Dear Authors,
Both Reviewers recommended acceptance of your manuscript subject to minor revisions. I agree with the assessment of the Reviewers, both among the top experts in the field, that your manuscript will be a significant contribution to the field of data assimilation -
Rev. 1:
"It is definitely an important contribution to data assimilation theory, with potentially important implications for practice of data assimilation."

Rev. 2:
"The context of the algorithm is very nice in that it explicitly recognizes and samples the errors in the EnKF due to the use of finite size ensembles and nonlinear dynamics. By using a hyperprior assumption to account for these errors, the authors are able to bring analytic insight to a number of more ad hoc techniques (inflation and localization) that are commonly used in EnKF applications."

Thank you for your careful response to the two reviews. I have a few minor suggestions, most of them related to Reviewer comments, for your consideration before publication of your work. In a number of instances referenced below, if I'm not mistaken, you offered an explanation in your response to the Reviewers but made no changes in the text of the manuscript. Would it be helpful to include comments related to your response to the Reviewers in the text of the manuscript, as some readers may have questions similar to those posted by the Reviewers?

1) Rev 1's first comment - some clarification in the text may be helpful?
2) Rev 2's comment #5 - add reference in text to relevant discussion in Section 4?
3) Rev. 2's comment #6 - add clarification into text?
4) Rev. 2's comment #8 - text referenced by the Rev. on p.17, "Assuming linear..." - this sentence is awkward and unnecessarily complicated. Can you please revise (break sentence into two if necessary)?
5) Rev. 2's point #8, his side comment - would you consider adding a reference here to the related discussion that you mention in your response is in Section 6?
5) Rev. 2's comment #9 - Text referenced by Rev. 2 here ("Varying model nonlinear...") is unclear, can you please reword?
6) Rev. 2's comment #14 - suggest you also insert "static" as you refer to fixed, climatological, not case dependent covariances (see you use that term elsewhere in the manuscript)
6) Rev; 2's comment #15 - can you please explain in the text (here and/or in conclusions) the applicability of ENKF-N (i.e., in peered model context, and that it may not be best in realistic, imperfect model setups)?
7) If you like, you could acknowledge the two anonymous Reviewers comments in the Acknowledgement section?
8) Second sentence in Abstract, "fixes" - can you use more formal and more informative language (e.g., "somewhat ad hoc procedures")

With your revised manuscript please include a point-by-point response to the comments above. I will take a quick look at your responses and the revised manuscript before a final decision on the publication of your work.
With regards
Zoltan Toth

Dear Authors,
Thank you for your careful response to the editorial comments. Congratulations on the final acceptance of your contribution and thank you for choosing NPG for publishing your research results. For a minor suggestion see separate (non-public) comment. With regards
Zoltan Toth