Dear authors,
after the second round of revisions, both referees still come up with a list of important suggestions for improvement of your paper. Since they acknowledge that the revised version is strongly improved compared to the original manuscript, I anticipate that you could satisfy the referee comments in a new revised version.
Therefore I ask you to take the referee comments into account and prepare a second revised version, together with a detailed response to the referee comments.
Best regards
Thomas Röckmann

Dear authors,
the referees have evaluated your revised manuscript again and both conclude that the paper is strongly improved now. Still they have suggested a number of changes that I would like to you to take into account. In particular the point of referee 1 (autocorrelation issues) and the comment of referee 2 on the definition of the DT and related arguments in the manuscript seem important.
Please revise the manuscript accordingly and submit it together with a rebuttal letter stating the applied changes and the answers to the referee comments.
Best regards
Thomas Röckmann

Dear authors,
I am glad to inform you that you have addressed the referee comments satisfactorily and your manuscript is now accepted for publication in ACP subject to technical corrections. Please consider the minor changes that I have suggested in the attached file and submit the final version to the editorial office.

I have one more request regarding the following referee comment:
Referee 2 : p.18, l.19 The numbers given here differ from the ones given in the
supplement.
ANSWER : It is normal : the O3 trends given in the manuscript take into
account the autocorrelation while the results reported (for information) in Table
S2 in the Supplement do not (the trend best estimates are the same but their
uncertainties are larger when the autocorrelation of the data is taken into
account).

Please add then in the supplementary table a note describing this difference.

Best regards
Thomas Röckmann