the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 21 Jan 2025
Incorporation of multi-phase halogen chemistry into Community Multiscale Air Quality (CMAQ) model
Abstract. Recent studies have revealed that halogen radicals (Cl, Br, and I) significantly influence atmospheric oxidation capacity, affecting O3 formation or destruction. However, understanding of halogen chemistries remains limited. To better investigate atmospheric halogen chemistries, we incorporated halogen processes into the Community Multi-scale Air Quality (CMAQ) model: (i) emissions of Cl2, HCl, Br2, and HBr from anthropogenic sources, and Br2, I2, HOI, and halocarbons from natural sources; and (ii) 177 multi-phase halogen reactions. After developing the model, we examined its performance against observed data. The results demonstrated significant improvements in simulating observed nitryl chloride (ClNO2) mixing ratios at supersites. The index of agreement (IOA) improved from 0.41 to 0.66, and the mean bias (MB) decreased from -159.36 ppt to -25.07 ppt. These improvements were driven by four atmospheric key reactions: (i) ClO + ClO → Cl2; (ii) HOBr + Cl- → BrCl; (iii) different parameterization of γN2O5; and (iv) 2NO2 + Cl- → ClNO + NO3-. We then examined the net Ox production rate (P(Ox)), which increased from 3.08 ppb/h to 3.33 ppb/h on land and decreased from 0.21 ppb/h to 0.07 ppb/h over ocean in the presence of halogen radicals. Further analysis of the impacts of halogen processes on key atmospheric species revealed that levels of OH, HCHO, and NOx increased by ~0.007 ppt (5.5 %), ~0.03 ppb (1.6 %), and ~0.29 ppb (2.9 %), respectively, while levels of HO2 and VOCs decreased by ~0.45 ppt (5.3 %) and ~0.71 ppb (5.9 %), respectively.
Competing interests: At least one of the (co-)authors is a member of the editorial board of Atmospheric Chemistry and Physics. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.- Preprint
(2406 KB) - Metadata XML
-
Supplement
(1029 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-23', Anonymous Referee #1, 10 Mar 2025
In Kim et al., 2025, the authors add and update the halogen (Chlorine, Bromine, and Iodine) chemistry scheme in the WRF-CMAQv5.2.1 setup for the Korean peninsula region. To implement the model, they first generate new emission datasets for anthropogenic Cl, Br species and natural Iodine species for the region. They run six simulations, CTRL (no halogens), EXPCl (where chlorine reactions are updated), EXPCl-Br (where Chlorine and Bromine are updated), EXPCl-Br-I (where all three species are updated), EXPCAM (where Saiz-Lopez, 2014 CAM-Chem halogen scheme is used), and EXPCMAQ (CMAQ model with halogen chemistry by Sarwar et al., 2015). First they compare EXPCl-Br-I, EXPCAM and EXPCMAQ against observed ClNO2, Cl2 values from the KORUS-AQ campaign at two sites, Olympic Park and Mt. Taehwa. Results show that the simulation with halogen chemistry performs better in matching observed ClNO2 and Cl2 values, although there is uncertainty for Cl2 both in terms of the simulated values and observations. Ozone changes in response to the new halogen reactions shows compensating changes with increased formation of ozone over land and increased destruction over the oceans. Finally, the impact of the halogen scheme on OH (5% increase), HO2 (5.3% decrease), VOCs (5.9% decrease), NOx (2.9% increase), and HCHO (1.6% increase) is discussed. I think the paper is well-written. The structure of the paper is straight-forward and easy to understand. This study highlights the importance of having a more complete halogen chemistry in chemistry models, as it affects OH, VOCs, nitrate, sulfates, PM2.5 and O3 and there is a need to understand these changes on a regional scale. Other than some changes that clarify and improve the readability of the paper, I think this paper is suitable for publication.
Major Suggestions:
- Can you provide more details on the model runs and how the statistics were calculated - How long was the model run for and what is the time-period of the runs? Which period were the observations taken, (summer or winter or which months) And how typical were these observations compared to the other sites? Maybe I missed it, but can you provide a reference(s) for the campaign?
- In section 3.1.1, there is an emphasis to state that adding the three-species halogen chemistry helped in bringing the model results closer to observed (lines 340-342 for example), compared to EXPCMAQ. But most of these changes (75% or so) are solely due to updating the chlorine scheme. Have you compared EXPCMAQ against EXPCl and EXPCl-Br? I think the emphasis in this section should be changed to how updating the Cl mechanism by adding new reactions and changing the parameterization of N2O5, brings down nighttime ClNO2 levels, followed by the impact of the addition of the HOBR reaction.
- In section 3.1.2, can you elaborate on the sensitivity test and how it was conducted? Were all reactions considered and these four reactions stand out or was there a reason to pick only these 4? Given the uncertainties in the partitioning of N2O5 onto chloride containing particles in the model being a big source for the change, have you tried testing any other parameterizations other than the one used?
- Can you write more about the differences between EXPCAM and either EXPCMAQ or EXPCl-Br-I? The ClNO2 is near-zero in EXPCAM. Is this because of missing reactions in the 2014 version of CAM-Chem that was adopted and tested against here? There may have been updates and modifications to that scheme as well, so maybe a more recent version (if changed) might be useful for discussion.
- I wonder if there can be some discussion on uncertainties in the simulations (could be in the supplementary), especially considering that some of the changes seem to be on the smaller side - such as changes in ozone (in terms of percentage changes). I appreciate that is because of the competing reactions that affect the formation and destruction of ozone. But are these changes significant?
- Is the updated version of the model code and the results going to be available to the public? If so, can you give the location for that as well in addition to the CMAQ web page.
Minor Suggestions:
Line 83: Change examines to examine.
Line 274: CCHO should be CH3CHO.
Line 381: Maybe use "Future" instead of "Further".
Line 578: I think "attempted to incorporate" can be modified because CMAQ modeling systems already exist with some halogen processes. Maybe being clearer that you added and updated three halogen species reactions in the CMAQ modeling system would be better.
Figure 3b: Perhaps this is a rounding error, but the sum of the percentages in the pie chart add up to 100.1%.
Figure 3: In the caption, can you add that the grey regions are night-time?
Figure 5: What do the DIFF's stand for? Why is it MAX for the top two and MIN for the bottom panel? Are the standard deviations and error bars for the observed?
Figure 8: Can you add the mean values of changes within or in top of the panel each of the species & simulations? It would be nice to see what the changes are for each of these experiments within the figure itself. At least for the full halogen process panel.
Citation: https://doi.org/10.5194/egusphere-2025-23-RC1 -
RC2: 'Review of the manuscript “Incorporation of multi-phase halogen chemistry into Community Multiscale Air Quality (CMAQ) model” by Kim et al., 2025.', Rafael Pedro Fernandez, 11 Mar 2025
The paper presents the implementation of halogen chemistry in the CMAQ model, including a comprehensive representation of chlorine, bromine and iodine sources and chemistry. Once the model updates are described, the study focus on reproducing ClNO2 and O3 observations over the Korean Peninsula during the KORUS-AQ campaign. The results focus on the improvements in the model-observation comparisons, including the index of agreement and other statistical parameters, particularly over two inland locations within a polluted / semi-polluted environment. Then, they identify the four main reactions in the model accounting for the largest fraction of the improvement. Based on this, the authors evaluate the influence of air quality in the whole modeled domain, highlighting the major and sometimes opposite differences observed over continental and oceanic domains, and provide general conclusions about the benefits of considering halogen chemistry in the study.
First of all, I would like to recognize the efforts from the group to implement their own version of halogen chemistry in CMAQ, which is of major importance as the community needs more modeling studies focused on the halogen influence on atmospheric chemistry and climate. However, I believe the current version of the work does not allow reaching a firm conclusion of the results obtained, probably because the paper attempts to address all at once the complete technical implementation, the observational improvements achieved over continental locations, and the overall impact and influences over different regions. While I leave for the authors to decide if it is convenient to present all of these developments within a single work or to partially split into independent companion papers, I recommend the authors to address the following major comments and submit a revised version for further consideration.
Major Comments:
P4,L93: In the introduction you clearly state that “We then investigate the formation of ozone using the new halogen processes”, which is what the paper focus first … but then at the end the work attempts to provide much wider conclusions of the halogen influence on atmospheric chemistry and air quality. You may want to focus here on ozone production driven by ClNO2 chemistry, and leave the more general discussion for a companion paper.
P5,L111: Comparison with ClNO2 observations are of mayor importance for this study. Indeed, it might be worth to mention in the title. Current title and abstract give the impression of a general halogen chemistry development, while the work mostly focus on ClNO2 and its role for ozone production.
P20,L326-328: Your results for EXP_CAM are not surprising to me as the heterogeneous ClNO2 formation through N2O5 was not considered in Saiz-Lopez et al., (2014), which is your reference. However, it should be mentioned that those studies focused on oceanic and pristine conditions, and not polluted areas with high NOx and inland chlorine emissions. Indeed, further research from the group lead by Dr. Saiz-Lopez considered enhanced HCl and ClNO2 production within continental areas (see for example Li et al. 2022), which clearly showed important implications when anthropogenic chlorine sources are considered. Therefore I recommend including the complete chlorine scheme from CAM-Chem in your analysis to avoid reaching erroneous conclusion, or at least to clarify why your EXP_CAM simulations do not reproduce ClNO2 observations.
P22,L356-367: Of all four dominant process mentioned here, R20 in Table 1 (ClO + ClO) is particularly surprising to me, as this reaction is typically considered in stratospheric ozone depletion, but not for boundary layer studies. How do you explain that the surface observations are sensible to this reaction? Is it because of the subsequent Cl2 photolysis? Is due to the coupling of ClO + NO2? Can you explain how this reaction can contribute to ClNO2 formation during the night, when ClO abundance is zero and in addition, any Cl2 formed would not be photolyzed until the next morning?
P33,L547-562: I follow the explanation about the changes in the OH/HO2 partitioning, but oceanic SLH (particularly iodine) has been clearly shown to reduce the OH abundance (not increase it), and consequently to increase the CH4 burden and lifetime (Li et al., 2022). Your results over the ocean seems to contradict that. How could this be? Indeed, the null cycle mentioned in your work results in a shift of partitioning from OH to HO2 (which is fine), but the total OH abundance is controlled by O3 + hv --> O1D. Given that oceanic halogens reduce O3, there is less O1D and therefore OH formation should decrease. Have you discarded any influence from the BC affecting the overall results? Note that O3 changes for the EXP_Cl_Br_I - EXP_ctrl are in line with Li et al. (2022) for oceanic domains, but the OH changes are not consistent with the changes in O3 (unless I missed something).
General Comments:
P2,L41: put this values in the context of equivalent changes reported in the literature, here and elsewhere.
P3,L65: You could also cite other previous works with the implementation of halogen chemistry in WRF-Chem, e.g. from Badia et al., (2019). P4,L75: Similarly, given that you compare your results with those of the CAM-Chem model, you should also cite some of the CAM-Chem studies focused on the impact of halogen chemistry over the oceans as Saiz-Lopez et al., (2014, 2023), Iglesias-Suarez et al., (2020), Li et al., (2022).
P5,L106: What version of CMAQ did you consider? Have you used and/or compared w.r.t the previous implementation of halogen chemistry in CMAQ? (e.g. Sarwar 2015).
P7,L149: You should cite and compare your methodology and emission values for anthropogenic emissions with other works to put your regional results into context of the current literature. In P8,L180 you compare with Kim et al., 2023, but should also compare with the anthropogenic emissions from Saiz-Lopez et al., 2023.
P10,L206-211: How did you validate the overall halocarbon emission inventory implemented in your model? Based on the Ordoñez et al. (2012) inventory, global scaling of chl-a bitmaps was necessary to reproduce observations. In addition, what type of diurnal profile did you apply to the emissions?
P10,L212-218: Similarly, how did you exactly implement the SSA-dehalogenation process? Note this process is very efficient and depends on many parameters that present a large spatio-temporal variability (see Ordoñez et al. 2012 and Fernandez et al., 2014). Could you please provide more details about the implementation and the net bromine flux from sea-salt.
P10,L220-223: Halogen Chemical reactions. Please, provide at least a general introduction of which are the important reference works considered in this study.
P23,L380-385: In relation to the four highlighted reactions, I can think of many other processes that could be important to evaluate: for example: i) are Cl2 or any other species assumed to be uptake into the aerosol phase and provide Cl- (aq)? Ii) Did you consider any hourly variation in the emission strength of anthropogenic halogens of HCl and Cl2 (Eq. 1) that could impact your night-time results?. What about the NO2 sources that are required for the ClNO2 formation, how is their spatial and temporal variation? (this apply also to P33,L541). If all of these were found to be irrelevant, at least a couple of sentences explaining why they are not important should be given. I completely agree that further studies are necessary to investigate the main factors causing these discrepancies (which should be highlighted in the conclusions).
P26,L434-439: The way the text is written seems to indicate that this is a result from this study, while the opposite effect between continental (polluted) and oceanic (pristine) has been previously described in the literature (e.g., Li et al., 2022, Saiz-Lopez et al., 2023). Please rephrase to make it clear that your results are in agreement with those of previous studies.
P28,L462-464: this sentence makes no sense. Please rephrase. Are you sure the SSA dehalogenation process for bromine is well implemented in your model?. P33,L545: I would expect larger impacts of bromine than chlorine over the oceans, which is not the case. Could this be possible due to a small efficiency of the SSA-dehalogenation process for bromine?
P30,Eq.8 and Eq.9: Please control the F(Ox) expression for missing production channels and explain in case some terms are not considered. For the case of D(Ox) note that the Cl+O3 term should not be considered as it results in the formation of ClO, which is part of Ox (see Saiz-Lopez et al., 2014 for a complete list of all halogen-driven OddOx loss rates).
P36,L586-594: Once again, the conclusions concentrate on the importance of halogens to improve the model representation of ClNO2 observations, which are mostly related to air-quality in polluted environments. However, the final part of the paper focus on the wider implications of halogens over continental and oceanic domains, which were not validated before. Indeed, in L602-606 you summarize model results for several species but omit mentioning the inconsistent results found for OH. In case you decide to keep all the analysis in a single paper, a detailed discussion of this important issue should be included.
Language editing comments and Typos:
P2,L26 and elsewhere: Please refer to halogen chemistry, not chemistries.
P22,L365: It appears à it is evident.
P35,L583: replace I by I2 inside the parenthesis.
Table 2: Check for typos and consistency in R1 and R3.
Figures, Tables and Captions
Table 3: given the importance of reaction R6 in your results, more details should be given in the text. Note that most model implement different versions of R1.
Figure 4: it is not clear for me if this comparison exercise considers only nightime or 24-hs model output. Given that the model is shown to underestimate Cl2 observations during the day ... then it is expected that if 24-hs is considered the presented results would imply a night-time over-estimation for the EXP_CL_BR_I case. Am I right? Could you please clarify in the text?
References
Ordoñez et al., 2012 (https://acp.copernicus.org/articles/12/1423/2012/)
Fernandez et al., 2014 (https://acp.copernicus.org/articles/14/13391/2014/)
Saiz-Lopez et al., 2014 (https://acp.copernicus.org/articles/14/13119/2014/)
Badia et al., 2019 (https://acp.copernicus.org/articles/19/3161/2019/)
Iglesias-Suarez et al., 2020 (https://www.nature.com/articles/s41558-019-0675-6)
Li et al., 2022 (https://www.nature.com/articles/s41467-022-30456-8)
Saiz-Lopez et al., 2023 (https://www.nature.com/articles/s41586-023-06119-z)
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 166 | 32 | 8 | 206 | 28 | 4 | 6 |
- HTML: 166
- PDF: 32
- XML: 8
- Total: 206
- Supplement: 28
- BibTeX: 4
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 80 | 39 |
| South Korea | 2 | 18 | 8 |
| China | 3 | 11 | 5 |
| Canada | 4 | 10 | 4 |
| United Kingdom | 5 | 8 | 3 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 80