Summary: The authors present reactive nitrogen measurements from the recent AQABA campaign, a ship-based summertime field campaign in the Mediterranean Sea, Red Sea, and the Arabian Gulf. Using a thermal dissociation cavity ring down spectrometer, they measure NOx (NO + NO2) and other reactive nitrogen species NOz (HNO3 + HONO + pNit + ClNO2 + N2O5 + others). Using these measurements and other auxiliary observations, they derive mean NO2 concentrations and lifetimes, cumulative NO2 losses during the day and night, and mean ozone production efficiencies. They also use back trajectories and ratios of NOz / NOy and NO2 / SO2 to attribute their observations to local ship emissions, near-coast industrial activities, and longer range transport.

This is a very well put together manuscript. The results are robust and thorough, and very useful for the community, especially given the lack of in situ data in the Red Sea and Arabian Gulf. I would recommend publication, following a few minor edits/suggestions for the authors first.

General comments:

In the abstract (i.e. page 1 line 15) and in the methods section (i.e. page 5, line 18), the TD-CRDS instrument is presented as the primary instrument for both NOx and NOz, but it seems that the NOx measurements are all from the CLD instrument, which should be clarified. Also, were the NOz measurements derived from NOy – NOx(CRDS) or NOy – NOx(CLD)?

What was the methodology used to select the sub-regions in each Sea? Were they grouped together by eye? By some kind of filter? By similar HYSPLIT back trajectories? I ask because it looks like there is some data that is not considered in any of the sub-regions. If the sub-regions were selected by eye, I worry that it would bias the results in the paper to more “extreme” values, because that’s what stands out most. Was there any attempt to do some kind of statistical analysis to group together regions with similar chemistry/emissions influence?

Related to the previous comment, do those sub-regions include an entire diurnal cycle? Do each of the sub-regions capture the same segment(s) of the diurnal cycle? If not, have the authors considered how this may change the relative contribution to NOz of the various species?

Specific comments:

Page 4, Line 21: How long is the inlet upstream of the TD? Is there any concern about “sticky” gases such as HONO or HNO3 being trapped in the inlet walls before entering the TD?

Page 5, Line 14: The authors later explain that PM1 comes from the AMS instrument, but since that’s in the next section, it’s a bit unclear here whether the PM1 measurement is coming from the OPC.

Page 5, Line 33: Does the CIMS instrument output a HNO3 or a HONO measurement, even if uncalibrated? Can that signal be correlated with the calculated HNO3 or the LOPAP HONO measurement?

Page 6, Line 11: What is the path length of the HONO instrument? Is it co-located with the other instruments?

Page 6, Line 26: Were HYSPLIT back trajectories calculated at all locations along the ship track, or just those representative locations shown in the figures?

Page 9, Line 23: Here and in a few other places, regions are described as being “influenced by land-based” pollution. However, it seems that depending on how populated the land is, “land-based” can either mean free from fresh emissions, or heavily influenced by fresh emissions. Can the authors clarify, or perhaps use terminology that describes the regions in terms of their influence from fresh emissions, either ship- or land-based?

Page 10, Line 11: Figure 4c doesn’t really show the averaged NOx lost per day and per night, just the cumulative. Can the authors add those numbers to the figure, perhaps under the legend, or remove the reference to that figure in this line?

Page 17, Line 20: How do the authors define “background mixing ratios of NOx” here?

Page 20, Line 5: How are OPE and NOPR defined differently here? It’s a little unclear how these two parameters relate to each other.

Figure 14: The fitted light for the Eastern Mediterranean (light blue) is very difficult to see over the experimental data.

Friedrich et al. report on measurements of 'reactive nitrogen' as part of the shipborne 2017 AQABA campaign. Data were acquired using a custom-built thermal dissociation cavity ring-down spectrometer recently described by Friedrich et al. (2020) supplemented by numerous auxiliary measurements. In this paper, the authors dive into some of the analysis of this rich data set, focussing on the conversion of NO_x to NO_z in 3 regions, on HONO budgets and on ozone production efficiencies.

The paper is written well. It is a bit too long, and there were some organizational shortcomings (see below) that should be addressed before the paper is accepted.

Major comments

1) Organizational / presentation issues  

(a) NO_z/NO_x ratios were investigated in 3 regions (3.1-3.3 - Mediterranean Sea, Red Sea, and Arabian Gulf). It is not sufficiently clear why the data set was divided in this way.

(b) A brief but incomplete analysis of HONO is presented in section 3.4. The introduction does not mention HONO (other than in R5) and this section appears "out of the blue".  Consider removing this section if the plan is to write a separate paper on HONO anyways.

(c) In section 3.5, ozone production efficiencies (surprise!) are calculated, which should be discussed in the introduction and perhaps also mentioned in the title.

(d) Information already presented elsewhere should be removed and the previous paper(s) cited - e.g., Fig 1 of this paper is similar to Fig 1 / Fig 3 of Tadic et al. (2020), Fig 9 and 13 and Table 1 of this paper and Fig 7 of Eger et al. (2019). Likewise, rather than describing the TD-CRDS over 2 pages (section 2.1 - page 4 and 5) I would suggest simply citing the earlier Friedrich et al. (2020) manuscript.

(e) Pages 2-3. The section on nitrogen oxide chemistry in the introduction is written well, but in my opinion is not needed - similar sections of text have been presented numerous times, including by the authors themselves in recent papers. It would have been more informative and interesting to tell the reader about what made the AQABA campaign interesting and worthwhile (e.g., effects of temperature/climate and mineral dust on nitrogen oxides, special/unique NOx sources in the regions etc.) and add more background on ozone production efficiencies (see g).

(f) A table summarizing the various measurements and techniques would help.

(g) The OPE values calculated need to put more into context of existing literature (pg 19 lines 24-30). Consider stating in the introduction what values are typical or would be expected and expand the discussion.

(2) (Perceived) lack of novelty.

There have already been at least 7 papers presenting results from the AQABA campaign, yet the introduction avoids telling the reader what was presented in the earlier papers. In the introduction, it should make clear to the reader what new information and/or analysis are presented in this paper and why this paper is worthwhile. In particular, it should be stated how this paper differentiates itself from Tadic et al. (2020), "Net ozone production and its relationship to nitrogen oxides" to avoid the perception of duplication (in particular of section 3.5).

(3) NOx lifetime - pg 9 line 30.

The equation given here is too simplistic in my opinion. Equation (1) should account for N₂O₅ formation, which can increase the k₁₃[O₃] term by a factor of up to 2 (see Brown, S. S., et al. (2004), Nighttime removal of NO_x in the summer marine boundary layer, Geophys. Res. Lett., 31(7), L07108, doi:10.1029/2004GL019412.). There are also sinks such as the heterogeneous conversion of NO₂ to HONO/HNO₃ that may need to be considered (mentioned on page 18, lines 8-).

Minor/Specific comments

In the future, please number continuously and do not restart numbering on each page. You are creating more work for the reviewer which is in nobody's interest.

Abstract, line 20 - HONO. Consider stating how the role of HONO was assessed (were there measurements?)

Abstract line 24 - OPE. Consider stating how OPE were calculated (plots of ΔO₃ vs ΔNO_z ?)

pg 1 line 8 - Please define "M"

pg 3 line 2 - "lifetimes of a few hours". The lifetime of PAN may be much longer aloft are in Arctic environments.

pg 4 line 19 - please specify the make/model of the 3-way valve and state what the internal surfaces are made of.

pg 4 line 31 "adding 19 ppmv of O3" - Please clarify if this mixing ratio refers to the amount of O3 (in O2?) added (in which case also state the flow) or if "19 ppmv" refers to the amount of O3 after addition to the sampled air.

pg 5 line 4. Does the TD-CRDS respond to nitrate associated with mineral dust which may occur in the study area (e.g., https://acp.copernicus.org/articles/16/1491/2016/)?

pg 5 line 20. Please state here how the detection limit was defined (move up from line 27) and also state how long data were averaged (longer averaging times => better detection limits).

pg 5 line 26. Why correct for humidity?

pg 6 line 14-16 How can the uncertainty of j data be 10% if upwelling radiation was not included?

pg 6 line 25. Which meteorological field was used for the HYSPLIT trajectories?

pg 17 section 3.4 "HONO formation"

The analysis appears to be only considering daytime processes in this section. How does the HONO budget during AQABA compared to the observations by Wojtal et al. (2011)?

pg 20 line 30. "HONO photolysis was as a significant source of NOx." It is not a net source if HONO is generated from NOx.

References

Wojtal, P., Halla, J. D., and McLaren, R.: Pseudo steady states of HONO measured in the nocturnal marine boundary layer: a conceptual model for HONO formation on aqueous surfaces, Atmos. Chem. Phys., 11, 3243-3261, 10.5194/acp-11-3243-2011, 2011.