Dear Colleague,

We have received for your manuscript comments from two referees who both found your work on the oxygen dynamics in the North Sea very carefully carried out and even setting a new standard in oxygen modeling. Modeling well oxygen behavior is a tricky business because of the intricate role of physical, biogeochemical and biological processes. Both referees mentioned it is suitable for publication but advised to improve material and findings so as to make your study much stronger and easier to read. They both recommended to shorten drastically the length of your article by moving more technical methodological material into an Appendix and to better articulate your findings around a clear research question. You should also clearly separate the Methodology from the Results/Discussion section.

Referee N°1 would also prefer to see the results and discussion section split up into two separate parts. I agree with all previous points above except the latter one. I would suggest you simply separate much better in each of the different subsections your own results description and the subsequent discussion.

I want to mention here that modellers often do not worry about validation because it might look too technical. For me, it represents the basics of a carefully performed modeling study. So I would like to congratulate you for having achieved this in detail. However it renders the paper too long. A reasonable compromise would be to keep the Taylor diagrams (which integrate the assessment of the model performance) within the main text along with a shortened qualitative validation allowing to provide the basic necessary insight into the North Sea oxygen dynamics. You can also follow easily Referee N°2 comments to shorten by referring to original publications, when applicable, for methods or description of in situ observations or monitoring programmes. Moving Section 3.1 “Evaluation of the stratification and MLD criterion” (including Fig. 3) to an appendix is a wise choice. Your suggestions to answer this comment from Referee N°2 seem suitable to me.

I fully agree with Referee N°2 about the misuse of hypoxia, please use rather oxygen deficiency throughout your manuscript.

As to identify a clearly formulated question, this needs to be stated at the very start of the paper. I agree that the scope of your study is to show that biogeochemical models are capable of providing a temporally and spatially consistent picture on oxygen conditions with respect to their occurrence and extent. In addition, they provide detailed information about the processes leading to low oxygen conditions and by this improve our understanding of the system, in our case the North Sea.
What are the controlling drivers of the oxygen dynamics in the various subregions of the North Sea Why low oxygen conditions do or do not occur under similar conditions of stratification and in different regions of the North Sea?
These could well be your overarching scientific question.

With respect to the choice of the subregions and their representativeness, I would recommend you do what you are suggesting that is to conduct a grid-cell based analysis using the key parameters for low oxygen conditions identified by your current analysis – similar to the approach by Druon et al. (2004), but based only on stratification period, water depth/sub-MLD volume and primary production.

Given this general guidance on the major revisions to be achieved, you would like of course to also consider the more specific comments and technical corrections raised by both referees for editing your manuscript.

Looking forward to receiving an improved updated version,
Best regards,
Dr Veronique Garcon

Dear Colleague,

We have received your detailed answers to both referees as well as your revised manuscript. We acknowledged the great care you took in addressing most of the comments they addressed to you and your co-authors. In particular contributing to shorten significantly the manuscript by reshaping/moving subsections in Appendix and putting forward a clearly formulated research question.

I would require, before final publication, a few minor revisions listed below to be undertaken. Numbers of lines refer to your last version of revised manuscript.

In the Abstract, you say on Lines 9-11: to quantify the impact …bottom: but it would be more instructive to the readers to actually say what are the processes playing the key role in controlling oxygen dynamics here. Alternatively, on Line 16: After Oxygen deficiency, a short sentence explaining why this higher susceptibility to O2 deficiency in type (2) areas of the North Sea would be beneficial in the Abstract. We should find in the Abstract the answers to your 2 well formulated questions on lines 122-124 of the Introduction section.

Section 2.2: Re-reading this sentence on lines 255-257: In addition…. grid cells occurs, I found it quite uneasy to understand the definition of MLD, can you rephrase?

Line 258: differ (without s)

You should move the definition of the Oxygen deficiency index (ODI) to section 2.4.2 in Section 2 Material and Methods. You argue in your response “that the problem is that the set of parameters, by which the ODI is described, and the equations actually are a result of the analysis based on Table 1 and the matching of simulated bottom O2.” It is puzzling since the definition of a robust index should not depend on downstream results.

Section 2.5: I would be cautious when you refer to interannual variability throughout the text, comparing 2 years (such as 2002 and 2010 line 372) does not constitute per se a proper interannual variability analysis , but simply a comparison of 2 years.

Section 3.1.1.: You can still reduce the length of the descriptive part on the evolution of bottom O2 at North Dogger in 2007-2008 and at MARNET stations Ems in 2010-2011 (lines 399 to 438).

Line 437: besides these short-term changes in the difference … are less than..

Section 3.1.3 Taylor diagrams for quantitative assessment of model performance are nowadays quite classical so I think you can omit having a subsection just on this. Rather insert at the end of section 3.1.1 the few lines on your figure 5 (time series) and end of section 3.1.2 the few lines on your figure 5 (spatially resolved data).

Line 616: Division of this mass…

On lines 615-617 you derive an estimate of the amount of time required for the consumption of the entire amount of O2 available in Vsub. A few lines above (lines 612-614), you compute the daily O2 consumption rate using a 6 months duration period. This is not clear to me.

Lines 687-688: Again I would not say the interannual variability here, but simply difference between years.

On Figures 8, 9 and 10 please enlarge all lines since some are hardly visible (yellow, green, light blue, light beige)

I think section 3.4 can be shortened significantly by not detailing all the numerical values of all physical and biological processes for both years 2002 and 2010.

Line 815-816 : do you mean : however it is also visible…?

Presently the conclusions section is far too long, there you should offer the reader a brief synthesis of the major findings of your study. Do not recall on which section (line 972) this was found for instance.

Lines 982-984 should go as explanation in the abstract.

Section from lines 1043 until line 1104 can be shortened without loosing much information.

Dear Colleague,

I first apologize for the slow handling of your last revised version, uploaded on March 14, 2016. I went through all the corrections you brought about and I think you have done a proper job. I hope you found this useful for the improvement of your manuscript. I would be tempted to say that you could further shorten the conclusions section (line 998 to line 1036) since the Marine Framework Directive and the response of the North Sea oxygen dynamics under climate scenario are a bit beyond the scope of your study but I leave to you to decide...

Best regards,
Dr Veronique Garcon