the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Radical chemistry in the Pearl River Delta: observations and modeling of OH and HO2 radicals in Shenzhen 2018
- 1State Key Joint Laboratory of Environmental Simulation and Pollution Control, College of Environmental Sciences and Engineering, Peking University, Beijing, China
- 2State Environmental Protection Key Laboratory of Atmospheric Ozone Pollution Control, Peking University, Beijing, China
- 3Institute of Energy and Climate Research, IEK-8: Troposphere, Forschungszentrum Juelich GmbH, Juelich, Germany
- 4School of Atmospheric Sciences, Sun Yat-Sen University, Zhuhai, China
- 5Laboratory of Atmospheric Observation Supersite, School of Environment and Energy, Peking University Shenzhen Graduate School, Shenzhen, China
- 1State Key Joint Laboratory of Environmental Simulation and Pollution Control, College of Environmental Sciences and Engineering, Peking University, Beijing, China
- 2State Environmental Protection Key Laboratory of Atmospheric Ozone Pollution Control, Peking University, Beijing, China
- 3Institute of Energy and Climate Research, IEK-8: Troposphere, Forschungszentrum Juelich GmbH, Juelich, Germany
- 4School of Atmospheric Sciences, Sun Yat-Sen University, Zhuhai, China
- 5Laboratory of Atmospheric Observation Supersite, School of Environment and Energy, Peking University Shenzhen Graduate School, Shenzhen, China
Abstract. The ambient OH and HO2 concentrations were measured continuously during the STORM (STudy of the Ozone foRmation Mechanism) campaign at the Shenzhen site, located in Pearl River Delta in China, in autumn 2018. The diurnal maximum OH and HO2 concentrations, measured by laser-induced fluorescence, were 4.5 × 106 cm−3 and 4.5 × 108 cm−3, respectively. The state-of-the-art radical chemical mechanism underestimated the observed OH concentration, similar to the other warm-season campaigns in China. The OH underestimation was attributed to the missing OH sources, which can be explained by the X mechanism. Good agreement between the observed and modeled OH concentrations was achieved when an additional numerical X equivalent to 0.1 ppb NO concentrations was added to the base model. The modeled HO2 could reproduce the observed HO2, indicating the HO2 heterogeneous uptake on HO2 chemistry was negligible. Photolysis reactions dominated the ROx primary production rate. The HONO, O3, HCHO, and carbonyls photolysis accounted for 29 %, 16 %, 16 %, and 11 % during the daytime, respectively. The ROx termination rate was dominated by the reaction of OH + NO2 in the morning, and thereafter the radical self-combination gradually became the major sink of ROx in the afternoon. The atmospheric oxidation capacity was evaluated, with a peak of 0.75 × 108 molecules cm−3 s−1 around noontime. A strong positive correlation between O3 formation rate and atmospheric oxidation capacity was achieved, illustrating the atmospheric oxidation capacity was the potential tracer to indicate the secondary pollution.
- Preprint
(1455 KB) -
Supplement
(451 KB) - BibTeX
- EndNote
Xinping Yang et al.
Status: final response (author comments only)
-
RC1: 'Comment on acp-2022-113', Anonymous Referee #1, 14 Mar 2022
Radical chemistry in the Pearl River Delta: observations and modeling of OH and HO2 radicals in Shenzhen 2018
The paper by Yang et al. presents results from a measurement campaign in September / October 2018 at Shenzhen in the Pearl River Delta. Unfortunately, the collected data-set is not state-of-the art and the subsequent interpretation suffers from this lack of data and thus the manuscript does not bring any useful new insight into atmospheric chemistry. The shortcomings in the data-set compared with what is current state-of-the art are:
- OH reactivity has not been measured: the analysis of the radicals budget is based on calculated OH reactivity, which is unsatisfactorily, especially given that kOH measurements are now widely available and add much confidence to the data set. Missing OH reactivity is widely observed under various conditions, and field campaigns quantifying OH and HO2 should also measure OH reactivity to unravel possible missing OH reactivity, rather than using the calculated OH reactivity as a lower limit to evaluate the experimental OH and HO2 data.
- The reference cell for stabilizing the laser wavelength did not work during the campaign: rather high NO concentrations have therefore been used in the FAGE for HO2 conversion, and it can be doubted that no interference from RO2 measurements occurred under these conditions. This is even more strange, as the authors indicate line 99, that there is no obvious difference in HO2 signal, when changing from 10 to 20 ppm. Does this mean the instrument works already under 100% HO2 conversion? Then, an RO2 interference seems very likely. However, 100% conversion efficiency is unlikely, as in a recent paper of the same group, describing a campaign carried out just a few months before in May / June 2018 (Ma et al., OH and HO2 radicals chemistry at a suburban site during the EXPLORE-YRD campaign in 2018, ACPD, doi.org/10.5194/acp-2021-1021), a conversion efficiency of 20% was obtained using 5 ppm. Or maybe the authors wanted to say that no obvious difference in HO2 concentration was observed. Then, the HO2 conversion rates under different NO concentrations need to be specified.
- Your model underestimates OH concentration under low NO conditions. I am very surprised that you claim that this is due to an unknown chemical X species, without even loosing a word about possible interferences in the OH measurements. Such increasing OH interferences with decreasing NO concentrations have been identified unequivocally with different FAGE instruments, and an experimental technique has been developed to quantify such possible interferences, this needs to be discussed. And even though some FAGE systems might be more prone to this interference than others, the FAGE community seems to agree on, that occasional measurements with such a pre-injector system are indispensable during field campaigns, especially when low NO concentrations are expected during the campaign. Looking at your above-mentioned paper describing a field campaign a few months before this one, it seems that you had already developed such a pre-injector system at the time of the campaign, because you had already used it. So why did you not use it in this campaign? In my opinion it is idle to discuss OH measurements that are underestimated by the model at low NO conditions, as long as OH-interference to an unknown species has been excluded by experiments using a pre-injector.
I understand that a field campaign is a lot of effort, and the authors want to publish the data, but I feel that the present data set is too uncomplete to justify a stand-alone paper in ACP. Maybe the data can be published as a complementary data-set together with another field campaign?
Besides these fundamental shortcomings I have no other comments.
-
AC1: 'Reply on RC1', Keding Lu, 26 May 2022
Thanks for your helpful comments which would help us to improve the manuscript. The comments and suggestions are valuable and very helpful for us. We have taken all these suggestions into account and have made corrections in the revised manuscript. The supplement is our responses. Thanks very much.
-
RC2: 'Comment on acp-2022-113', Anonymous Referee #2, 02 Apr 2022
General comments:
This manuscript describes OH and HO2 measurements in Shenzhen during Autumn 2018 and the skills of a photochemical box model to reproduce the observed radical levels when constrained with simultaneous observations of key reactants, to test the atmospheric photochemistry theory. The results showed that the model underestimated OH levels while it reproduced HO2 levels well. Missing recycling from HO2 to OH by species X was suggested, while the required levels of X were not very high (0.1 ppb). The atmospheric oxidation capacity was calculated on the observation basis and compared with the O3 formation rate. The results, though not including OH reactivity measurements, are worth to be added to those from numbers of previous field studies, for future diagnostics of missing processes in the model. Though the number of field studies in China has been increasing, the chemical conditions are very different among the studies and more field evidence is necessary. However, for this purpose, clarification is necessary at several points. First, uncertainty analysis needs to be provided for both observations and model simulations, particularly when the authors claim introduction of X to explain the model’s underestimation of the OH levels. Second, the authors cited values or results from previous studies for comparison but in-depth analysis/discussion across the studies were not provided in search for missing processes. The authors should specify the important characteristics of the conditions studied during this campaign and what is enabled with the observational results. Third, I am afraid that the major OH term of the atmospheric oxidation capacity is not very innovative; it is just the OH reactivity multiplied to the OH concentrations, i.e., OH loss rate, and thus it is very natural that it correlates with F(O3), i.e., the HO2/RO2 + NO rate, when the OH loss produces RO2/HO2 and the peroxy radicals undergo reactions with NO. Overall, I would suggest major revisions regarding the points above and the following specific comments.
Specific comments:
- Page 1, Line 26. Definition of the atmospheric oxidation capacity should be briefly mentioned in Abstract.
- Page 1, Line 26. x -> times character
- Page 2, Line 56. What are the important chemical conditions for this STORM campaign, in terms of differences from previous studies done in PRD, for example, city center/rural, NOx/BVOC levels, seasons etc.? From the inset map of Figure 1, I was not able to see if the site was in the city or in a rural region.
- Page 3, line 68. The coordinate should be 22.60 deg N and 113.97 deg E? (decimal points)
- Page 4, lines 89-96. Any literature to which the readers refer for further information of the specific FAGE instrument? Also, the uncertainty in OH and HO2 measurements should be quantified.
- Page 4, line 107. Did the author mean latest isoprene chemistry?
- Page 4, line 110. What are the "long-lived" species? Readers may think CO2 or CH4 as long-lived, which are not surely supposed in this context. I believe they are the modeled carbonyls/peroxides etc. But did they reach steady state within 2 days of integration? Did the authors assume a fast turnover time constant (dilution constant) for them? If so, any justification of the assumption?
- Page 4, lines 118. It is confusing to mention the observed k_OH from other studies, as that measurement was not available for this particular study.
- Page 8, Line 181. More discussion is preferred; what are the similarities and what are the differences to/from the previous studies in PRD?
- Page 8, Figure 3c and d. The fraction of the modeled OVOCs is fairly large. More explanation is needed what these species are and how their concentrations are justified.
- Page 8, line 188. Were the aerosol surface concentrations measured? Can the authors discuss maximum possible uptake coefficient from the surface concentrations?
- Page 8, line 193. It is worth mentioning where the study (Stevens et al. 1997) took place and add more information.
- Page 8, lines 193-194. I did not understand what the authors meant with the sentence "The comparison of the measured HO2/OH ratio…”.
- Page 10, line 232. The unknown OH source NEEDS TO explain
- Page 10, lines 251-252. Species other than NO played a significant role to explain the model’s OH underestimation
- Page 13, Line 301, Table 1. AOC includes all combination of pollutants i and oxidant j (j= OH, O3, NO3)? As the equation (1) does not include j, this in unclear.
- Page 14, Figure 7c. Why different units are used for the AOC and F(O3)? They can be both in ppb h-1 for example and should have close values.
- Page 14, Line 322. Why a fixed value (9x10^-12) is used for the rate constants of the RO2 + NO reactions? They should be variable in RACM depending of R and therefore these values should be used.
- Page 15, lines 351-352. Rewording is necessary.
- Page 16, line 362. What are the gradients here?
- Page 16, line 363. makes the quantification of F(O3) achieved – rewording is necessary.
-
AC2: 'Reply on RC2', Keding Lu, 26 May 2022
Thanks for all the valuable comments and suggestions which are helpful for the important guiding significance to us. We have taken all these comments and suggestions into account. Please refer to the attached supplement for the reply. Thanks very much.
-
RC3: 'Comment on acp-2022-113', Anonymous Referee #3, 02 Apr 2022
This paper presents measurements of OH and HO2 radicals during the during the STORM campaign in the Pearl River Delta and compare their measurements to model predictions. The authors conclude that the model underestimates the measured OH concentration but can reproduce the measured HO2 concentrations. The authors propose that the “X” mechanism can explain the discrepancy, similar to that proposed in previous studies. The proposed mechanism involves an unmeasured species “X” that converts RO2 to HO2 and HO2 to OH similar to NO. The authors conclude that a mixing ratio of “X” equivalent to 0.1 ppb of NO is needed to bring the measured OH concentrations into agreement with the measurements.
However, it is not clear that their measurements support their conclusion that the model significantly underestimates the measured concentrations, as it appears that the model agrees with the measurements to within the uncertainty of the technique. This is in contrast to the previous measurements highlighted in the paper, where the discrepancy between models and measurements were found to be much greater, such as the factor of 3-5 found by Hofzumahaus et al. (2009). While the addition of the X mechanism does improve the agreement with the measurements, there is no discussion as to why the measurements reported here are in better agreement with the model compared to the previous measurements discussed in the paper. The paper would benefit from an expanded discussion of the measurement-model agreement taking the uncertainties associate with both into account. In addition, the paper would benefit from an expanded discussion of a comparison of their results with the previous measurements mentioned in the manuscript, especially the difference between their measurements and those at the Backgarden and Heshan sites in the PRD (Hofzumahaus et al., 2009; Tan et al., 2019). Such a discussion could provide more information about the source of the model-measurement discrepancies at all these sites.
The measurements of OH and HO2 appear to be high quality and are of interest to the atmospheric chemistry community. In addition to addressing the major comment described above, I believe the paper would be publishable after the authors also address the following in a revised manuscript.
- The authors state that the base model agrees with the measurements to within the uncertainties of the measurements and the model (line 177), but then states that the model underestimates the measurements after 10 am when NO decreases. However, based on the information provided in Figure 3a, it appears that the model still agrees with the measurements to within the combined uncertainty of both the model and the measurements. This should be clarified. Addition of uncertainty estimates in Figure 3 would help to illustrate the agreement.
- Similarly, the base model predictions at low NO shown in Figure 5, although lower than the median measurements, appear to be within the combined uncertainty of model and measurements. The authors should quantify the discrepancy between the measurements and the model at each NO bin and reassess whether there is significant disagreement at low NO.
- The analysis of the OH measurements assumes that there are no interferences associated with the LIF-FAGE measurements. However, there is no discussion of whether the authors tested for unknown interferences with their measurements through a chemical modulation technique similar to that described in Tan et al. (2019). This should be addressed, as a significant interference would suggest that the model overestimation of OH could be more significant.
- I assume that the higher NO flow that was used in the HO2 measurements was required to increase the signal to allow for adjusting the laser wavelength given the failure of the reference cell. Were these measurements included in the data? While the authors claim that the NO concentrations were still low enough to minimize RO2 conversion to OH, did the authors perform calibrations of some RO2 conversion efficiencies to confirm this? What HO2 to OH conversion efficiencies did these two NO flows correspond to? Providing more details on the potential for RO2 interferences with the HO2 measurements would improve the reader’s confidence in the measurements.
- The authors should clarify that the rate of ozone production shown in equation 2 (line 322) represents the gross instantaneous rate of ozone production rather than the net rate of ozone production, as it does not take into account any NO2 formed that does not lead to O3 production through the formation of HNO3 from the OH + NO2 reaction. In contrast Tan et al. (2017) appear to use the net rate of ozone production in their analysis of the chemistry at the Wangdu site. As a result, the comparison of the rate of ozone production between the sites shown in Figure 7c may not be an appropriate comparison. This should be clarifed.
- AC3: 'Reply on RC3', Keding Lu, 26 May 2022
- The authors state that the base model agrees with the measurements to within the uncertainties of the measurements and the model (line 177), but then states that the model underestimates the measurements after 10 am when NO decreases. However, based on the information provided in Figure 3a, it appears that the model still agrees with the measurements to within the combined uncertainty of both the model and the measurements. This should be clarified. Addition of uncertainty estimates in Figure 3 would help to illustrate the agreement.
Xinping Yang et al.
Xinping Yang et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 389 | 140 | 16 | 545 | 31 | 8 | 19 |
- HTML: 389
- PDF: 140
- XML: 16
- Total: 545
- Supplement: 31
- BibTeX: 8
- EndNote: 19
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1