The Characteristics of Water-Soluble Inorganic Ions in PM_1.0 and Their Impact on Visibility at a Typical Coastal Airport

Abstract

Water-soluble inorganic ions (WSIIs) can increase the hygroscopicity of aerosols, which will transform aerosols into larger sizes and reduce visibility by enhancing light scattering. To explore the characteristics of WSII concentrations and their impacts on visibility in a coastal airport, in this study, PM_1.0 samples at two monitoring sites (including airport site and background site) were collect in spring and summer, and 12 species of ions were detected. In general, secondary water-soluble inorganic ions (SNA, including ${\mathrm{SO}}_4^{2-}$, ${\mathrm{NO}}_3^{-}$ and ${\mathrm{NH}}_4^{+}$) and ${\mathrm{Ca}}^{2+}$ were the dominant WSIIs in PM_1.0, contributing about 89% to 95% of the total measured ions. The continental contributions of ${\mathrm{SO}}_4^{2-}$, ${\mathrm{K}}^{+}$, and ${\mathrm{Ca}}^{2+}$ accounted for more than 60% during the whole period, while Na⁺ and ${\mathrm{Cl}}^{-}$ were mainly from marine sources. The source identification showed that airport emissions were a major source at the sampling site and significantly contributed to the levels of sulfate, nitrate, and ammonium. Agricultural activities were the dominant sources impacting visibility in spring, while airport emissions and secondary inorganic aerosols were the main components affecting visibility in summer. Therefore, improving atmospheric visibility in coastal airport areas should focus on reducing the precursors of secondary particulates and reducing biomass-burning activities.

1. Introduction

Airport pollution has received increasing attention in recent years because of the rapid growth of air transport and the expected expansion of capacity to meet these needs [1]. The particulate matter emitted from aircraft exhaust is of small particle sizes and consists mainly of volatile sulfates and nitrates [2,3]. Zhu et al. indicated that the size distributions of ultrafine particles collected at the Los Angeles International Airport showed significantly high concentrations, with a peak particle size of approximately 14 nm [4]. In the airport, fine particles (PM_1.0, with an aerodynamic diameter of $\le$1 μm) is one of the crucial air pollutants [5]; these particles have a smaller size and larger surface area, thus causing greater harm to air quality [6]. Acharja et al. revealed that ${\mathrm{Cl}}^{-}$, ${\mathrm{NH}}_4^{+}$, ${\mathrm{NO}}_3^{-}$, and ${\mathrm{SO}}_4^{2-}$ were the dominant ions in PM_1.0 and PM_2.5 at the Indira Gandhi International Airport [5].

Aerosols can significantly decrease visibility by absorbing and scattering light. The extinction effect of an aerosol depends on its composition and structure. The light-scattering efficiency of aerosols is closely related to the particle size, and aerosols with a diameter at 0.5–2 μm have the highest efficiency of scattering visible light. Hygroscopic species such as water-soluble inorganic ions (WSIIs) can increase the water absorption of aerosols, which, in high humidity conditions, transforms aerosols into larger sizes and reduces visibility by enhancing light scattering [7]. The formation of hydrophilic aerosols and their hygroscopic growth are considered to be the main causes of haze formation in polluted areas. Higher concentrations of secondary water-soluble inorganic ions in fine particulate matter likely cause poor visibility at airports, thus affecting flight safety [2,8].

The atmospheric conditions of coastal regions differ greatly from those of inland regions due to the land-sea interface, temperature variations, and consequent local circulation development. In particular, sea salt is involved in atmospheric chemistry by interacting with atmospheric pollutants. For instance, sodium chloride in polluted areas (such as cities or areas with flight tracks or ship routes) reacts with sulfuric and nitric acid to form sodium sulfate, nitrate, and hydrochloric acid gas [9]. Tianjin Binhai International Airport (TSN) is an international airport bordering the Bohai Sea in China [10]. In addition to emissions from aircraft exhaust, the air pollution in TSN is subjected to multiple influences from sea salt, agricultural activities, secondary inorganic aerosols, and fugitive dust. Based on daily observation data from meteorological stations, the TSN area experienced several days of low visibility during the spring and summer of 2021, and a minimum visibility of less than 0.4 km was observed during this period. However, few studies have quantified the contribution of flight activities to WSII concentrations in PM_1.0 [11,12,13]. Likewise, there are still research gaps on the pollution characteristics of particulate ions in airport areas and their impact on visibility.

To explore the characteristics of WSII concentrations and their impacts on visibility in the coastal airport of our study, PM_1.0 samples at two monitoring sites were collected in spring and summer. Daily and seasonal variations in the WSIIs in PM_1.0 were analyzed. The marine and continental fractions of some chemical ions (${\mathrm{SO}}_4^{2-}$, ${\mathrm{Cl}}^{-}$, ${\mathrm{K}}^{+}$, ${\mathrm{Ca}}^{2+}$, and ${\mathrm{Mg}}^{2+}$) were estimated to separate the contributions of these ions in term of local continental sources and marine sources. Furthermore, the principal component analysis (PCA) was applied to identify the sources of WSIIs in the TSN area. The factors influencing visibility were analyzed, considering the WSII concentrations, meteorological factors, and flight activities. This study can provide valuable reference data for controlling air pollution and maintaining good visibility in airport areas for other large airports located in coastal cities worldwide.

2. Materials and Methods

2.1. Sampling and Monitoring

The sampling sites were situated on the rooftops of two school buildings on the Civil Aviation University of China campus in the southwest of the TSN (117°20′48″ E, 39°07′28″ N) (Figure 1). The airport site was located in a building about 600 m away from Runway 34 L of the TSN. This site is mainly impacted by aircraft emissions since it is a short distance from the airport. In addition, it may also be affected by road traffic as it is located near a major highway to the south of the airport (around 550 m). The background site was located approximately 1.70 km from the end of Runway 34 L. In addition to the influence of road traffic sources, background site may also be affected by the source of biomass burning or soil/construction dust due to the construction sites and fields. Other anthropogenic activities (e.g., coal burning) may also be conducive to particulate matter (PM) formation, since a university and residential areas with relatively high population density are located around these two monitoring sites. Meteorological data were obtained from the Meteorological Airport Database (METAR) of the China Meteorological Administration, including direction, wind speed, temperature, visibility, and atmospheric pressure.

PM_1.0 samples were collected from 8:00 to 20:00, in the spring period from the 29 April to the 16 May 2021, and in the summer period spanning from the 18 July to the 4 August 2021 (sampling was not conducted from 29 to 31 July due to heavy rain). Particles were collected on quartz filters (PALLFLEX, 90 mm) using medium-volume samplers (HY-100SFB) at a flow rate of 100 L·min⁻¹. Before sampling, the filters were baked at 600 °C in a muffle furnace for 4 h to remove organic artifacts or impurities. Before and after sampling, the filters were equilibrated under constant temperature (22 ± 1 °C) and relative humidity (35 ± 1%) for 72 h. The sampled filters were wrapped with annealed aluminum foil and stored in a refrigerator at −18 °C. The time from sampling to analysis were not exceed 2 weeks.

2.2. Methodology and Quality Control

For WSIIs analyses, one-eighth of each quartz filter was placed into a specific vessel with scale and then placed in an ultrasonic water bath (Type AS3120A, AutoScience Inc., Tianjin, China) using 8 mL distilled deionized water for 20 min. The extraction liquid was filtered using a cellulose syringe filter (diameter: 13 mm, pore size: 0.45 µm) and was subsequently injected into a Dionex DX-120 Ion Chromatograph with a flow rate of 1.0 mL/min, which consisted of a separation column, a guard column, a self-regenerating suppressed conductivity detector, and a gradient pump. Cations were detected using a weak acid eluent (18 mmol/L MSA) and anions were detected using a gradient weak base eluent (KOH: 5 mmol/L (0–8 min), 5–30 mmol/L (8.1–15 min), 30 mmol/L (15.1–25 min), 5 mmol/L (25.1–30 min)). The recovery of each ion was in the range of 80–120% [14]. The analytes included cations (${\mathrm{Na}}^{+}$, ${\mathrm{NH}}_4^{+}$, ${\mathrm{K}}^{+}$, ${\mathrm{Mg}}^{2+}$, ${\mathrm{Ca}}^{2+}$) and anions (${\mathrm{F}}^{-}$, ${\mathrm{Cl}}^{-}$, ${\mathrm{NO}}_2^{-}$, ${\mathrm{Br}}^{-}$, ${\mathrm{NO}}_3^{-}$, ${\mathrm{PO}}_4^{3-}$, and ${\mathrm{SO}}_4^{2-}$) [14].

The routine monitoring of background contamination was performed using an operational blank (unexposed filters stored in an aluminum foil bag before and after sampling until analysis), which was processed simultaneously with the field samples. Before and after the sampling, each filter was weighted 3 times, and the average value was used as the weight variation was less than 0.05 mg. During the sample analysis, all the glassware and filter assemblies were acid-washed and oven-dried to avoid contamination among the samples. Standard solutions were prepared and detected three times, showing that the relative standard deviation of the ions was less than 2%. The method detection limits (MDLs) were within the range of 0.01–0.04 μg/m³ and 0.03–0.07 μg/m³ for cations and anions. Quality control in the sampling period was shown in Text S1.

2.3. Data Processing

Multivariate statistical approaches, including Pearson’s correlation analysis, the stepwise regression model (SRM) and the principal component analysis (PCA), were applied in this study to explore the relationships and potential sources of WSIIs.

2.3.1. Calculation of Sea Salt (ss) and Non-Sea Salt (nss) Fractions

The methodology assumes that ${\mathrm{Na}}^{+}$ purely originates from the sea, and the concentration ratios of ionic species and Na⁺ for sea water are taken to estimate the sea salt (marine) fraction and non-sea salt (continental) fraction. The formulas used to estimate the ith ion’s concentrations in the jth sample for sea salt concentrations (${SS}_{ij}$) and non-sea salt concentrations ($N{SS}_{ij}$) t are given by Equations (1) and (2):

$${SS}_{ij}=C_{{\mathrm{N}\mathrm{a}}^{+},j}\times {Ratio}_{i/{\mathrm{Na}}^{+}}$$

(1)

$$N{SS}_{ij}=C_{i,j}-{SS}_{ij}$$

(2)

where, $C_{{\mathrm{Na}}^{+},j}$ is the concentration of ${\mathrm{Na}}^{+}$ in the jth sample; $C_{i,j}$ is the concentration of the ith ionic species in the sample; ${Ratio}_{i/{\mathrm{Na}}^{+}}$ is the ratio of ${\mathrm{SO}}_4^{2-}$, ${\mathrm{Cl}}^{-}$, ${\mathrm{K}}^{+}$, ${\mathrm{Ca}}^{2+}$ and ${\mathrm{Mg}}^{2+}$ to ${\mathrm{Na}}^{+}$ for the sea water, with values of 0.125, 1.16, 0.0218, 0.0439, and 0.227, respectively [15,16].

2.3.2. Assessment of Source Apportionment

To identify the sources of WSIIs in detail, we used the PCA model, a multivariate analysis tool for receptor modeling in environmental source apportionment studies. The PCA analysis could reduce the dimensionality of a dataset containing a large number of interrelated variables, while remaining as much variability as possible in the meantime [17]. The process involves converting the initial variables into uncorrelated (orthogonal) principal components, which are weighted linear combinations of the original variables [18].

In this study, the concentration of 12 species of WSIIs (airport site in spring and summer) were used as variables for PCA. Data were tested by Kaiser-Mayer-Olkin criteria and the Bartlett’s test of sphericity prior to the analysis. It is suitable for PCA analysis when the value of KMO is greater than 0.5 and the significance level of Bartlett’s test is less than 0.05 [19]. A Varimax orthogonal rotation was employed to calculate the rotated factor loadings so that loadings values were clustered from 0 to 1, making them more physically interpretable. Pollution sources were identified based on the representation of factor loadings [20].

2.3.3. Stepwise Regression Model

The stepwise regression model (SRM) is an effective statistical method that is mainly used to select the most influential predictor variables from a large number of potential predictor variables while eliminating those with the smallest predictive influence. This method greatly simplifies the construction of linear regression models through the automatic variable selection process, making it particularly suitable for dealing with complex and high-dimensional data sets [21].

The stepwise regression model is shown in Equation (3):

$$\gamma ={\beta }_0+{\beta }_1{X}_1+{\beta }_2{X}_2+{\beta }_3{X}_3+\dots +{\beta }_i{X}_i+\epsilon$$

(3)

where $\gamma$ is the dependent variable, ${\beta }_0$ is the intercept, ${\beta }_1$, ${\beta }_2$, ${\beta }_3$,…, ${\beta }_i$ are the coefficients of the explanatory variable, and $\epsilon$ is the error term.

3. Results and Discussion

3.1. Concentrations of PM_1.0 and Water-Soluble Inorganic Chemical Ions

The ambient concentrations of PM_1.0 and ions are shown in

Table 1. The average concentrations of PM_1.0 and WSIIs during the sampling period (μg/m³).

PM1.0 and Ions	Spring		Summer
	Airport Site	Background Site	Airport Site	Background Site
PM1.0	39.08 ± 16.27	31.79 ± 16.27	26.86 ± 23.85	29.94 ± 23.85
${\mathrm{Na}}^{+}$	0.19 ± 0.15	0.14 ± 0.03	0.20 ± 0.05	0.19 ± 0.03
${\mathrm{NH}}_4^{+}$	2.97 ± 2.30	3.07 ± 2.41	6.15 ± 3.08	5.68 ± 2.65
${\mathrm{K}}^{+}$	0.03 ± 0.03	0.02 ± 0.01	0.04 ± 0.02	0.05 ± 0.03
${\mathrm{Mg}}^{2+}$	0.18 ± 0.10	0.24 ± 0.08	0.12 ± 0.07	0.13 ± 0.06
${\mathrm{Ca}}^{2+}$	2.59 ± 1.42	1.49 ± 0.44	0.51 ± 0.20	1.24 ± 2.32
${\mathrm{F}}^{-}$	0.09 ± 0.12	0.14 ± 0.09	0.08 ± 0.03	0.09 ± 0.07
${\mathrm{Cl}}^{-}$	0.28 ± 0.47	0.14 ± 0.07	0.11 ± 0.08	0.11 ± 0.08
${\mathrm{NO}}_2^{-}$	0.05 ± 0.05	0.02 ± 0.01	0.04 ± 0.04	0.01 ± 0.01
${\mathrm{Br}}^{-}$	0.01 ± 0.01	0.01 ± 0.01	0.01 ± 0.01	0.01 ± 0.01
${\mathrm{NO}}_3^{-}$	2.56 ± 2.48	2.66 ± 2.90	1.56 ± 2.52	1.66 ± 2.74
${\mathrm{PO}}_4^{3-}$	0.13 ± 0.06	0.16 ± 0.07	0.13 ± 0.07	0.24 ± 0.27
${\mathrm{SO}}_4^{2-}$	2.04 ± 1.47	1.91 ± 1.38	5.87 ± 2.45	5.90 ± 2.50

. During spring and summer, the total mass concentrations of water-soluble ions in PM_1.0 were 11.13 and 14.82 μg/m³ at the airport site and 9.99 and 15.32 μg/m³ at the background site. The secondary ions, including ${\mathrm{SO}}_4^{2-}$, ${\mathrm{NO}}_3^{-}$, and ${\mathrm{NH}}_4^{+}$, were the main water-soluble ions and were separately converted from gas precursors SO₂, NO_X, and NH₃. Calcium ions were also present in a relatively high amount. During the campaigns, ${\mathrm{Ca}}^{2+}$ and SNA contributed ~89% to 95% of the total measured ions (Table 1).

SNA in PM_1.0 plays an important role in ambient air quality and visibility. Guo et al. (2014) [22] investigated the process of PM episodes in Beijing, finding that they were characterized by efficient nucleation and continuous particle growth, and were dominated by local secondary formation. They attributed the continuous growth in the particle size and constant accumulation of particle mass concentrations to the highly elevated concentrations of gaseous precursors such as NO_X, SO₂, and volatile organic compounds (VOCs), while the contributions from primary emissions and regional transport were negligible. Zhang et al. (2015b) [23] found that the concentrations of SO₂, NO_X and anthropogenic VOCs in Beijing and other cities in developing countries were significantly higher than those in urban areas of developed countries, resulting in a large amount of secondary generation of sulfate, nitrate, and secondary organic aerosol. Therefore, a high proportion of SNA in particulate matter may cause reduced visibility.

3.2. Seasonal Variation of WSIIs

PM_1.0 in the TSN area showed an average 17% higher mass trend of mass concentration in spring than in summer, which is similar to the result of most other studies performed in general urban environments [24,25]. The intense solar radiation in summer helps to destroy the stable boundary layer and favors pollution dilution. By contrast, higher PM_1.0 levels in spring could be attributed to mild conditions arising from the lower wind speed. In spring, the background concentrations in both seasons showed a higher trend than the airport concentrations, indicating that airport activities strongly impact on the PM_1.0 concentration level in the airport area. Frequent traffic and industrial activities in summer explain the high concentration in the background in summer.

Figure 2 shows daily variations in the daytime concentration of typical chemical ions in PM_1.0. The concentration of ${\mathrm{Ca}}^{2+}$ in spring was 2.59 ± 1.42 μg·m⁻³ (airport site) and 1.49 ± 0.44 μg·m⁻³ (background site) in PM_1.0. However, the ${\mathrm{Ca}}^{2+}$ concentration in summer was nearly six times lower than that in spring, with values of 0.51 ± 0.20 μg·m⁻³ (airport site) and 1.24 ± 2.32 μg·m⁻³ (background site) in PM_1.0. The springtime peak in ${\mathrm{Ca}}^{2+}$ is most reasonably attributed to high concentrations of fugitive dust, which is in accordance with several previous studies [26,27,28,29] Indeed, during springtime, there was a considerable increase in flight activities due to the Labor Day holiday. Considering the close relationship between the ${\mathrm{Ca}}^{2+}$ concentration and flight activities in TSN, as discussed in Section 3.5, the intensive flight activities during this spring period could also cause a higher ${\mathrm{Ca}}^{2+}$ concentration. Mazaheri et al. (2013) collected PM_1.0 samples on filters 200 m from an airfield runway for elemental analysis using energy-dispersive X-ray spectroscopy. Their results showed that particulate ${\mathrm{Ca}}^{2+}$ derived not only from engine exhaust but also from various sources such as tire wear, dust and automotive traffic [30]. Abegglen et al. noted that calcium was detected in both kerosene and lubricating oil in a relatively large concentrations. Both kerosene and lubricating oil might be responsible for the presence of calcium in exhaust particles [31]. This could also explain the higher spring ${\mathrm{Ca}}^{2+}$ concentration found at the airport site than the background site.

SNA, as the dominant ion species in PM_1.0, has been found in many other previous studies [32,33]. The concentrations of ${\mathrm{SO}}_4^{2-}$ and ${\mathrm{NH}}_4^{+}$ were 0.9~1.1 times higher in summer than in spring. Nevertheless, the concentration of ${\mathrm{NO}}_3^{-}$ exhibited the opposite seasonal trend, as the ${\mathrm{NO}}_3^{-}$ concentration decreased by 37.6%~39.1% in summer. The seasonal trends in these ions were mainly due to several reasons. In the atmosphere, ${\mathrm{NH}}_4^{+}$ was primarily generated from the reaction between NH₃ (g) and H₂SO₄ (g), leading to the concentration of ${\mathrm{NH}}_4^{+}$ presented the same seasonal trend as ${\mathrm{SO}}_4^{2-}$. Several studies have pointed out that ${\mathrm{SO}}_4^{2-}$ is primarily generated from SO₂ via heterogeneous or multiphase oxidation in the gaseous phase, which is influenced by OH radical concentrations and the presence of clouds. Thus, high RH in summer could facilitate an aqueous heterogeneous reaction, and secondary conversion in the atmosphere would be accelerated under this condition, leading to increased ${\mathrm{SO}}_4^{2-}$ production [34,35]. High temperatures in summer also accelerate the dissociation of NH₄NO₃ to ammonia and nitric acid that remain in the gas phase, inhibiting ${\mathrm{NO}}_3^{-}$ generation [5,33,36,37,38,39].

3.3. Influence of Meteorology on the Concentration of WSIIs

The weather conditions, including the temperature, relative humidity, wind speed, and atmospheric pressure during the sampling period, are shown in Tables S1 and S2, while the relationship between PM_1.0 and meteorology was shown in Text S2. For WSIIs in PM_1.0, correlation analysis showed that the relative humidity and wind speed had an obvious impact on the ion concentration in the TSN region, while no significant trend was found in the influence of temperature and ambient air pressure (

Table 2. Correlation matrix between WSII concentrations in PM_1.0 and meteorological conditions.

Meteorological Conditions	T	RH	WS	P0
airport Site	0.157	0.598 **	−0.478 **	0.003
background Site	0.308	0.640 **	−0.570 **	0.008

** Significant at p < 0.01.

).

Specifically, relative humidity showed a positive relationship (airport site: r² = 0.60; background site: r² = 0.64) with WSIIs. With an increase in relative humidity, the SNA concentrations varied consistently. A similar phenomenon in the relationship between relative humidity and ion concentrations was also found at Indira Gandhi International Airport in New Delhi, in Shanghai, and in the North China Plain regions of China [5,40,41]. High relative humidity is conducive to the formation of particles; this is due to the fact that high RH in a stable and low boundary layer thickness greatly enhances the oxidation of SO₂ and NO_x to form ${\mathrm{SO}}_4^{2-}$ and ${\mathrm{NO}}_3^{-}$ aerosols, respectively [5].

The wind speed significantly negatively correlated (airport site: r² = −0.478; background site: r² = −0.570) with WSIIs. This indicated that high wind speeds usually favored the dilution of pollutants and thus decreased WSII concentrations. In this study, lower wind speeds mainly occurred in the summer, with correspondingly high ion concentrations. This negative relationship between ion concentrations and wind speed was also found in other urban areas (e.g., Guangzhou, Hangzhou, and the Beijing–Tianjin–Hebei region) [10,41,42,43,44].

Wind direction was another essential component influencing ion concentrations [43]. Different predominant wind directions may alter the background pollutant loadings in the research area [42]. As shown in Figure 3, compared to conditions where the dominant wind originated from the north/northwest, the average concentrations of WSIIs during spring significantly increased (from 6.90 to 14.22 μg·m⁻³) under east/southeast wind conditions. East/southeast winds in spring could bring traffic-related or other sources-related air pollutants flowing through the monitoring sites.

3.4. Source Identification of Water-Soluble Inorganic Chemical Ions

The marine and continental fractions of the measured chemical ions in PM_1.0 were estimated in this research to identify the WSII sources. As illustrated in Figure 4, the continental contributions of ${\mathrm{SO}}_4^{2-}$, ${\mathrm{K}}^{+}$ and ${\mathrm{Ca}}^{2+}$ accounted for more than 60% during the whole period. In comparison, ${\mathrm{Cl}}^{-}$ was mainly from marine sources, while the marine fraction of ${\mathrm{Mg}}^{2+}$ varied in a comparatively large range from 27.68% to 82.99%. These ion results were similar to those in the Indira Gandhi International Airport [16].

To investigate the potential sources of WSIIs at the airport site, PCA was applied to this research. The varimax rotated factor loadings matrix of WSIIs is shown in

Table 3. Varimax rotation factor loadings on WSIIs concentrations (bold numbers are typical component with high loading values).

Ions	Spring				Summer
	Factor 1	Factor 2	Factor 3	Factor 4	Factor 1	Factor 2	Factor 3	Factor 4
Na+	0.072	0.966	0.026	0.059	0.749	0.515	0.325	0.011
${\mathrm{NH}}_4^{+}$	0.928	0.221	−0.224	0.124	0.940	0.113	0.127	0.243
K+	0.358	0.874	−0.255	0.007	0.869	−0.028	−0.285	−0.247
Mg2+	−0.150	−0.204	0.738	−0.013	−0.065	0.020	0.927	−0.132
Ca2+	−0.365	0.116	0.782	−0.159	0.497	−0.028	0.842	0.139
F−	−0.052	−0.072	0.086	−0.075	0.118	0.737	0.257	−0.192
Cl−	0.089	0.976	0.016	−0.101	0.368	0.805	0.057	0.009
${\mathrm{NO}}_2^{-}$	−0.012	−0.193	0.284	−0.824	0.124	−0.082	0.008	0.949
Br−	0.238	−0.228	0.114	0.824	−0.112	0.866	−0.229	0.030
${\mathrm{NO}}_3^{-}$	0.936	0.149	−0.165	0.082	0.711	0.488	−0.133	0.289
${\mathrm{PO}}_4^{3-}$	0.651	−0.292	0.164	0.019	−0.293	0.478	0.557	0.300
${\mathrm{SO}}_4^{2-}$	0.864	0.303	−0.278	0.148	0.800	−0.114	0.518	0.082

. High loading values are highlighted in bold font. In both spring and summer, four principal components (PCs) were identified, explaining 80.2% and 86.1% of the data variance, respectively.

In spring, the WSII sources at the airport site were identified as four factors. Factor 1 was secondary inorganic aerosols with high contributions of ${\mathrm{NO}}_3^{-}$, ${\mathrm{NH}}_4^{+}$, and ${\mathrm{SO}}_4^{2-}$ [45,46]. Secondary nitrates and sulfates are formed via the oxidation of NO_x and SO₂ through photochemical reactions. Oil, coal, and biomass combustions are the typical sources of NO_x and SO₂, which can interact with NH₃ to form NH₄NO₃ and (NH₄)₂SO₄ [47]. Factor 2 included high loadings of ${\mathrm{Cl}}^{-}$, ${\mathrm{Na}}^{+}$ and ${\mathrm{K}}^{+}$, and was identified as the mixed source of sea salt and biomass burning [48,49]. As illustrated by previous studies, ${\mathrm{K}}^{+}$ is usually used as the most important tracer for biomass burning in receptor models [39,47]. Factor 3 was calculated to contain high loadings of ${\mathrm{Mg}}^{2+}$ and ${\mathrm{Ca}}^{2+}$. Yin et al. [50] indicated that ${\mathrm{Mg}}^{2+}$ and ${\mathrm{Ca}}^{2+}$ can be regarded as airport emission tracer species; thus, this factor was recognized as airport emissions. Factor 4 was regarded as other sources. For ${\mathrm{NO}}_2^{-}$, As for ${\mathrm{NO}}_2^{-}$ its source is complicated. Studies have found that gaseous HONO can be quickly formed through the redox reaction of NO₂ on black carbon [51]. On the other hand, NO and OH radicals can react homogeneously to generate HONO, which then attaches to the surface of particles [52]. However, Br⁻ was the tracers from the seawater. Considering the complexity of atmospheric sources, factor 4 was identified as other sources.

Four factors were also recognized in summer. Factor 1 was identified as secondary inorganic aerosols due to the high loadings of ${\mathrm{NH}}_4^{+}$, ${\mathrm{NO}}_3^{-}$ and ${\mathrm{SO}}_4^{2-}$. Factor 2 included high loadings of ${\mathrm{Cl}}^{-}$, ${\mathrm{Na}}^{+}$ and ${\mathrm{Br}}^{-}$, and was identified as the source of sea salt. Factor 3, with high concentrations of ${\mathrm{Mg}}^{2+}$, ${\mathrm{Ca}}^{2+}$, ${\mathrm{PO}}_4^{3-}$ and ${\mathrm{SO}}_4^{2-}$, was characterized by airport emissions. Factor 4 was regarded as other sources.

Airport emissions were a major source of pollution for the sampling site and contributed significantly to sulfate, nitrate, and ammonium levels. Sea salt sources also made a significant contribution to WSIIs. During monsoon seasons, abundant precipitation and winds from the Bohai Sea can contribute to sodium and magnesium salt source loading in the sampling region [53]. In the spring and summer, secondary inorganic aerosol is also an important source of particulate matter in the airport area. Due to the stronger hygroscopicity of secondary organic aerosols, atmospheric visibility could be significantly reduced. Therefore, the management of secondary organic aerosols in airport requires more attention.

3.5. Relationship Between WSII Concentrations and Visibility

Visibility degradation is not only a visual problem but may also pose a safety hazard to human activities, especially for the civil aviation industry. As expected, a negative correlation was displayed between WSII concentrations and visibility at the airport site, with correlations of 0.67 (p < 0.05) in spring and 0.93 (p < 0.01) in summer. The meteorological parameters, individual ions, and visibility were analyzed further (

Table 4. Correlation matrix between individual WSII concentrations and meteorological parameters and visibility for airport site during the spring and summer.

Parameters	${\mathbf{Na}}^{+}$	${\mathbf{NH}}_4^{+}$	${\mathbf{K}}^{+}$	${\mathbf{Mg}}^{2+}$	${\mathbf{Ca}}^{2+}$	${\mathbf{Cl}}^{-}$	${\mathbf{NO}}_3^{-}$	${\mathbf{PO}}_4^{3-}$	${\mathbf{SO}}_4^{2-}$	T (℃)	RH (%)	WS (m/s)
Spring	−0.260	−0.662 *	−0.700 *	0.421	0.586	−0.296	−0.641 *	−0.487	−0.393	0.326	−0.722 *	0.275
Summer	−0.472	−0.909 **	−0.588 *	0.297	−0.082	−0.427	−0.917 **	0.190	−0.466	0.349	−0.618 *	0.431

** Significant at p < 0.01. * Significant at p < 0.05.

). The relative humidity and SNA contributed greatly to the degradation of visibility in the TSN area, and this result was similar to those for Shanghai [54], Fuzhou [55], Guangzhou [56], and Beijing [57]. Hygroscopic species such as sulfate and nitrate can enhance the water absorption of aerosols, which, in high-humidity conditions, transform into larger particles and reduce visibility by enhancing light scattering [54,55,56,57]. Additionally, a close relationship was found between ${\mathrm{K}}^{+}$ and atmospheric visibility, indicating that biomass burning accounted for a considerable decrease in daytime visibility, since ${\mathrm{K}}^{+}$ is consistently identified as a tracer for biomass burning [39,47]. The considerable role of biomass combustion in decreasing visibility has been confirmed in many previous studies, and biomass burning contributes to a deterioration in visibility due to its high extinction ratio of sulfate and organic carbon production [58,59]. Unlike in the source apportionment results, the contributions of ${\mathrm{Na}}^{+}$ and ${\mathrm{Cl}}^{-}$ ions to the visibility reduction are not obvious. However, the relatively high humidity in coastal cities is an adverse factor in reducing visibility. Thus, improving atmospheric visibility in the area should focus on reducing precursors that accelerate atmospheric oxidation and the production of secondary particulates, along with reducing biomass-burning activities.

A statistical approach using stepwise multiple linear regression (SMLR) was used to evaluate the correlation of visibility with WSII concentrations, in order to further understand the relationship between visibility and the major air pollutants. The SMLR formulas for spring and summer are expressed by Equations (4) and (5), respectively.

$$\mathrm{V}\mathrm{i}\mathrm{s}\mathrm{i}\mathrm{b}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}=-2.095\left[{\mathrm{K}}^{+}\right]+1.016{[\mathrm{Na}}^{+}]-0.570{[\mathrm{Mg}}^{2+}]+0.484\left[{\mathrm{Cl}}^{-}\right]-0.099\left[{\mathrm{NH}}_4^{+}\right]-5.760\times {10}^{-16}$$

(4)

$$\mathrm{V}\mathrm{i}\mathrm{s}\mathrm{i}\mathrm{b}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{t}\mathrm{y}=-0.804{[\mathrm{NO}}_3^{-}]-0.275{[\mathrm{Ca}}^{2+}]-0.200\left[{\mathrm{K}}^{+}\right]+0.155{[\mathrm{Mg}}^{2+}]-0.111{[\mathrm{PO}}_4^{3-}]+4.823\times {10}^{-16}$$

(5)

The regression coefficients estimated in the above equations can express the degree of pollutant contribution to reduced visibility. Based on the coefficients in these equations, it could be concluded that ${\mathrm{K}}^{+}$ played an important role in reducing visibility in spring, while ${\mathrm{NO}}_3^{-}$ was the main factor affecting visibility in summer. This result shows that agricultural activities (with ${\mathrm{K}}^{+}$ as their tracer) were the dominant source impacting visibility in spring, which is consistent with the source apportionment results. In summer, airport emissions and secondary inorganic aerosols were the main components affecting visibility. Thus, reducing the emissions from specific pollution sources in spring and summer could effectively elevate the visibility at TSN.

4. Conclusions

To explore the characteristics of WSII and their impacts on visibility at a coastal airport, PM_1.0 samples at two monitoring sites were collected in spring and summer, and 12 species of ions were detected via ion chromatography. The daily and seasonal variations in WSIIs and the relationship between pollution sources and visibility were investigated.

The average PM_1.0 mass concentrations in the TSN area showed a trend of being high in spring and low in summer, this result is similar to that of most other studies performed in general urban environments. ${\mathrm{Ca}}^{2+}$ and SNA were the dominant WSIIs in PM_1.0, contributing ~89% to 95% of the total measured ions. The concentrations of ${\mathrm{Ca}}^{2+}$ and ${\mathrm{NO}}_3^{-}$ were higher in spring than in summer. The ${\mathrm{Ca}}^{2+}$ concentration in particular was nearly six times higher in spring than in summer. The springtime peak in ${\mathrm{Ca}}^{2+}$ is most reasonably attributed to high concentrations of fugitive dust during this spring monitoring time.

A correlation analysis between the meteorological parameters and WSII concentrations showed that relative humidity and wind speed had an obvious impact on the ion concentrations in the TSN region, while no significant trend was observed for the influence of temperature and ambient air pressure.

The continental contributions of ${\mathrm{SO}}_4^{2-}$, ${\mathrm{K}}^{+}$ and ${\mathrm{Ca}}^{2+}$ accounted for more than 60% during the whole period. In comparison, ${\mathrm{Cl}}^{-}$ was mainly from marine sources, while the marine fraction of ${\mathrm{Mg}}^{2+}$ varied in a comparatively large range from 27.68% to 82.99%. Airport emissions were a major source of pollution for the sampling site and contributed significantly to sulfate, nitrate, and ammonium levels. Sea salt sources also made a significant contribution to WSIIs, and the contribution showed an increase of 9.33% from spring to summer. Greater water vapor convergence due to the high temperature and rain in summer is the main reason for this phenomenon. The contribution of secondary inorganic aerosols showed a slightly increased tendency in summer, with values of 20.26% in spring and 21.42% in summer. Therefore, the management of secondary organic aerosols in summer requires more attention.

Our research on visibility showed that agricultural activities were the dominant factor impacting visibility in spring. In summer, the airport emissions and secondary inorganic aerosols were the main components affecting visibility. In addition, the relatively high humidity in coastal cities is an adverse factor in reducing visibility. Thus, improving atmospheric visibility in the area should focus on reducing precursors that accelerate atmospheric oxidation and the generation of secondary particulates, as well as reducing biomass-burning activities.