the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Abundance of volatile organic compounds and their role in ozone pollution management: Evidence from multi-platform observations and model representations during the 2021–2022 field campaign in Hong Kong
Abstract. Volatile organic compounds (VOCs) are a diverse group of species that contribute to ozone formation. However, our understanding of VOC dynamics and their effect on ozone pollution is limited by the lack of long-term, continuous, and speciated measurements, especially of oxygenated compounds. To address this gap, this study for the first time integrates on-land, shipborne, and spaceborne measurements from a field campaign in Hong Kong during 2021–2022, analyzing 45–98 VOC species over land and water. Results show that oxygenated VOCs (OVOCs) account for 73 % (37 ppbv) of the total VOC concentration and 56 % of the total ozone formation potential (OFP), underscoring their indispensable role in VOC chemistry. Despite such importance, OVOCs are underestimated by 45 %–70 % in the CMAQ model, while non-methane hydrocarbons (NMHCs) face a lesser underestimation of 47 %–48 % (i.e., “model underestimation”). Meanwhile, the model does not currently account for 17–56 species of the total measured VOCs (i.e., “model omission”). According to this, we break down the observed overwater VOC concentration of 51 ppbv into three components: 9 ppbv (18 %) successfully represented, 35 ppbv (69 %) underestimated, and 7 ppbv (14 %) omitted in the model. For OFP, the breakdown shows 26 % successful representation, 54 % underestimation, and 20 % omission. Together, both “omission” and “underestimation” reveal the overall “VOC underrepresentation” in the model, which partly results in greater ozone sensitivity to VOCs than observed by spaceborne TROPOMI in polluted areas. The findings provide valuable insights into regional pollution dynamics, and inform VOC-related model development and air quality management.
- Preprint
(5205 KB) - Metadata XML
-
Supplement
(12317 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-3227', Anonymous Referee #1, 02 Sep 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-3227/egusphere-2025-3227-RC1-supplement.pdfCitation: https://doi.org/
10.5194/egusphere-2025-3227-RC1 -
RC2: 'Comment on egusphere-2025-3227', Anonymous Referee #2, 04 Sep 2025
Over view
This manuscript integrates on-land, shipborne, and spaceborne measurements from a field campaign in Hong Kong during 2021-2022, analyzing 45-98 VOC species over land and water. This study presents OVOCs dominate total VOC concentration and OFP, highlighting their critical role in ozone chemistry. The CMAQ model significantly underestimates both OVOCs and NMHCs, and omits 17-56 measured species entirely. Model underrepresentation of VOCs leads to overestimated ozone sensitivity to VOCs, potentially misguiding pollution control strategies. The study integrates TROPOMI satellite data with surface measurements to evaluate spatial and temporal patterns of ozone precursors and diagnose model performance. The study provides valuable empirical evidence of VOC underrepresentation in models and underscores the growing importance of OVOCs in urban ozone formation. However, the QAQC for the VOC measurement is not comprehensive enough, especially for the OVOC species (e.g. formaldehyde). There are a few details that should be addressed, notably pay close attention to the readability of most figure needs to be improved throughout the manuscript.
General comments
Page 4 Line 17: What is the difference between solvent use and consumer products and volatile chemical product (line 30, page 3). The literature cited here is mostly based on simulation studies conducted in the United States. Is there any research specifically targeting China? In fact, many studies have proposed that OVOCs emitted from VCP are the key to improving simulation. It is suggested to compare and analyze the OVOCs measured in this study with VCP emission related research.
Page 5 line 18-24: This study emphasizes the importance of OVOCs, but the measurement of OVOCs only uses offline sampling. Is this sampling done every day? How to ensure that the concentration of OVOCs data collected offline is at the same level as that of NMHC data collected online, or how to carry out quality control work? I don't think it's enough to simply add up two measurement methods to determine the importance of OVOCs. Please provide evidence of the reliability of this data. Regarding the reliability of the measurement methods, at least it needs to be given in the SI.
Page 5 Line26: I think HCHO should be formaldehyde here, and it would be better not to directly use its chemical formula as a name in the main text or explain it when first used. Is this “40 sites snapshot” only analyzing the concentration of formaldehyde? Has the reliability of formaldehyde concentration been demonstrated by comparing other OVOC species sampled by DNPH cartridges? Why are the two HPLC selections different, and does this have an impact on the analysis results? Meanwhile, what other work were taken to ensure the representativeness of this 40 sites snapshot?
Page 5-6 Section 2.1.1-2.1.2: I think the description of the instrument here is pointless. Suggest summarizing and rewriting these sentences. In the methodology section, it should not only list the instruments used, but also express more about the measurement ability of VOCs species in this study, the differences between different measurement methods, and the addition of QAQC explanations to verify the reliability of the data used in this study. In the two types of measurements, the quality control process for species measured by GC-MS needs to be provided (species in Table 1), at least in the SI section, and cannot be solely based on references. Additionally, the molecular formula of ethylbenzene and m/p- xylene (C8H10) is incorrect in Table 1.
Page 10 Section 2.4: How is the species time resolution of NO2 and VOCs (especially formaldehyde) unified among different measurement methods in this study? Is there significant uncertainty, especially when conducting FNR value analysis?
Page13 Line14-27: When conducting OFP analysis, it is suggested to propose the MIR value of formaldehyde species to highlight that formaldehyde also contributes significantly to OFP at high concentrations, distinguishing it from other species that only contribute high concentrations or only contribute high OFP. Meanwhile, it is difficult to see from the time series in Figure 3 that the correlation performance between formaldehyde and measured total VOC reaches 0.72-0.85. It is recommended to add scatter plots and linear relationship fitting graphs.
Page 13 Line 28-Page 14 Line 10: I think formaldehyde as a reliable indicator of ambient VOCs is not convincing enough. This study only selected 3 sites tested in Hong Kong, and the number of VOCs species measured is limited (45 species). It is suggested to add more comparisons with existing literature, such as adding test results from Guangzhou/Shenzhen urban areas in the PRD region, to verify the regional representativeness of this statement.
Page 14 Figure 3: The overlapping species names here are difficult to distinguish. It is recommended to only display the species with higher contributions, and other species should be presented separately in the form of legends or by category. It would be better if NMHC and OVOCs could be clearly distinguished. At the same time, it is necessary to unify the naming convention for formaldehyde and HCHO on the figure. The naming of species should be standardized, such as for aldehydes 3C+ without any explanation and non-standard, at least using expressions like C3-C6 aldehydes.
In addition, all VOCs concentrations are daily average concentrations, which are easily affected by nearby emission sources or human factors, resulting in significant data deviations. Has this part been checked and corrected accordingly?
Page19 Line8-15: Need to compare with more research on Hong Kong or the PRD region to better illustrate the sources of these underestimated species from models, or refer to more localized emission inventory data.
Page 22 Line 7-15: These sentences read a bit awkwardly. What is the FNR threshold used in this study to divide VOC-limited and transitional regime sensitive to both VOC and NOx, and which season in Figure 9 corresponds to different months? What is the threshold in other literature and how much does it differ from the PRD? These information need to be provided by the author in the text here, rather than simply expressing regional specificity. Meanwhile, it is difficult to obtain the specific values of FNR values for each month on the vertical axis of Figure 9c. It is recommended to modify it.
Furthermore, the FNR threshold used in this study is based on research in PRD region. It is necessary to verify whether it is fully applicable in different seasons and weather conditions, and to discuss the influence of the satellite inversion uncertainties of NO₂ and HCHO on the FNR classification. It is suggested to conduct an uncertainty propagation analysis of the FNR threshold to assess its impact on the ozone sensitivity classification. At the same time, it is suggested to compare the data from Hong Kong with that from other cities in the PRD region, in order to verify the regional consistency in the composition of VOCs and the ozone sensitivity.
Page 24 Line1-15: Many of the text here are meaningless. Please add some quantitative summary of the data, the existing research's understanding of underestimation, and the differences from this study. Also, summarize how future improvements in the model can optimize this situation. Additionally, the underestimation of OVOCs in the model has been proposed by many studies. The CMAQ model in this study is based on very few species, and underestimation is inevitable. It is suggested to add more detailed chemical mechanisms of OVOCs (such as MCM or updated aromatics oxidation mechanisms) and conduct comparative analysis.
Specific Comments
Please check that the full names of many abbreviations do not require capitalization of the first letter. Pay attention to the format of the SI document, at least the title, author, and other information are required.
Page 2 Line 4: would suggest remove “first time”. This manuscript talked a lot about first time, if so, why wasn't previous research conducted? Is there any difficulty? All of these need to be raised. But in fact, this has already been done elsewhere.
Page2 Line 18: TROPOMI here should be modified as full name as TROPOspheric Monitoring Instrument (TROPOMI).
Page4 Line 21: Please check the references here. Wang et al., 2024, have 2 articles in the references section.
Page 3 Line 7: The full name of HKEPD appears repeatedly, please review the entire manuscript for similar issues, such as the full name of the VCDs.
Page 3 Line 10: Fig S1 needs to add captions indicating the types of stations marked differently on the figure.
Page 4 Line24-31: I think the description here is pointless. Would suggest summarize and rewording these sentences. Some of the content should be in Section 2.
Page 4 line36: I think it is not advisable to use the results of this article to represent the entire southern China, would suggest modified to PRD region in China.
Page 6 Line 10-11: GC-MS/ECD/FID, need to provide the full name here.
Page 8 Line 17: The full name of CMAQ is not reflected and the abbreviation of line 26 CMAQ should be before the Modeling System.
Page 12: Is the NMHCs in Figure 1d missing the data for December? Even so, the horizontal axis on the graph should be 1-12 instead of not displaying 12.
Page 13 Line 6: HCHO-to-NO2 ratio (FNR), it needs to be clarified that F refers to formaldehyde.
Page 14: It is recommended to use the form of year/month or month number instead of the word 'month' for the horizontal axis in Figure 3 and Figure S4. Additionally, please do not use screenshot format in Figure (unnecessary lines appear in Figures S4 and S5).
Page 17: The wavy line in Figure 5 maybe an incorrect format.
Citation: https://doi.org/10.5194/egusphere-2025-3227-RC2
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
1,479 | 39 | 13 | 1,531 | 26 | 23 | 20 |
- HTML: 1,479
- PDF: 39
- XML: 13
- Total: 1,531
- Supplement: 26
- BibTeX: 23
- EndNote: 20
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1