Quantifying the nitrogen isotope effects during photochemical equilibrium between NO and NO2: implications for δ15N in tropospheric reactive nitrogen

Li, Jianghanyang; Zhang, Xuan; Orlando, John; Tyndall, Geoffrey; Michalski, Greg

doi:https://doi.org/10.5194/acp-20-9805-2020

Articles | Volume 20, issue 16

https://doi.org/10.5194/acp-20-9805-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/acp-20-9805-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 20, issue 16

Research article

|

21 Aug 2020

Research article |

| 21 Aug 2020

Quantifying the nitrogen isotope effects during photochemical equilibrium between NO and NO₂: implications for δ¹⁵N in tropospheric reactive nitrogen

Jianghanyang Li, Xuan Zhang, John Orlando, Geoffrey Tyndall, and Greg Michalski

Download

Final revised paper (published on 21 Aug 2020)
Preprint (discussion started on 24 Jan 2020)
Supplement to the preprint

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'An important addition to the field that will inspire further studies to validate the model.', Matthew Johnson, 23 Feb 2020
- AC1: 'Response to RC1', Jianghanyang Li, 02 Apr 2020
SC1: 'link between atmosphere and snow', Joel Savarino, 02 Mar 2020
- AC2: 'Response to SC1', Jianghanyang Li, 02 Apr 2020
RC2: 'Review of Quantifying the nitrogen equilibrium and photochemistry-induced kinetic isotopic effects between NO and NO2, by Li et al.', Anonymous Referee #2, 16 Mar 2020
- AC3: 'Response to RC2', Jianghanyang Li, 02 Apr 2020

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Jianghanyang Li on behalf of the Authors (02 Apr 2020) Author's response Manuscript

ED: Referee Nomination & Report Request started (03 Apr 2020) by Jan Kaiser

RR by Anonymous Referee #2 (18 May 2020)

Suggestions for revision or reasons for rejection

The manuscript’s abstract, introduction and conclusions are much improved based upon responses to the reviewer comments. A few important issues remain.

It’s a bit of a circular argument to suggest that because they can “fit” the Freyer et al 1993 observational value for the combined LCIE and EIE of ~ -18 per mil that this justifies their experimental value for the combined LCIE of -10 per mil. There are several ways in which the study needs to be more quantitative in its approach to evaluating the experimental results. Currently, overall, I am unimpressed with the application to environmental data because the data under heavily controlled experimental conditions shows a great deal of variability that is not discussed. How will we ever understand this in the field, with a wide variety of complications versus the chamber experiments if we cannot even fully address the uncertainties associated with the controlled experiments?

Figure 1A and 1B should be dissected QUANTITATIVELY. Best fit terminology is used in the manuscript in a here qualitative manner. For Figure 1A, the actual best linear fit to the experimental results should be quantified and compared against the forced intercept line that is currently pictured. The standard error for both the slope and intercept should be calculated for the best linear fit to the data and this should be compared to assess whether the experimental results are statistically significantly different than (0,0). It appears it might be significantly different and this should be discussed if it is (i.e., what causes the difference from (0,0)? how might this be better addressed in future experiments?).

The response to reviewers includes discussion of the error bars in Figures 1A and 1B, but this also needs to be discussed in the main text and should be indicated in the caption what the error bars represent.

For Figure 1B, the -10 line (which is actually -10.5 in the figure) is simply not the best qualitative fit to both the experimental data and the Freyer observation. Even qualitatively the language should be changed to better address what you are trying to prove here. Better yet, I suggest quantitatively calculating what line actually fits the experimental data…you could even do this while including and excluding the Freyer datapoint. Then address how the actual best fit through the data compared with the theoretical (case-based) lines drawn in Figure 1B.

The lack of correction for isobaric influence needs to be directly addressed in the Methods section. This could in fact make a difference on the order of 0.5-0.8 per mil under the equations used here. While the non-zero D17O correction to d15N is probably not making a difference in the dark experiments, it could make a difference in the photochemical experiments. It is clear in the literature regarding measurement of d15N from N2O that this correction is important and there is no reason for it not at least be addressed here.

The addition of Figure 1C is nice, but I would suggest adding it as Figure 1A and referring to it much earlier on (when defining EIE, LCIE and PHIFE) and then referring to it again after quantifying the values and noting that they are on the figure – the figure caption could also include the values in the figure are determined “in this work”.

Lastly, I am confused by mention (in the response to reviewers) of Walters and Michalski assuming alpha1-alpha2 = 0. My understanding is that the alpha1 is assumed to be 1 such that the fractionation factor is 0; alpha2 is theoretically determined with an epsilon ~ -7 per mil. So that would suggest that a1-a2 does not = 0. Please revisit this to be sure that this misinterpretation is not confusing the results/discussion in the manuscript.

Minor comments:
Line 65 – remove the word “be”
Line 74 – approximations is misspelled
Line 81 – notable should be notably
Line 124 – ranged should be ranging
Line 147 – remove the word “break” before NO2

Line 419 – why mention nitrate aerosols specifically? Why would you not also compare with wet-deposited nitrate?

Line 423-426 – need to be careful here. This addition was in response to the reviewers’ who raised concerns about this. All of the calculations done in Figure 2 are based only on changes in NO and NO2 concentrations and ratios, using an alpha value that is constant. So it is important to consider that it is possible that temperature could have an impact on the interpretation of field data, and this should be investigated based upon comparisons between field collections and the laboratory experimental determinations.

Figure 3 caption – you might add a note that points out the different y-axis for figure 3D.

Hide

ED: Publish subject to minor revisions (review by editor) (18 May 2020) by Jan Kaiser

AR by Jianghanyang Li on behalf of the Authors (20 May 2020) Author's response Manuscript

ED: Publish subject to minor revisions (review by editor) (07 Jun 2020) by Jan Kaiser

AR by Jianghanyang Li on behalf of the Authors (08 Jun 2020) Author's response Manuscript

ED: Publish subject to minor revisions (review by editor) (17 Jun 2020) by Jan Kaiser

AR by Jianghanyang Li on behalf of the Authors (23 Jun 2020) Author's response Manuscript

ED: Publish subject to minor revisions (review by editor) (29 Jun 2020) by Jan Kaiser

AR by Jianghanyang Li on behalf of the Authors (29 Jun 2020) Author's response Manuscript

ED: Publish as is (21 Jul 2020) by Jan Kaiser

Dear Dr Li

I agree that the 17O correction is not an issue, but nonetheless, I thought that forcing a regression through 0 makes a pretty strong statement about the accuracy of your measurement model. Such confidence might be justified for a fundamental physics experiment, but in the "real world" of lab chemistry, it might be a but of an overstatement.

I also agree with your notion that including the dark experiments in the regression and not forcing it though 0 would give a similar result to the one given in the manuscript. However, there are obviously experimental differences in the light and dark experiments.

I'd like to add one final thought: If you do include the three dark experiments in the regression, then it would make sense to rearrange equation 8 so that δ(NOx) becomes a parameter (the intercept). If you do this with all 8 data points, the slope remains (28.5±1.1) ‰ (n = 8; ε = (29.4±1.2) ‰). If you pool the three dark experiments to avoid giving undue weight to the initial conditions, the slope increases to (29.1±1.5) ‰ (n = 6; ε = (30.0±1.6) ‰). Within 1σ uncertainty, this agrees with your result of ε = (28.9±1.9) ‰ and is more than 3σ lower than the result of Walters et al. 2016 (ε = (35.6±1.5) ‰).

Our latest discussion (as well as the previous review history) will be published together with the final paper so that future readers will be able to follow it.

I'll accept the paper as is.

Please ensure that the supplement is removed. Apparently, there was some misunderstanding about it in the Editorial Office, which meant the paper got held up for 3 weeks at the validation stage.

Thank you for choosing ACP for your publication!

Kind regards
Jan Kaiser

Hide

Short summary

Nitrogen isotopic compositions of atmospheric reactive nitrogen are widely used to infer their sources. However, the reactions between NO and NO₂ strongly impact their isotopes, which was not well understood. We conducted a series of experiments in an atmospheric simulation chamber to determine the isotopic effects of (1) direct isotopic exchange between NO and NO₂ and (2) the isotopic fractionations during NO_x photochemistry, then developed an equation to quantify the overall isotopic effect.