DearAuthors,

thank you for carefully going over the manuscript and accommodating many of the reviewers’ suggestions to improve it. By and large, myself and both reviewers find the manuscript publishable now. However, there are two important concerns raised referee #1 that I fully share, namely the use of multiples of the Kettle et al. budget terms in the model, and the inclusion of SH sites:

1. Kettle et al. multiples
The suggestion made in several recent publications to significantly rivise some of the Kettle et al. source and sink terms upward is partly based on constraints from atmospheric measurements and partly on a somewhat revised mechanistic understanding. If both plant uptake and ocean emissions were indeed higher by a factor of 2 or 3, then it seems not unlikely that the regional and seasonal distribution would also need to be different. You have accounted for this to some extent by increasing the ocean source strength more in the tropics and less in southern high latitudes, but I could imagine that some fairly small changes to the seasonality in the NH plant uptake and ocean fluxes would have very similar effects on your modeled OCS as the massive overall increases you applied.
I would think that just showing the simple unscaled Kettle et al, 2002 inventory (combined magnitudes and spatiotemporal variations of sources and sinks) to obviously not being able to reproduce your observations might be a much stronger and safer statement to make. And with the SiB model you have something that improves the fit dramatically in many cases, and that is much more closely linked to CO2 and the carbon cycle than any Kettle et al. scaling.

2. Inclusion of SH sites
The main reason you bring up for not including data from tropical and SH stations are concerns about the presence of water vapour compromising the retrievals at SH sites. I find this not entirely convincing, given that in a recent paper on S.H. OCS trends by more or less the same group of authors (GRL, doi: 10.1002/2015GL065879), water vapour was not pointed out at all as a potential problem for the retrieval and resulting data quality. The retrieval approaches in both papers appear very similar (except for the use of one of the four OCS lines used, namely p18 instead of p19), and the trend paper states that water vapour is corrected for as an interfering species. Plus, I don't think that there is so much more water vapour in the atmosphere at, for example, Arrival Heights than at some of the NH sites you use. The second argument presented against including SH sites is the relatively lower sensitivity to the land fluxes that are the main focus of your study. But even in the NH you do need to include modified ocean fluxes to balance the budget, and I don't think that the simulated seasonal cycles are completely insensitive to the applied ocean fluxes. The comparisions for SH sites would provide a great way to isolate the seasonality resulting from the ocean fluxes (at least to some extent) and I think it would add valuable information to the interpretation of your NH results.
You promise in your outlook that you will eventually include SH sites in a similar study, but the case why you can't include this right here and now is not strong! Including these sites would greatly strengthen your paper in terms of scientific significance and quality, and make the value to the community much larger.


Now, I will not reject your paper if you decide to keep the simulations with the Kettle et al. multiples and to not include the SH sites. Nevertheless, I ask you to carefully consider taking these extra efforts that would in my view turn a fair paper into a very good and highly valuable paper. I'm sure that many others would appreciate this just as much as myself and at least one of the referees.

Best regards,
Marc

Dear authors,

thank you for sharpening the model results section with respect to using Kettle et al. multiples. With respect to using southern hemisphere data, you're obviously not in the position to include them in the present manuscript. This is a real pity, but it seems the way it is.

I hereby accept the manuscript for publication in ACP in its present form.

Best regards,
Marc von Hobe