the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 06 Nov 2025
Heat and continental transport shape the variability of volatile organic compounds in the Eastern Mediterranean: Insights from multi-year observations and regional modeling
Abstract. Volatile organic compounds (VOCs) are key precursors in tropospheric ozone formation and secondary organic aerosol formation, thereby influencing regional air quality and climate. This study investigates the seasonal, diurnal, and temperature-driven variability of VOCs at a rural background site in Cyprus, located in the Eastern Mediterranean. VOC measurements (April 2022–June 2024) conducted with a proton transfer reaction time-of-flight Mass spectrometer (PTR-ToF-MS), along with HYSPLIT airmass back-trajectory analysis and WRF-Chem atmospheric chemistry simulations. Seventy-six VOCs were quantified and categorized into classes such as oxygenated VOCs (OVOCs), aliphatic hydrocarbons, aromatic hydrocarbons, terpenes, and nitrogen- and sulfur-containing compounds. Most VOCs exhibited distinct diurnal cycles, observed highest during 08:00–14:00 UTC due to photochemical activity and temperature-driven emissions. Biogenic VOCs, particularly isoprene, peaked in summer within the 35–38 °C range but declined under extreme heat, suggesting emission suppression from thermal stress. Monoterpenes showed elevated levels both day and night, reflecting contributions from biogenic and anthropogenic sources. Dimethyl sulphide (DMS) increased during warmer months, indicating enhanced marine microbial activity. OVOCs displayed strong seasonal and thermal enhancement during hot, dry summers due to both primary emissions and secondary formation. Benzene rose above 35 °C from evaporative and potential stress-related biogenic sources, whereas toluene and xylene were higher in colder months, linked to combustion processes. While WRF-Chem captured seasonal trends, most VOCs were underestimated, highlighting missing emission sources and oxidation pathways. Overall, the study emphasizes temperature and regional transport as key drivers of VOC variability in the Eastern Mediterranean.
- Preprint
(1965 KB) - Metadata XML
-
Supplement
(1982 KB) - BibTeX
- EndNote
Status: open (until 18 Dec 2025)
- RC1: 'Comment on egusphere-2025-5124', Arnaud P. Praplan, 28 Nov 2025 reply
-
RC2: 'Comment on egusphere-2025-5124', Anonymous Referee #2, 02 Dec 2025
reply
General comments:
This manuscript "Heat and continental transport shape the variability of volatile organic compounds in the Eastern Mediterranean: Insights from multi-year observations and regional modelling" presents VOCs data measured by PTR-ToF-MS between 2022 and 2024 at a rural background site in Cyprus experiencing heat waves. A detailed discussion is presented on the oxygenated VOCs, the dominant class of VOCs at this site. Long-term time series of OVOCs are relatively rare, yet the authors provided valuable insights by presenting both seasonal and diurnal patterns of these compounds alongside other chemical classes, and by clearly summarizing the contributions from biogenic, anthropogenic, and secondary emissions. In addition, the contribution of the clustered air masses to VOC mixture at this site is presented, including the seasonal variations. Nevertheless, the modelled VOCs by WRF-Chem model simulation overestimate measured VOCs highlighting the limitations of this model under those meteorological conditions.
The manuscript deserves publication given the unique dataset acquired and the additional insights it provides. However, several points outlined below require further clarification and detail
Specific comments
- L361: Is there an explanation for why concentrations in summer 2022 (Fig. S4), show a less pronounced seasonal pattern in comparison to 2023 and 2024 in summer? The higher VOC concentrations observed in spring 2024 in comparison to the other spring campaigns are also linked to ambient temperature?
- L437-440: What is the CO diurnal pattern? As CO is mostly influenced by combustion and traffic-related emissions, it could help to give further explanations for the the morning and evening peaks.
- L460: I'm wondering why ethanol does not show a temperature dependance as it is the case for methanol and acetone. In general, I'm surprised to see in the hierarchical clustering (Fig. 6) that ethanol is not closely related to methanol or acetone.
- L565: "The dominant clusters originated from Northwest Asia (C4, 34.3%) and Europe (C3, 33.6%), jointly accounting for more than two-thirds of air mass transport." By looking at the wind rose (Fig. 2d), N-W and S-W are the most dominant wind directions while N-E winds don't seem to be predominant. Fig. 7 showed that the C4 is coming from N-E. Could you explain, for example, why C1 and C2 do not account for a greater proportion to the air mass contribution?
- Section 3.7 on WRF-Chem model simulation. I understand the importance of being able to simulate VOCs mixing ratio. Nevertheless, I felt that this section did not do justice to this dataset, as only good adequations are shown between modeled and measured for isoprene (despite the model can capture the seasonal variations for the other VOCs). It is important to highlight the limitations and the need to improve the model but the authors should give more details on how to improve future estimations, especially as such heat waves even might occur more frequently in the future.
- Wildfires have been mentioned a few times, but it would be an asset to this manuscript if further details could be provided, as the wildfires that occurred in 2022 and 2023 could potentially have been captured at this monitoring site. Did the authors see a typical VOC signature from those wildfires?
Technical corrections
- L53-54 "these compounds are frequently classified into the broader categories of oxygenated VOCs (OVOCs), biogenic VOCs (BVOCs), and anthropogenic VOCs (AVOCs), each of the species with distinct sources, reactivities, and atmospheric lifetimes" I found a bit misleading to classify OVOCs as their own category as they can be emitted both from biogenic and anthropogenic sources as well.
- L141: What is the concentration of the certified gas standard? How do you calibrate the VOCs that are not included in the gas standard?
- L210-211: "atmospheric research-hemispheric transport of air pollution (EDGAR-HTAP) version 2 emission inventory. while BVOCs were simulated online": Should be "(EDGAR-HTAP; version 2 emission inventory)? And a comma instead of a period.
- L213, L629: wrong format for the citation
- In Table SI, the units for the measured VOCs should be indicated.
- L229: add a subscript for NO2.
- L297-298: "Elevated levels during summer (Fig. 2b) suggest increased biogenic activity and enhanced photochemical production". Fig 2b refers to the seasonal variations of RH, should it be 2a? Additionally, the elevated levels during summer could be linked to solvent evaporation and/or fuel evaporation.
- L393-394: It would be useful to indicate the concentrations for the monoterpenes. By looking at Table 2, it is not clear if the row containing CAO, Cyprus; Winter 2022-24 belongs to this study (likely) or to Debevec et al (2017).
- L434, 459: It should be Mediterranean Sea.
- L511: It would be great to indicate the MVK/isoprene ratios for future comparisons.
- Figure 6; on the y and x axis, label "Cluster" is indicated but it is not clear to me what it means.
Citation: https://doi.org/10.5194/egusphere-2025-5124-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 184 | 47 | 17 | 248 | 21 | 12 | 12 |
- HTML: 184
- PDF: 47
- XML: 17
- Total: 248
- Supplement: 21
- BibTeX: 12
- EndNote: 12
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
In the manuscript 'Heat and continental transport shape the variability of volatile organic compounds in the Eastern Mediterranean: Insights from multi-year observations and regional modelling' submitted to Atmospheric Chemistry and Physics by Grag et al., the authors present conclusions based on a very large dataset spanning over two years of mainly VOC measurements with PTR-ToF-MS at a rural background site in Cyprus, supported by meteorological and air pollution data, as well as modelling with WRF-Chem.
The data analysis provided is extensive and detailed, and my main comment is related to the inconsistent grouping of VOCs and I have other small minor comments related to some interpretation of the results and wording.
1. It appears throughout the manuscript that the 76 "identified" VOCs (even though for many it is simply the chemical composition that has been identified, not a specific VOC) have been grouped in inconsistent ways. I understand that grouping is difficult can be done in various ways depending on what the authors want to discuss, but I would still recommend either unifying the way it is done in the manuscript or alternatively justify each time why the grouping is done a certain way for each section when presenting results.
The abstract and conclusions mention three classes (biogenic, anthropogenic, and secondary/oxygenated) that seem to be the 'main' classification, but it is not really used elsewhere in the manuscript and also, I would argue that is it not a good choice of classification for the following reasons:
- On the use of 'biogenic' and 'anthropogenic' for compounds: There are several instances in the manuscript where 'biogenic' is used as shorthand for isoprene and monoterpenes and 'anthropogenic' for e.g. aromatics compounds, even though the authors acknowledge themselves in the abstract that monoterpenes have contributions from biogenic and anthropogenic sources, and also mention that aromatic compounds might stem from stress emissions of vegetation. Therefore, this should be made less ambiguous throughout the manuscript and I recommend the authors to use classification in chemical families for compounds and only discuss sources as 'biogenic/anthropogenic'.
- On the use of 'secondary/oxygenated' (this is only done in the conclusions, not the abstract): This is also misleading as it is clear that there are primary sources of oxygenated compounds, so it is probably best not to conflate the two.
On lines 283-286, the authors mention six categories, even though five are listed afterwards. Are N- and S-containing compounds two separate categories? Based on Fig. S3, there seem to be lumped together in the analysis. Then, the 'groups' in Table 1 are a mix of chemical class (aldehyde, aromatics, etc.) and sources (biogenic) and Figure 2 also use a finer grouping of the compound as well as present some compounds with high concentrations individually.
On page 10, Table 1 contains 18 selected VOCs, but the Table does not seem to be mentioned in the text, so that it is not clear what is the basis for the selection (are those calibrated compounds?). Line 320 mention the 20 most abundant VOCs in Fig. S3, but those are then again a different subset of compounds.
Table 2 use 'terpenes' (so does title of section 3.3.1), which I would recommend to use elsewhere in the manuscript too. However, for someone who is not familiar with VOCs, the group separation in Table 2 is not the clearest. Maybe vertical lines would help? Also, it introduces the group 'Nitrile' which could be 'N-containing compound' to remain more consistent with the rest of the manuscript. Then, Figure 4 is almost consistent with Figure 3, even though e.g. aliphatic hydrocarbons are not included.
Then, Figure 5 show the temperature dependence of 16 compounds. The authors could maybe have used the selected compounds of Table 1 for consistency. I understand that using groups of compounds might be less appropriate here for the discussion. These same 16 compounds are then used in the following figures. I was wondering if each or some of these compounds are used as proxy for/represent a whole group of compounds. This could be stated explicitly and the selection of compounds to focus on argued more precisely.
On top of that, I understand that for modelling purpose the grouping and classification has to be done according to the compounds in the model (section 3.7).
All this to say that I would recommend using a consistent grouping of compounds throughout the manuscript, which will automatically improve the presentation of the results and the discussion and make it easier to follow for the reader. It will also make the wording of the conclusions clearer, in my opinion. For instance, the authors could use the groups and (representative) compounds from the model and add additional groups from Figure 3. Alternatively, the authors could focus on selected VOCs (Table 1), explaining their decision either because they are the most abundant (or most abundant within their class), or they are the ones for which calibration is available, or they represent a group of compounds or a source (e.g. marine, traffic), or any other clearly-defined reason. I understand the difficulty of presenting clearly such a large dataset and that many people are involved in the data analysis, however, I believe that the manuscript would benefit from streamlining the presentation of the results to tell a consistent story rather than stick together various pieces.
More specific comments:
- Lines 140-143: The authors mention automatization of the blank measurements and calibration once a day. Can the authors comment if these daily measurements have shifted throughout the measurement period, depending on when the instrument was started? Are they distributed more or less equally, or might they influence diurnal patterns? Have some specific hours less data than others due to that?
- Sections 2.1 and 2.3: I would include the information regarding the location of the meteorological measurements in section 2.1 and give it a clear label that then can be used in section 2.3 to make it even clearer that the meteorological data is taken from that location, while air pollutants are co-located at the CAO-AMX site (within 20m).
- Section 2.5: There is a small clarification needed when it comes to the description of the emissions. The authors mention EDGAR-HTAP and MEGAN, but then in the outer model domain mention that EDGAR was used. Is it meant for both anthropogenic and biogenic sources or is it still specifically for anthropogenic emissions (similary to EDGAR-HTAP)
- Line 239: the authors mention 'xylene (XYL), representing xylene and more reactive aromatic species'. There are three xylene isomers, so I would suggest writing 'xylene (XYL), representing xylenes (or 'xylene isomers') and more reactive aromatic species' and use 'xylenes' when appropriate and not referring to the modelled species in WRF-Chem. In addition, xylenes have the same mass as ethylbenzene and I'm not sure if it shows up with the same m/z.
- Lines 299-301: The authors write that organic acids 'mainly arise from biogenic emissions, anthropogenic emissions and secondary photochemical oxidation of VOCs' and it made me wonder what other sources there might be.
- Lines 466-467: The authors suggest that a 'probable cause' for increased methanol levels could be 'increased fire activity with average temperature beyond 38°C'.
- Section 3.5: While this is an 'Inter-species relationship' section, when the authors write for example that 'isoprene, monoterpenes, and MVK correlate strongly [...] indicating secondry photochemical formation and biogenic influence during high radiation periods', this seems to mean implicitly that this is the case for isoprene, so compounds correlating with it follow a similar pattern. However, the correlation matrix contains such correlations with environmental conditions, including solar radiation, so the authors could use that information as well to support their statements.