Spatiotemporal Patterns and Regional Transport Contributions of Air Pollutants in Wuxi City

Abstract

In recent years, with the rapid socioeconomic development of Wuxi City, the frequent occurrence of severe air pollution events has attracted widespread attention from both the local government and the public. Based on the real-time monitoring data of criteria pollutants and GDAS (Global Data Assimilation System) reanalysis data, the spatiotemporal variation patterns, meteorological influences, and potential sources of major air pollutants in Wuxi across different seasons during 2019 (pre-COVID-19) and 2023 (post-COVID-19 restrictions) are investigated using the Pearson correlation coefficient, potential source contribution function (PSCF), and concentration-weighted trajectory (CWT) models. The results demonstrate that the annual mean PM_2.5 concentration in Wuxi decreased significantly from 39.6 μg/m³ in 2019 to 29.3 μg/m³ in 2023, whereas the annual mean 8h O₃ concentration remained persistently elevated, with comparable levels of 104.6 μg/m³ and 105.0 μg/m³ in 2019 and 2023, respectively. The O₃ and particulate matter (PM) remain the most prominent air pollutants in Wuxi’s ambient air quality. The hourly mass concentrations of criteria pollutants, except O₃, exhibited characteristic bimodal distributions, with peak concentrations occurring post-rush hour during morning and evening commute periods. In contrast, O₃ displayed a distinct unimodal diurnal pattern, peaking between 15:00 and 16:00 local time. The spatial distribution patterns revealed significantly elevated concentrations of all monitored species, excluding O₃, in the central urban zone, compared to the northern Taihu Lake region. The statistical analysis revealed significant correlations among PM concentrations and other air pollutants. Additionally, meteorological parameters exerted substantial influences on pollutant concentrations. The PSCF and CWT analyses revealed distinct seasonal variations in the potential source regions of atmospheric pollutants in Wuxi. In spring, the Suzhou–Wuxi–Changzhou metropolitan cluster and northern Zhejiang Province were identified as significant contributors to PM_2.5 and O₃ pollution in Wuxi. The potential source regions of O₃ are predominantly distributed across the Taihu Lake-rim cities during summer, while the eastern urban agglomeration adjacent to Wuxi serves as major potential source areas for O₃ in autumn. In winter, the prevailing northerly winds facilitate southward PM_2.5 transport from central-northern Jiangsu, characterized by high emissions (e.g., industrial activities), identifying this region as a key potential source contribution area for Wuxi’s aerosol pollution. The current air pollution status in Wuxi City underscores the imperative for implementing more stringent and efficacious intervention strategies to ameliorate air quality.

1. Introduction

The issue of atmospheric pollution in China has received considerable academic attention, driven by rapid economic growth, industrialization, and urbanization [1,2]. Empirical studies demonstrate that air pollution significantly undermines public health, causing ~1.3 million excess deaths per year, while concurrently incurring gross domestic product (GDP)-related economic costs of 1–8% in China [3,4,5]. To control air pollution, the Chinese government implemented the revised Ambient Air Quality Standards (GB3095-2012) in 2012 (Table S1). This policy mandated real-time monitoring of six criteria pollutants, including inhalable particulate matter (PM₁₀ and PM_2.5), sulfur dioxide (SO₂), nitrogen dioxide (NO₂), carbon monoxide (CO), and ozone (O₃), across all 338 municipal cities. By 2013, China had established a nationwide air quality monitoring network capable of hourly measurements, achieving comprehensive coverage. The enforcement of various pollution control interventions has led to progressive air quality enhancements. Recent years have witnessed sustained improvements in air quality, attributable to the implementation of multiple governance measures [6,7]. The analysis of monitoring data from 366 urban stations across China revealed 13.6–30.5% declines in ambient levels of PM₁₀, PM_2.5, SO₂, and CO between 2015 and 2017 [6].

The spatiotemporally heterogeneous nature of emission control policies necessitates systematic examination of pollutant dispersion dynamics and driving factors to quantify policy impacts. Van et al. systematically analyzed the decadal (2005–2015) emission trajectories of NO_x and SO₂ across China, providing a policy-relevant assessment of air quality management efficacy [8]. Hu et al. conducted a comprehensive analysis of the spatiotemporal distribution characteristics and evolving patterns of criteria air pollutants in China’s Yangtze River Delta (YRD) metropolitan cluster between 2006 and 2019 [9]. Recent observations reveal divergent trends in China’s air quality from 2013 to 2017, with particulate matter (PM) levels demonstrating measurable declines, whereas surface O₃ concentrations showed statistically significant increases [10]. Li et al. newly identified the persistence of O₃ pollution into late winter haze seasons, challenging the conventional understanding of seasonal air pollution patterns [11]. Naeem et al. conducted a comparative analysis of air quality variations in 12 cities across Pakistan, India, China, and Korea before and during the COVID-19 outbreak using satellite remote sensing data [12]. Furthermore, urban air quality is influenced by the coupled effects of multiple factors, including emission sources, meteorological conditions, topographic features, and atmospheric chemical transformations [13,14]. Li et al. demonstrated significant negative correlations between mean pollutant concentrations and daily temperature, precipitation, and relative humidity [15]. Complementing this, Wang et al. identified that persistent weak surface winds, combined with shallow boundary layers, consistently triggered severe pollution episodes [16]. Chemical interactions between pollutant species are well-documented; for example, empirical evidence from Zhang et al. revealed NO_x and SO₂ not only function as key precursors for PM_2.5 through gas-to-particle conversion, but also catalyze heterogeneous reactions during severe haze events [17].

As the primary engine of China’s economic growth, the YRD urban agglomeration continues to face significant atmospheric pollution challenges as a prominent environmental bottleneck in its urbanization process, despite implementing a series of pollution control measures. The persistent issue stems from two key factors, including consistently high energy consumption intensity, with 12,000 metric tons of standard coal per square kilometer in 2021, and sustained rapid growth in vehicle ownership, with an annual growth rate of 8.7% [18,19,20,21]. Moreover, the 2019 Outline of the YRD Integration Development Plan mandates that the YRD region must ensure a continuous reduction in total emissions of major air pollutants and improve regional air quality.

Wuxi, located in the southeastern part of Jiangsu Province, serves as a pivotal central city within the YRD region. It boasts a strategic geographical position and serves as a convergence point for multiple national initiatives. The implementation of rigorous atmospheric pollution control measures constitutes a fundamental determinant for sustaining Wuxi’s socioeconomic progression. To date, comprehensive studies elucidating the spatiotemporal characteristics and key drivers of atmospheric pollution in the Wuxi region remain scarce. To address this research gap, this study has the following research goals: (1) to elucidate the temporal variations and spatial distribution patterns of major atmospheric pollutants during two distinct periods: the pre-COVID-19 year (2019) and the post-COVID-19 restriction year (2023) using temporally resolved regulatory air monitoring observations coupled with Global Data Assimilation System (GDAS) reanalysis data; (2) to investigate the interactions among pollutants and their associations with meteorological factors; (3) building upon the backward trajectory, to integrate the potential source contribution function (PSCF) and concentration-weighted trajectory (CWT) analysis to conduct quantitative numerical simulations of significant potential pollution sources affecting atmospheric pollutants across different seasons, aiming to identify the major potential source regions contributing to pollutant transport to Wuxi.

The structure of this paper is organized as follows: Section 2 describes the study area and data sources. Section 3 first analyzes the seasonal, monthly, and diurnal variations of pollutants, followed by investigations of the spatial distributions, inter-pollutant relationships, pollutant–meteorology interactions, and potential sources. Finally, Section 4 summarizes the main conclusions. This study aims to provide support for air quality forecasting and pollution control strategies in Wuxi and its surrounding regions.

2. Data and Methodology

2.1. Site and Data Description

Wuxi City is situated in the south of Jiangsu Province and is mainly composed of plains, with scattered low mountains and residual hills, bordering the Yangtze River in the north and Taihu Lake in the south (Figure 1). Wuxi is a central city of the YRD, with an urban area of approximately 4627 km², a permanent population of 7.5 million, and a total of 2.7 million vehicles. According to the 2023 Statistical Bulletin, the GDP of Wuxi has climbed to 1.55 trillion yuan, ranking 14th in China. Wuxi City belongs to the subtropical humid monsoon climate zone, with southeast winds prevailing in the hot, humid summer and northerly winds prevailing in the cold, dry winter. The mean temperature (T) is up to 17 °C, the wind speed (WS) averages 2 m/s, and the mean relative humidity (RH) is 70%. The surface meteorological parameters, including T, the maximum temperature (T_max), the minimum temperature (T_max), WS, RH, and precipitation (Pre), were obtained from Wuxi Meteorological Bureau. The concentrations of PM_2.5, PM₁₀, SO₂, NO₂, CO, and O₃ were measured at a 1 h time resolution by the air quality monitoring stations, and corresponding data were downloaded from the Wuxi Environmental Monitoring Center. The 8h O₃ represents the maximum moving average concentration of ozone over any consecutive 8 h period. Two years of monitoring data (2019 and 2023) are employed in this study. As illustrated in Figure 1, the eight monitoring sites are Xue Lang (XL, 31.49° N, 120.27° E), Huang Xiang (HL, 31.62° N, 120.28° E), Cao Zhang (CZ, 31.56° N, 120.29° E), Qi Tang (QT, 31.50° N, 120.24° E), Dong Ting (DT, 31.59° N, 120.35° E), Wang Zhuang (WZ, 31.55° N, 120.35° E), Rong Xiang (RX, 31.56° N, 120.25° E), and Yan Qiao (YQ, 31.68° N, 120.30° E). The six criteria pollutant concentrations in Wuxi were calculated as the mean values obtained from the eight sampling sites mentioned above. By averaging the hourly data, the daily, monthly, seasonal, and annual concentrations of criteria pollutants were obtained.

2.2. Methods

The PSCF [22] analysis reflects the probable source region by statistically evaluating 24 h backward trajectories (24 times per day with a 1 h simulation period) combined with pollutant concentration data using conditional probability calculations. The meteorological database was extracted from the GDAS [23]. The selected simulation height of 500 m (mid-boundary layer) ensures the representation of large-scale flow features while reducing ground friction interference. The study domain extends from 24° N to 42° N and from 112° E to 130° E, encompassing over 95% of the total trajectory coverage area. The target region was discretized into 32,400 grid cells with a spatial resolution of 0.1° × 0.1°.

The PSCF is mathematically defined as follows:

$$PSC{F}_{ij}={\displaystyle \frac{m_{ij}}{n_{ij}}}$$

(1)

where PSCF_ij represents the potential source contribution value for grid cell (i, j), n_ij represents the total number of trajectory endpoints passing through grid cell (i, j), and m_ij represents the number of endpoints in grid cell (i, j) associated with polluted trajectories. This study defines trajectories associated with daily average PM_2.5 concentrations >35 μg/m³ and 8h O₃ concentrations >100 μg/m³ (Grade I standard limits of the Chinese guidelines) as PM_2.5-polluted trajectories and O₃-polluted trajectories, respectively.

As a conditional probability metric, the PSCF is subject to certain uncertainties. The magnitude of error depends on both the total trajectory residence time and the value of n_ij. When air masses exhibit brief residence times in a grid cell, or when n_ij is small, the PSCF results may contain significant errors. To address this limitation, researchers have introduced a weighting function, W(n_ij), when n_ij falls below three times the average number of trajectory endpoints per grid cell across the study domain [24,25]. To mitigate potential uncertainties arising from low n_ij values, we introduced an empirical weighting function, W(n_ij), defined as follows:

$$WPSC{F}_{ij}={\displaystyle \frac{m_{ij}}{n_{ij}}}\times W(n_{ij})$$

(2)

where WPSCF_ij represents the weighted PSCF value for grid cell (i, j). The weighting function W(n_ij) was set as follows [26,27]:

$$W(n_{ij})=\left\{\begin{array}{l}1.00, 80<n_{ij}\\ 0.70, 20<n_{ij}\le 80\\ 0.42, 10<n_{ij}\le 20\\ 0.05, n_{ij}\le 10\end{array}\right.$$

(3)

The PSCF method identifies the spatial distribution of potential pollution sources by calculating the proportion of polluted trajectories in each grid. However, it cannot quantitatively reflect the differences in contribution magnitudes between various source regions. The CWT model is employed to quantitatively evaluate the pollution intensity associated with different air mass pathways by calculating trajectory concentration weights [28]. The specific computational procedure is as follows:

$$C_{ij}={\displaystyle \frac{{\displaystyle {\sum }_{h=1}^M{C}_h\times {\tau }_{ijh}}}{{\displaystyle {\sum }_{h=1}^M{\tau }_{ijh}}}}\times W(n_{ij})$$

(4)

where C_ij represents the mean concentration weight for grid cell (ij); h denotes the index of the trajectory; M is the total number of trajectories; τ_ijh indicates the residence time of trajectory h in grid (i, j), approximated by the count of trajectory endpoints within the grid; and C_h is the measured pollutant concentration corresponding to trajectory h when passing grid (i, j). Similar to the PSCF method, the CWT analysis incorporates the same weighting function W(n_ij) to reduce uncertainties in C_ij calculations:

$$WCW{T}_{ij}=C_{ij}\times W(n_{ij})$$

(5)

where WCWT_ij represents the weighted CWT value for grid cell (i, j).

Both PSCF and CWT analysis is run by the MeteoInfo soltware-TrajStat Plugin [29]. The PSCF and CWT models have been important tools for identifying potential source areas of atmospheric pollutants. More information can be found elsewhere [24,25,26,27,28,29,30].

3. Results and Discussion

3.1. Annual Variation

The COVID-19 pandemic and its subsequent control measures had a profound impact on air quality worldwide. This study compares air pollution levels during two distinct periods: 2019 (pre-COVID-19), representing baseline air quality under normal socioeconomic conditions, and 2023 (post-COVID-19 restrictions), reflecting air quality after China’s full lifting of social control measures.

Statistical analysis was conducted on air pollutant concentrations and the Air Quality Index (AQI) in Wuxi for the years 2019 and 2023. The results indicate that both years had complete datasets with 365 valid days of AQI records. As shown in

Table 1. Summary of AQI and the concentrations of six criteria air pollutants in Wuxi during 2019 and 2023 ^a.

	Year	AQI	PM2.5 (µg/m3)	PM10 (µg/m3)	SO2 (µg/m3)	NO2 (µg/m3)	CO (mg/m3)	8h O3 (µg/m3)	Substandard Ratio b
Annual	2019	83.4 ± 33.5	39.6 ± 21.7	71.8 ± 37.2	8.3 ± 3.0	39.9 ± 16.8	0.9 ± 0.2	104.6 ± 54.4	28.2%
	2023	74.1 ± 32.1	29.3 ± 18.4	56.3 ± 37.4	7.8 ± 1.7	31.6 ± 17.2	0.8 ± 0.2	105.0 ± 46.6	17.5%
Spring	2019	82.8 ± 30.4	41.5 ± 16.4	79.2 ± 25.0	8.9 ± 2.4	41.5 ± 10.2	0.8 ± 0.2	115.9 ± 47.9	23.9%
	2023	82.3 ± 29.7	30.6 ± 11.7	73.1 ± 47.2	8.3 ± 1.5	29.8 ± 11.9	0.7 ± 0.1	122.0 ± 40.7	10.9%
Summer	2019	90.6 ± 41.5	26.3 ± 10.5	46.3 ± 16.4	6.4 ± 1.5	25.8 ± 7.9	0.7 ± 0.2	143.1 ± 53.0	39.1%
	2023	76.6 ± 39.4	17.5 ± 8.2	30.4 ± 13.1	6.5 ± 0.9	19.0 ± 5.9	0.7 ± 0.1	127.3 ± 49.4	23.9%
Autumn	2019	77.8 ± 28.3	34.2 ± 15.9	74.2 ± 44.2	8.5 ± 3.3	43.5 ± 17.5	0.9 ± 0.2	104.6 ± 45.4	20.9%
	2023	70.0 ± 26.8	27.5 ± 15.8	51.0 ± 26.9	8.0 ± 1.4	37.1 ± 15.9	0.9 ± 0.2	106.2 ± 41.5	19.8%
Winter	2019	82.2 ± 31.3	56.6 ± 27.8	87.9 ± 41.9	9.3 ± 3.6	49.9 ± 19.0	1.1 ± 0.2	53.5 ± 22.2	28.9%
	2023	67.3 ± 29.3	41.6 ± 24.9	70.7 ± 36.4	8.4 ± 1.9	40.7 ± 22.1	0.9 ± 0.3	63.6 ± 20.4	15.6%

^a The error denotes standard deviation, p < 0.05. ^b The percentage of days with AQI > 100.

, 262 days (71.8% of the total) were classified as “excellent/good” (AQI < 100) in 2019, whilst 301 days (82.5% of the total) reached “excellent/good” standards in 2023. Substandard days (AQI > 100) decreased from 103 days (28.2%) in 2019 to 64 days (17.5%) in 2023. The comparative analysis revealed a consistent improvement pattern in Wuxi’s air quality, with a progressive increase in annual ‘excellent/good’ air quality days and a corresponding reduction in heavily polluted days.

The annual mean PM_2.5 (PM₁₀) concentrations in 2019 and 2023 were 39.6 ± 21.7 µg/m³ (71.8 ± 37.2 µg/m³) and 29.3 ± 18.4 µg/m³ (56.3 ± 37.4 µg/m³), respectively (Table 1), exceeding or approaching the Grade II limits of the China Ambient Air Quality Standards (CAAQS) (annual mean: 35 µg/m³ for PM_2.5 and 70 µg/m³ for PM₁₀) (Table S1). The standard deviation in 2019 was generally higher than that in 2023, likely due to more frequent severe pollution episodes (AQI > 200: 22 days in 2019 vs. 8 days in 2023). Regarding gaseous pollutants, the annual mean concentrations of SO₂, NO₂, CO, and 8h O₃ in Wuxi in 2019 (2023) were 8.3 (7.8), 39.9 (31.6), 0.9 × 10³ (0.8 × 10³), and 104.6 (105.0) µg/m³, respectively. The statistical analysis revealed significant decreases (p < 0.05) in annual mean concentrations for PM, SO₂, NO₂, and CO in 2023 relative to the 2019 levels, with O₃ being the sole exception. Strengthening legislative and regulatory measures is recommended to further control O₃ and PM pollution. In recent years, Wuxi has implemented measures such as clean energy substitution, clean coal technologies, and advanced boiler emission treatment, leading to further improvement in air quality.

Figure 2 presents the proportions of the predominant pollutants in Wuxi during the two study periods. The predominant pollutant is defined as the one with the highest contribution to the Air Quality Index (AQI) when the AQI exceeds 50. The term “Clean” denotes periods when the AQI remains below 50, corresponding to the excellent air quality classification according to CAAQS. During both 2019 and 2023, 8h O₃, PM_2.5, PM₁₀, and NO₂ were the most frequently observed predominant pollution, while neither SO₂ nor CO emerged as predominant pollution. Specifically, the proportion of 8h O₃ showed a modest increase from 39% in 2019 to 41% in 2023, whereas PM₁₀ levels remained stable. In contrast, PM_2.5 exhibited a marked declining trend, decreasing from 19% in 2019 to 7% in 2023. Wuxi maintained relatively low NO₂ pollution levels, and further reduced during 2023.

3.2. Temporal Variations

3.2.1. Seasonal and Diurnal Variation

Further analysis was conducted on the seasonal, monthly, and diurnal variation patterns of pollutant concentrations in Wuxi over these two periods, with the results illustrated in Figure 3.

The seasonal variations in PM_2.5 and pollution levels in Wuxi during 2019 and 2023 reveal that the highest concentrations occurred in winter, followed by spring and autumn (with spring slightly higher than autumn), while the lowest levels were observed in summer (Table 1). This variation pattern aligns with the seasonal trends of PM_2.5 reported in other regions [31,32,33]. The mean PM_2.5 mass concentration in winter was more than twice that observed in summer. The highest PM_2.5 was recorded in winter 2019, at 56.6 μg/m³, while the lowest was observed in summer 2023, at 17.5 μg/m³. The diurnal variation curve exhibits the most pronounced fluctuations in winter, while the amplitude is the most subdued in summer (Figure 3). This phenomenon can be attributed to the prevalence of stagnant weather conditions during the cold winter, characterized by a stable atmospheric structure and a lower planetary boundary layer height, which facilitate pollutant accumulation [34]. Additionally, pollutants are advected from more polluted areas to monitoring stations by prevailing winter winds, resulting in local air quality degradation. Summer precipitation rises markedly, accumulating up to 441.1 mm in 2019 and 737.9 mm in 2023 (

Table 2. Statistics of meteorological parameters at Wuxi *.

Meteorological Parameters	Year	Total	Spring	Summer	Autumn	Winter
T (°C)	2019	17.4	16.7	27.4	19.4	6.0
	2023	17.8	17.7	28.2	19.5	5.9
Tmax (°C)	2019	21.8	21.8	31.5	24.0	9.8
	2023	22.7	22.9	32.5	24.6	10.9
Tmin (°C)	2019	13.6	12.1	23.9	15.4	2.9
	2023	13.7	13.2	24.7	15.2	1.9
RH (%)	2019	72.1	65.0	74.6	72.1	76.6
	2023	70.5	63.2	76.1	73.2	69.4
WS (m/s)	2019	2.0	2.1	2.2	1.9	1.9
	2023	2.0	2.4	1.9	1.8	1.9
Pre (mm)	2014	1030.5	158.1	441.1	178.5	252.8
	2015	1266.0	179	737.9	194.2	154.9

* Note: mean values for air temperature (T), maximum temperature (T_max), minimum temperature (T_min), relative humidity (RH), wind speed (WS), accumulated value of precipitation (Pre).

), respectively. The abundance of hygroscopic particles in the atmosphere enhances the effective dilution and deposition of airborne PM_2.5 [35]. Concurrently, intensified solar radiation during summer reduces the probability of temperature inversion formation. The improved atmospheric turbulence, horizontal transport, and vertical diffusion collectively inhibit the accumulation of air pollutants.

The diurnal variation patterns of air pollutants are critically important for elucidating the influences of potential emission sources, including domestic cooking, vehicular traffic, industrial activities, and associated meteorological parameters in urban environments. The diurnal variation of PM_2.5 concentrations in Wuxi exhibits a bimodal pattern with dual peaks and dual troughs, showing significant fluctuations during daytime and relatively stable levels at night (Figure 3a). The peak and trough timings of the four seasons vary slightly. The diurnal cycle exhibits a morning peak (9:00–10:00) and an evening peak (21:00–22:00), with daytime (15:00–18:00) and nocturnal (02:00–04:00) troughs. Winter morning peaks lag by 1 h relative to the other seasons. The diurnal variation characteristics of PM_2.5 in Wuxi are consistent with other existing studies [31]. The daily fluctuation of PM_2.5 mass concentration is primarily influenced by human activities, atmospheric diffusion conditions, photochemical reactions, and traffic emissions. In the early morning, atmospheric stability and frequent temperature inversions lead to weak turbulent diffusion, trapping pollutants near the surface. PM_2.5 levels begin to rise from 6:00, peaking around 9:00 due to increased human activity, temperature-driven photochemical reactions, and traffic emissions. In the afternoon, the elevation of the atmospheric boundary layer, enhanced solar radiation, and stronger convective turbulence promote pollutant dispersion and dilution, causing PM_2.5 concentrations to decline further, reaching their lowest point around 17:00. Subsequently, with increased emissions from evening rush-hour activities, combined with atmospheric stabilization and weakened turbulent transport, PM_2.5 levels rise again, peaking around 21:00.

The mass concentrations of PM₁₀ in Wuxi exhibited distinct seasonal variations, with higher levels observed during spring (79.2 μg/m³ in 2019 vs. 73.1 μg/m³ in 2023) and winter (87.9 μg/m³ in 2019 vs. 70.7 μg/m³ in 2023) compared to summer (46.3 μg/m³ in 2019 vs. 30.4 μg/m³ in 2023) and autumn (74.2 μg/m³ in 2019 vs. 51.0 μg/m³ in 2023). Wuxi is located in the southern part of Jiangsu Province. According to data from online statistics websites [36], the pollutant concentrations in central and northern Jiangsu are higher than those in Wuxi. The prevailing northerly winds during the spring and winter seasons facilitate the long-range transport of dust particles from northern regions to Wuxi. The diurnal variation of PM₁₀ concentrations displayed a bimodal distribution, with peak concentrations occurring during the morning (09:00–11:00) and evening (20:00–22:00) periods, while the lowest values were typically recorded in the afternoon (around 16:00) (Figure 3b).

SO₂ and NO₂, as gaseous precursors of PM_2.5 and serve as characteristic tracers for emission sources from fossil fuel combustion and vehicular activities, providing important insights into PM_2.5 sources. The diurnal variation of SO₂ concentrations in Wuxi exhibited a unimodal pattern across all seasons, with consistently higher daytime levels compared to nighttime values. Peak concentrations occurred around 10:00, with 13 μg/m³ in 2019 and declining to 9–10 μg/m³ in 2023, indicating persistent daytime SO₂ emission sources in the surrounding areas, likely associated with industrial activities requiring fossil fuel combustion (Figure 3c). The moderate elevation in SO₂ concentrations after 18:00 during the autumn and winter months likely originates from residential heating activities, particularly domestic coal combustion for space heating. The characteristic winter-maximum pattern in SO₂ shows strong temporal coherence with PM mass concentration peaks during winter, demonstrating the substantial contribution of coal-fired heating emissions to particulate pollution.

The diurnal variation of NO₂ exhibited a distinct bimodal pattern, with morning (08:00) and evening (20:00) peaks, occurring approximately 1–2 h earlier than the corresponding PM_2.5 mass concentration maxima (Figure 3d), which implicates traffic-related emissions as a dominant contributor to PM_2.5 in Wuxi, with vehicular exhaust representing one of the most significant anthropogenic sources. The diurnal variation pattern of CO mass concentration exhibits similar characteristics to that of NO₂ (Figure 3e).

The near-surface O₃ concentrations in all four seasons exhibited a unimodal diurnal variation pattern (Figure 3f). This is attributed to the fact that ground-level O₃ is a secondary pollutant formed through photochemical reactions involving its precursors, e.g., NO_x, CO, and volatile organic compounds (VOCs), with solar radiation intensity being the dominant factor driving its diurnal variation [37,38]. During the pre-sunrise period, O₃ concentrations exhibited minimal temporal variation, with a gradual declining trend due to weak solar radiation. The morning traffic rush hour led to increased emissions of O₃ precursors, which acted as reducing agents and consumed O₃, resulting in the daily minimum concentration occurring between 06:00–07:00. Subsequently, as solar radiation intensified, O₃ levels progressively increased, peaking between 15:00–16:00. Surface O₃ concentrations maintained consistently low levels, with gradual declining trends during pre-dawn hours across all four seasons. NO₂ concentrations reach their daily minimum at 15:00, followed by a gradual increase. This pattern further confirms that NO₂ is a key participant in O₃ formation, as the period between 12:00 and 15:00 represents the optimal window for O₃ production. During this time, substantial nitrogen oxides (NO_x) engage in photochemical reactions, leading to a significant increase in O₃ levels and a corresponding depletion of NO₂. The O₃ concentrations decrease following sunset in response to diminishing solar radiation intensity and declining temperatures, with this phenomenon being particularly pronounced during nighttime hours in the cold season.

Significant seasonal variations were observed for criteria air pollutants in 2019, with winter concentrations of PM_2.5, PM₁₀, NO₂, CO, and SO₂ exhibiting higher levels compared to other seasons, highlighting particularly severe winter pollution episodes. By 2023, overall reductions were observed in ambient concentrations of PM_2.5, PM₁₀, NO₂, CO, and SO₂, primarily attributable to a suite of air quality control policies implemented by municipal authorities. Maximum 8h O₃ concentrations were observed during summer in both 2019 and 2023. Of particular significance is the minimal difference (merely 5 µg/m³) between spring and summer concentrations in 2023. Despite overall pollutant reductions, O₃ concentrations in 2023 remained statistically indistinguishable from the 2019 levels, reflecting counterbalancing influences from persistent local anthropogenic emissions and/or enhanced regional transport effects.

Additionally, the proportions of predominant pollutants exhibited significant seasonal variations (Figure 2). During spring, O₃ and PM₁₀ were the dominant pollutants, accounting for 42% and 22% in 2019 and 53% and 26% in 2023, respectively. During summer, O₃ contributed to over 70% of pollution, making it the most frequently dominant pollutant. During autumn, O₃ and PM₁₀ were the predominant pollutants in 2019, with contributions of 40% and 23%, respectively. However, in 2023, only O₃ remained dominant (43%), while PM10’s contribution declined significantly to 5%. By contrast, during winter, PM_2.5 and NO₂ were the predominant pollutants, contributing 56% and 20% in 2019, and 24% and 20% in 2023, respectively. O₃ and PM are the main factors causing air pollution in Wuxi.

3.2.2. Monthly Variation

A consistent U-shaped monthly trend was observed across all criteria pollutants, except O₃, while the secondary pollutant O₃ showed an anti-phase pattern. Taking 2019 as an example, concentrations of PM_2.5, PM₁₀, SO₂, NO₂, and CO peaked during the cold months, reaching 60.7 μg/m³ (January), 99.1 μg/m³ (January), 10.7 μg/m³ (January), 57.1 μg/m³ (December), and 1.14 mg/m³ (January), respectively. During the warm months, these pollutants reached their lowest levels: 22.1 μg/m³ for PM_2.5 (August), 41.2 μg/m³ for PM₁₀ (August), 6.3 μg/m³ for SO₂ (June), 23.9 μg/m³ for NO₂ (August), and 0.7 mg/m³ for CO (July). The 8h O₃ concentrations peaked in June at 151.8 μg/m³. During summer, the average regional temperature gradually increased to 27–28 °C (Table 2), accompanied by extended sunlight duration. These conditions enhanced photochemical reaction rates in the atmosphere, consequently promoting O₃ formation and accumulation. Statistical analysis revealed a modest negative correlation (r = −0.20), between daily precipitation amounts and O₃ levels, suggesting limited precipitation-mediated O₃ suppression in this region (detailed in Section 3.4). Maximum 8h O₃ concentrations in Wuxi peaked on Sundays (110.5 μg/m³) and reached minimum levels on Tuesdays (98.9 μg/m³). This phenomenon primarily results from the ‘O₃ weekend effect’ and lower NO_x emissions [39,40,41].

In brief, compared with 2019, Wuxi made some progress in overall air quality by 2023. However, several issues remain in air pollution control. This is primarily reflected in the fact that the major pollutants in Wuxi have gradually shifted from particulate matter to O₃, demonstrating that the measures implemented in Wuxi, including construction dust management, the promotion of new energy vehicles, and the regulation of heavy-duty diesel trucks, have shown preliminary effectiveness in mitigating particulate matter pollution.

3.3. Spatial Distribution

Table 3. Average diurnal concentrations of six criteria air pollutants at eight stations in Wuxi during 2023 (the units are µg/m³ for PM_2.5, PM₁₀, SO₂, NO₂, and 8h O₃; and mg/m³ for CO).

Station	Description	PM2.5	PM10	SO2	NO2	CO	8h O3
urban district
HX	Huang Xiang	29.3	65.3	7.7	39.8	0.9	105.0
CZ	Cao Zhang	29.5	55.8	7.8	34.5	0.9	101.2
DT	Dong Ting	30.8	58.1	8.5	35.8	0.9	105.8
WZ	Wang Zhuang	29.3	55.8	7.1	35.3	0.9	105.4
YQ	Yang Qiao	29.0	57.9	7.7	31.9	0.9	105.5
the north shore of Taihu Lake
XL	Xue Lang	27.0	54.9	6.7	24.8	0.8	110.3
QT	Qi Tang	27.9	50.8	7.4	24.6	0.8	109.3
RX	Rong Xiang	28.3	49.8	9.8	25.9	0.8	107.0

summarizes the detailed average concentrations of ambient pollutants at different monitoring sites in 2023. The highest daily mean PM concentrations were recorded at the DT (PM_2.5: 13.5 μg/m³) and HX (PM₁₀: 65.3 μg/m³) monitoring sites, while the XL and QT sites exhibited the peak 8h O₃ concentrations, with 110.3 μg/m³ and 109.3 μg/m³, respectively. The HX site registered the highest NO₂ levels (39.8 μg/m³), whereas the SO₂ and CO concentrations remained relatively consistent across all eight monitoring sites. The monitoring sites exhibited distinct spatial distribution characteristics due to variations in emission sources. The frequencies of O₃ and PM exceeding the Grade II CAAQS were significantly higher than those of other pollutants, indicating that atmospheric pollution in this region was primarily driven by O₃ and PM exceedances.

Overall, the urban district sites (HX, CZ, DT, WZ, and YQ) exhibited the highest pollution levels, whereas the north shore of Taihu Lake (XL, QT, and RX) recorded the cleanest air. This spatial pattern can be attributed to several factors, such as higher population density in urban areas, more intensive activities (e.g., vehicular traffic, construction projects, and catering services), and significant fugitive dust emissions, collectively leading to elevated anthropogenic emissions of PM and NO₂ (PM_2.5: 29.0–30.8 μg/m³, PM₁₀: 55.8−65.3 μg/m³, NO₂: 31.9−39.8 μg/m³). In contrast, the urban center exhibited lower 8h O₃ concentrations, particularly 101.2 μg/m³ in CZ. Furthermore, the northern shore of Taihu Lake in Wuxi is characterized by elevated moisture levels and reduced anthropogenic emission compared to urban core areas. The favorable topographic conditions further enhance atmospheric dispersion in this region, consequently leading to a corresponding reduction in the pollutant concentrations (PM_2.5: 27.0−28.3 μg/m³, PM₁₀: 49.8−54.9 μg/m³, NO₂: 24.6−25.9 μg/m³). The predominance of biogenic VOC emissions from substantial vegetation cover constitutes the major driver of observed O₃ elevation (107.0−110.3 μg/m³) in the northern Taihu Lake basin.

3.4. Correlations Between Air Pollutants and Meteorological Factors

Based on the preceding analysis, except for O₃, the concentrations of various pollutants in Wuxi exhibit similar temporal trends, suggesting intrinsic relationships among them. This study quantified inter-pollutant relationships using Pearson correlation coefficient (r) analysis, with the resultant correlation matrix presented in

Table 4. Pearson correlations between six pollutants and meteorological elements ^a: 0.50–1.00 (strong positive correlation), 0.25–0.49 (moderate positive correlation), 0–0.25 (weak positive correlation), −0.25 to 0 (weak negative correlation), −0.50 to −0.25 (moderate negative correlation), and −1.00 to −0.50 (strong negative correlation).

PM2.5

−

1 > R ≥ 0.5

PM10

0.86 **

−

0.5 > R ≥ 0.25

SO2

0.62 **

0.69 **

−

0.25 > R ≥ 0

NO2

0.67 **

0.72 **

0.63 **

−

0 > R ≥ −0.25

CO

0.80 **

0.62 **

0.47 **

0.56 **

−

–0.25 > R ≥ −0.5

8h O3

−0.11 *

0.00

0.09

−0.25 **

−0.24 **

−

−0.5 > R ≥ −1

T

−0.49 **

−0.41 **

−0.32 **

−0.49 **

−0.46 **

0.65 **

−

Tmax

−0.42 **

−0.30 **

−0.21 **

−0.40 **

−0.43 **

0.72 **

0.98 **

−

Tmin

−0.54 **

−0.51 **

−0.43 **

−0.56 **

−0.47 **

0.52 **

0.97 **

0.91 **

−

WS

−0.33 **

−0.30 **

−0.34 **

−0.50 **

−0.38 **

−0.03

0.15 **

0.12 *

0.19 **

−

RH

−0.08

−0.34 **

−0.51 **

−0.13 *

0.18 **

−0.45 **

−0.04

−0.15 **

0.09

0.00

−

Pre

−0.20 **

−0.28 **

−0.31 **

−0.18 **

−0.04

−0.20 **

0.05

−0.01

0.12 *

0.15 **

0.42 **

−

PM2.5

PM10

SO2

NO2

CO

8h O3

T

Tmax

Tmin

WS

RH

Pre

^a Note: mean values for air temperature (T), maximum temperature (T_max), minimum temperature (T_min), wind speed (WS), relative humidity (RH), and precipitation (Pre). * p < 0.05, ** p < 0.01.

.

PM_2.5 and PM₁₀ demonstrated a highly significant positive correlation, with a coefficient of 0.86. Concurrently, PM exhibited strong positive correlations with SO₂, NO₂, and CO. The key contributing factors to these relationships include the following aspects: Firstly, SO₂ undergoes atmospheric oxidation to form sulfate aerosols (SO₄²⁻), which serve as critical precursor species in PM nucleation and growth processes. Secondly, the NO₂ accumulated in the near-surface atmospheric boundary layer participates in photochemical chain reactions, yielding nitrate compounds (NO₃⁻) that constitute a major fraction of secondary aerosols. Thirdly, CO predominantly originates from combustion-derived emissions, e.g., vehicular exhaust and industrial flue gases, which are inherently associated with concomitant PM emissions. Furthermore, SO₂ and NO₂ exhibit a strong correlation (r = 0.63), attributable to their shared emission sources and concurrent atmospheric removal processes in Wuxi. The observed correlations between CO and SO₂/NO₂/PM arise from their co-emission characteristics during fossil fuel combustion processes. Specifically, while CO primarily originates from incomplete combustion in power plants, industrial facilities, and vehicular engines, these same emission sources simultaneously release SO₂, NO₂, and PM, resulting in consistent atmospheric covariation patterns. The correlation between CO and PM_2.5 (r = 0.80) was significantly stronger than that between CO and PM₁₀ (r = 0.62), indicating that CO emission processes are more closely associated with fine particulate matter emissions.

O₃, as a secondary atmospheric pollutant, exhibits complex formation dynamics wherein NO₂ and CO serve as essential precursors [42]. The observed moderate negative correlations (r = −0.25 for NO₂-O₃; r = −0.24 for CO-O₃) reflect the non-linear photochemical processes governing O₃ production. While O₃ formation is fundamentally dependent on NO_x and CO concentrations through the Chapman cycle and hydrocarbon oxidation pathways, its net production is substantially modulated by multiple environmental variables: solar irradiance intensity, which regulates photolysis rates; ambient temperature, which influences reaction kinetics; relative humidity, which affects radical chemistry; and aerosol loading, which alters heterogeneous reaction pathways. Consequently, these competing physicochemical mechanisms result in non-stationary precursor–O₃ relationships that cannot be adequately characterized by simple correlation metrics alone.

Atmospheric environments exhibit significant regional characteristics, with local meteorological conditions substantially influencing ambient pollutant concentrations. Table 4 presents the statistical relationships between atmospheric pollutants and key meteorological parameters, including T_max, T_min, T, RH, WS, and Pre, in Wuxi City.

Temperature demonstrated moderate negative correlations with PM_2.5 (r = −0.49), PM₁₀ (r = −0.41), SO₂ (r = −0.32), NO₂ (r = −0.49), and CO (r = −0.46). Elevated temperatures enhance vertical convective mixing, promoting efficient pollutant dispersion through increased boundary layer height and improved ventilation coefficients. Conversely, thermal inversion layers frequently form under low-temperature conditions, characterized by suppressed atmospheric turbulence and reduced mixing depth, which collectively inhibit vertical pollutant transport and facilitate the accumulation of primary and secondary species in the surface layer.

Elevated temperatures exert a positive forcing effect on photochemical processes, leading to increased O₃ concentrations, as evidenced by their strong positive correlation (r = 0.65). Under conditions of stable local emissions, WS serves as the dominant regulator of atmospheric pollutant dispersion dynamics. Low WS facilitates atmospheric pollutant accumulation, while stronger winds promote effective dispersion. All measured pollutants, except O₃, exhibited moderate negative correlations with WS (r = −0.33 for PM_2.5-WS; r = −0.30 for PM₁₀-WS; r = −0.34 for SO₂-WS; r = −0.50 for NO₂-WS; r = −0.38 for CO-WS).

Wuxi’s dense hydrographic network and persistently high atmospheric humidity (70.5% for 2019 vs. 72.1% for 2023) facilitated enhanced pollutant removal. Relative humidity exhibited the strongest correlation with SO₂ concentrations (r = 0.51) among all meteorological factors. This observed relationship can be attributed to enhanced aqueous-phase partitioning of SO₂ under high-humidity conditions: increased atmospheric water vapor promotes the dissolution and subsequent oxidation of SO₂ to form sulfate aerosols (SO₄²⁻), thereby reducing gaseous SO₂ concentrations through phase transition. The negative correlations between PM and RH primarily result from the hygroscopic growth and subsequent wet deposition of aerosols. Atmospheric water droplets exhibit stronger adsorption effects on larger particulate matter than on smaller particles. Consequently, the size-selective removal process leads to a more pronounced humidity-mediated depletion of PM₁₀ (r = −0.34) relative to PM_2.5 (r = −0.08). High-humidity conditions promote O₃ depletion through aqueous-phase reactions, resulting in a significant negative correlation between RH and O₃ concentrations (r = −0.45). Moreover, precipitation demonstrated weak to moderate negative correlations with all measured pollutants (r = −0.31 to −0.04). This reflects precipitation’s dual role in atmospheric cleansing: direct wet deposition of PM and soluble gases through below-cloud scavenging processes and indirect modulation of pollution concentrations via humidity modification, as evidenced by the significant positive correlation between precipitation and relative humidity (r = 0.42).

3.5. Pollutants Source Analysis Based on PSCF and CWT

As shown in the seasonal analysis of criteria pollutants in Figure 2, O₃, PM_2.5, and PM₁₀ were the most frequently observed dominant pollutants. Compared to PM₁₀, PM_2.5 exhibits smaller particle sizes, rendering it more susceptible to regional transport via air masses. To further investigate the potential source regions of major pollutants in Wuxi City, we calculated the WPSCF for PM_2.5 and 8h O₃, respectively (Figure 4 and Figure 5). The potential source simulation point was set in urban Wuxi (31.59° N, 120.35° E, elevation: 4 m), with the trajectory starting height set at 500 m above ground level. The 24 h backward trajectory calculation was primarily designed to focus the study area on the YRD regional joint prevention and control zone.

During the spring of 2019, the high-value potential source areas of PM_2.5 (WPSCF > 0.6) were mainly concentrated in urban Wuxi, east of Changzhou, and southeast Zhenjiang (Figure 4a). Additionally, some high-value areas were distributed in northwestern Suzhou. The sub-high-value areas (WPSCF > 0.5) expanded to cover most of Zhenjiang, Changzhou, and Suzhou in Jiangsu Province, as well as southeastern Nanjing and parts of Huzhou, Jiaxing, and Ningbo in Zhejiang Province, and Xuancheng in Anhui Province. As the PM_2.5 concentration decreased from 41.5 μg/m³ during the spring of 2019 to 30.6 μg/m³ in 2023, the corresponding WPSCF values during the spring of 2023 dropped below 0.4 (Figure 4e). The potential source areas were primarily localized in Wuxi and its surrounding regions, indicating that PM_2.5 pollution was dominated by local accumulation. Local emission control should be prioritized in spring, along with joint prevention and control measures with neighboring cities.

Summer exhibited the best air quality throughout the year, with mean PM_2.5 concentrations of 26.3 μg/m³ in 2019 and 17.5 μg/m³ in 2023, respectively. Unlike the spatial patterns observed in spring, no significant high-value zones (WPSCF > 0.5) were detected (Figure 4b,f), as the WPSCF values predominantly remained below 0.5 in 2019 and 0.3 in 2023. These results indicate minimal influence from transboundary particulate matter transport, highlighting the dominance of local meteorological conditions, e.g., enhanced dispersion and precipitation in reducing pollution levels.

In autumn 2019, the high-value potential source areas (WPSCF > 0.6) primarily clustered in central Wuxi, parts of Suzhou, and northern Jiaxing (Figure 4c). The sub-high-value regions (WPSCF > 0.5) showed a concentric distribution adjacent to the WPSCF > 0.6 source regions, extending to southwestern Suzhou and the border area between Zhenjiang and Changzhou, with minor distributions in Taizhou and Yancheng. The autumn 2023 data revealed a contraction of zones with WPSCF > 0.6 to localized areas, concurrent with the PM_2.5 concentration reduction from 34.2 to 27.5 μg/m³ (Figure 4g). Spatial analysis indicates that high-PSCF regions during autumn were principally confined to local and proximate areas, mirroring the spring distribution pattern. This spatial consistency underscores the imperative for intensified local emission controls, and cross-jurisdictional air quality management initiatives with bordering municipalities.

Winter exhibited the highest PM_2.5 concentrations, with mean values reaching 56.6 μg/m³ (41.6 μg/m³) in 2019 (2023), demonstrating greater influence from regional transport (Figure 4d,h). The WPSCF values > 0.6 in winter (2019) were primarily distributed in local Wuxi, Changzhou, eastern Zhenjiang, southern Nanjing, southern Taizhou, and southwestern Suzhou. Notably, Wuxi, Taizhou, and Changzhou exhibited even higher WPSCF values exceeding 0.7. The PM_2.5 concentrations in Taizhou and Changzhou reached 64.2 μg/m³ and 66.7 μg/m³ during winter in 2019 [36], significantly exceeding the level observed in Wuxi (56.6 μg/m³). This substantial disparity highlights the non-negligible influence of regional transport on local air quality. The WPSCF > 0.5 areas were predominantly distributed around the high-value zones, extending to encompass the municipal areas of Nantong, Yancheng, and Yangzhou in Jiangsu Province, along with the north of Zhejiang Province. Wuxi predominantly experiences northerly winds in cold winter, while the central-northern Jiangsu urban cluster exhibits elevated PM_2.5 concentrations, resulting in significant transport impacts on Wuxi’s air quality. These findings highlight the imperative for Wuxi to both intensify local emission controls and enhance coordinated air quality management with upwind cities in central-northern Jiangsu Province.

The annual average 8h O₃ concentrations in 2019 and 2023 were approximately equal. Excluding summer, the 8h O₃ concentrations in spring, autumn, and winter of 2023 were all higher than those during the corresponding seasons of 2019, indicating that Wuxi’s joint prevention and control measures for O₃ pollution need further strengthening. The 24 h potential source regions of O₃ were predominantly concentrated within a 200 km radius of Wuxi.

The regions with WPSCF values > 0.7 for O₃ during spring were primarily clustered in the Wuxi metropolitan area (Figure 5a,e). Among the four seasons, summer exhibited higher WPSCF values, indicating stronger influence from regional transport (Figure 5b,f). The PSCF > 0.6 region encompassed southern Jiangsu (e.g., Suzhou, Changzhou, Zhenjiang, and Nanjing), northern Zhejiang (e.g., Huzhou and Jiaxing), and Shanghai. The high-value areas displayed a southeastward spatial trend, paralleling the dominant summer wind vectors. The air masses passing through these high-emission regions transported substantial amounts of O₃ precursors, which were subsequently converted to O₃ through photochemical reactions, significantly impacting downwind areas. The spatial distribution of high WPSCF values showed good consistency with Pang et al.’s conclusion that the O₃ high-emission areas of YRD were primarily concentrated in economically developed cities along the Yangtze River [43]. The autumn WPSCF pattern (values > 0.6) delineated a distinct eastern transport corridor encompassing Suzhou, Nantong, Shanghai, and Jiaxing (Figure 5c,g). In contrast, winter exhibited minimal regional transport influence on O₃ concentrations, as evidenced by WPSCF values below 0.2 (Figure 5d,h).

To further validate the PSCF analysis results, Figure 6 and Figure 7 present the CWT analysis, which quantitatively assesses the pollution potential of source regions. The spatial distributions of potential source regions identified by concentration contributions and weighting factors showed strong consistency, both indicating that the predominant high-value zones were concentrated in local Wuxi and adjacent areas. Achieving compliance for PM and O₃ pollution standards demands focused local interventions, supported by regional emission coordination.

The high-concentration zones for PM_2.5 in spring (WCWT > 40 μg/m³) exhibited spatial congruence with elevated WPSCF areas, collectively identifying Wuxi local sources, Zhenjiang, Changzhou, and Suzhou as dominant potential source regions (Figure 6a,e). During summer, the WCWT values generally remained below 30 μg/m³ (Figure 6b,f), indicating consistently low PM_2.5 concentrations throughout the area. The spatial distributions of autumn WCWT hotspots (WCWT > 40 μg/m³) demonstrated strong agreement with elevated WPSCF regions (Figure 6c,f), jointly identifying local Wuxi and southern Jiaxing as dominant potential source areas for PM_2.5. Subsidiary WCWT hotspots (>30 μg/m³) showed concentrated distribution across the Changzhou–Zhenjiang urban corridor, parts of Suzhou, and Taizhou.

The distribution range of WCWT is significantly larger in winter than in other seasons (Figure 6d,h). The high-value areas (WCWT > 60 μg/m³) are mainly distributed in northern Wuxi, Changzhou, and the border regions of Suzhou and Zhenjiang. The sub-high regions (WCWT > 35 μg/m³) are primarily located in the urban area of Wuxi, eastern Zhenjiang, most of Changzhou, parts of Taizhou, Yangzhou, Yancheng, and Nantong, indicating that PM_2.5 in Wuxi during winter is significantly influenced by transport from cities in the central and northern regions of Jiangsu Province.

A comparison of the WCWT distributions of O₃ across the four seasons reveals that the contribution of surrounding regions to O₃ levels is higher in summer than in other seasons (Figure 7b,f). In spring, regions with WCWT values of 8h O₃ > 100 μg/m³ are primarily concentrated in southern Wuxi and a small northwestern portion of Zhejiang Province (Figure 7a,e). The high-WCWT regions (8h O₃ > 100 μg/m³) during autumn demonstrated perfect spatial concordance with elevated WPSCF areas (>0.6), mainly located in the eastern part of Wuxi (Figure 7c,g). In winter, O₃ pollution remained relatively subdued, with all WCWT values below 70 μg/m³ (Figure 7d,h). Our findings highlight the considerable role of intercity transport from the Taihu Lake-rim urban agglomeration in exacerbating Wuxi’s O₃ pollution, underscoring the urgency of coordinated emission controls on O₃ precursors, such as VOCs and NO_x, across this industrialized zone.

In brief, the source apportionment results reveal pronounced seasonal variability. To mitigate the effects of cross-boundary transport, comprehensive joint prevention and control mechanisms for air pollution should be established and rigorously enforced in the Wuxi region.

4. Conclusions

This study conducted the first statistical analysis of ambient air quality data in Wuxi City for the years 2019 (pre-pandemic) and 2023 (after the full lifting of COVID-19 restrictions). The spatiotemporal variation trends of atmospheric pollutant concentrations, the correlations among different pollutants, the relationships between pollutants and meteorological factors, as well as the potential source regions of pollutants were examined. The main conclusions are as follows:

(1)

The exceedances of O₃ and PM concentrations occurred with significantly higher frequency compared to other pollutants, marking them as the dominant contaminants. SO₂ and CO concentrations never exceeded national standards throughout the study period. Compared with 2019, a 26% reduction in PM_2.5 concentrations was observed in 2023, decreasing from 39.6 μg/m³ to 29.3 μg/m³. However, 8h O₃ pollution exhibited no significant improvement, remaining at persistently elevated levels (104.6 μg/m³ in 2019 vs. 105.0 μg/m³ in 2023).

(2)

The diurnal profiles of criteria pollutants, except O₃, exhibited consistent bimodal concentration patterns. Winter demonstrated significantly greater amplitude in mass concentration fluctuations relative to other seasons. NO₂ concentrations peaked during morning and evening traffic rush hours, consistent with vehicular emission patterns. PM exhibited delayed peak occurrences, lagging NO₂ by 1–2 h, suggesting secondary aerosol formation processes. SO₂ displayed a unimodal distribution. O₃ also demonstrated a distinct single-peak diurnal profile, reaching maximal values during mid-afternoon, with summer exhibiting highest peak amplitudes compared to other seasons.

(3)

From the spatial distribution pattern of air pollutants, it can be observed that in Wuxi’s central urban area, the mass concentrations of PM_2.5, PM₁₀, and NO₂ are relatively high, while O₃ concentrations are relatively low. In contrast, monitoring sites near the northern shore of Taihu Lake exhibit lower PM concentrations but higher O₃ levels. Additionally, O₃ demonstrates a weekend effect.

(4)

Correlation analysis indicates that PM concentrations exhibit significant correlations with other air pollutants. Temperature demonstrated moderate negative correlations with PM and NO₂, while exhibiting a strong positive correlation with O₃. Both wind speed and precipitation showed weak to moderate negative correlations with pollutant concentrations. Relative humidity displayed a particularly strong negative correlation with SO₂, along with moderate negative correlations with PM₁₀ and O₃.

(5)

PSCF and CWT analyses showed high consistency in identifying major potential source regions. In spring, the dominant potential source regions of O₃ and PM_2.5 in Wuxi were concentrated in the Suzhou–Wuxi–Changzhou metropolitan area, southern Nanjing, Zhenjiang, and northern Zhejiang Province. During summer, high-potential source areas of O₃ were mainly distributed around Taihu Lake, including Changzhou, Wuxi, Suzhou, and Huzhou, while southeastern cities such as Jiaxing, Ningbo, Huzhou, and Shanghai also contributed to O₃ pollution transport in Wuxi. In autumn, the dominant potential sources of O₃ were located in the eastern part of Wuxi. In winter, the major source regions of PM_2.5 originated from central and northern Jiangsu Province.

By integrating multi-source observational data with model simulations, we established a causal framework for complex air pollution in this typical YRD industrial city. The findings provide critical scientific evidence to support regional coordinated air quality management strategies.