Rainwater and aerosol samples were collected from a coastal urban area (Qingdao) and remote islands (Qianliyan and Shengsi) and along cruise tracks in the Yellow Sea and East China Sea from 1997 to 2005. The samples were analyzed for nitrogen species (NO3−, NO2−, NH4+, and organic nitrogen) and other important elements. The nitrogen species concentrations showed considerable temporal and spatial variations for wet as well as dry atmospheric depositions. In addition, there was a dramatic reduction in the influence of anthropogenic emissions on nitrogen species with increasing distance from coastal urban stations to remote areas across the Yellow Sea and East China Sea. The monsoon climate of East Asia also had prominent effects on the atmospheric composition of nitrogen, with higher loadings in northerly (i.e., winter) than southerly (i.e., summer) monsoon periods, owing to strong emissions from the East Asian landmass. Dust storms in spring dramatically reduced the periodically high concentrations of atmospheric pollutants (e.g., nitrogen species) across the NW Pacific Ocean, but this was accompanied by a twofold-to-fourfold increase in the temporal deposition flux, which showed broad spatial dimensions. Finally, our study identified a strong gradient of wet as well as dry nitrogen deposition fluxes from East Asia to the interior of the North Pacific Ocean. The gradient reflected changes in emission sources and chemical reactions (e.g., forming secondary aerosols), rainfall and scavenging, and change in air mass trajectory.
 Atmospheric inputs of plant nutrients (e.g., nitrogen) and trace elements are an important consideration in the science of climate change. Biogeochemical cycles and interactions between the atmosphere, the continents, and the ocean are strong sources of feedback in the evolution of climate [Duce et al., 2008; Gruber and Galloway, 2008]. For example, atmospheric wet (e.g., rain and snow) and dry (e.g., aerosol) depositions impact the mechanism by which nitrogen is released and exerts effects on earth systems remote from the source region over a time scale as short as days but with vast spatial dimensions [Galloway et al., 2004; Jickells, 2006].
 Studies have shown that nutrient depositions from the atmosphere on to the Earth surface have increased as a result of anthropogenic perturbations over the last 100 years. This, in turn, is thought to affect carbon fixation in the upper part of the ocean, balancing, at least in part, the increase in the atmospheric supply of carbon dioxide [Herut et al., 1999; Baker et al., 2003; Carbo et al., 2005]. Moreover, it has been shown that atmospheric sources of nutrients may contribute considerably to new production in the surface ocean, particularly in oligotrophic areas, and/or trigger algal blooms [Zhang, 1994; Paerl, 1997; Zhang et al., 2004]. The NW Pacific Ocean at midlatitudes is an important receptacle of land-derived materials (e.g., from rivers) in addition to having the highest aeolian fluxes of soil dust and anthropogenic emissions because of its proximity to the Gobi Desert in Central East Asia and large, rapidly modernizing population centers located upwind.
 The Yellow Sea and the East China Sea are located 22.5–40°N of the NW Pacific Ocean and both are under the influence of East Asia monsoon climate. Northerly (i.e., northeast and northwest) winds begin to prevail over the Yellow Sea and East China Sea in late September each year. The winter monsoon reaches a maximum in December–February, depending on the interactions between the Siberian-Mongolian High and the Aleutian Low. The summer monsoon starts in late April, such that by August the Yellow Sea and East China Sea are subject to predominantly southerly (i.e., southwest and southeast) winds, which originate from the tropical Indian and West Pacific Oceans.
 Although the global flux of nitrogen has been simulated in several models, the details of atmospheric depositions in marine recipient environments are still not well understood due to measurement difficulties and thus a lack of data. At the regional scale, the marginal seas of the NW Pacific Ocean have been scarcely studied despite the fact that air–sea interactions in this region are of critical importance in terms of global climate change. In this study, atmospheric wet and dry depositions were sampled at three sites along the coast and on the islands of the Yellow Sea and East China Sea; these data were supplemented with shipboard measurements made between 1997 and 2005. The samples were analyzed for nitrogen species and major ions and the results interpreted in the context of information derived from meteorological records and air mass back trajectories, which together enabled us to identify different sources of the atmospheric depositions and to estimate their wet versus dry fluxes; particularly, the following questions were addressed: (1) gradient of nitrogen species across marginal region of NW Pacific Ocean and impact of anthropogenic emissions (section 4.1), (2) impact of monsoon climate on the variability of aerosols and spring dust bloom events (sections 4.2 and 4.3), (3) comparison of deposition fluxes between different parts of NW Pacific Ocean (section 4.4), and (4) contribution of inorganic versus organic nitrogen in the atmospheric deposition (section 4.5). These findings were used to assess the significance of atmospheric depositions in nitrogen cycling in the marginal seas of the NW Pacific Ocean.
2. Materials and Methods
2.1. Collection of Rainwater and Aerosol Samples
 The atmospheric sampling of rainfall and aerosols was performed during the period 1997–2005 at three stations: for the Yellow Sea, at Qianliyan Island and Qingdao, and for the East China Sea, at Shengsi Island, (Figure 1 and Table 1). Samples from these sites were obtained on a north-south transect along the rim of the Yellow Sea and East China Sea. Rain and aerosol samples were also obtained during seagoing cruises using techniques discussed elsewhere [Zhang et al., 2005].
Table 1. Observational Sites of This Work, Including the Geographical Features of Qianliyan, Qingdao, and Shengsi, With a Summary of Meteorological Information
Description of Sampling Sites
Qingdao (Yellow Sea)
Located on top of a building that is 500 m from the coastline, later moved to the meteorological station on a small hill 1500 m from the coast. Air pollution is supplied from traffic, domestic combustion, and transport from urban sources upwind.
Qianliyan (Yellow Sea)
Located on an island of 50–60 km from the mainland, with 20–30 habitants. The aerosol and rain collectors were installed in a meteorological observatory. Air pollution is derived from the nearby fishing harbor and navigation, as well as anthropogenic sources from the mainland to the north and west with high densities of population and manufacturing.
Shengsi (East China Sea)
Located on an island of 60 km from the mainland, with 11,000 habitants distributed over an area of 5.12 km2. Air pollution is derived from a nearby fishing harbor, anthropogenic emissions from the village and remote mainland sources to the west.
 Rainwater samples were collected on an event basis in precleaned plastic funnels with a surface area of 400 cm2 [Zhang et al., 1999]. Subsamples were used for pH measurements. Rainwater samples were filtered through Nuclepore filters (pore size: 0.4 μm), treated by the addition of chloroform to a concentration of 0.4% (V/V), and then stored in the dark at 4°C. At Qianliyan and Shengsi, aerosol samples were collected on Nuclepore filters using a low-volume (flow rate: 120 l/min) pump prior to 2005, due to logistical constrains [cf. Zhang et al., 2002], while at Qingdao a high-volume (flow rate: 1 m3/min) pump was used. The sampling time ranged from 24 to 48 h per filter, depending on loading of the aerosols. This sampling procedure was consistent with the long-term collection of data sets that began in the early 1990s [Zhang et al., 1993, 1999; Liu et al., 2002].
 Gases species of nitrogen (e.g., NH3 and HNO3) were not measured in this study, because of lack of instrumentation. Field campaigns were stopped at Qingdao (1999–2003), Qianliyan (1999) and Shengsi (2004–2005), when the electricity supply on the island was shut down and sampling system (e.g., pump) has to be repaired and/or replaced.
2.2. Chemical Determinations
 In the laboratory, the filters were weighed before and after sample collection to estimate aerosol loading (μg/m3) and the amount of particles in rainwater (μg/L). Nitrate, NO2−, and NH4+ in rainwater were determined photometrically using an autoanalyzer (Model: Skalar SANplus) with a precision of <5–10%. National seawater references were run with each batch of samples to determine the precision of nutrient analysis. The concentration of dissolved inorganic nitrogen (DIN) was defined as the sum of the concentrations of NO3−, NO2−, and NH4+. Total dissolved nitrogen (TDN) was measured according to the methods of Grasshoff et al. . The dissolved organic nitrogen (DON) concentration was estimated by subtracting the DIN from TDN. Anion concentrations (Cl−, F−, and SO42−) in rainwater were determined by ionic chromatography (Model: EP-2000), and cation (Ca2+, K+, Mg2+, and Na+) concentrations by atomic absorption spectrometry (Model: WFX-1C). Aerosol samples were ultrasonically extracted for 1 h with purified water obtained from a Milli-Q Plus (resistivity > 18 MΩ cm, Millipore Co.), filtered through Nuclepore filters, and the filtrates analyzed photometrically to determine the amounts of NO3−, NO2−, and NH4+ using an autoanalyzer (Model: Skalar SANplus).
 All laboratory wares, including those for field sample collection and laboratory chemical analyses, were first soaked in 1:5 HCl (v/v), followed by thorough rinses with distilled and then Milli-Q waters to avoid contamination. Statistical analysis of data was carried out using the Statistical Package of the Social Sciences Software for Windows [Ma, 2001].
2.3. Estimate of Deposition Fluxes
 The wet deposition flux (Fw) was calculated from precipitation (P) data and the concentration (Cw) of the substance of interest on a rain event basis, such that:
The monthly deposition flux was obtained by summarizing the data for all individual rain events for a given month after correction for the difference between the amount of rainfall and the samples that were used for chemical analysis [Zhang et al., 1999].
 The dry deposition flux (Fd) was calculated using the concentration of the nitrogen species leached from aerosols (Cd) and the deposition velocity (V):
In equation (2), the term V varied depending on the aerosol size and composition, i.e., gravitational settling for large particles (i.e., soil dust and sea salts), and impaction and diffusion for particles of submicrometer size (e.g., anthropogenic emissions). Meteorological conditions in the troposphere were also important factor affecting the deposition velocity. After comparison of the deposition velocity for ammonium and nitrate aerosols based on the data from literature, size-fraction weighed deposition velocity in calculation of dry deposition fluxes was used, that is 1.15 cm/s for nitrate plus nitrite and 0.5 cm/s for ammonium aerosols (cf. Shi et al.  and Table S1 in the auxiliary material).
 The concentrations of inorganic nitrogen species (NO3−, NO2−, and NH4+) at the three coastal and island stations varied considerably, ranging from 0.1 μM to 700–800 μM for nitrate and ammonia and 0.1 μM to 25–30 μM for nitrite, and the latter is usually a minor component of DIN. There was an overall trend in rainwater composition with respect to inorganic nitrogen species, with a higher concentration in coastal urban areas than on the remote islands, and higher nitrate and ammonia levels at Yellow Sea stations (Qianliyan) than in the East China Sea station (Shengsi).
 At Qianliyan, the rainwater pH was 3.25–7.50, with a volume-weighted average of 4.59. A pH ≤ 5.6 (i.e., acid rain) was measured in 65.1% of samples, reflecting the dominant influence of upwind land source anthropogenic emissions. The nitrate and ammonia concentrations in the rainwater samples were lower (1–10 μM) in summer, when rainfall is the most abundant. By contrast, extremely high levels (up to 300–400 μM) in rainwater were typical in winter and/or early spring (Figure 2). Occasionally, a single rain event in summer resulted in an elevated concentration (250–300 μM) of NO3− and NH4+; this most often happened after a considerable dry period marked by limited rainfall. At Qianliyan, the concentrations of inorganic nitrogen species decreased with increasing rainfall and remained at <30–40 μM when individual rainfall events were >40 mm.
 Although data on the DON for rainwater are rather limited for the Yellow Sea, measurements at Qianliyan in 2004–2005 showed a DON of 3.5–268 μM and a DON to TDN ratio of 0.05–0.5; on average ca. 25% of the TDN at Qianliyan consisted of organic forms.
 High concentrations of inorganic nitrogen, especially nitrate (>150 μM), were measured in winter and spring at the coastal urban areas west of the Yellow Sea (Qingdao), ammonia levels were high in summer (200–250 μM) (Figure 2). The seasonal variation at Qingdao was less than that at the remote island of Qianliyan, located further out in the Yellow Sea, presumably, at least in part, because the strong local emissions are imposed on a rainwater background that is regulated by weather conditions [Zhang et al., 1999]. The pH of the rainwater at Qingdao was 2.80–7.74, with a rain-volume-weighted average of 4.81, owing to the proximity of this station to the upwind emission sources. The concentration of DON in the rainwater of Qingdao was 5–215 μM, with a DON/TDN of 0.07–0.65. Thus, on average, 30% of the TDN in rainwater is accounted for by DON.
 In the East China Sea, at Shengsi Island, the rainwater pH was 3.45–7.04, with a rain-volume-weighted average of 4.37. Again, inorganic nitrogen species showed important seasonal and interannual variations, but their concentrations were statistically lower at Shengsi than from Qingdao and Qianliyan of Yellow Sea. High levels of nitrate and ammonia were almost exclusively detected in winter, due to the limited amount of rainfall. Indeed, the concentrations were as much as 1 order of magnitude higher in winter than in summer, indicating a reduction in the concentration with increasing rainfall. At Shengsi, the concentration of nitrite in rainwater was 0.1–22 μM, comparable to that at Qinaliyan in the Yellow Sea. The DON concentration in rainwater at Shengsi ranged from 2 to 200 μM. Unlike inorganic nitrogen species, DON in rainwater was high in summer or winter, or both and showed a trend toward an exponential decrease with higher amounts of rainfall (r2 = 0.30). The DON/TDN ratio of rainwater was 0.03–0.7 at Shengsi, for which the interevent variability was more important than the seasonal variations.
 The composition of the aerosol samples was indicative of their different sources, i.e., dust storms from East Asia, with a dominance of soil particles in spring, sea-salt particles from the Pacific Ocean in summer and autumn, and strong anthropogenic emissions including fly ash from upwind landmasses in winter. These findings were similar to those previously reported for trace elements [Liu et al., 2002].
 At Qianliyan, in the Yellow Sea, the aerosol levels ranged from 1 to 5 μg/m3 in summer to 250 μg/m3 in spring, when dust storms prevailed. Also in spring, the aerosol concentration of NO3− + NO2− at the Yellow Sea reached 150–200 nmol/m3 and even higher a few days after the outbreak of wind storms in central East Asia, while in winter the levels were ≥200 nmol/m3 following strong upwind emissions from northern China, winter values were 2 orders of magnitude higher than those recorded in summer (Figure 3). Ammonia levels in the aerosols at Qinaliyan varied over 2 orders of magnitude, between 2 and 240 nmol/m3. In addition to the high concentrations typical of spring and winter, high ammonia concentrations (100–200 nmol/m3) were occasionally measured in summer. Moreover, when expressed in units of aerosol mass (unit: nmol/g of aerosols), the NO3− + NO2− and NH4+ concentrations in air showed a trend toward reduction with increasing aerosol levels. This was especially the case when the amount of aerosols was >100–200 μg/m3 (Table S2).
 Coastal urban areas of the Yellow Sea (Qingdao) had much higher concentrations of nitrogen species: 47.0–973 nmol/m3 for NO3− + NO2− (mean: 267 nmol/m3) and 59.3–1115 nmol/m3 for NH4+ (mean: 339 nmol/m3), presumably owing to the proximity of station to anthropogenic emission sources from inland. At Qingdao, elevated concentrations of NO3− + NO2− were recorded in winter (≥400–600 nmol/m3) and at times were 10-fold higher than those in summer (∼50 nmol/m3), when the southerly monsoon prevailed. As was observed for NH4+ in aerosols in summer, concentrations as low as 50–100 nmol/m3 were occasionally measured, i.e., much lower than the ≥ 500 nmol/m3 typical for winter samples. In this region, the ratio of NO3− + NO2− to total water-soluble inorganic nitrogen (WSIN) was 0.13–0.69 (mean: 0.42), indicating that, on average, NH4+ was the major component. When this value was compared to those obtained from the remote station of the Yellow Sea (Qianliyan), the ratio of NO3− + NO2− to WSIN was 0.01–0.89, with a similar mean of 0.40.
 Farther southward, at Shengsi, in the East China Sea, aerosol NO3− + NO2− was lower and NH4+ higher than the corresponding measurements for the Yellow Sea stations. At Shengsi, aerosol concentrations were 1.8–235 nmol/m3 for NO3− + NO2− (mean: 31 nmol/m3) and 1.2–372 nmol/m3 for NH4+ (mean: 53 nmol/m3). As in the Yellow Sea, elevated NO3− + NO2− and NH4+ levels in aerosols usually occurred in winter, due to bursts in anthropogenic emissions in East Asia with subsequent transport seaward by northerly winds; however, occasionally, the concentration was also high in summer (Figure 3). When nitrogen species levels were normalized by the mass of aerosols (units of nmol/g of aerosols) the NH4+ level, but not the NO3− + NO2− concentration, tended to decrease with higher amounts of aerosols (Table S2). The ratio of NO3− + NO2− to WSIN at Shengsi was 0.04–0.90 (mean: 0.41).
4.1. Nitrogen Species and Anthropogenic Emissions
 The NO3− + NO2− concentration in rainwater, measured at the atmospheric boundary layer, was 44.6 ± 3.48 μM (Qianliyan) and 32.7 ± 3.69 μM (Qingdao) at the Yellow Sea, and 22.7 ± 1.63 μM (Shengsi) at the East China Sea. For NH4+, averages of 54.2 ± 4.96 μM (Qianliyan), 63.9 ± 11.8 μM (Qingdao), and 35.6 ± 2.37 μM (Shengsi), were determined. Chung et al.  and Park and Lee  measured the concentration of nitrogen species in rainwater samples from the east coast of the Yellow Sea (i.e., South Korea) and reported values of 19.5 ± 1.61 μM for NO3− + NO2− and 43.4 ± 6.20 μM for NH4+. These concentrations are generally comparable to those reported in this study, although the observation period was shorter in those two studies. Further eastward in the Pacific Ocean and off the Japan Archipelago, the contribution of nitrogen species to the rainwater composition was less, 12.6 ± 0.35 μM for NO3− + NO2− and 17.3 ± 0.36 μM for NH4+ [Fujita et al., 2003], due to the significant influence of the marine atmosphere and to the disposal/dilution of continentally derived materials.
Nakamura et al.  recently reported aerosol concentrations of 17–557 nmol/m3 (average: 164 ± 136 nmol/m3) for NH4+ and 26–336 nmol/m3 (average: 128 ± 86 nmol/m3) for NO3−, measured from open areas of the East China Sea. These values are well within the range of our aerosol composition measurements made at the two island-based stations (i.e., Qianliyan and Shengsi); however, since the samples of Nakamura et al.  were collected from the open East China Sea and its eastern boundary, their values are rather high compared to those obtained in our study (cf. Table 2 and Figure S1). Based on the extended observation period (1997–2005) and the diversity of the geographic settings of Yellow Sea and East China Sea as compared to other works from East Asia, it can be assumed that our data represent a reasonably wide coverage of concentrations, as shown in Table 2 and Figure S1. It has been suggested that “pure marine” and “continentally affected” aerosols differ by a factor of 4–5 for inorganic nitrogen, i.e., an average of 42.9 ± 10 nmol/m3 for NH4+ and 48.6 ± 28.6 nmol/m3 for NO3− for pure marine aerosols and 221 ± 129 nmol/m3 for NH4+ and 165 ± 86 nmol/m3 for NO3− for continentally affected aerosols in this region [Nakamura et al., 2005]. A mean concentration of 31.1 ± 17.9 nmol/m3 was recorded for NO3− + NO2− and 46.4 ± 27.3 nmol/m3 for NH4+ at the Yellow Sea (Qianliyan) and 33.1 ± 16.1 nmol/m3 for NO3− + NO2− and 58.2 ± 36.3 nmol/m3 for NH4+ at the East China Sea (Shengsi). Thus, on an interannual scale, the mean aerosol compositions for inorganic nitrogen species in this study tended to be low, approaching the so-called pure marine aerosol composition, despite the influences of land source anthropogenic emissions and continental outflow events arising from the nature of the monsoon climate in the East Asian continental margins (e.g., Yellow Sea and East China Sea).
Table 2. Concentration of Nitrogen Species in Rainwater and Aerosols From the Yellow Sea and East China Seaa
NO3− + NO2−
NO3− + NO2−
Values in parentheses are the concentration range.
44.7 ± 72.1 (0.143–435)
53.5 ± 76.8 (0.340–1363)
30.5 ± 43.2 (0.720–583)
45.0 ± 45.6 (0.453–241)
23.1 ± 32.6 (0.143–191)
35.6 ± 53.6 (0.181–507)
22.1 ± 38.3 (1.81–235)
37.0 ± 82.7 (1.19–608)
32.9 ± 49.6 (0.143–242)
63.9 ± 76.2 (0.360–397)
244 ± 193 (47.0–973)
306 ± 195 (59.3–1120)
 As for the mass of aerosols in the atmospheric boundary layer, concentration ranges of 10–100 μg/m3 are typical for the Yellow Sea and East China Sea. As shown in Table 2 and Figure S1, the concentrations of nitrogen species from aerosols was lower in the East China Sea than in the Yellow Sea, presumably owing to the effect of the open continental shelf, whereas the Yellow Sea is surrounded by China on its northern and western sides and by Korea Peninsula on its eastern side. The values measured in this study are consistent with previously published data. For example, Lee et al.  reported aerosol nitrogen concentration of 143.8 nmol/m3 in the eastern part of Yellow Sea. A substantial increase in this level ensued following the outbreak of dust storms in East Asia [Carmichael et al., 1996]. Moreover, the concentrations of nitrogen species in the western East China Sea were in the range of 200–400 nmol/m3, comparable to the higher end of shipboard measurements of aerosols [Shimohara et al., 2001; Nakamura et al., 2006].
 Thus, it can be expected that wet and dry atmospheric depositions over the continental margins (i.e., Yellow Sea and East China Sea) are characterized by a strong concentration gradient for nitrogen species between East Asia and the North Pacific Ocean Interior. On the one hand, this gradient is explained by the elevated but highly varying levels coming from land sources, specifically, natural emissions (soil-dust storms) and anthropogenic perturbations, particularly in winter, when northerly winds dominate. On the other hand, there are predominating marine boundary layer processes (i.e., marine end-member composition) that reflect the monsoon climate of the continental margins at midlatitudes, especially in summer, when southerly winds prevail. This results in a dilution of the high concentrations coming from land sources.
 Data of ammonium and nitrate plus nitrite show a weak but positive correlation, but with different NH4+/(NO3− + NO2−) ratio between Qingdao, Qianliyan and Shengsi. Statistically, NH4+/(NO3− + NO2−) ratio at these three stations is highly dependent upon the trajectory of air mass. For instance, air mass across Japan, South Korea and/or Taiwan in summer can have also high NH4+ concentrations relative to NO3− + NO2− in rainwaters, which may not be true for dry deposition, while air mass across East Asia in winter and spring usually have higher NO3− + NO2− relative to NH4+ and hence a reduction in NH4+/(NO3− + NO2−) ratio. At regional scale, NH4+/(NO3− + NO2−) ratio can be 0.48–4.7 in aerosols and 0.5–6.9 in rainwater, with a trend that higher values occur at inland and coastal urban areas than remote islands in wet deposition while this may not be true for aerosol samples. It has been indicated that NH4+ can be from emission of chemical fertilizer application and livestock breeding, whereas high NO3− is mainly owing to fossil fuel combustion [Shi et al., 2010]; this is consistent with higher NH4+/(NO3− + NO2−) in spring–autumn than in winter at Qianliyan and Shengsi. Moreover, it has been observed that in summer and fall NH4+ is combined with fine aerosols, while NO3− + NO2− are concentrated in coarse mode; in winter both NH4+ and NO3− + NO2− are mainly in fine mode aerosols [Shi et al., 2010].
4.2. Monsoon Climate and the Chemical Compositions in the Marine Boundary Layer
 The nitrogen species composition in winter (December–February), when northerly winds dominate the mass transport in air of land source emissions, and during the summer monsoon (June–August), when southerly aspects prevail over the entire Yellow Sea and East China Sea, reveals distinct atmospheric depositions (Table 3 and Figure S2). With respect to wet deposition, in the Yellow Sea during the winter monsoon, NO3− + NO2− concentrations in rainwater can be as much as 50% to fourfold higher than NH4+ concentrations, whereas for the same period in the East China Sea, NH4+ is on average 50–100% as high as NO3− + NO2−. In the marine boundary layers of the Yellow Sea and East China Sea, NH4+ levels are higher than those of NO3− + NO2− when the summer monsoon is at its maximum, differing by 50–100% or twofold-to-threefold (Table 3). With the exception of Qingdao, where local emissions are important contributors to the chemical composition of wet depositions, the concentrations of nitrogen species in rainwater were higher during the northerly than during the southerly periods. Higher pollutant levels in the winter atmosphere have been attributed to strong anthropogenic emissions along the upwind side of East Asia (i.e., Northern China) and to the dominance of northerly winds, which carry materials from land sources to marine recipients. Moreover, the limited number of rain events in winter may induce the accumulation of pollutants in the atmosphere by increasing the residence time, in addition to a more efficient scavenging effect [Zhang et al., 1999].
Table 3. Comparison of Nitrogen Species in Wet (μmol/L) and Dry (nmol/m3) Depositions Between Northerly (December–February) and Southerly (June–August) Monsoon Periods in the Yellow Sea and East China Seaa
The data in parentheses are the concentration range.
NO3− + NO2−
NO3− + NO2−
 Regarding atmospheric dry deposition, NH4+ levels in aerosols at Qingdao and Shengsi in winter may be up to twofold higher than those of NO3− + NO2−, as northerly winds prevail over the East Asian marginal seas. NO3− + NO2− levels at Qianliyan may be 50–100% higher than those of NH4+, presumably because of the influence of land source emissions from the upwind side (Figure S2). In summer, when southerly winds prevail, aerosol samples show a trend toward higher NH4+ relative to NO3− + NO2−, with a reduction in the molar ratio between NH4+ and NO3− + NO2− from the open East China Sea to coastal urban areas. Specifically, NH4+/(NO3− + NO2−) in aerosols decreased from 2.79 at Shengsi, to 2.61 at Qianliyan, and further to 1.62 at Qingdao (Figure S2), which suggests a gradual dilution in the atmospheric loadings of nitrogen species from anthropogenic emissions in marine boundary layers when southerly winds blow from the open ocean to the East Asian continent. As mentioned before, the monsoon climate has an important impact on both atmospheric loading and the deposition of nitrogen species (Figure 4). Moreover, when nitrogen species are expressed in terms of aerosol mass, then the concentration of NH4+ at Qianliyan in winter is comparable to that of NO3− + NO2−, whereas at Shengsi it is 50% higher. A similar phenomenon was noted at Qingdao, indicating the involvement of different aerosol populations (Figure S2). In summer, when the southerly monsoon prevails, aerosol-mass-normalized concentrations of NH4+ can be 50–100% higher than those of NO3− + NO2−; this was true for the remote islands and for the coastal urban station. Overall, this induces a higher dry deposition flux of reductive (ammonium and organic nitrogen) relative to oxidized (nitrate) nitrogen at the Yellow Sea and East China Sea, with potential profound effects on competition among phytoplanktonic species that use different nitrogen forms in photosynthesis. Unlike the wet depositions, level of nitrogen species of dry deposition events can be comparable to or even higher during the southerly monsoon period than in winter, as shown for aerosol-mass-normalized ammonium (Figure 4).
4.3. Effect of Dust Storms on the Atmospheric Loading of East Asian Marginal Seas
 Spring dust events are not uncommon in East Asia and its adjacent midlatitude marginal seas. On average, 10–20 dust events are recorded between March and May each year, with occasional dust storms in June as well [Zhang et al., 1993; Liu et al., 2002].
 The Yellow Sea is below the major trajectory of East Asian dust storms affecting the interior of the Northwest Pacific Ocean. Nonetheless, at least for the period 1997–2005, aerosol concentrations during the spring nondust period of spring were 40 μg/m3 on average and can be a factor of fourfold-to-fivefold higher (100–200 μg/m3) when dust storms from the East Asian inland burst toward the NW Pacific Ocean (Figure 5 and Table S3). With respect to the nitrogen species in aerosols, the NO3− + NO2− concentration in atmospheric loads increase by twofold-to-fivefold during dust storms, while average atmospheric NH4+ concentrations are either comparable or even higher during nondust weather conditions than when dust storms pass over the Yellow Sea (Figure 5). This underlines the observation that the behaviors of oxidized (NO3−) and reductive (NH4+) nitrogen species in marine boundary layers can become decoupled when soil-dust storms passes over the NW Pacific Ocean and hence have different impacts on the biogeochemical cycles (e.g., a change in the NO3−/NH4+ ratio) of the surface ocean, due to the different sources of these species, or to the formation of secondary aerosols in atmosphere, or both. For example, data in Figure 5 and Table S3 show that atmospheric loading of nitrate to the surface ocean is proportional to the amount of aerosols in the marine boundary layer, i.e., an enrichment effect, while this may not necessarily be true for NH4+, as its level in aerosols can decrease relative to nitrate when dust storms prevail over the NW Pacific Ocean.
 Another important feature is that when atmospheric nitrogen species are normalized to the mass of aerosols, their concentrations tend to decrease with higher amounts of aerosols in the atmosphere. This is the case for NO3− + NO2− and for NH4+. Indeed, as shown in Figure 5, concentrations of nitrogen species, especially of NH4+, after normalization to the amount of aerosols can be twofold-to-fivefold higher in nondust than in dust storm events. This implies that changes in air mass trajectory between soil dust and non dust periods (dynamic process) as well as dilution effect by soil particles (chemical reaction) can have a dramatic influence on pollutants that have accumulated in the atmosphere, presumably through renewal of air mass and surface-area-related reactions (e.g., nucleation and formation of secondary particles). For instance, nitrate in aerosols can be produced by gas-to-particle conversion and/or gas-phase reactions with hydroxyl radicals; ammonium aerosol can be the product of aerosol- or droplet-phase reactions that involve nitric and sulfate acids in air [Fosco and Schmeling, 2006; Galindo et al., 2008]. When dust storms prevail in marine boundary layers, it can be expected that the dilution effect of pollutants by soil dusts will be proportional to the amount of particles in the atmosphere. The observed negative relationship between nitrogen species and aerosol mass provides evidence that aerosols are not saturated regarding their combination with nitrate and ammonia during the outbreaks of spring dust storms over the NW Pacific Ocean, after being remobilized from desert areas including the Gobi Desert in East Asia.
4.4. On the Deposition Fluxes and Application to Northwest Pacific Ocean
 The extensive data sets of this study allow an update of nitrogen flux through atmospheric wet and dry depositions in the Yellow Sea and East China Sea, in addition to comparisons to other regions of the Northwest Pacific Ocean (Figure 6). Our data show that, on an annual basis, there is a strong gradient of deposition fluxes from the landmass of East Asia to the North Pacific Ocean, i.e., a spatial dimension of 102–103 km. The dry deposition flux for nitrate and ammonium is 5–10 mmol m−2 yr−1 in the Yellow Sea proper and the open shelf of the East China Sea, but as high as 50–60 mmol m2 yr−1 for nitrate and 40–50 mmol m−2 yr−1 for ammonium in the atmospheric boundary layers of the eastern part of China at the coast (Figure 6). The dry deposition flux can even be higher in the Bohai, further to the north, than in the East China Sea, with a twofold difference. This is due to the fact that the Bohai is a semienclosed marine area surrounded by the North China mainland, while the East China Sea has an open shelf downwind, toward the Northwest Pacific Ocean (Figure 6). Moreover, the dry deposition flux for nitrogen species (i.e., nitrate, nitrite and ammonium) in the Yellow Sea and East China Sea is comparable to that of other coastal and marine provinces of the Northwest Pacific Ocean, for example, Japan/East Sea and offshore area from Japan at 20–30°N and 130–140°E (Figure 6).
 The wet deposition flux for nitrogen species is distinctive in this region, because, unlike dry deposition, it is regulated by both source materials and rainfall, with precipitation increases from 600 to 800 mm yr−1 in the coastal areas of North China to ca. 3000 mm yr−1 in the Northwest Pacific Ocean. Thus, the wet deposition flux for nitrate is 10–20 mmol m−2 yr−1 for the Yellow Sea and Japan/East Sea, and 20–30 mmol m−2 yr−1 for the East China Sea and other regions of the Northwest Pacific Ocean at 130–140°E (Figure 7). However, the wet deposition flux for ammonium shows a strong gradient, from 50 mmol m−2 yr−1 at the east coast of China to 10–20 mmol m−2 yr−1 for the Northwest Pacific Ocean at subtropical regions (130–140°E). The wet deposition fluxes of the Yellow Sea and Japan/East Sea are comparable for ammonium, i.e., 30 mmol m−2 yr−1 (Figure 7). Moreover, at comparable amounts of rainfall, the deposition fluxes of nitrogen species from coastal urban areas are always higher than those of marine provinces further offshore, indicating the footprint of anthropogenic emissions (Figure 7).
 It should be kept in mind, however, that NH3 and HNO3 as well as other gas species of nitrogen, which can form a substantial contribution to the dry deposition, were not measured in this study. Hence our calculation of fluxes may represent an underestimate of real dry deposition, and over estimate of wet to dry deposition ratios. In fact, concentrations and deposition fluxes for NH3 and HNO3 is highly variable in spatial and temporal dimensions. In the North Sea for example, the deposition flux of NH3 + HNO3 may account for 30–50% of total dry deposition [Jickells, 1995; De Leeuw et al., 2003]. If we assume a similar proportion of NH3 + HNO3 for the Yellow Sea and East China Sea, the total dry deposition flux can be increased by a factor of up to two with a reduction of wet to dry depositions by 1/2.
4.5. Contributions of Organic Nitrogen
 Another approach to understanding the effects on the marine biogeochemical cycle of anthropogenic emission from land sectors is to compare the composition of nitrogen species along the gradient, from coastal urban areas (Qingdao), to the Yellow Sea (Qianliyan), the East China Sea (Shengsi), and the remote interior of the Pacific Ocean (e.g., the Hawaiian Islands). Figure 8 shows the ratio DON to DIN in rainwater and its gradual reduction, from 0.4 to 0.5 in coastal urban areas of East China to 0.1–0.2 at Hawaii [Cornell et al., 2001], with the Yellow Sea and East China Sea proper in between. In terms of aerosols, the ratio of water-soluble organic nitrogen to inorganic nitrogen averaged 0.4–0.5 at Qingdao and declined to 0.2 at Hawaii, while samples from East China Sea showed different organic to inorganic nitrogen ratios (0.12–0.32) for aerosols [Nakamura et al., 2006]. This indicates that in coastal urban areas organic forms of nitrogen can account for 30% of total water-soluble nitrogen, presumably owing to the proximity of these areas to the emission sources. By contrast, in the remote stations and/or interior of the ocean, the proportion of organic forms in the total nitrogen pool dropped to 15% in the marine boundary layers.
 Organic nitrogen in atmospheric wet depositions was increased with higher DIN, but different correlations at Qingdao, Qianliyan, and Shengsi (Figure 9). Statistically, the linear regression slopes between DON and DIN in rainwater were 0.4–0.5 in the Yellow Sea but 0.2–0.3 in the East China Sea. As mentioned earlier, some of the organic nitrogen forms can be derived from anthropogenic emissions. The data in Figure 9 imply that the contribution of organic nitrogen to the total wet deposition fluxes can be more important in the marine boundary layer dynamics of the Yellow Sea than that of the East China Sea. This is in agreement with the fact that anthropogenic perturbations (e.g., change in composition of nitrogen deposition flux) have a more profound impact on the Yellow Sea than on the East China Sea.
5. Summary and Concluding Remarks
 In this study, rainwater and aerosol samples were collected from coastal urban and remote island stations and along the trajectory of various cruises in the Yellow Sea and East China Sea in 1997–2005. Analyses of nitrogen species, including nitrate, nitrite, ammonium, and organic forms, were carried out to examine their temporal and spatial variations in the marginal seas located on the downwind side of East Asia that are under the influence of westerlies at midlatitudes.
 As shown in this study, the concentration of nitrogen in atmospheric wet and dry depositions is higher in winter than in summer along temporal scales; samples from the Yellow Sea usually had higher concentrations of nitrogen species than those from the East China Sea. This spatial variability reflects differences in the source materials and the weather conditions. Moreover, concentrations of nitrogen species can be quite different between consecutive sampling events, indicating the episodic nature of rainwater and aerosol compositions in this region.
 Under the monsoon climate, samples obtained in winter, when northerly winds prevail over the Yellow Sea and East China Sea, had, on average, higher concentrations of nitrogen species, presumably due to strong anthropogenic emissions and air mass boost trajectories over the East Asian continent, whereas concentrations were lower in summer, during the southerly monsoon period.
 In spring, soil-dust storms initiated in East Asia can considerably increase aerosol loading of the marine atmosphere of the Yellow Sea and, to a lesser extent, the East China Sea. However, concentrations of nitrogen species were found to be diluted in marine boundary layers when the absolute concentration was normalized to the aerosol mass.
 Consequently, for dry deposition fluxes, strong (difference by a factor of up to 5) gradients for nitrogen species are established via atmospheric pathways, from the coastal urban areas of East Asia to the remote and interior regions of the open Northwest Pacific Ocean. These gradients are related to differences in emission sources and the transformation of nitrogen species. The wet deposition of nitrogen demonstrates a pattern of flux consistent with regulation by emission sources and rainfall, and with a species-specific character.
 Compared to other marginal seas of the Northwest Pacific Ocean, the concentrations of nitrogen species in marine boundary layers are higher in the Yellow Sea and East China Sea. Moreover, the molar ratio between atmospheric oxidized (nitrate) and reductive (ammonium) nitrogen in the study areas was twofold lower in these regions than in marine recipients downwind in Europe and North America, probably a result of extensive agriculture, with strong emissions of NH4+, and other human activities in the source region (i.e., atmospheric pollutants) of the East Asia, where ammonium emission can be more significant than that of nitrate.
 The present study was funded by the Ministry of Science and Technology of China, through contract 2006CB400601, and by the Natural Science Foundation of China (40721004). We express our sincere gratitude to the captain and crew of R/V “Beidou” and to our colleagues at the meteorological stations of Qingdao, Qianliyan, and Shengsi for their kind support in sample collections. Two anonymous reviewers and the Editor are acknowledged for their suggestions and comments, which helped to improve the original manuscript.