Dear Kenneth,

thank you very much for your replies the reviewers and my concerns. these seem quite reasonable and I thus invite you to send in a revised version, which will then be sent out to the reviewer.

Best, Helge

Dear Kenneth,

your MS has no been re-reviewed positively, but there are two issues that remain from the reviewer's point of view (which I share). Before accepting the MS for publication, I would thus like to ask you to work on them. Below you'll find the reviewer's comments. If you feel confident that you can address the remaining issues, I would welcome a revised version of your MS.

With the best wishes, Helge



Re-Review of manuscript BG-2018-61-AC1

I have read the revised manuscript and the responses from the authors. Most of the have been adequately addressed. However, I two remaining requests:

1. Regarding the core of my main concern – that the suggested ventilation method takes long time and requires substantial amounts of power to reach background levels. This is acknowledged in the response and mentioned more briefly in the revised main text, but it is important that readers get more information about the nature of this issue and the consequences than given in the present revised version.

This is a rather serious issue because it means that the quality of the measurements will depend on the extent of previous gas accumulation – in turn depending on all factors affecting gas accumulation levels e.g. pCO2 and the piston velocity during the previous accumulation period which both can change rapidly (meaning that the effect is not constant over repeated measurements and can therefore not be corrected for easily). Hence there will be a varying carry-over-effect making subsequent measurements partly depending on eachother. This is a fundamental statistical problem as we prefer to work with independent measurements – not measurements being interdependent. There will therefore be differences in measurement quality/precision among different systems and over time in ways that are difficult to predict or correct for. In some cases, this issue may be less important, but in other cases it will be a serious issue, and if a very large fraction of the future data on aquatic CO2 emission would suffer from these issues, it would be a concern.

The above issue is a large enough concern to make others abandon this venting approach previously. Nevertheless I think it would be good if this manuscript is published to clarify this alternative as it may be fine in some cases and stimulate to further development. In the published paper it is crucial to not only highlight how easy and inexpensive this approach to vent chambers are, but also to thoroughly inform the readers about the concerns and the consequences for the data produced.

I therefore request implementation of the below in the main text:
- Please clearly state the pumping time and power consumption needed to really ensure reaching within 5 % of background levels in the described system.
- Please provide a figure showing the chamber headspace gas concentrations over time during the venting period is needed to facilitate reader understanding of this issue.
- Provide a clear explanation that the carry-over effect and the starting level reached after a specified pumping time will vary over time and in space depending on multiple factors, that users must be aware of varying data quality and that the best practice would be to pump long enough time the get close to the natural background levels.
- Please clearly in brief explain this issue and need for careful consideration of pumping time in the abstract.

2. On the comment regarding the potential of gas transport through the open tube during measurement periods: A test with a sealed chamber amended with CO2 was made in the lab as far as I understand showing no gas exchange over the open tube. A question was if it could not be different outdoor due to e.g. wind effects. I interpreted the response that not problems had been encountered. I think this may very well be the case and this may likely not be a big problem. However, I am also not sure gas exchange causing a continuous bias would be easily detected from the measurements, and we should be careful to test these small things before many start to use the approaches. As I think this would be easily tested by just using the same approach reported from the lab, but outdoor (and maybe even by enforcing potential wind effects using e.g. a fan) during an hour or so, I would recommend to do this test and thereby confirm our belief that this is not a problem with solid data.

Provided these suggestions are implemented adequately in the main text I congratulate the authors to an inspiring paper that can become valuable for increased future data collection.

Dear Kenneth and co-author team,

thank you very much again for submitting your interesting work to BG. After the successful implementation of suggestions/criticism raised by the reviewer(s), I would hereby like to accept your MS for publication.

With the best wishes, Helge