This is a difficult amount of information to include in a single publication. As written, the details overpower to overall message of the article. Details regarding the preparation of in-house standards (preparation methods, overall mixture stability, surface treatment) could be moved to an appendix section to allow readers to focus on the direct measurement and value assignment. 

Authors were very careful to warn readers of the broad distribution of value assignment and instability in these mixtures. This does raise questions about the application of these results within participating institutions, discussions of next steps or guidance in analyses of these compounds would be constructive. An additional question, did those with large offsets identify the cause and perform follow up analyses to demonstrate capabilities?  

Questions to authors:

Line 195- 200: "recovery due to initial loss", what is recovery? Was there an initial decrease and then increase on observed amount fraction for particular analytes? Can that be attributed to instrument performance as described later on or is this based on wall effects for the analyte of interest?  

Lines 235-240: Cylinder wall passivation is a significant challenge. Further detail regarding the testing of the passivation approach, with water in particular, would be helpful. While this section described the amount of water injected into the cylinder there is no quantification or stability of water vapor described. Additionally, there is potential for offsets or bias of analytes like MVK in the presence of water. 

General question: How will this publication impact the CMC's for participating institutions? In cases like 31% coverage factor, this would not be useful within the user community.

This manuscript, “Towards a high quality in-situ observation network for oxygenated volatile organic compounds (OVOCs) in Europe: transferring traceability to the International System of Units (SI) to the field” describes 2 approaches to create SI-traceable standards for acetaldehyde, acetone, methanol and methyl ethyl ketone in detail. The motivation for this work is described nicely in the introduction. There are three scientific goals of the paper: to create reference gas mixtures (RGMs) in dry nitrogen at amount fractions of ~100 nmol mol-1 that are stable in time, to create working standards at amount fractions on the order of 10 nmol mol-1 by diluting the RGMs, and to create working standards at amount fractions on the order of 10 nmol mol-1  by spiking whole air samples. This work is highly collaborative and detailed. The sections are well-organized.

General Comments:

1. In general, the paper is highly detailed. I think a schematic describing the overarching process for creating these working standards could assist with the clarity of the paper, and anywhere a schematic could reduce details included as text would be appreciated (i.e. for the filling of the parent and subsample cylinders, or the diluting system).
2. It is difficult to compare the two methods used to create working standards. For diluted RGMs, the results are shown in percent difference compared to the assigned amount fractions, and for the spiked whole air samples the results are shown in amount fractions plotted alongside the assigned value. Do the authors have a reason for providing the results in this way? It may be more helpful to future users of these techniques to present both in units of percent differences.
3. In section 4.1, the authors describe the methods used to assess the working standards based on whole air samples. My understanding is that the cylinders were large enough to be sent to all the laboratories, but canisters only had enough volume “for one analysis”. Could the authors comment on their recommendations for going forward with this technique? Would monitoring stations be expected to purchase a cylinder (for long term use) or a canister (for one time use) of working standard? Does the cost of manufacturing these standards seem worth the trouble for monitoring stations?
4. Could the authors expand on what they mean by “molar masses” as a source of uncertainty during the preparation of RGMs (line 192)?
5. On line 397 the authors write, “Compound loss after preparation due to surface effects might explain relative differences of around -5 % for ethanol.” Do the authors mean that ethanol had relative differences of around -5 %, and the most likely reason for that is surface effects? On first read, I thought the authors were suggesting they had done a calculation for surface effects and could explain -5 % of the ethanol differences.
6. On line 430 the authors write, “vessel material may play a role in the lack of homogeneity…”. Was there a material that performed better than others? This is an interesting result.
7. On line 549 the authors write, “Assessment results for amount fraction levels around 10 nmol mol-1 suggests that SI-traceable working standards based on dilution of RGMs can be used as calibration standard at monitoring stations for key OVOCs, such as acetone. For other OVOCs, like methanol, acetaldehyde and MEK, suitability is not so clear due to the large uncertainty and result dispersion, particularly at amount fractions < 5 nmol mol-1.” However, in the conclusion, line 610 is, “The first type, based on RGMs diluted with dry nitrogen, using for that a portable dilution system that ensures SI-traceability after dilution, seems to be suitable for calibration of acetone, MEK and methanol at amount fractions around 10 nmol mol-1.” These statements seem conflicting as written.

Technical Notes:

Line 67 - typo “atmosphere and play and important” should be “atmosphere and play an important”

Line 68 - “the main sinks are the oxidation with the OH radical and…” would be better worded as “the main sinks are oxidation with OH radicals and…”

Line 81 - typo “emphasises” should be “emphasise” or “emphasize”

Line 118 - should “metrological accepted term” be “metrologically accepted term”?

Line 214 - “PTFE” and “SS” are used for the first time, but not defined as acronyms

Line 233 - typo “consisted on certified” should be “consisted of certified”

Line 245 - typo, no need for comma after “both”

Lines 320 and 799 - could not find an Eq. (B1); should a different equation be referenced?

Line 326 - CRF is used for the first time, but not defined as an acronym

Line 553 - “contributed to that”, replace ‘that’ with the true meaning

Line 671 - typo “the carrier was helium” should be “the carrier gas was helium”

Table B1 - the subscripts appear in the order “2,3,1”, consider rearranging to “1,2,3”

The current Table B3 appears in section B2, and the current Table B2 appears in section B3. It seems like renaming the tables (and their text references) would fix this.

Line 738 - typo, missing a close-parenthesis

Line 828 - typo “Prior instrument calibration, …” should be “Prior to instrument calibration, …”