3.1.1 Anhydrosugars

Levoglucosan, a thermal degradation production of cellulose (Simoneit, 2002) and a specific indicator for biomass burning (Simoneit et al., 1999), can largely modify the chemical composition of atmospheric aerosols on a regional to global scale. Due to its water solubility, levoglucosan contributes to water-soluble organic carbon in aerosols, significantly enhancing the hygroscopic properties of atmospheric aerosols (Mochida and Kawamura, 2004). The abundance of levoglucosan detected in the present study (0.09–41.2 ng m⁻³, mean 6.1 for day and 0.10–64.3 ng m⁻³, mean 8.4 ng m⁻³ for night) was not the highest among measured sugar species (Fig. 3 and Table 1). The average concentration of levoglucosan is close to that reported at Gosan, Jeju Island, in summer (mean 8.0 ng m⁻³) (Fu et al., 2012). During the whole sampling period, higher concentrations of levoglucosan were found in the samples under the effect of continental air masses based on 5-day backward trajectories (Figs. S2–S6), consistent with our previous report about lipids, PAHs, and phthalates (Kang et al., 2017). Mannosan and galactosan, isomers of levoglucosan, are produced by pyrolysis of cellulose and hemicellulose and can also act as biomass burning tracers (Fabbri et al., 2009; Simoneit, 2002). Mannosan and galactosan were detected in the samples with similar variation trends to levoglucosan (Table 1 and Fig. 2), further indicating the strong impact of continental biomass burning activities on the marine atmosphere. As shown in Table 1 and Fig. 3, the nighttime average concentrations of all three anhydrosugars were higher than the daytime ones, probably attributable to enhanced biomass combustion and the lower height of the PBL as well as the land breeze during the night.

In addition, hardwood tends to contain higher levels of cellulose than hemicellulose; therefore the mass concentration ratios of levoglucosan to mannosan (L ∕ M) can be utilized as a diagnostic parameter for diverse biomass burning substrates (Zhu et al., 2015). Previous studies found that L ∕ M ratios from softwood are in a range of 3–10 and those from hardwood are 15–25, while the ratios from crop residues are even higher (25–40) (Zhu et al., 2015). In this study, the L ∕ M ratios range from 0.71 to 32 with an average of 4.2, and higher values were observed in the terrestrially influenced aerosols, especially in the coastal areas near Fujian Province (Fig. S7). The lower L ∕ M ratios in our study (average 4.2) suggest that the aerosols were mainly associated with the burning of softwood, consistent with the lower L ∕ M ratios (2.1–4.8) observed on Okinawa Island in May–June (Zhu et al., 2015) and Mt Fuji (4.6–7.6, mean 5.5) (Fu et al., 2014). It is worth noting that a couple of samples near Fujian Province in Southeast China were characterized by a higher L ∕ M ratio (> 20) (Fig. S7), agreeing well with the values of straw burning smokes (Chen et al., 2013). Zhu et al. (2015) also reported combustion of agricultural residues, peat, and wood could contribute to high L ∕ M ratios. Therefore, these higher values probably indicate emissions from terrestrial burning of straw residues, which could affect the chemical components of aerosols in the western North Pacific via long-range atmospheric transport as confirmed by backward trajectories (Fig. S5).

3.1.2 Sugar alcohols

Sugar alcohols detected in these samples consist of arabitol, mannitol, inositol, and erythritol, which had similar temporal patterns (Fig. 2e–h). Higher abundances of sugar alcohols observed in the terrestrially influenced aerosol samples suggest a significant contribution from terrestrial sources (Figs. S2–S6), while the marine fungi and algae, which can also release fungal spores into the marine atmosphere via bubble bursting, contribute little to the sugar alcohols in the coastal aerosols. Obviously, mannitol and arabitol were the most abundant sugar alcohols detected in the present study (Fig. 3 and Table 1), ranging from 0.03 to 169 ng m⁻³ (16.3 ng m⁻³) and 0.02–51.0 ng m⁻³ (5.1 ng m⁻³), respectively. A strong positive correlation between arabitol and mannitol was found in marine aerosols (p < 0.001, r=0.996, N=51), suggesting a similar origin. Mannitol and arabitol are very common in fungi, and are the most frequently occurring sugar alcohols in plants (Burshtein et al., 2011), which can be utilized to assess the contribution of fungal spores to the aerosol OC (Bauer et al., 2008). For instance, mannitol is particularly abundant in algae (Burshtein et al., 2011). The peak concentrations of sugar alcohols were found in the nighttime samples, which can be attributed to the increased activities of yeasts and fungi at night (Graham et al., 2003). Moreover, the maxima concentrations of these sugar polyols, especially arabitol and mannitol, were observed in offshore regions in early summer (June) (Fig. 2e–h), likely due to the more active microbial activities resulting from warmer temperature in June and more biota in coastal regions (Fig. 1).

3.1.3 Sugars

Glucose, fructose, sucrose, and trehalose are the primary saccharides measured in marine aerosols over the ECS. Glucose was the dominant sugar compound (0.31–209, 18.8 ng m⁻³), followed by sucrose (0.09–216, 11.7 ng m⁻³) (Figs. 2i–l and 3). Glucose and fructose originate from plant materials, such as pollen, fruits, and their fragments (Fu et al., 2012; Graham et al., 2003). Both glucose and fructose presented higher levels in the coastal areas, demonstrating a great contribution from terrestrial vegetation. Medeiros and Simoneit (2007) reported that a high abundance of glucose associated with lower molecular weight fatty acids (mainly C₁₆) was attributed to the spring bloom of algae. In our study, glucose correlated well with measured C_16:0 fatty acid (p=0.001, r=0.46, N=51), which were mainly emitted from the ocean surface via sea spray (Kang et al., 2017), suggesting that marine sources also contributed to the particulate glucose in the oceanic atmosphere. It was noted that fructose had a strong correlation with glucose (p < 0.001, r=0.94, N=51), indicating they share similar sources. Sucrose, as a predominant sugar species in the phloem of plants as well as developing flower buds, is reported to be the richest and dominant component of airborne pollen grains (Graham et al., 2003). The high abundance of sucrose along with fructose and glucose observed in our study implies a large emission of airborne pollen grains into the marine atmosphere over the ECS. Significantly, the concentrations of glucose, fructose, and sucrose were highest at the beginning of June in early summer (Fig. 2i–l), probably attributable to an enhanced pollen emission, because pollen counts tend to be highest in late spring/early summer in temperate zones (Graham et al., 2003). Such peak concentrations in early June are in accord with a previous study by Pashynska et al. (2002). Trehalose, a fungal metabolite, is present in a variety of microorganisms (fungi, bacteria, yeast, and algae) and a few higher plants as well as invertebrates (Medeiros et al., 2006a). Thus, trehalose can be used as a microbial biomarker and stress protectant. Furthermore, trehalose was reported to be the most abundant sugar in soil (Rogge et al., 2007; Medeiros et al., 2006b; Jia and Fraser, 2011); thus the enrichment of trehalose in aerosols can be indicative of soil resuspension and unpaved road dust (Simoneit et al., 2004a; Fu et al., 2012). The trehalose in our study was positively correlated with non-sea-salt calcium ($\mathrm{nss}-{\mathrm{Ca}}^{\mathrm{2}+}$) (p < 0.01, r=0.43, N=51), the best tracer for soil dust (Virkkula et al., 2006), suggesting that the atmospheric trehalose over the ECS was mainly derived from the resuspension of soil particles. The higher abundance of trehalose in coastal regions (Fig. 2l) may be in connection with the outflow of Asian dust primarily occurring in winter/spring (Fu et al., 2012) and inland soil resuspension/dust aerosols.

3.2.1 Isoprene SOA tracers

Isoprene is a reactive biogenic hydrocarbon and primarily originates from terrestrial photosynthetic vegetation (e.g. trees and plants). Marine phytoplankton and seaweed can also emit isoprene (Yokouchi et al., 1999; Shaw et al., 2010). Moreover, bacteria produce isoprene as well, and the bacterial isoprene production is temperature-dependent (Kurihara et al., 2010). In spite of much lower emission strength in the ocean region, more recent research has suggested that the oceanic sources of isoprene significantly impact atmospheric chemistry and cloud microphysical properties in the remote marine boundary layer because of its high reactivity (about 1–2 h lifetime) (Hackenberg et al., 2017).

Figure 4Temporal variations in biogenic SOA tracers detected in marine aerosols over the East China Sea during May to June 2014. The open and shaded circles represent daytime and nighttime values, respectively.

Download

Figure 5Average concentrations of SOA tracers detected in marine aerosols over the East China Sea.

Download

Six isoprene SOA tracers, including 2-methylglyceric acid, three C₅-alkene triols, and two 2-methyltetrols (2-methylthreitol and 2-methylerythritol), were identified in the marine aerosols over the ECS. Isoprene SOA tracers showed diurnal variations, with higher average concentrations at night (Figs. 4a–c, 5, S8), consistent with the report by Fu et al. (2010). The higher abundance during the night can be explained by the enhanced gas–particle partition at cooler temperatures during the nighttime and/or increased input of continent-originated isoprene SOA into the oceanic atmosphere via land–sea breeze circulations at night. However, a t test showed that the difference between daytime and nighttime concentrations for 2-methyltetrols, C₅-alkene triols, and 2-MGA was not that significant (p > 0.05). As expected, much higher concentrations of isoprene tracers were observed in coastal regions, where continental outflows exert larger effects from spring to early summer. For the remote ocean, terrestrial sources have weak impacts because of the short atmospheric lifetime of isoprene and the dilution effects during long-range atmospheric transport. In addition to the effect of continental outflow, more nutrients in the coastal and estuarine regions could be another factor responsible for the higher levels of isoprene SOA tracers compared to the pelagic areas. Because nutrient-rich surface water can promote the development of phytoplankton blooms, it increases chlorophyll-a concentrations. Chlorophyll-a is a measure of phytoplankton, or algal, biomass (Quinn et al., 2014) and currently most widely used proxy for predicting isoprene concentrations in water (Hackenberg et al., 2017). Numerous studies reported the positive relationship between isoprene emission and chlorophyll-a in the surface seawater (Hackenberg et al., 2017; Zhu et al., 2016). In the present study, the temporal and spatial distributions of chlorophyll-a at the ECS surface during the whole sampling period are characterized by higher coastal levels, such as waters near Zhejiang and Fujian provinces, but lower abundance in eastern Taiwan and the remote sea (Fig. 1). Thus, high chlorophyll-a waters in coastal locations mean more isoprene emissions than remote open waters. However, the isoprene in the remote ocean may mainly originate in situ from biological production by marine biota at the ocean surface.

The low-NO_x products 2-methyltetrols with mass concentrations ranging from 0.03 to 41.9 ng m⁻³ (4.8 ng m⁻³) were the major species among the isoprene SOA tracers, in line with an earlier report about summer aerosols in China (Ding et al., 2014). In particular, concentration ranges of 2-methyltetrols were 0.11–17.0 ng m⁻³ (3.8 ng m⁻³) during the daytime and 0.03–41.9 ng m⁻³ (5.7 ng m⁻³) at night, with 2-methylerythritol being about 2.1-fold more abundant than 2-methylthreitol. This ratio is similar to the ratios calculated in previous studies (Fu et al., 2010; Ion et al., 2005; Cahill et al., 2006). The atmospheric levels of 2-methyltetrols are comparable to those reported in marine aerosols collected during a round-the-world cruise (0.07–15, 2.4 ng m⁻³) (Fu et al., 2011). However, these values are small compared to those from terrestrial emissions, such as mountain (Fu et al., 2014; Cahill et al., 2006) and forest aerosols (Miyazaki et al., 2012; Fu et al., 2010; Claeys et al., 2004a). A significant positive relationship between 2-methyltetrols and levoglucosan (p < 0.001, r=0.87, N=51) in our study suggests that biomass burning may also generate the precursors of 2-methyltetrols followed by photochemical reactions (Xie et al., 2014). In addition, 2-methyltetrols correlated with C₂₉ n-alkane (the dominant species of terrestrial higher plant waxes) as well (p < 0.001, r=0.74, N=51), suggesting these organic tracers originate from higher plants or from similar source regions as well. The correlations above illustrate that terrestrial emissions (e.g. biomass burning and higher plants) contributed significantly to the levels of 2-methyltetrols in the marine atmosphere.

C₅-Alkene triols are formed from the photooxidation of isoprene under low-NO_x conditions (Surratt et al., 2006; Lin et al., 2013). They were detected in all the samples ranging from 0.03 to 14.6 (2.2 ng m⁻³), which are higher values than those reported in Arctic aerosols (0.01–0.15 ng m⁻³) (Fu et al., 2009) and other marine aerosols (0.002–4.6, 0.65 ng m⁻³) (Fu et al., 2011), but lower than the atmospheric levels of mountain aerosols (Fu et al., 2010, 2014) and subtropical urban aerosols from Hong Kong (Hu et al., 2008). Such a difference illustrates the outflow of continental aerosols, which can be confirmed by the significant positive correlations between C₅-alkene triols and tracers of terrestrial emissions, such as C₂₉ n-alkanes (p < 0.001, r=0.78, N=51) and levoglucosan (p < 0.001, r=0.87, N=51). It was found that 2-methyltetrols correlated well with C₅-alkene triols in marine aerosols over ECS (p < 0.001, r=0.86, N=51), suggesting a similar formation mechanism or common sources. However, the mass concentration ratios of C₅-alkene triols to 2-methyltetrols showed significant variation in different sampling sites (Fig. 6a), indicating different formation pathways, consistent with a previous report (Fu et al., 2010).

Concentrations of 2-methylglyceric acid (2-MGA), formed by photooxidation of isoprene under high-NO_x (NO_x = NO+NO₂) conditions (Surratt et al., 2006), were in the range of 0.09–8.3 ng m⁻³ (1.4 ng m⁻³), being greatly lower than those in mountain aerosols (Cahill et al., 2006; Fu et al., 2010, 2014), implying a much stronger influence of NO_x on isoprene SOA formation in continental aerosols. 2-MGA is a possible further oxidation product of methacrolein and methacrylic acid, which are two major gas-phase oxidation products of isoprene (Claeys et al., 2004b; Edney et al., 2005; Fu et al., 2009). 2-MGA was related to C₂₉ n-alkanes (p < 0.001, $r=\mathrm{0.84},$ N=51) and levoglucosan (p < 0.001, r=0.81, N=51) as well, again suggesting a terrestrial input. These isoprene SOA tracers in marine aerosols over the ECS may stem from terrestrial higher plants and biomass burning, and then were oxidized during the transport to the oceanic atmosphere. Similar correlations between isoprene-derived SOA tracers and levoglucosan were mentioned in a previous study, which reported that biomass burning enhanced the isoprene SOA formation (Li et al., 2018). Additionally, 2-methyltetrols are higher generation products than 2-MGA (Fu et al., 2014), but the ratios of 2-MGA to 2-methyltetrols did not vary significantly with cruise track. However, a couple of high ratios were observed in the ocean–air-influenced aerosols, suggesting the importance of marine sources for fresh isoprene-derived SOAs in the atmosphere (Fig. 6b).

${\mathrm{NO}}_{\mathrm{3}}^{-}$ was found to be related to 2-MGA, 2-methyltetrols, and C₅-alkene triols (p < 0.001, r=0.60–0.71, N=51). The ratio of 2-MGA to 2-methyltetrols was found to be negatively correlated with particulate ${\mathrm{NO}}_{\mathrm{3}}^{-}$ (p < 0.05, $r=-\mathrm{0.31}$, N=51). These relationships reveal that there may be a close connection between the formation of isoprene SOA and ${\mathrm{NO}}_{\mathrm{3}}^{-}$ in the marine boundary layer.

Figure 6Temporal variations in the concentration ratios of isoprene and α∕β-pinene oxidation products in the marine aerosols over the East China Sea. The open and shaded circles represent daytime and nighttime samples, respectively.

Download

3.2.2 Monoterpene SOA tracers

The detected α∕β-pinene oxidation products in the study consist of pinonic, pinic acids, 3-hydroxyglutaric acid (3-HGA), and 3-methyl-1,2,3-butanetricarboxylic acid (MBTCA). They are derived from the photooxidation of α∕β-pinene with O₃ and OH radicals (Hoffmann et al., 1997; Yu et al., 1999; Glasius et al., 2000; Iinuma et al., 2004). Since monoterpenes account for around 35 % of the global biogenic VOCs' emissions, these compounds have been utilized to estimate the role of monoterpene oxidation in the SOA formation (Griffin et al., 1999). Monoterpenes were chiefly emitted from needle leaf trees. In this study, monoterpene SOA tracers were found to be positively correlated to levoglucosan with p < 0.001 (r ranges from 0.68 to 0.82), indicating terrestrial biomass burning made substantial contributions to the formation of monoterpenes, then were transported to the oceanic atmosphere. Generally, total monoterpene SOA tracers in our study showed a major peak in spring (Fig. S1g), in agreement with a previous study (Zhu et al., 2016). Pinonic and pinic acids, well-known tracers for α/β-pinene, ranged from 0.02 to 1.6 ng m⁻³ (0.47 ng m⁻³) and from 0.16 to 14.9 ng m⁻³ (3.4 ng m⁻³), respectively (Table 1). Their concentrations were higher than those in high Arctic aerosols (average 69 and 514 pg m⁻³, respectively) (Fu et al., 2009). In addition, concentrations of pinic acid were 7 times higher than pinonic acid, similar to an earlier report (Fu et al., 2009). The vapour pressure of pinic acid is about 2 orders of magnitude lower than pinonic acid (Bhat and Fraser, 2007). Thus, pinic acid can saturate and readily nucleate, being expected to have a higher fraction in the aerosol phase than pinonic acid.

Two novel monoterpene SOA tracers, 3-hydroxyglutaric acid (3-HGA) and 3-methyl-1,2,3-butanetricarboxylic acid (MBTCA), were also detected in these marine aerosols. Both of them are reported to be higher generation products of α-pinene photooxidation (Kourtchev et al., 2009; Szmigielski et al., 2007). The concentration ranges of 3-HGA in marine aerosols over the ECS were lower than those of aerosols in Mt Tai, central east China (Fu et al., 2010), but comparable to other studies about marine aerosols (Fu et al., 2011). The average abundance of atmospheric MBTCA for all samples was 5.6 ng m⁻³, with mean concentrations of 6.8 and 4.5 ng m⁻³ during the day and night, respectively, more abundant than 3-HGA (2.2 ng m⁻³). Interestingly, these mean values of MBTCA were comparable to those of mountain aerosols (Fu et al., 2010), but still higher than the marine aerosols collected in the remote seas (Fu et al., 2011). Since the ECS is adjacent to mainland China, the atmospheric boundary layer in these regions would inevitably be affected by continent-derived air masses, as supported by the spatial pattern of individual SOA tracers (Figs. 4 and S1), in particular of the peaks occurring in the coastal areas. Furthermore, in contrast to isoprene SOA tracers, monoterpene SOA tracers commonly exhibit higher average daytime concentrations (Table 1 and Fig. 5), but a t test showed no significant difference (p > 0.05) between daytime and nighttime concentrations for both total and individual monoterpene SOA tracers. The concentration ratio of 3-HGA plus MBTCA to pinic acid ((3HGA+MBTCA)/pinic) showed higher values in the terrestrially influenced aerosols (Fig. 6c). In addition, these mass concentration ratios were mostly higher during the daytime (mean 2.0) than nighttime (1.9), indicating that aged aerosols are more abundant in the daytime.

3.2.3 Sesquiterpene SOA tracers

The analysis of sesquiterpenes is a great challenge due to their high reactivity and relatively low vapour pressure. However, the aging of β-caryophyllene-derived SOA has proved its significant effects on all CCN-relevant properties (Asa-Awuku et al., 2009). β-Caryophyllene is one of the most abundant species among sesquiterpenes originating from plants (Duhl et al., 2008). β-Caryophyllinic acid is an ozonolysis or photooxidation product of β-caryophyllene (Jaoui et al., 2007). Moreover, sesquiterpenes accumulated in leaves and woods can be emitted during biomass combustion (Ciccioli et al., 2014).

The atmospheric levels of β-caryophyllinic acid were 0.16–17.2 ng m⁻³ (mean 2.9 ng m⁻³), with 0.17–17.2 ng m⁻³ (3.5 ng m⁻³) during the daytime and 0.16–9.6 ng m⁻³ (2.3 ng m⁻³) during the nighttime, respectively (Table 1). As expected, β-caryophyllinic acid correlated well with levoglucosan (p < 0.001, r=0.61, N=51) in agreement with an earlier report (Ding et al., 2016), indicating the substantial contribution of terrestrial biomass burning to sesquiterpenes' loading in the marine atmosphere. On the other hand, biomass burning processes (e.g. crop straw combustion and forest fires) can raise ambient and/or leaf temperature to some extent, consequently probably enhancing the emission of sesquiterpenes from trees and plants.

Many studies have proved that sesquiterpene emissions increase significantly with increasing ambient temperatures (Tarvainen et al., 2005; Jaoui et al., 2007). Ambient temperature seems to be the dominant factor controlling temporal variation in sesquiterpene emission, although other factors contribute too (Duhl et al., 2008). An earlier study reported that emission rates of sesquiterpene were 1.2–3 times higher in the daytime as well (Duhl et al., 2008). All the reports described above perfectly interpret much higher concentrations of β-caryophyllinic acid during the daytime in our study. The abundance of β-caryophyllinic acid in this study is higher than that reported in the remote marine aerosols during a round-the-world cruise, but is comparable to the maximum concentration of 2.5 ng m⁻³ observed on the Californian coast (Fu et al., 2011); however, the concentrations of β-caryophyllinic acid were much lower than those of Mt Tai aerosols, central east China (average 12 ng m⁻³ for both daytime and nighttime aerosols) (Fu et al., 2010). The spatial distributions of β-caryophyllinic acid also presented higher levels in coastal regions than other sampling sites (Fig. 4h), again proving the contribution of terrestrial aerosols.

Table 2Concentrations of organic carbon (OC) (ngC m⁻³) from biogenic primary emission (biomass burning OC and fungal spore OC) and biogenic SOC and their contributions to aerosol OC (%) in marine aerosols over the East China Sea.

^* The total mass concentrations of SOC produced by isoprene (2-methylglyceric acid and 2-methyltetrols were used), α/β-pinene, and β-caryophyllene were estimated using the tracer-based method by Kleindienst et al. (2007).

Download Print Version | Download XLSX

3.4.1 Analysis of source profiles

After testing runs with a different number of factors (5–9), eight factors were chosen on the basis of the minimum value of Q (goodness of fit parameters) and a probable source profile expected from the study region.

Figure 10Source profiles identified by PMF. Blue bars represent the mass contribution with the y axis on the left, while black dots stand for the percentage of species to the sum shown by the y axis on the right.

Download

Figure 10 illustrates that the first source had high loadings of Na⁺ (90.8 %), suggesting a contribution from sea salt. This profile also contained a significant amount of ${\mathrm{SO}}_{\mathrm{4}}^{\mathrm{2}-}$, which can react with sea salt particles and release HCl gas, leaving a lower Cl⁻ ∕ Na⁺ ratio (0.1) than that of seawater (1.8) (Boreddy et al., 2014).

The second source shows high loadings of levoglucosan (66.3 %), arabitol (55.0 %), mannitol (51.1 %), and trehalose (50.4 %), representing mixed sources of biomass burning and fungal spores (Simoneit et al., 1999; Bauer et al., 2008; Medeiros et al., 2006a). Yang et al. (2012) found an enhanced abundance of fungal tracers on account of biomass burning activities, during which large numbers of fungi could be dispersed into the surrounding atmosphere or be carried upward with the warm plume to other fields via long-distance atmospheric transport.

The third source can be interpreted as crustal dust because of high loadings of Mg²⁺ (66.1 %) and Ca²⁺ (64.2 %), characteristic elements of soil/crustal dust (Xu et al., 2016). This source may include airborne road dust, construction dust, and windblown soil particles, which are derived primarily from terrestrial sources.

Figure 11Contributions of different sources to organic carbon (OC) in marine aerosols over the East China Sea.

Download

The fourth factor is characterized by the dominance of ${\mathrm{NH}}_{\mathrm{4}}^{+}$ (58.4 %) and ${\mathrm{SO}}_{\mathrm{4}}^{\mathrm{2}-}$ (49.7 %), which can be classified as secondary ammonium sulfate. The molar ratio of ${\mathrm{NH}}_{\mathrm{4}}^{+}$ and ${\mathrm{SO}}_{\mathrm{4}}^{\mathrm{2}-}$ was 2.9 in this profile, suggesting that (NH₄)₂SO₄ was the dominant sulfate form in the marine atmosphere over the ECS. Due to its long lifetime in the atmosphere, terrestrial ${\mathrm{SO}}_{\mathrm{4}}^{\mathrm{2}-}$ could be transported over long distances to coastal areas and even to the remote sea (Itahashi et al., 2017), affecting the chemical composition of oceanic atmosphere.

The fifth factor exhibits high loadings of di-isobutyl (DiBP) and di-n-butyl (DnBP), dominant species of phthalates in the marine aerosols over the ECS, which is assigned to plastic materials' emission likely from industry, agriculture, and domestic application in coastal regions.

The sixth factor has high loadings of sucrose (85.3 %) followed by fructose (50.9 %), implying a significant emission from airborne pollen grains to the marine atmosphere over the ECS during late spring to early summer (Fu et al., 2012).

The seventh profile presents high loadings of Cl, likely to be associated with coal combustion, which provides a significant releases of chlorine (McCulloch et al., 1999; Sun et al., 2013). Another confirmation is the high concentration of ${\mathrm{SO}}_{\mathrm{4}}^{\mathrm{2}-}$ from this profile, since coal consumption can produce mass sulfate as well.

The eighth factor illustrates high loadings of ${\mathrm{NO}}_{\mathrm{3}}^{-}$ (67.5 %) and biogenic SOA tracers, especially the monoterpene SOA tracers, i.e. PA (67.5 %) and MBTCA (65.2 %). This factor could be attributed to the photochemical oxidation products stemmed from the emission of vehicle and biogenic VOCs. Formation of secondary nitrate depends on NO_x, which is mainly produced from power plants and mobile sources (Heo et al., 2009; Kim et al., 2006). Anthropogenic NO_x could also enhance biogenic SOA formation via nitrate radical oxidation of monoterpenes (Xu et al., 2015). The positive correlations between ${\mathrm{NO}}_{\mathrm{3}}^{-}$ and isoprene-derived SOA tracers as discussed before in our study also suggest ${\mathrm{NO}}_{\mathrm{3}}^{-}$ and BSOA may share common formation and/or transport pathways. Previous studies have reported that nitrogen-containing species play a pivotal part in the formation and fate of SOA through varying radical and oxidant regimes and particle properties, such as volatility and hygroscopicity (Chen et al., 2017), agreeing well with our results.

Overall, the eight sources based on PMF were sea salt, biomass burning and fungal spores, crustal dust, secondary sulfate ammonia, plastic emission, pollen grains, coal combustion, and secondary nitrate and BSOA, which contributed to the TSP over the ECS by 16.9 %, 3.5 %, 8.0 %, 28.5 %, 6.3 %, 2.0 %, 14.3 %, and 20.4 %, respectively. The results of PMF show that secondary origin and marine natural emissions could be the main sources for the aerosols over the ECS. Figure 11 shows the contributions of different sources to OC in marine aerosols. On the whole, secondary nitrate and BSOA (25.5 %) and biomass burning and fungal spores (19.5 %) contributed more to OC than other sources during the whole sampling periods, elucidating the significant influence of biogenic primary and secondary sources on marine organic aerosols.

Figure 12Temporal variation in sources which contributed to marine aerosols (µg m⁻³).

Download

3.4.2 Temporal and spatial variation in sources

Figure 12 shows the temporal and spatial variation in each profile during the day and the night. The contribution of each source changes over time and varies with distance from the continent. In general, higher levels contributed by biomass burning, fungal spores, crustal dust, pollen grains, coal combustion, secondary nitrate, and BSOA were basically observed in coastal aerosols and/or terrestrially influenced aerosols, suggesting the strong influence of continental air from East Asia and Southeast Asia in light of the back trajectories and wind directions during the sampling periods (Figs. S2–S6 and S10). However, the contributions from sea salt particles, indicative of oceanic emission (organic components can be emitted from the ocean surface together with sea salt particles via sea spray or bubble bursting), tend to be higher in the aerosols mainly affected by marine air masses. Our study demonstrates that primary and secondary OM of terrestrial origin plays an important role in the marine aerosol chemistry over the western North Pacific through long-range atmospheric transport in addition to natural emission of the ocean.