant biases and improvements. Are these statements based on a significance test? What kind of test was used? – If not either the wording should be changed or a significance test should be performed to confirm. In general, I think it would be interesting to see which improvements are significant.

P14, Section 4.1.3: It is discussed that Rp3 overall shows a better performance in predicting the blowing snow frequency overall, while RpNew mainly improves the prediction of higher magnitude blowing snow events. It is great to see that RpNew improves for the higher magnitude events. It would, however, be interesting to see both effects discussed? Why is there an improvement for the higher magnitude events and why does it perform worse overall? What are the impacts of the decrease in performance when looking at the overall blowing snow frequency.

P15, L367: “… 5(a)).This analysis…” -> Missing space after .

P21, L476: “…in under-prediction blowing snow particle concentration.” -> Should probably read: “…in an under-prediction of the blowing snow particle concentration.”

P22, Table 4: From just reading the table caption and the formula, it is not clear that ER_ds only refers to the eroded snow blown off the continent. Please clarify.

P22, Table 4: For “(a) Difference between RpNew and RP3, and (b) SMB difference between RpNew (2000-2010) with CRYOWRF (2010–2020)”: Add information about the two differences shown. E.g. … in Gt (% of xxx). To which value is the percentage related?

P22, Table4b and P23, L509ff: The time periods of the RACMO and CRYOWRF simulations are not the same (RACMO: 2000-2010; CRYOWRF: 2010-2020). The difference in the between the simulated time periods is not discussed. How much might the different periods influence the difference in the results between the two models?

P23, L516ff: The authors state that they can see a trend that the surface sublimation is reduced in the presence of blowing snow sublimation. This is very interesting, can this hypothesis be discussed in more detail. And as a second step, how can it be shown that this trend is not present in CRYOWRF, compared to what? (Same for P24, L556ff).

P23, L518: The authors state that the difference in erosion due to snow being blown off Antarctica is “significant”. Is this based on a significance tests? What test has been used? Please specify or reformulate.

P23, L527: I think it should read “radius classes” instead of “radii classes”

P24, L536: “insitu observation from site D47”: observation -> observations

P25, L570: The authors state that the RACMO data is available, but they don’t say where. Please add a link to the repository from which the data can be downloaded.