the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 09 Sep 2025
Elevated Anthropogenic Contributions to Trace Elements in Marine Aerosols Compared to Coastal Qingdao in Eastern China
Abstract. Trace elements (TEs) in aerosols over offshore eastern China originate from both terrestrial and marine emissions. However, their variations with source regions remain poorly understood. During spring and summer 2018, PM2.5 samples were collected at Qingdao, a coastal city in eastern China, and adjacent Bohai and Yellow Seas. TEs were quantified and analyzed by source region, followed by source apportionment. In spring, TE concentrations were significantly higher over land. Crustal dust contributed 39.2–77.8 % of Fe, Mn, Cr and Ni; while waste and industrial emissions contributed 29.4–70.1 % of Cu, Zn and Pb. Westerly winds conveyed anthropogenic TEs offshore, with coal combustion contributing 25.9–61.4 % to As, Cd, Pb, Zn and Cr, and oil combustion contributing 58.6–84.4 % to V and Ni in marine aerosols, indicating efficient long-range pollutant transport. In summer, dust influence declined. Biomass burning contributed 38.2–46.3 % of Zn, Cd, Pb and Cr, while vehicular emissions dominated As and Cu (41.7–57.3 %) at Qingdao. Over marine areas, anthropogenic elements (Zn, As, and Cd) occasionally exceeded coastal levels, with coal combustion remaining dominant (40.8–75.5 %). Ship emissions became especially prominent, contributing 79.3 % of V and 63.3 % of Ni offshore. Southeasterly winds transported ship-derived pollutants coastward, markedly increasing Fe (21.2 %) and Mn (14.0 %) compared to spring (1.9 % and 1.8 %, respectively). These results reveal distinct seasonal shifts in TE source across land-sea gradients, highlighting growing anthropogenic impacts, particularly from coal combustion and maritime shipping on marine aerosols. Quantifying these contributions helps assess marine biogeochemical impacts and supports targeted pollution control.
- Preprint
(1294 KB) - Metadata XML
-
Supplement
(4520 KB) - BibTeX
- EndNote
Status: open (until 21 Oct 2025)
- RC1: 'Comment on egusphere-2025-4005', Anonymous Referee #1, 20 Sep 2025 reply
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,619 | 23 | 5 | 1,647 | 6 | 13 | 10 |
- HTML: 1,619
- PDF: 23
- XML: 5
- Total: 1,647
- Supplement: 6
- BibTeX: 13
- EndNote: 10
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This study addresses the emission sources of trace metals in PM2.5 collected in Qingdao, the Bohai Sea, and the East China Sea. The content appears to fit within the scope of Atmospheric Chemistry and Physics. However, there are many major concerns that need to be addressed before this manuscript can be published. Across the manuscript, discussions on the key methods and findings of this study are very brief (see Major Concerns), while less important sections tend to be redundant. For instance, this manuscript addresses the factors controlling the metal solubility in aerosol particles. However, the discussion is somewhat speculative because no water-soluble metal concentrations were measured. Accordingly, the processes suggested to influence metal solubility may not necessarily represent the actual conditions of the samples analyzed. It would therefore be advisable either to include data on water-soluble metal concentrations and solubility, or to limit the discussion accordingly.
I am also concerned about the interpretation of the correlation analysis. The use of phrases like "strong correlation" and "significant correlation" is misleading, as these terms should be reserved for cases where the correlation coefficient (r) is at least 0.6 or 0.7. However, the reported r values in this manuscript are frequently below 0.4. For instance, in Line 237, the correlation coefficient (r) is only 0.29, yet it is described as a “strong correlation. Since the section on PMF source apportionment is arguably the most important part of this work, such an overstatement of the results compromises its scientific integrity. I strongly recommend revising the language to more honestly and cautiously reflect the actual numerical data.
The manuscript also contains many other issues that I have outlined in the "Major Concerns" and "Specific Comments". Should the authors be able to make comprehensive and substantial revisions to address all of these points, I would be open to reconsidering its publication. However, if such major revisions are not feasible, the paper should be rejected. Therefore, I cannot recommend the current version for publication in ACP.
Major comments
1. The methodology section is too brief. The following information is necessary to ensure the quality of the data obtained from the samples.
Sample duration: Please provide sampling duration for each aerosol collection or the total integrated volume should be provided. From the result and discussion section, it appears that sampling was conducted separately during the day and night, but it is unclear whether this applies only to Qingdao or also to the marine aerosol samples.
Selection of wind direction: For marine aerosol sampling, it is often insufficient to collect samples only while the vessel is underway to avoid contamination from the ship itself. It is therefore crucial to clarify whether sampling was restricted to specific wind directions (e.g., -90° to 90°, with 0° being the bow). If selection of wind-direction was not performed, the potential impact on the samples should be clearly explained. This is particularly important since one of the key findings of this study is the increasing influence of ship-derived aerosols in recent years.
Filter blank: Please clarify the levels of trace metal blanks on the filters after cleaning and indicate the impact of the filter blanks on the measured samples. Given the short sampling time, the influence of filter blanks on the PM2.5 and marine aerosol samples from Qingdao should be clearly stated.
Experimental conditions: It should be clarified whether the acid digestion of aerosols, especially the evaporation to dryness, was carried out in a clean room or under ambient air conditions. Given that metal concentrations in marine aerosols are low compared to continental aerosols, if the experiments were performed under ambient conditions, the potential impact of contamination during evaporation and drying should be addressed.
OC and EC: Please provide more detailed information. I am sorry that I cannot get any information. If authors feel method section is too long, please provide details in Supplemental Information.
PMF: It is stated that 81 samples obtained from both Qingdao and the ship-based marine observations were input together into the PMF analysis. If so, this would mean that samples collected on the same day at different sampling sites were included in the same model, which does not seem to align with the typical application of PMF. In particular, the emission sources and controlling factors for metal elements are expected to differ substantially between Qingdao and marine aerosols. In this case, including both types of samples could result in the outcomes being somewhat averaged, potentially failing to reflect the actual conditions at either Qingdao or the marine sites. If the input file indeed included both Qingdao and marine aerosol samples, the validity of the resulting source apportionment should be explicitly addressed.
By contrast, if PMF analyses were conducted separately for Qingdao and the marine aerosols (or for spring and summer), this should be clearly stated. In that case, the relatively small number of marine aerosol samples raises the question of whether the dataset is sufficiently large to support an 8-factor solution.
Considering that the PMF results form the core of this study, in either scenario, the current Methods description does not make it clear how the temporal variations in source contributions for the two sites (e.g., Figure S5) were obtained, nor does it provide information to evaluate the validity of these results.
2. In this manuscript, terms such as “coastal area,” “offshore area,” and “marine area” are frequently used to indicate aerosol sampling sites in Qingdao, the Bohai Sea (BS), and the Yellow Sea (YS). “Coastal” presumably refers to Qingdao. In the case of “offshore” and “marine”, it was often unclear whether they referred exclusively to either the Bohai Sea or the Yellow Sea, or to both regions together. As a result, it took considerable effort to determine (and in many cases I could not determine) which area the terms “marine/offshore aerosol” referred to throughout the manuscript, which significantly reduced readability. Therefore, I strongly suggest that the sampling sites be explicitly specified as Qingdao, BS, and YS, in order to improve clarity.
3. In the discussion of source contributions to PM2.5, it is excellent that both relative contributions (percentages) and absolute contributions (atmospheric concentrations: μg/m³) are presented. Unfortunately, the discussion for the source apportionment of metal elements in Section 4.3 seems to be based solely on relative contributions. When assessing the supply of metal elements to the ocean, the absolute deposition is more important than the relative contribution of anthropogenic emissions. For instance, while the contribution of coal combustion-derived Fe to PM2.5 is higher in summer (60%) than in spring (about 20%) over the Yellow Sea, the Fe concentration itself is more than three times higher in spring than in summer (Fig. 2). Therefore, the deposition of coal combustion-derived Fe to the ocean surface is likely higher in spring or at least comparable to summer. In my view, the regional and seasonal variations in relative source contributions are considerably smaller than the variations observed in absolute concentrations. Thus, as demonstrated earlier for Fe emitted from coal combustion, similar cases are likely to occur in the discussions of other elements as well. Consequently, the current description could lead to significant misunderstandings about the seasonality of metal element inputs into the ocean from anthropogenic aerosols. I therefore strongly recommend that a discussion of absolute concentration be included.
Specific comments
L20: Confirm whether Ni is truly from nearby sources. In the following text, it is stated that most Ni originates from heavy oil combustion.
L24: In the abstract, vehicular emissions are described as important for As and Cu, but in the main text, vehicular emissions are reported to be important for OC/EC, Cu, and Zn. Please clarify which is correct.
L26–29: The abstract states that ship emissions had a significant impact on Fe and Mn, whereas the main text indicates that the impact was from coal combustion (in fact, the main text does not explicitly mention Fe from ship or heavy oil combustion).
L49–51: If 20% of d-Fe is anthropogenic Fe, this indicates that more than half of the d-Fe originates from mineral dust. In that case, if anthropogenic Fe is indeed more efficient than mineral-derived Fe in promoting primary production, it would be advisable to explain the reasons for this in the discussion.
L65–69: This study performed the source apportionment of total metals, not dissolved metals. Consequently, it is difficult to address whether atmospheric processes affect metal solubility or not. Therefore, it seems inconsistent to present such discussion as if it were directly linked to the objectives of this study.
L71: Typo: contributs → contributes.
L83–84: Are the 18 and 9 samples the combined totals from the Bohai Sea (BS) and Yellow Sea (YS) in spring and summer, respectively? If that is the case, please specify the number of samples from each region.
L108–109: Please clarify the rationale for analyzing filter blanks every 10 samples.
L145–146: Please elaborate the reasons for concluding that Zn is of anthropogenic origin. And what does it mean that anthropogenic zinc shows the highest concentration? Enrichment factor relative to average upper continental crust (EF = (X/Al)aerosol/(X/Al)crust, X is your target elements) is helpful to discuss source of metal elements in aerosol particles.
L152–154: Figure 2 appears to plot the average metal concentrations of aerosols collected from each region. However, average values are highly sensitive to outliers with extremely high concentrations, particularly when the number of samples is small. Consequently, the higher mean concentrations of Zn observed in summer compared to spring at YS are likely influenced by samples such as SU005. Therefore, it is important to carefully consider whether SU005 should be included in the calculation of summer average concentrations (or consider plotting the median values instead of the mean) when comparing with spring samples.
Furthermore, this sample frequently appears in discussions of key aspects of this study. Omitting detailed information about SU005 in the manuscript would therefore be highly unhelpful, and its characteristics should be explicitly addressed within the discussion.
L158–159: V and Ni were employed as tracer elements for heavy oil combustion in L147–148; however, heavy oil combustion does not appear to be included here. Since this is also important for source apportionment using PMF, a consistent discussion is required.
L165–167: This discussion is highly speculative. If only the anthropogenic tracers (Zn, V, Ni, Cd, etc.) in YS aerosols showed lower concentrations in 2018 compared with 2011, the decrease could be attributed to emission regulations. However, the concentrations of mineral dust tracers, such as Fe and Al in the PM2.5 samples collected in your study, are also significantly lower than in 2011. Therefore, rather than reflecting the effect of emission controls, the difference may simply indicate that less East Asian aerosols were transported to the Yellow Sea in 2018 than in 2011. If you wish to emphasize the impact of emission regulations, a more detailed discussion is required.
L175–181: The paragraph is difficult to follow and requires revision. Specifically, (i) what does the lower Fe concentration in Qingdao compared to other megacities imply? Does it indicate a smaller contribution from anthropogenic Fe? (ii) Please explicitly indicate the anthropogenic sources for Zn, Pb, and As. (iii) what does “sampling time” refer to? Does it mean the year when samples were collected, the time of day (e.g., daytime vs. nighttime), or the sampling duration? In any case, a more detailed discussion is necessary.
L188–190: It is likely that the correlation between Factor 1 and NO₂ or CO indicates that the normalized contribution of Factor 1 obtained from PMF correlates with NO₂ or CO. However, for readers who are not familiar with PMF, this is not easy to understand. A more detailed explanation is needed.
L204: Please correct Ca2+ to Ca2+.
L204–205 What does the anthropogenic origin of Co in the dust factor indicate?
L205–207 While many studies employ elemental ratios with respect to Al for source estimation, what is the reason for using ratios relative to Ca in your study? In addition, Ca in the denominator means Ca2+? Although Ca²⁺ was measured by ion chromatography after water extraction, is it certain that all Ca present in the aerosols is dissolved?
L213: Please correct PM2.5 to PM2.5.
L215–216: At least to me, Cu and Zn do not appear to contribute to this factor.
L224–227: The text is extremely difficult to understand. The term “marine area” in the first sentence likely refers to either the Bohai Sea or the Yellow Sea. However, the time series trend of sea salt in marine area (probably represented by the dashed line and open circles in Figure S5) does not resemble that of the dust factor in either summer or spring. In contrast, for the coastal site (presumably Qingdao) shown in the bar graph, the time series trends of sea salt and dust appear to be similar in spring, but not in summer. However, spring is the season with the highest emissions of Asian dust throughout the year, and it is also the period when the influence of air masses from inland regions is strongest (Fig. 1). As a result, dust concentrations are elevated in spring. Nevertheless, it is difficult to understand why the time series trends of dust and sea spray would be similar and attributed to reverse transport from the marine area. If reverse transport is indeed the cause, a more convincing explanation should be provided.
L:240–242: It seems that coal combustion plays an important role in the subsequent discussion; however, the explanation provided is rather brief. A more detailed discussion is warranted.
L300–305: Although the differences from 10 years ago appear to be small (25.2%→30.7% for secondary nitrate, 25.7%→19.3% for secondary sulfate, and 10.0%→11.9% for vehicular emissions), are these changes statistically significant?
L309: Previous sections treated the coastal area similarly to urban areas. Why is it now referred to as a 'background' source, especially in a context where residual oil combustion is found to be low? This framing seems to be a rhetorical choice.
L316–317: Please explain the reason why the transition to low-sulfur fuels led to an increase in the Ni/V ratio.
L333: sulfur cycle → carbon cycle?
L338–347: In the abstract, the importance of heavy oil combustion as a source of Fe in marine aerosols was mentioned; however, this does not seem to be the case here. Please clarify the content.
L352–353: This discussion is highly speculative. Actually, internal mixing of mineral dust with sea spray aerosol has been reported by previous studies. However, the effect on fractional Fe solubility is usually not positive. For instance, Hsu et al. (2010) showed negative impact of sea salt on fractional solubility of Fe. Sakata et al. (2022) showed that organic carbon in sea spray has potential to suppress Fe precipitation under weakly-acidic conditions, but not further Fe dissolution from mineral dust. Furthermore, according to Wu et al. (2023), when attempting to extract Fe from aerosols collected in East Asia using seawater, the fractional Fe solubility decreased compared to ultrapure water extraction in the absence of organic matter, whereas in the presence of organic matter, the fractional Fe solubility was comparable to that obtained with ultrapure water extraction. Please elaborate how Fe dissolution is enhanced by internal mixing between mineral dust and sea spray aerosols.
L406–407 Why internal mixing of mineral dust and sea spray is the synergistic interaction between mineral dust and anthropogenic matters.
Figure S5. This figure is frequently mentioned in the text. Could you please move it into the main body of the paper?
References
Hsu, S.-C., Liu, S. C., Arimoto, R., Shiah, F.-K., Gong, G.-C, Huang, Y.-T., Kao, S.-J., Chen, J.-P., Lin, F.-J., Lin, C.-Y., Huang, J.-C., Tsai, F., Lung, S.-C. C.: Effects of acidic processing, transport history, and dust and sea salt loadings on the dissolution of iron from Asian dust, J. Geophys. Res., 115, D19313. https://doi.org/10.1029/2009JD013442, 2010.
Sakata, K., Kurisu, M., Takeichi, Y., Sakaguchi, A., Tanimoto, H., Tamenori, Y., Matsuki, A., and Takahashi, Y.: Iron (Fe) speciation in size-fractionated aerosol particles in the Pacific Ocean: The role of organic complexation of Fe with humic-like substances in controlling Fe solubility, Atmos. Chem. Phys., 22, 9461–9482, https://doi.org/10.5194/acp-22-9461-2022, 2022.
Wu, H. Y., Hsieh, C. C., Ho, T. Y.: Trace metal dissolution kinetics of East Asian size-fractionated aerosols in seawater: The effect of a model siderophore. Mar. Chem., 254, 104277. https://doi.org/10.1016/j.marchem.2023.104277. 2023.