the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 12 Sep 2025
Response relationship between atmospheric O3 and its precursors in Beijing based on smog chamber simulation and a revised MCM model
Abstract. Ozone (O3) pollution has been receiving increasing attention, but its simulation performance in models remains unsatisfactory. This study characterized the response relationship between O3 and its precursors in the atmospheric relavant condition through a combination of smog chamber experiments and Master Chemical Mechanism (MCM) box model. By adding chamber wall related reaction mechanisms, the model achieved significant improvement in simulating O3 with an Normalized Mean Bias (NMB) value changing from -76.1 % to -12.7 %. The enhanced model was subsequently extended to the ambient atmospheric conditions in the Daxing District of Beijing, incorporating the parameterization of ground related reactions, heterogeneous reactions of Nitrogen Dioxide (NO2), and unidentified NO2 sinks. Compared to basic model, the resulting revised model demonstrated substantially enhanced accuracy in simulating ambient O3 concentrations with an Normalized Mean Bias (NMB) value changing from 113.8 % to -5.2 % and enhanced O3 formation sensitivity to Volatile Organic Compounds (VOCs) in Daxing District. These findings underscore that incorporating interface mediated chemical processes and accounting for unidentified NO2 sinks into model is critical for determining the sensitivity of O3 formation and optimizing regional O3 pollution control strategies.
- Preprint
(1248 KB) - Metadata XML
-
Supplement
(1035 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2025-3956', Anonymous Referee #1, 02 Oct 2025
-
AC1: 'Reply on RC1', Jialin LuLu, 20 Jan 2026
We sincerely appreciate the your careful review and valuable guidance. The manuscript has been thoroughly revised according to the your suggestions, and all changes have been clearly highlighted using the Track Changes mode in the revised version. Enclosed please find our point-by-point responses to your comments for your kind consideration.
-
AC1: 'Reply on RC1', Jialin LuLu, 20 Jan 2026
-
RC2: 'Comment on egusphere-2025-3956', Anonymous Referee #2, 26 Nov 2025
This manuscript outlines attempts to improve the modelling of O3 concentrations, and the production of NOx-VOC isopleths, by box modelling simulations. The paper has two halves. The first half describes a series of chamber experiments in which a chamber-specific mechanism is produced to improve O3 concentration predictions. The second half attempts to use the developed chamber-specific mechanism to account for model biases in a box model of the ambient environment, suggesting that heterogeneous reactions could account for over-predictions in ozone concentrations from the base simulation.
The experimental procedures for the chamber experiments seem sound, and while there is some room for improvement in the explanation of the method by which the chamber-specific mechanism rates are selected (as explained later), the produced mechanism appears to be in line with existing literature and reproduces the observations well.
However, I believe that there are numerous issues with the implementation of the ambient box models, including the author’s focus on NO2 uptake as the sole explanation for model discrepancies. Some of these issues may be clarified with additional context in the explanation, but others may require deeper consideration. I have organised my comments in the attached document into 4 sections: NO2 Sink, General Comments, Specific Comments, and Minor Comments. I do believe there is value in efforts to improve O3 predictions and that the potential role of NO2 uptake is an important consideration, and I hope that addressing these comments can provide the improvements in the manuscript that I believe are necessary for publication.
-
AC2: 'Reply on RC2', Jialin LuLu, 20 Jan 2026
We sincerely appreciate the your careful review and valuable guidance. The manuscript has been thoroughly revised according to your suggestions, and all changes have been clearly highlighted using the Track Changes mode in the revised version. Enclosed please find our point-by-point responses to your comments for your kind consideration.
-
AC2: 'Reply on RC2', Jialin LuLu, 20 Jan 2026
-
RC3: 'Comment on egusphere-2025-3956', Anonymous Referee #3, 09 Dec 2025
This manuscript presents the relationship between O3 and its precursors using smog chamber experiments and a revised MCM box model. The authors improve O3 simulation by accounting for chamber wall effects under experimental conditions and unidentified NO2 sinks under ambient conditions, highlighting the sensitivity of O3 formation to VOCs and the implications for mitigation strategies in Daxing, Beijing.
However, the mechanisms related to chamber wall effects appear to have negligible influence when applied to real atmospheric conditions, whereas the unidentified NO2 sinks required for model-measurement agreement are unrealistically large. As a result, the revisions offer limited insights for ambient applications. Moreover, the simplified mechanisms used in the box model do not adequately represent interactions with meteorology or emissions. Therefore, I do not recommend the publication of this paper in ACP.
Major Comments
- The revised model (O3 SVR) shows better agreement with chamber results than the base model, primarily due to the introduction of OH generation associated with chamber wall effects. While such a mechanism may be justified within a chamber, there is minimal physical basis for applying this wall-induced radical source to ambient atmospheric conditions. Ground-related reactions on atmospheric chemistry do not mimic chamber wall reactions, and extending this mechanism to the atmosphere is inappropriate.
- Reaction rate constants associated with wall effects in Table 2 are key parameters, but their optimization process is insufficiently described (Page 6, Line 159; Page 7, Line 162 and Line 167). For example, how the reported 𝐽NO2 = 0.0015 ppbv s−1 (range 0.00075-0.0030 ppbv s⁻¹) in Angove et al. is converted to 1.2 × 106 molecule cm−3 s−1 in this study (reaction 1: hν + wall → OH).
The optimized rate constants may be chamber-specific. A controlled validation experiment using a small, clean plastic chamber (e.g., 1 m3) under similar conditions is therefore recommended. Is the revised model, with these parameters, able to reproduce results from such test-chamber conditions?
Upper and lower limits for relevant parameters should also be provided, and a sensitivity analysis is recommended to assess how uncertainties in each parameter affect model-measurement discrepancies.
- Other experiments beyond Iso&Tol02 and Iso&Tol04 should be briefly described in the Supplement, including their experimental design and key results beyond simply displaying NMB values (Page 8, Line 185).
- The authors introduced a constant NO2 sink to correct the model-measurement bias (Page 12, Line 270). However, the magnitude of this sink lacks physical justification and may simply force agreement for the wrong reasons. The authors should discuss alternative explanations for the observed bias beyond invoking an NO2 uptake, and evaluate whether other processes could plausibly account for the discrepancy.
- The attribution of O3 overestimation in the SVR NO2-sink model on 16-17 Aug to prevailing winds and associated air mass changes (Page 12, Line 277) requires further validation. The model overestimates O3 continuously from late afternoon on Aug 15 through early morning Aug 16, during which wind speed decreased by more than a factor of four. Statistical analysis is needed to support the proposed explanation.
Minor Comments
- Subtitles for each section should be more concise.
- Table 1. Please explain why NOx,0 is even lower than NO0 under some experimental conditions?
- Please clarify what J1-J56 correspond to in Table S1, and explain why J4 is selected in AtChem2-MCM in Table S2 (Supplement Page 6).
- Units should be provided for all variables in equations (1) and (2).
- Please discuss whether effects other than wall loss may contribute to model-measurement discrepancies (Page 6, Line 144).
- Figure 1. Please explain why, despite the large base-model/measurement differences for both Iso&Tol02 and Iso&Tol04 throughout the process, the base-model/measurement difference at 6 h is much smaller for Iso&Tol04 than for Iso&Tol02?
- Figure 2 and Figure 4. Using a consistent colorbar range across subplots is recommended to improve comparability.
- Table 3. Please clarify whether 0.007 s⁻¹ represents 𝐽NO2 at noon and provide full definitions for kgn, kan, Sa, and other variables.
Technical Corrections
- Grammar issues should be corrected, including “the complex of atmospheric conditions” (Page 2, Line 39), “studying how secondary pollutants like O3 formation” (Page 2, Line 48), and “details information” (Page 5, Line 116).
Citation: https://doi.org/10.5194/egusphere-2025-3956-RC3 -
AC3: 'Reply on RC3', Jialin LuLu, 20 Jan 2026
We sincerely appreciate the your careful review and valuable guidance. The manuscript has been thoroughly revised according to your suggestions, and all changes have been clearly highlighted using the Track Changes mode in the revised version. Enclosed please find our point-by-point responses to your comments for your kind consideration.
Status: closed
-
RC1: 'Comment on egusphere-2025-3956', Anonymous Referee #1, 02 Oct 2025
This manuscript presents a well-structured and scientifically rigorous study that combines smog chamber experiments with a revised MCM box model to improve the simulation of O3 formation and its sensitivity to precursors. The work is highly relevant to current air quality challenges, particularly in regions like Beijing suffering from severe O3 pollution. The methodological approach is sound, the results are clearly presented, and the conclusions are well-supported by the data. The inclusion of chamber wall effects and unidentified NO2 sinks represents a valuable contribution to the field. I recommend acceptance after minor revisions.
General comments
- The authors note that the NO2 sink rate constant used is higher than values reported in previous studies (Page 12, Line 285). A brief justification or speculation on why this might be the case (e.g., unaccounted surfaces, synergistic effects) would strengthen the discussion.
- The authors should briefly discuss the potential uncertainties in the uptake coefficients (γ) used for aerosols. Could the use of a constant γ, which may vary with aerosol composition and phase state, partly explain the need for such a large additional sink? A sentence or two on these limitations would be helpful.
- The discussion of the sensitivity shift (Section 3.6) could be enhanced by more explicitly linking the increased radical concentrations to the enhanced VOC sensitivity. A concise explanation could be: "The introduction of the NO2 sink reduces NO2 levels, which in turn lowers NO concentrations due to photo-stationary state relationships. Lower NO levels diminish the titration of O3 and, more importantly, reduce the scavenging of HO2 and RO2 radicals by NO. This increases the radical chain length and amplifies the role of VOC oxidation in O3 production, thereby shifting the system towards greater VOC sensitivity."
- In the conclusion or discussion, it would be valuable to explicitly state that while the box model is excellent for isolating chemical mechanisms, the identified NO2 sink rate constant is an "effective" rate that may also compensate for the lack of physical processes like advection and vertical dilution. A recommendation for future work using a 3D model with these revised chemical mechanisms to validate and spatially contextualize the findings would be a logical and strong ending point.
Technical comments
- “the complex of atmospheric conditions” (Page 2, Line 39) might be better expressed as “the complexity of atmospheric conditions”.
- Page 3, Table 1, NOx concentration values are lower than those of NO, resulting in a negative NO2 Correction should be applied to NOx concentrations (including both NO and NO2). Additionally, the symbol “–” can be used to indicate that a reactant was not added to the chamber.
- Page 6, Line 151: change “revised model in experiment Iso&Tol02 were” to “revised model in experiment Iso&Tol02 are” for grammatical agreement.
- Page 7, Table 2: “Refered” should be “Referred”.
- “the slope of the ridge line of the EKMA curves” (Page 9, Line 203) is correct, but consider using “ridgeline” as one word for consistency.
- “the uptake coefficient in the chamber wall is equal to that in the atmospheric ground” (Page 9, Lines 218-219) – consider rephrasing to “on the chamber wall” and “on the ground surface” for precision.
Citation: https://doi.org/10.5194/egusphere-2025-3956-RC1 -
AC1: 'Reply on RC1', Jialin LuLu, 20 Jan 2026
We sincerely appreciate the your careful review and valuable guidance. The manuscript has been thoroughly revised according to the your suggestions, and all changes have been clearly highlighted using the Track Changes mode in the revised version. Enclosed please find our point-by-point responses to your comments for your kind consideration.
-
RC2: 'Comment on egusphere-2025-3956', Anonymous Referee #2, 26 Nov 2025
This manuscript outlines attempts to improve the modelling of O3 concentrations, and the production of NOx-VOC isopleths, by box modelling simulations. The paper has two halves. The first half describes a series of chamber experiments in which a chamber-specific mechanism is produced to improve O3 concentration predictions. The second half attempts to use the developed chamber-specific mechanism to account for model biases in a box model of the ambient environment, suggesting that heterogeneous reactions could account for over-predictions in ozone concentrations from the base simulation.
The experimental procedures for the chamber experiments seem sound, and while there is some room for improvement in the explanation of the method by which the chamber-specific mechanism rates are selected (as explained later), the produced mechanism appears to be in line with existing literature and reproduces the observations well.
However, I believe that there are numerous issues with the implementation of the ambient box models, including the author’s focus on NO2 uptake as the sole explanation for model discrepancies. Some of these issues may be clarified with additional context in the explanation, but others may require deeper consideration. I have organised my comments in the attached document into 4 sections: NO2 Sink, General Comments, Specific Comments, and Minor Comments. I do believe there is value in efforts to improve O3 predictions and that the potential role of NO2 uptake is an important consideration, and I hope that addressing these comments can provide the improvements in the manuscript that I believe are necessary for publication.
-
AC2: 'Reply on RC2', Jialin LuLu, 20 Jan 2026
We sincerely appreciate the your careful review and valuable guidance. The manuscript has been thoroughly revised according to your suggestions, and all changes have been clearly highlighted using the Track Changes mode in the revised version. Enclosed please find our point-by-point responses to your comments for your kind consideration.
-
AC2: 'Reply on RC2', Jialin LuLu, 20 Jan 2026
-
RC3: 'Comment on egusphere-2025-3956', Anonymous Referee #3, 09 Dec 2025
This manuscript presents the relationship between O3 and its precursors using smog chamber experiments and a revised MCM box model. The authors improve O3 simulation by accounting for chamber wall effects under experimental conditions and unidentified NO2 sinks under ambient conditions, highlighting the sensitivity of O3 formation to VOCs and the implications for mitigation strategies in Daxing, Beijing.
However, the mechanisms related to chamber wall effects appear to have negligible influence when applied to real atmospheric conditions, whereas the unidentified NO2 sinks required for model-measurement agreement are unrealistically large. As a result, the revisions offer limited insights for ambient applications. Moreover, the simplified mechanisms used in the box model do not adequately represent interactions with meteorology or emissions. Therefore, I do not recommend the publication of this paper in ACP.
Major Comments
- The revised model (O3 SVR) shows better agreement with chamber results than the base model, primarily due to the introduction of OH generation associated with chamber wall effects. While such a mechanism may be justified within a chamber, there is minimal physical basis for applying this wall-induced radical source to ambient atmospheric conditions. Ground-related reactions on atmospheric chemistry do not mimic chamber wall reactions, and extending this mechanism to the atmosphere is inappropriate.
- Reaction rate constants associated with wall effects in Table 2 are key parameters, but their optimization process is insufficiently described (Page 6, Line 159; Page 7, Line 162 and Line 167). For example, how the reported 𝐽NO2 = 0.0015 ppbv s−1 (range 0.00075-0.0030 ppbv s⁻¹) in Angove et al. is converted to 1.2 × 106 molecule cm−3 s−1 in this study (reaction 1: hν + wall → OH).
The optimized rate constants may be chamber-specific. A controlled validation experiment using a small, clean plastic chamber (e.g., 1 m3) under similar conditions is therefore recommended. Is the revised model, with these parameters, able to reproduce results from such test-chamber conditions?
Upper and lower limits for relevant parameters should also be provided, and a sensitivity analysis is recommended to assess how uncertainties in each parameter affect model-measurement discrepancies.
- Other experiments beyond Iso&Tol02 and Iso&Tol04 should be briefly described in the Supplement, including their experimental design and key results beyond simply displaying NMB values (Page 8, Line 185).
- The authors introduced a constant NO2 sink to correct the model-measurement bias (Page 12, Line 270). However, the magnitude of this sink lacks physical justification and may simply force agreement for the wrong reasons. The authors should discuss alternative explanations for the observed bias beyond invoking an NO2 uptake, and evaluate whether other processes could plausibly account for the discrepancy.
- The attribution of O3 overestimation in the SVR NO2-sink model on 16-17 Aug to prevailing winds and associated air mass changes (Page 12, Line 277) requires further validation. The model overestimates O3 continuously from late afternoon on Aug 15 through early morning Aug 16, during which wind speed decreased by more than a factor of four. Statistical analysis is needed to support the proposed explanation.
Minor Comments
- Subtitles for each section should be more concise.
- Table 1. Please explain why NOx,0 is even lower than NO0 under some experimental conditions?
- Please clarify what J1-J56 correspond to in Table S1, and explain why J4 is selected in AtChem2-MCM in Table S2 (Supplement Page 6).
- Units should be provided for all variables in equations (1) and (2).
- Please discuss whether effects other than wall loss may contribute to model-measurement discrepancies (Page 6, Line 144).
- Figure 1. Please explain why, despite the large base-model/measurement differences for both Iso&Tol02 and Iso&Tol04 throughout the process, the base-model/measurement difference at 6 h is much smaller for Iso&Tol04 than for Iso&Tol02?
- Figure 2 and Figure 4. Using a consistent colorbar range across subplots is recommended to improve comparability.
- Table 3. Please clarify whether 0.007 s⁻¹ represents 𝐽NO2 at noon and provide full definitions for kgn, kan, Sa, and other variables.
Technical Corrections
- Grammar issues should be corrected, including “the complex of atmospheric conditions” (Page 2, Line 39), “studying how secondary pollutants like O3 formation” (Page 2, Line 48), and “details information” (Page 5, Line 116).
Citation: https://doi.org/10.5194/egusphere-2025-3956-RC3 -
AC3: 'Reply on RC3', Jialin LuLu, 20 Jan 2026
We sincerely appreciate the your careful review and valuable guidance. The manuscript has been thoroughly revised according to your suggestions, and all changes have been clearly highlighted using the Track Changes mode in the revised version. Enclosed please find our point-by-point responses to your comments for your kind consideration.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,942 | 127 | 38 | 2,107 | 67 | 28 | 39 |
- HTML: 1,942
- PDF: 127
- XML: 38
- Total: 2,107
- Supplement: 67
- BibTeX: 28
- EndNote: 39
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This manuscript presents a well-structured and scientifically rigorous study that combines smog chamber experiments with a revised MCM box model to improve the simulation of O3 formation and its sensitivity to precursors. The work is highly relevant to current air quality challenges, particularly in regions like Beijing suffering from severe O3 pollution. The methodological approach is sound, the results are clearly presented, and the conclusions are well-supported by the data. The inclusion of chamber wall effects and unidentified NO2 sinks represents a valuable contribution to the field. I recommend acceptance after minor revisions.
General comments
Technical comments