Thank you for the revisions of your manuscript and for addressing most points raised by the reviewers. As one reviewer has requested major revisions I have sent the revised manuscript out for review again.
However, to speed up the processes once that review has been received, I would already like to make a few additional suggestions as the Editor of your manuscript and would be glad if you could consider them in your revisions.
This is an interesting paper and very promising approach, and the material is generally well presented.
However, the text is quite compact and I feel that some aspects are still unclear. In particular, you should still better describe what the uncertainty reduction is and how it can be interpreted, despite the new paragraph on page 11, l7-13. Please could you expand this paragraph and better explain what uncertainty reduction really is and how it relates to the terms in equation 1. Maybe you could also include an example, e.g. that uncertainty reduction of xxx % compared to the initial uncertainty of xxx means that parameter xxx is known better by ??? How does this relate to the parameters in table 1?
Your use of the term observations is also confusing. Particularly with regard to page 7, l14-17, page 11, l20, and p12, l5. Please check that these sentences are consistent and maybe include more explanatory text.
Similarly, it is unclear when and why you have used ice concentration and when not, e.g. p9, l14-16, and p11, l17. Why was ice concentration not used? Or was it not used as input variable, but as predicted variable?
The abstract is very general. Can you include some quantitative results? Also I wonder if NASA OIB needs to be mentioned in the abstract. Aren’t the results likewise applicable to other thickness data from the same transects, e.g. from submarine missions?
The long flight lines raise another question. How would the results be different were the flight tracks much shorter and only focused on the study regions and regions slightly further upstream? Can you comment on that? In contrast to one of the reviewers, I do believe that regional studies with less extensive observational data would be feasible too? At least when relative short prediction times are considered. Wouldn’t only ice conditions be relevant in regions and ice fields which could drift into the study region during the prediction period?
Some minor comments:
P3, l24: what is the name of the satellite mission?
P4, l 14: approach THAT operates…
P4, l24: is the southern boundary the Atlantic? Needs to be clarified.
P5, l11: rephrase “precipitation is added to the snow mass”? What happens anyways when air temperatures are at or above freezing, i.e. during the summer?
Spell out PHC (p5), TAF (p8).
P6, l18: why does this guarantee full consistency?
P7, l22: WAS perfect
P10, l12, and figure 2: Why do you describe 9 regions but only use 2 (or 3?) in the paper?
P12, l7: should the 30 cm in the text not be the same as the value in table 1, i.e. 0.5 m for ice and 0.2 m for snow thickness?
P13, l22: Do you mean LOW impact? Why is it remarkable?
P14, l10: do you mean INCREASES ice concentration?
P15, l4: compare to FIGURE 9, or cf Figure 9.
Table 1: precip, given in m/s?
All table and figure captions. Please spell out the acronyms/abbreviations of variables etc. and just include them in parentheses such that the captions are self-explaining and do not require searching the text for explanation.
Fig 1: BS2PB is light blue? How would results be different if flight tracks were shorter?
Figure 12: add annotation with case description to each panel, such that differences between their contents become immediately obvious.