the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 26 Aug 2025
Atmospheric nitrogen deposition fluxes into coastal wetlands and their impacts on ecosystem carbon sequestration in East Asia
Abstract. Coastal wetlands serve as critical sinks for both carbon and nitrogen within regional ecosystems, playing an essential role in mitigating atmospheric greenhouse gases and nutrient enrichment. This study integrates high-resolution wetland type data, AIS-based ship emission inventories, and regional nitrogen deposition simulations to quantify nitrogen inputs to East Asian coastal wetlands from the perspective of source–sink coupling. Firstly, the atmospheric nitrogen deposition fluxes in coastal wetland areas of East Asia were simulated and evaluated with WRF-CMAQ. Nitrogen deposition fluxes were spatially coupled with classified wetland maps in ArcGIS. Net primary productivity (NPP) was estimated using a modified CASA light-use efficiency model, incorporating solar radiation and FPAR from remote sensing. Carbon sequestration and oxygen release were then quantified using stoichiometric relationships based on NPP. The results indicate that total nitrogen deposition across East Asian coastal wetlands follows a general gradient of “high in the south, low in the north” and “strong in urban-industrial clusters, weak in remote coastal zones.” On average, ship emissions contribute 10.13 % and 15.22 % to NO3--N and NH4+-N deposition, respectively, while their contribution to gaseous NH3-N is negligible. Among wetland types, salt marshes receive the highest nitrogen input per unit area (654.99 mg NO3--N·m-2·yr-1), although tidal flats dominate total regional nitrogen input due to their extensive spatial coverage. Dry and wet deposition exhibit significant seasonal variation: wet deposition consistently prevails during the spring and summer months due to frequent rainfall, while dry deposition becomes increasingly prominent in autumn and winter. For instance, in the Korean Peninsula, the wet-dry gap in nitrate deposition reaches 0.17 g N·m-2·yr-1, while the Yangtze River Delta exhibits relatively balanced ammonium inputs (dry-wet difference of only 0.05 g N·m-2·yr-1). Carbon sequestration capacity shows strong spatial and temporal coupling with nitrogen deposition. Mangrove forests exhibit the highest annual NPP (~776.16 g C·m-2·yr-1 in summer), supported by high FPAR and solar radiation (1749.29 MJ·m-2), followed by salt marshes and tidal flats. Seasonal patterns reveal a summer peak in carbon uptake across all wetland types, with mangrove NPP in summer being two times higher than winter values. Nitrogen deposition primarily enhances carbon sequestration during warm seasons; for instance, in the mangroves of the Pearl River Delta, nitrogen inputs increase summer CO2 fixation by 6.85 g C·m-2, while the effect is negligible in winter (<0.06 %) or in nitrogen-saturated regions. These findings provide a scientific foundation for understanding how coastal ecosystems respond to anthropogenic activities and long-range nitrogen transport. Furthermore, the results serve as an important reference for wetland conservation, nitrogen cycle management, and the development of regional carbon neutrality strategies.
- Preprint
(1488 KB) - Metadata XML
-
Supplement
(394 KB) - BibTeX
- EndNote
Status: open (until 27 Nov 2025)
-
RC1: 'Comment on egusphere-2025-3801', Anonymous Referee #1, 12 Nov 2025
reply
-
AC1: 'Reply on RC1', Yan Zhang, 26 Nov 2025
reply
We greatly appreciate the reviewer’s thorough evaluation and constructive feedback. The comments have helped us significantly improve the scientific robustness and presentation of the manuscript. We have addressed each point in detail and revised the manuscript accordingly. A point-by-point response is provided below, with all modifications clearly indicated in the revised version. More details of the response could be found in the supplement file.
1.The number of references cited throughout the manuscript is insufficient, and many places that should be supported by citations are not. For example, in lines 215–230, many key concepts and mechanisms are presented without referencing the relevant literature.
R: Thank you for your suggestion. We conducted a full-text review and added citations of references in some places where literature support was needed. The number of references has increased from 69 to 108 now.
2.The discussion is not sufficiently in-depth.
R: Thank you for your suggestion. Additional content has been added to our Results and Discussion, such as comparing the simulation results of the impact of ship emissions of nitrogen deposition in this study with other studies (Section 3.1). The reasons for the seasonal differences in nitrogen deposition in different coastal wetland areas were further discussed and summarized (Section 3.2). The changes of NPP in coastal wetlands of East Asia were discussed from the perspectives of seasonal differences, regional differences and differences among different types of wetlands (Section 3.3). For specific modifications, please refer to Results and Discussion. For the specific revisions in the manuscript, please refer to Supplement.
3.The manuscript lacks comparison with other related studies.
R: Thank you for your suggestion. In Section 3.1, we added the contribution of ship emission inventories to other studies for comparison. Also, we have added comparisons with other studies in the discussion sections of Sections 3.1 and 3.2.
4.In Figure 2, the N and E (north and east) indicators are missing from the map.
R: The N and E (north and east) indicators are added in the Figure 2.
5.In Table 1, there is an extra horizontal line under “Anthropogenic”.
R: Thank you for your suggestion. The table seems to have extra line segments due to the issue of page spread. We have adjusted this table.
6.Does “total nitrogen deposition” refer only to inorganic nitrogen deposition, or to the sum of inorganic and organic nitrogen deposition?
R: The relevant explanations have been added to the methodology.
In this study, total nitrogen deposition refers exclusively to total inorganic nitrogen (TIN), defined as the sum of oxidized inorganic nitrogen species (NO2, NO, NO3⁻) and reduced inorganic nitrogen species (NH3 and NH4+).
7.Why choose 1, 4, 7, 10 to represent spring, summer, autumn, and winter instead of 12-2 to represent winter, and 3-5 to represent spring? Because this is not the result of a post-sampling experiment, so I think it's better to choose three months to represent a season.
R: On one hand, in this study we have to do a series of sensitive simulations for different emission categories in our experimental designs, the simulation of continuous months would require a lot of computational costs, which was not feasible in practical study. On the other hand, the selection of January, April, July and October to represent winter, spring, summer and autumn follows a widely adopted practice in regional atmospheric modeling studies that require seasonally representative simulations rather than continuous multi-month runs. These four months are commonly used as seasonal proxies because they correspond to the midpoint of each climatological season and minimize transitional effects associated with monsoon onset and withdrawal. Our goal was to capture the characteristic meteorological conditions, emission patterns and deposition behaviors of each season through independent simulations. January, April, July and October provide the most stable representation of seasonal atmospheric states while avoiding inter-month variability associated with early or late seasonal transitions. This approach is consistent with numerous WRF-CMAQ and regional climate studies that adopt single-month seasonal proxies when performing factorial or scenario-based experiments rather than long continuous runs. For this reason, we have added the following explanations of limitations in the methodology:
The use of a single representative month for each season is a methodological simplification relative to full three-month seasonal simulations. This choice was dictated by the factorial experimental design, which required independent simulations under two emission scenarios, and by the associated computational demands. Although this single-month representation is widely adopted in regional atmospheric modelling studies and has been demonstrated to capture the characteristic meteorological and chemical features of each season(Wu et al., 2021; Li et al., 2018), it inevitably introduces some degree of uncertainty related to intra-seasonal variability. Future work involving continuous multi-month simulations for each season would help further constrain this uncertainty.
8.I think the methods section does not clearly explain how different emission sources are distinguished. This may cause readers to question the robustness of your source-specific results.
R: Thank you for your suggestion. We have added explanations on how to distinguish different emission source areas in the methodology. Specifically as follows:
In developing the emission inventories, this study explicitly separated terrestrial anthropogenic sources from marine ship emissions to enable a source-specific attribution of atmospheric nitrogen inputs to coastal wetlands. The land-based anthropogenic emissions for China were derived from the Multi-resolution Emission Inventory for China (MEIC) , while emissions for other Asian regions were based on the MIX Asian inventory(Yue et al., 2017). Ship emissions were calculated using a bottom-up method based on AIS data with a fine resolution, following established practices for high-resolution marine emission modelling (Jiang et al., 2024; Fan et al., 2016). Based on the inventory of land-based anthropogenic, two parallel emission scenarios were constructed. The first scenario included both land-based anthropogenic and marine ship emissions (with shipping scenario). The second scenario excluded all ship emissions (without-shipping scenario). The contribution of shipping to nitrogen deposition was quantified by comparing the deposition fields simulated under these two scenarios.
9.The seasonal differences are not discussed in sufficient detail. For example, in lines 307–309: “In other words, even within regions that share similar latitudinal positions and ecological characteristics, the ratio of dry to wet deposition can exhibit significant variation due to differences in atmospheric circulation patterns, precipitation regimes, and land-sea interactions.” You could try to incorporate backward trajectory analysis and precipitation data (if available) to support and deepen this discussion. There are also no relevant references cited here.
R: We thank the reviewer for the insightful suggestion. We have now expanded the seasonal interpretation by incorporating mechanisms involving monsoon circulation, precipitation patterns, and marine air-mass transport, supported by relevant literature. Although backward-trajectory analyses were not conducted within the current study, we now cite previous trajectory-based studies over Kyushu and coastal Japan that demonstrate the influence of clean maritime inflow on nitrogen deposition patterns. This revision strengthens the mechanistic understanding of seasonal variability and improves the robustness of our interpretation. The specific modifications are as follows:
In other words, even within regions that share similar latitudinal positions and ecological characteristics, the ratio of dry to wet deposition can exhibit significant variation due to differences in atmospheric circulation patterns, precipitation regimes, and land-sea interactions (Zhao et al., 2017; Zhang et al., 2006). For example, the annual dry and wet sedimentation fluxes of nitrate nitrogen and ammonium nitrogen in Kyushu of Japan are 183.92 tons, 337.05 tons, 15.45 tons and 49.24 tons respectively. Nitrogen deposition over Kyushu is generally lower than that observed in heavily industrialized regions of China, a pattern supported by nationwide monitoring records showing comparatively modest wet and dry nitrogen inputs across Japan’s coastal zones (Itahashi et al., 2021; Morino et al., 2011). These studies attribute the lower levels of nitrogen deposition largely to weaker local emission sources and the dominant influence of maritime air masses, which dilute atmospheric reactive nitrogen prior to deposition. This relatively low input of nitrogen is likely a consequence of limited industrial emissions coupled with the predominant influence of clean maritime air masses originating from upwind oceanic regions. (Hayashi et al., 2021; Kiriyama et al., 2021).
10.it is better to use “Autumn” rather than “Fall.”
R:Thank you for your suggestion. The relevant expressions in the manuscript have been corrected in full.
-
AC1: 'Reply on RC1', Yan Zhang, 26 Nov 2025
reply
-
RC2: 'Comment on egusphere-2025-3801', Anonymous Referee #2, 13 Nov 2025
reply
This study used WRF-CMAQ with integrated high-resolution wetland type data, AIS-based ship emission inventories, and regional nitrogen deposition simulations to quantify nitrogen inputs to East Asian coastal wetlands from the perspective of source–sink coupling. The findings provide a scientific foundation for understanding how coastal ecosystems respond to anthropogenic activities and long-range nitrogen transport. This study has a certain level of innovation and logic. However, major revisions are still needed.
- In this study, the simulated nitrogen deposition flux is the most important model result. However, there is no information about how the flux is simulated. I suggest adding the process in the Methods section.
- The ship emissions inventory used in this study only considered NOₓ, NH₃, PM₂.₅, and PM₁₀. However, commonly used ship emissions inventories include SO₂, NOₓ, PM₂.₅, CO, hydrocarbons, and GHG species (Yi et al., 2025). This study highlights the impact of nitrogen species on coastal wetlands; however, this limitation still needs to be mentioned, as these species can interact with each other.
- In the process of calculating carbon sequestration, the authors did not mention which parameters are based on model results and which are based on literature.
- In Figure 1, the unit for nitrogen emissions is missing. Besides, the nitrogen emissions are not clearly defined: does the nitrogen here only include NO and NO₂, or does it contain other species?
- The first paragraph in Section 3.1 did not cite any figures, tables, or references. It is not clear where the results come from.
- There is a distinct mistake in Line 66 of the Supplementary Information (SI).
Yi, W., Wang, X., He, T., Liu, H., Luo, Z., Lv, Z., and He, K.: The high-resolution global shipping emission inventory by the Shipping Emission Inventory Model (SEIM), Earth Syst. Sci. Data, 17, 277–292, https://doi.org/10.5194/essd-17-277-2025, 2025.
Citation: https://doi.org/10.5194/egusphere-2025-3801-RC2 -
AC2: 'Reply on RC2', Yan Zhang, 26 Nov 2025
reply
We sincerely appreciate the reviewer thoughtful feedback and the time they dedicated to evaluating our work. Their comments have greatly helped us enhance the quality of our study and its presentation. We have carefully addressed each comment in the responses that follow and have implemented all changes in the revised manuscript.
1.In this study, the simulated nitrogen deposition flux is the most important model result. However, there is no information about how the flux is simulated. I suggest adding the process in the Methods section.
R: Thank you for your suggestion. We added the process of nitrogen deposition simulation to the methodology. Specifically as follows:
In this study, total nitrogen (TN) refers exclusively to total inorganic nitrogen (TIN), defined as the sum of oxidized inorganic nitrogen species (NO2, NO, HNO3-/NO3⁻) and reduced inorganic nitrogen species (NH3 and NH4+). TIN was simulated for four representative months of 2017 (January, April, July and October), corresponding to winter, spring, summer and autumn. Selecting single representative months has been widely adopted in regional modelling to capture climatological seasonal characteristics under factorial experimental designs (Li et al., 2019; Qi et al., 2017). Each simulation used a five-day spin-up period to minimize the influence of initial conditions. In total, eight model experiments were conducted, consisting of four months and two emission scenarios. This simulation framework provided a consistent basis for evaluating both seasonal variations and the source-specific contributions of nitrogen deposition in East Asian coastal wetlands. The nitrogen deposition flux was directly output from the model. The dry deposition flux for each nitrogen species was calculated by the model based on the dry deposition velocity multiplied by the simulated surface-layer concentration. The wet deposition flux was simulated by scavenging nitrogen species from the atmosphere through both in-cloud and below-cloud processes.
2.The ship emissions inventory used in this study only considered NOₓ, NH₃, PM₂.₅, and PM₁₀. However, commonly used ship emissions inventories include SO₂, NOₓ, PM₂.₅, CO, hydrocarbons, and GHG species (Yi et al., 2025). This study highlights the impact of nitrogen species on coastal wetlands; however, this limitation still needs to be mentioned, as these species can interact with each other.
R: Thank you for the reviewers' suggestions. During the simulation process, the list used actually did contain SO2, NOx, PM2.5, CO, hydrocarbons and types of greenhouse gases, but this was not expressed clearly in the previous manuscript. Therefore, we supplemented the relevant explanations in the manuscript and increased the citations of the literature. The added content in the manuscript is as follows:
The resulting ship emissions inventory includes nitrogen oxides (NOx), ammonia (NH3), and particulate matter (PM2.5, PM10), sulphur dioxide (SO₂), NOₓ, carbon oxygen (CO), hydrocarbons, and greenhouse gas (GHG) species (Yi et al., 2025).
3.In the process of calculating carbon sequestration, the authors did not mention which parameters are based on model results and which are based on literature.
R: Both Section 2.3 in the manuscript and Section 1.3 in the attachment elaborate on the calculation method of NPP. SOL is calculated from the Global High-Resolution (3-hourly, 10 km) Surface Solar Radiation Dataset (1983-2018, monthly) described in the attachment. The values of FPAR and ε are derived from the literature. All other parameters are derived from the model results. Relevant explanations have been supplemented in Section 2.3 of the manuscript. The added content in the manuscript is as follows:
In the CASA model, biome‐specific constant FPAR values were assigned to different coastal wetland types to reflect their contrasting canopy structures and vegetation cover. Specifically, an FPAR of 0.85 was used for mangroves, consistent with satellite‐derived APAR estimates for dense mangrove forests (Zheng and Takeuchi, 2022). A moderate FPAR of 0.65 was adopted for salt‐marsh wetlands, in line with typical growing‐season FPAR (≈0.4–0.7) reported for marsh vegetation. For sparsely vegetated tidal flats, an FPAR value of 0.10 was chosen to represent the dominance of water and bare sediment and the low emergent leaf area during most tidal cycles (Hawman et al., 2023).
4.In Figure 1, the unit for nitrogen emissions is missing. Besides, the nitrogen emissions are not clearly defined: does the nitrogen here only include NO and NO₂, or does it contain other species?
R: Thank you for your suggestion. In the original figure, N refers to the nitrogen element. To reduce ambiguity, we have added the explanation of this part in our hands. The added content in the manuscript is as follows:
Overall, the nitrogen (N element) emission inventory and wetland type distribution in East Asia adopted in this study are shown in Fig. 1.
5.The first paragraph in Section 3.1 did not cite any figures, tables, or references. It is not clear where the results come from.
R: Thank you for your suggestion. We added the citations of Table 1 and the references in the first paragraph of Section 3.1 of the manuscript.
6.There is a distinct mistake in Line 66 of the Supplementary Information (SI).
R: Thank you for your suggestion. The reference error that existed here has been corrected.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,921 | 73 | 23 | 2,017 | 40 | 31 | 40 |
- HTML: 1,921
- PDF: 73
- XML: 23
- Total: 2,017
- Supplement: 40
- BibTeX: 31
- EndNote: 40
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This study applies a WRF–CMAQ modelling framework combined with multi-source emission inventories and high-resolution wetland maps to quantify nitrogen deposition in East Asian coastal wetlands and assess its impacts on carbon sequestration in different wetland types. The manuscript is logically organized and provides valuable model-based insights into source-specific nitrogen inputs and spatiotemporal deposition patterns. However, several methodological details, particularly the separation of ship and anthropogenic sources, the diagnosis of source-specific deposition fluxes, and the treatment of uncertainties, require clearer description before publication. At the same time, I think that the spatial distribution of nitrogen deposition should include the whole region, e.g. Yellow Sea, East China Sea. The manuscript has some originality and significance, but the writing is not rigorous enough, the arguments are not comprehensive enough, and the discussion is not in-depth enough. Careful revision is recommended.