Dear authors,

Thank you for responding to all of the reviewers' requests. I congratulate you to your very succinct and clear contribution, which can be almost accepted as is.

Please find herebelow some technical recommendations that need to be taken care of before the article is published.

With best regards

Christof JANSSEN


Technical comments (page (p) and line (l) numbers correspond to the author response version of the manuscript)

p. 2, l. 7: "must be minimised which implies consistent data series with traceability to common reference standards." Consider "reduced" instead of "minimised". The relative clause is not entirely clear. Do you mean "in order to obtain consistent" instead of "which implies" ? Please reformulate.
p. 2, l. 11, 14: system- -> system
p. 4, l. 7: by WCC-Empa to CO, CO2, and CH4 -> to CO, CO2, and CH4 by WCC-Empa
p. 4, l. 9: supplies on -> supplies to
p. 6, l. 26-27: the method ... in analogy -> a previously described method for CO2 and CH4 (Zellweger et al (2016)) was applied to CO and N2O.
p. 7, l. 17: "This was the case at the 95 % confidence level for the slope, which deviated with a mean value of 0.994 ± 0.068 (1 sigma) ..."
The 1 sigma value is likely the dispersion of the slope values and not the error of the mean. Why is the 95 % confidence level mentioned here ?
Consider replacing the phrase by "This was the case for the slope with a mean value of 0.994 and dispersion (1 sigma) of 0.068, which is not significantly different from one (t-Test, p = 0.49)."
p. 7 l. 18: The previous remark also applies to the bias. It should be made clear when the error of the mean or the standard deviation is meant.
p. 9: change corresponding mean sigma notation on this page as well.
p. 10 l. 1: Round Robin Experiment -> round robin experiment
p. 10 l. 6: still made based -> still based
p. 10 l. 6: which is in the case of N2O -> which in the case of N2O is
p. 11 l. 15: what is 1 sigma of the bias ?
p. 11 l. 20: factory water vapour correction -> factory implemented water vapour correction
p. 12 l. 8: check bias notation 1.20 ± 0.57. Prefer bias of 1.20 nmol/mol with dispersion 0.57 ,nmol/mol (1 sigma). What is the error of the mean ? Is the bias significant ? There are other instances of this misleading notation on this page. Please correct.
p. 13, l. 15-16. Double use of the word "use": were successfully used -> were employed successfully.
p. 13, l. 28. the standard gases must also pass though the dryer to compensate for a potential loss over the dryer. -> the standard gases must be passed through the dryer to compensate for a potential loss in the dryer.
p. 13, l. 21. which are essential -> which is essential
p. 13, l. 24. though drying of the sample air with -> through drying of the sample air with either
p. 25, l 3 : Boxplot of the slopes uncertainties of the of the regression analysis for the CO performance audits for ... -> Boxplot of slope uncertainties from the regression analysis of the CO performance audits for ...
p. 25, l 4: denotes to -> denotes
p. 27, Figs. 5 and 6: use amount fraction instead of mole fraction in the figure legend to be consistent with text.