|2nd review of the manuscript "Satellite-derived sea ice export and its impact on Arctic ice mass balance" by Ricker et al.|
The manuscript improved from the last version and I now find it suitable for publication in The Cryosphere after some minor points are addressed.
I had three mayor concerns.
The first about the ice volume flux calculation got resolved in the new version. I only have a minor comment left (see below).
My second concern was that the three ice volume export estimates presented here do not agree within their uncertainty estimates. The authors did not change that point but they now more clearly state that there are biases between the datasets, which are not taken into account here.
My third point was was only partially addressed: it was questioned if the residual between MYI export and MYI volume can really be attributed to MYI growth or mainly is an effect of the coarse, binary MYI/FYI classification used in this study. I would have hoped for a more critical discussion of the uncertainties caused by the MYI mask used here. Most areas in the Arctic are a mixture of MYI and FYI. The authors now added a sentence acknowledging the new ice growth in leads but make no efforts to quantify or at least estimate the effect of ice volume growth in leads to the overall mass gain in MYI areas. For example, ice growth of thin ice in winter easily is a magnitude larger than ice growth for MYI and therefore already small percentages of lead area can significantly contribute to the ice volume change within a 25km grid cell, which still would be classified as MYI here. Anyway, we can agree to disagree. In the end it's their conclusion and not mine and the methodology used is explained correctly and this point is not as prominent as before in the conclusions now. As said before, the most basic information was now added.
I have a few minor comments left which should be addressed before publication.
Pages and lines refer to the new version of the manuscript with changes marked.
p1,l5: ice drift
p2,l2: what do you mean with "this period"? Oct-Apr? Please clarify. Then this would be inline with my previous criticism that MYI volume change is comparably small.
p2,l14-31: I guess somewhere in this paragraph also the FS ice volume export time series obtained from ULS should be mentioned (Vinje et al., Kwok et al.).
p4,l7: "IFREMER" according to your own definition
p5,l4-5: Here one could cite one of the OIB studies supporting this 50% over FYI assumption, e.g., Kurtz et al., 2011 or Newman et al., 2014
p5, l16: The OSI-SAF ice type is the same product used for the CS2 thickness retrieval. What for are you using it here in addition? Why are two different user manuals cited (Aaboe et al. and Eastwood et al.). Could this be made consistent? Or are you indeed using two different ice type products?
p5,l21: do you mean "new ice forming in openings"? I rather would call them leads
p5,l24: "could be a result"
Fig. 1: "Means of Arctic sea ice volume fluxes ..." sounds strange.
p7, eq. 2 & 3: Okay, now I understand the equations. You write "... through the defined gate...", which is correct as the gate only varies between -12° and 20°E. However, maybe it is still worth mentioning that the equations are only valid between -45° and 45°E.
p7,l25: I guess you mean "Zonal uncertainties ..."
Table 2: Maybe some exceptionally high and low values could be marked and the monthly average for all years could be added to help the reader
p13,l6: They also would be larger for larger scales. I would remove that half-sentence. It is enough to remind the reader that these numbers are for a 25km scale.
p16,l16: "Sea ice" (I actually do not understand why you have removed the hyphen from all sea-ice volume, sea-ice export etc. word combination. In my view it is correct with hyphen if two expressions like "sea ice" and "volume" are combined).