l. 162, 168, 186, 189 and elsewhere: Please replace 1/2", 1/4", 1/8", 1/16" with 1/2 inch, 1/4 inch, etc. for clarity.

l. 163, 213, 219 and elsewhere: Reference gas and pressure for the volumetric flow rate should be given (273 K, 1 bar?).

l. 171: Hasn't this been published?
Brand WA, Rothe M, Sperlich P, Strube M, Wendeberg M (2016) Automated simultaneous measurement of the δ13C and δ2H values of methane and the δ13C and δ18O values of carbon dioxide in flask air samples using a new multi cryo-trap/gas chromatography/isotope ratio mass spectrometry system Rapid Commun Mass Spectrom 30:1523 10.1002/rcm.7587

l. 231 + 237: This use of the ± symbol (and the view of referee #1) is incorrect. Repeatability (and uncertainty) is always a positive, so the ± sign should be deleted. See also the Guide to the Expression of Uncertainty in Measurement (GUM), http://www.bipm.org/en/publications/guides/gum.html.

l. 250 + 251: What are the uncertainties of these values?

l. 277: What is the uncertainty associated with the value of 1.58 ‰?

l. 284 + 287: What was the matrix gas of these calibration gases? 80 % N2 + 20 % O2? And what gas was used for dilution?

l. 300: What was the matrix gas of these target gases?

l. 310: Please correct spelling of Eulerian.

l. 311, 322 and 332: Please clarify what tracers are the model - CH4, 13CH4 & CDH3; CH4, δ(13C) & δ(D); 12CH4, 13CH4 and 12CDH3? Strictly speaking only the latter combination be correct, but if the model is linear, the error made due to replacing 12CH4 with CH4 is small (about 1 % of the variations in the δ(13C) value, so probably less than 0.01 %).

l. 356: Electronic state symbols such as 1D in O(1D) should be written in roman (upright) font. http://iupac.org/cms/wp-content/uploads/2016/01/ICTNS-On-the-use-of-italic-and-roman-fonts-for-symbols-in-scientific-text.pdf

l. 444: This was changed back to the original pre-ACPD manuscript. As I pointed out then, "intercomparison" is a pleonasm and does not have any metrological meaning. It should be replaced with "Comparison" or "Intercalibration". The latter term would imply that one dataset is designated as a more reliable "reference" dataset. If no such assignment is made, then the term "comparison" should be chosen.

l. 511: Space missing before "well". Please add a root-mean-squared deviation (RMSD) and/or R^2 value or similar figure of merit to quantify "well" better.

Before the acknowledgements, please add a section on data availability, as per the ACP Data Policy (http://www.atmospheric-chemistry-and-physics.net/about/data_policy.html#data_availability). The supplementary information seems to contain the data from only one of the two instruments used. The data from the other instrument should be added as separate files.