Analysis of the Effect of Vegetation Types in Industrial Heat Source Radiation Areas on PM_2.5 Concentration Reduction in the Beijing–Tianjin–Hebei Region

Highlights

What are the main findings?

• This study proposes a systematic analysis approach based on industrial heat source radiation areas, PM_2.5, land cover, and digital elevation model data to systematically evaluate the PM_2.5 reduction effects of different vegetation types within industrial heat source radiation areas.
• Validated in the Beijing–Tianjin–Hebei region, the results indicate that vegetation can significantly reduce PM_2.5 in industrially influenced areas, with varying effects among different vegetation types.

What are the implications of the main findings?

• They provide a reference for ecological buffer design and urban ecological planning in industrial zones.
• They offer a scientific basis for vegetation optimization, emission mitigation, and sustainable air quality management.

Abstract

Industrial heat source (IHS) radiation areas are key accumulation zones for fine particulate matter with an aerodynamic diameter of 2.5 μm or less (PM_2.5) in heavily industrialized regions, but PM_2.5 concentration reduction rates by vegetation types have not been systematically assessed. A systematic analysis of PM_2.5 concentration reduction rates by vegetation types using IHS, PM_2.5, land-cover, and digital elevation model (DEM) data was conducted to assess PM_2.5 concentration reduction rates by vegetation types within IHS radiation areas. First, this study adopted the published IHS radiation-area dataset developed by Xin Sui et al. to define the spatial extent of industrially influenced areas. Second, PM_2.5 concentrations were extracted within IHS radiation areas and areas covered by different vegetation types to support the calculation of PM_2.5 concentration reduction rates. Third, PM_2.5 concentration reduction rates by vegetation types were evaluated through masking and regional statistical analysis. Results for Beijing–Tianjin–Hebei (BTH) in 2015 and 2020 show that: (1) the average PM_2.5 decreased from 71.70 to 39.60 µg/m³, corresponding to an overall reduction of 44.8%; (2) PM_2.5 concentration reduction rates varied substantially among vegetation types; open deciduous broadleaved forest showed the highest reduction rate of 39.08%, while rainfed and irrigated cropland showed negative reduction rates of −9.35% and −6.71%; (3) city-scale and case analyses show denser vegetation in radiation zones generally lowers PM_2.5 even under ongoing industrial activity. The study supports vegetation greening, IHS control, regional air quality improvement, and sustainable industrial development strategies.

1. Introduction

Particulate matter with a diameter of 2.5 micrometers or less, commonly referred to as PM_2.5, has significant adverse effects on the atmospheric environment, climate change, and human health [1,2,3]. When these particles enter the human bronchial system, they can severely irritate the respiratory tract and significantly impair pulmonary gas exchange, causing symptoms such as coughing, difficulty breathing, asthma, and bronchitis [4]. Industrial heat source (IHS) emissions are one of the main contributors to regional PM_2.5 pollution [5]. These emissions originate from high-energy-consumption industries such as cement plants, steel mills, coal chemical plants, and oil and gas development, forming concentrated radiation zones. At the same time, vegetation, as a critical component of urban ecosystems, can regulate urban microclimates and significantly reduce PM_2.5 concentrations [6]. Therefore, clarifying and assessing the role of different vegetation types in reducing industrial PM_2.5 has important implications for environmental improvement.

Surface vegetation is an important factor influencing PM_2.5 concentrations. It can reduce atmospheric PM_2.5 levels and effectively improve air quality, as confirmed by numerous studies [7,8]. For example, Wang et al. [8] indicated that forest vegetation can reduce atmospheric PM_2.5 concentrations by covering the ground, reducing dust emissions, and adsorbing particles on leaves. Li et al. [9] used curve regression analysis to examine the spatiotemporal distribution of PM_2.5 and its relationship with six land-use types, finding that urban expansion and increased cropland area contribute to higher PM_2.5 concentrations, whereas increased forest area helps reduce PM_2.5 levels. Zhuo [10], using 2021 PM_2.5 concentration, the Normalized Difference Vegetation Index (NDVI), and land-use data in Fujian Province, applied continuous buffer analysis and found that forests and water bodies significantly decrease PM_2.5 concentrations, while built-up land, bare land, and sparse vegetation increase them. Feng et al. [11] employed Sen trend analysis, Mann–Kendall significance testing, and Pearson correlation to investigate the impact of vegetation cover on the spatiotemporal distribution of PM_2.5, revealing a negative correlation between the two. Zhao et al. [12], using geographic information system (GIS) spatial analysis, generalized additive models, and geographically weighted regression, studied PM_2.5 responses to land-use changes and found that increased forest and grassland areas significantly reduced PM_2.5 concentrations. Zhang et al. [13] quantified contributions of different land-use types to seasonal PM_2.5 variations using boosted regression trees, showing that built-up areas are the main drivers of PM_2.5 in autumn and winter. In summary, vegetation cover can effectively reduce atmospheric PM_2.5 concentrations and improve air quality. However, studies focusing on the relationship between vegetation cover and PM_2.5 within IHS radiation areas remain relatively scarce, especially concerning quantitative comparisons of PM_2.5 concentration reduction rates among different vegetation types. Current studies still have three main limitations: limited quantitative assessment of PM_2.5 concentration reduction rates by vegetation type within existing IHS radiation areas, insufficient integration of IHS radiation-area data with land-cover information, and insufficient long-term analyses that overlook policy-driven vegetation impacts on PM_2.5 dynamics. Therefore, accurately analyzing the spatiotemporal variations in PM_2.5 within IHS radiation areas and quantifying PM_2.5 concentration reduction rates by vegetation type are crucial for optimizing emission control policies and improving regional air quality.

To address the above research gaps, this study takes the BTH IHS radiation areas as the research object and integrates multi-source remote sensing and geospatial data to conduct a systematic analysis of PM_2.5 concentration reduction rates by vegetation types. First, this study adopted the published IHS radiation-area dataset developed by Xin Sui et al. to define the spatial extent of industrially influenced areas. Second, PM_2.5 concentrations were extracted within IHS radiation areas and areas covered by different vegetation types to support the calculation of PM_2.5 concentration reduction rates. Third, PM_2.5 concentration reduction rates by vegetation type were evaluated through masking and regional statistical analysis. Compared with previous studies that mainly focused on general urban areas, broad land-use categories, or vegetation indices, this study further links IHS radiation areas, PM_2.5 concentrations, and detailed land-cover classes to quantify vegetation-associated PM_2.5 differences in industrially influenced areas. The results provide more targeted evidence for vegetation planning and air pollution mitigation in industrial regions.

The study is structured as follows: Section 2 introduces the study area, data sources, and methods for calculating PM_2.5 reduction rates; Section 3 analyzes the spatiotemporal variations in PM_2.5 and the PM_2.5 concentration reduction rate by vegetation types; Section 4 explores the impact of industrial activity changes through case studies of continuously operating, discontinued, and newly established industrial plants; Section 5 summarizes the findings and discusses their implications and limitations.

2. Materials and Methods

2.1. Study Area

The BTH region is situated on the North China Plain (36.0°N to 42.6°N, 113.5°E to 119.8°E), encompassing two municipalities directly under the central government (Beijing and Tianjin) and 11 prefecture-level cities in Hebei Province [14] (Figure 1). Home to over 8% of China’s total population [15], the BTH region serves as a core economic growth pole in northern China, one of the most industrialized regions in the country, and an important national base for energy and steel production [16]. Dense energy-intensive industries within the region have formed large-scale IHS radiation pollution zones. The BTH region features complex topography that slopes from northwest highlands to southeast lowlands, along with relatively stable meteorological conditions; these factors hinder the dispersion of pollutants and further exacerbate regional air pollution [17], with pollutant emissions reaching four times the national average [15,18]. In 2022, the annual average PM_2.5 concentration reached 44 μg/m³, 16 μg/m³ higher than the national average, rendering air pollution control an urgent task [19].

2.2. Data Sources

Multi-source data, including PM_2.5 concentrations, land cover, and IHS distribution, were collected for this study. The years 2015 and 2020 were selected to represent two stages of China’s air pollution control process. The year 2015 represents an early stage after the implementation of the Air Pollution Prevention and Control Action Plan issued in 2013, which aimed to improve air quality and reduce heavy pollution in key regions, including the Beijing–Tianjin–Hebei region. The year 2020 represents the final assessment year of the Three-Year Action Plan for Winning the Blue Sky Defense Battle, which was issued in 2018 and set a timeline and roadmap for further air pollution control. Therefore, comparing 2015 and 2020 allows this study to evaluate changes in PM_2.5 concentrations and vegetation-associated PM_2.5 concentration reduction rates across a period of intensified air pollution control. Although 2013 is an important policy starting point, it was not selected because this study focuses on post-policy changes and requires consistent PM_2.5, land-cover, and IHS datasets for comparative analysis. All data were standardized to the World Geodetic System 1984 (WGS-1984) coordinate system [20] to ensure spatial consistency. Detailed information on the datasets is shown in

Table 1. Summary of data sources and their specifications used in this study.

No.	Variable	Dataset	Period	Spatial Resolution	Temporal Resolution	Data Source
1	PM2.5 Concentration (PM2.5)	Comprehensive High-Resolution Air Pollution	2015, 2020	1 km	Annual	National Tibetan Plateau/Third Pole Environment Data Center (https://doi.org/10.5281/zenodo.3539349, accessed on 1 January 2026)
2	Land Cover (LC)	GlobeLand30	2015, 2020	30 m × 30 m	Annual	(https://www.cbas.ac.cn/yjjz/202403/t20240327_489908.html, accessed on 1 January 2026)
3	Industrial Heat Source Radiation Area (IHS)	A dataset of in-operation industrial heat source objects in BTH	2012–2021	1 km × 1 km	Annual	Science Data Bank (https://doi.org/10.57760/sciencedb.j00001.00430, accessed on 1 January 2026)

.

2.2.1. PM_2.5 Concentration Data

The PM_2.5 concentration data were obtained from the China High-Resolution High-Quality Ground-Level Air Pollution (CHAP) dataset (https://zenodo.org/records/6398971, accessed on 1 January 2026). The dataset is renowned for its long-term, full coverage, high spatial resolution (1 km), and high-quality data on ground-level air pollutants in China [21,22,23]. It is generated using artificial intelligence from big data sources, including ground-based strategies, satellite remote sensing products, atmospheric reanalysis, and model simulations, while considering the spatial and temporal heterogeneity of air pollution. The dataset has a temporal resolution of one year and a spatial resolution of 1 km, with units in µg/m³.

For this study, data from 2015 and 2020 were selected, with the coordinate system being WGS-1984. The data used Kriging interpolation to fill the spatial gaps in the satellite Moderate Resolution Imaging Spectroradiometer Multi-Angle Implementation of Atmospheric Correction Aerosol Optical Depth products, resulting in seamless nationwide ground-level PM_2.5 data.

2.2.2. Land-Cover Data

The land-cover data are sourced from the GlobeLand30 data product, specifically the 2020 release of the global 30 m fine land-cover classification product (GLC_FCS30) (https://data.casearth.cn/thematic/glc_fcs30/37, accessed on 1 January 2026). These data monitor land cover and its dynamic changes in global terrestrial areas (excluding Antarctica) from 1985 to 2022, with a spatial resolution of 30 m [24]. It is the longest-duration and most finely classified global 30 m land-cover remote sensing product internationally. The GLC_FCS30 uses a detailed classification system covering 35 land-cover categories, including grass and tree vegetation, sparse vegetation, and open deciduous broadleaf forests. In the GLC_FCS30 dataset, fc denotes vegetation cover fraction and is used to distinguish open and closed vegetation classes.

Due to the high spatial resolution and long temporal coverage of the GLC_FCS30 dataset, the 2015 and 2020 land-cover data were selected for this study. To support the analysis of PM_2.5 concentration reduction rates by land-cover and vegetation type, 12 land-cover classes were retained for the comparative analysis: open deciduous broadleaved forest, closed deciduous broadleaved forest, closed evergreen needle-leaved forest, sparse vegetation, closed deciduous needle-leaved forest, grassland, shrubland, open evergreen needle-leaved forest, bare areas, herbaceous cover, irrigated cropland, and rainfed cropland. Among them, bare areas were included as a non-vegetated comparison class, while wetland and water body classes, with labels 180 and 210, respectively, were excluded because this study focuses on terrestrial land-cover types within IHS radiation areas. These selected classes were used consistently in the subsequent zonal statistics and PM_2.5 concentration reduction-rate analysis. Figure 2 shows the spatial distribution of GLC_FCS30 land-cover classes in the BTH region in 2020.

2.2.3. Industrial Heat Source (IHS) Data

The IHS dataset is sourced from the study by Ma et al. (https://doi.org/10.57760/sciencedb.j00001.00430, accessed on 1 January 2026) [25]. In the original dataset, Xin Sui et al. [26] used an improved adaptive K-means clustering algorithm to identify IHS objects, which were further verified using high-resolution remote sensing imagery and point of interest (POI) data. The surrounding IHS radiation areas were then delineated using a region-growing algorithm. The dataset covers the BTH region from 2012 to 2021 with a spatial resolution of 1 km × 1 km and provides spatial information on IHS objects and their surrounding affected areas. The total area of the IHS radiation areas is 7287 km². In this study, the published IHS radiation areas were directly adopted as the spatial basis for subsequent PM_2.5 extraction, land-cover analysis, and evaluation of PM_2.5 concentration reduction rates by vegetation type.

2.3. Method

A systematic analytical framework (Figure 3) was established to evaluate PM_2.5 concentration reduction rates by vegetation type within IHS radiation areas of the BTH region from 2015 to 2020. Multi-source data were uniformly preprocessed through coordinate-system standardization, spatial-resolution unification, missing-value treatment, and outlier removal. The published IHS radiation-area dataset delineated by Xin Sui et al. was adopted as the spatial basis for defining industrially influenced areas. A total of 12 selected land-cover and vegetation classes were used for the comparative analysis, including open deciduous broadleaved forest, closed deciduous broadleaved forest, closed evergreen needle-leaved forest, sparse vegetation, closed deciduous needle-leaved forest, grassland, shrubland, open evergreen needle-leaved forest, bare areas, herbaceous cover, irrigated cropland, and rainfed cropland. Mask extraction and zonal statistics were then used to extract PM_2.5 concentrations within IHS radiation areas and selected land-cover and vegetation classes. Impervious surfaces were used as the reference baseline to calculate PM_2.5 concentration reduction rates by vegetation type, thereby supporting the quantitative assessment of vegetation-type differences and spatiotemporal variations. The detailed research steps are described in the following sections.

2.3.1. Data Preparation

In this study, both PM_2.5 concentration data and land-cover data were standardized to the WGS_1984 coordinate system. Subsequently, the data for the target area were clipped based on the BTH administrative boundaries. Due to differences in data sources, the temporal resolution of all datasets was unified to an annual scale, while the spatial resolution was consistently set to 30 m × 30 m. To ensure the completeness and reliability of the PM_2.5 data, outliers and missing values were systematically excluded. Finally, the Kriging interpolation method was used to fill in the gaps in the PM_2.5 concentrations. All spatial preprocessing and analyses, including coordinate-system transformation, raster clipping, spatial-resolution unification, mask extraction, Kriging interpolation, and zonal statistics, were performed using ArcMap 10.8 (Esri, Redlands, CA, USA).

2.3.2. Extraction of PM_2.5 Concentrations Within IHS Radiation Areas and Selected Land-Cover Classes

First, the IHS radiation-area data delineated by Xin Sui et al. using a region-growing algorithm based on IHS objects and DEM data were adopted as the spatial basis for defining industrially influenced areas. The original delineation considered the spatial diffusion characteristics of industrial emissions and terrain constraints [25]. Based on these data, the IHS radiation areas were used as spatial masks to extract PM_2.5 concentration values from the annual PM_2.5 raster datasets for 2015 and 2020, thereby excluding the influence of PM_2.5 concentrations outside IHS radiation areas on the subsequent analysis. Subsequently, the land-cover raster data were spatially matched with the extracted PM_2.5 concentration data within IHS radiation areas, providing the data basis for calculating PM_2.5 concentrations under different land-cover and vegetation classes. Finally, zonal statistics was performed using each vegetation-type zone as a statistical unit to calculate the mean PM_2.5 concentration within each zone. Through the above operations, the PM_2.5 concentration values of the entire IHS radiation areas and the corresponding PM_2.5 concentration values of each specific vegetation type within these areas in 2015 and 2020 were systematically obtained, laying a foundation for the subsequent evaluation of PM_2.5 concentration reduction rate by vegetation types.

2.3.3. Calculation of PM_2.5 Concentration Reduction Rates by Vegetation Type

To quantify differences in PM_2.5 concentration reduction among vegetation types within IHS radiation areas, the PM_2.5 concentration reduction rate was calculated for each selected land-cover and vegetation class. This calculation provided a quantitative basis for comparing PM_2.5 concentration differences among different land-cover and vegetation classes within IHS radiation areas.

Impervious surfaces were selected as the reference baseline for PM_2.5 concentration, as they lack vegetation cover and cannot intercept or adsorb PM_2.5, thus objectively representing the background PM_2.5 pollution level without vegetation intervention [27]. The PM_2.5 concentration reduction rate of each vegetation type was calculated to quantitatively evaluate its purification capacity. The specific calculation steps are as follows: First, the basic data were the PM₂.₅ concentration values of the entire IHS radiation area and the corresponding values of each specific vegetation type within the area in 2015 and 2020 obtained in Section 2.3.2. Second, the relative reduction rate was quantitatively calculated using the following formula:

$${\mathrm{P}\mathrm{M}}_{2.5\mathrm{a}\mathrm{r}}={\displaystyle \frac{{\mathrm{V}}_0-\mathrm{V}}{{\mathrm{V}}_0}}\times 100\%$$

(1)

where PM_2.5ar represents the reduction rate of PM_2.5 concentration (%) of a specific vegetation type; V₀ is the mean PM_2.5 concentration (μg/m³) of impervious surfaces in the IHS radiation area. V is the mean PM_2.5 concentration (μg/m³) in the area covered by a specific vegetation type. When the reduction rate is positive, it indicates that the vegetation type has a reducing effect on PM_2.5. When the reduction rate is negative, it indicates that the PM_2.5 concentration in the vegetation area is higher than that of the impervious surface, meaning that the vegetation has not reduced PM_2.5 concentrations and may even contribute to higher PM_2.5 levels.

3. Results

3.1. Temporal Variation in PM_2.5 Concentrations in IHS Radiation Areas of the BTH Region (2015 vs. 2020)

As shown in Figure 4, PM_2.5 concentrations in IHS radiation areas of the BTH region exhibited a significant declining trend from 2015 to 2020. Overall, PM_2.5 concentrations displayed a spatial pattern of higher levels in the central–southern areas and lower levels in the northern regions. In 2015, pixel-level concentrations ranged from 29.45 µg/m³ to 113.95 µg/m³, with an average of 71.70 µg/m³; by 2020, the range decreased to 19.10 µg/m³–60.09 µg/m³, with an average of 39.60 µg/m³, representing an overall reduction of 44.8%. Spatially, industrial cities and surrounding areas in central–southern Hebei remained the high-concentration zones, but concentrations declined markedly, and the number of high-value pixels decreased. In contrast, northern and mountainous cities maintained relatively low PM_2.5 levels with minor changes. This trend reflects the effectiveness of national environmental protection policies and industrial emission reduction measures implemented between 2015 and 2020, indicating that the impact of IHS activities on PM_2.5 concentrations has gradually diminished and that industrial restructuring has achieved measurable results.

3.2. Spatial Differentiation of PM_2.5 Concentration Reduction at Multiple Scales: BTH Region, Provincial and Municipal Levels

Following the same methodology, city-level PM_2.5 concentration reduction rates by vegetation types were calculated for Beijing, Tianjin, and the 11 prefecture-level cities in Hebei Province, as shown in Figure 5. The average reduction rate of vegetation in each city was computed and classified based on the reduction rate.

From the spatial distribution map of vegetation’s PM_2.5 reduction rates in the BTH region in 2015 and 2020, it can be observed that over the five years, the vegetation’s ability to regulate air pollution in the region showed a significant trend of “overcoming weaknesses and improving overall.” In 2015, the vegetation’s PM_2.5 reduction pattern was clearly concentrated in the industrially dense areas of the western and eastern parts, where cities such as Shijiazhuang, Baoding, Handan, and Tangshan were regions with relatively better vegetation-related PM_2.5 reduction. These cities were also among the areas with the highest concentration of IHSs in the BTH region. However, the higher PM_2.5 concentration reduction rate by vegetation types indicated that, at that time, local vegetation coverage had certain pollution regulation capabilities. In contrast, Hengshui and Cangzhou showed poor vegetation-related PM_2.5 reduction effects, with negative reduction rates indicating that the vegetation in these areas not only failed to effectively reduce PM_2.5 but may have contributed to the accumulation of air pollution due to insufficient vegetation coverage and weak ecological functions, making them weak spots for PM_2.5 reduction in the BTH region in 2015.

By 2020, the vegetation’s PM_2.5 reduction ability in the BTH region generally improved. Shijiazhuang, Baoding, and Tangshan maintained high reduction rates, indicating that despite the continued industrial activities in these cities, their vegetation ecological functions had not degraded and they continued to show a strong ability to reduce. The weaker regions also showed improvements, with Langfang showing a noticeable increase in its reduction rate and Cangzhou’s reduction rate also increasing significantly, turning from negative to positive. This indicates that over these five years, the environment in the BTH region improved. PM_2.5 concentration control was effective, and vegetation played an effective role in absorbing the PM_2.5 produced by IHS.

3.3. PM_2.5 Concentration Reduction Rate by Vegetation Type in IHS Radiation Areas

As shown in Figure 6, there are significant differences in PM_2.5 concentration reduction rates by vegetation types within the IHS radiation areas in the BTH region. Among all vegetation types, deciduous broadleaved forest, evergreen needle-leaved forest, and sparse vegetation show the highest reduction rates, with most exceeding 30%. Specifically, in 2015, the reduction rate of open deciduous broadleaved forests reached 39.08%. This is mainly related to the morphology and structure of their leaves. Broadleaved forests typically have large and wide leaves, which can cover a larger area and provide more surface area to capture and adsorb air pollutants, including PM_2.5 particles. In addition, the leaves of broadleaved trees are usually soft and have many small structures on the surface, increasing the leaf surface area and enhancing their ability to adsorb pollutants. Moreover, broadleaved trees have higher evaporation rates and stronger transpiration, which helps increase air humidity and promotes particle deposition, contributing to the filtration of other harmful substances in the air. Their well-developed root systems also help stabilize the soil and reduce dust pollution.

On the other hand, the leaves of needle-leaved forests are usually smaller and narrower, resulting in a relatively smaller surface area. Although needle-leaved forests may have a large number of leaves, their particle-capturing ability is lower due to the smaller leaf surface area and weaker transpiration, making their overall purification capacity generally less significant than that of broadleaved forests. However, in the IHS radiation areas of the BTH region, needle-leaved forests and some shrubs, with their higher vegetation density, also exhibit strong PM_2.5 reduction capacity.

In contrast, vegetation types such as rainfed cropland and irrigated cropland are relatively simple, and since the BTH region is located in northern China, vegetation coverage is low, especially in winter. The reduction rate of cropland is negative, which aligns with actual conditions. The burning of straw during sowing and harvesting seasons, as well as dust generated during soil tillage, leads to an increase in PM_2.5 concentrations, making PM_2.5 levels in cropland higher than those on impervious surfaces. However, from 2015 to 2020, the negative reduction effect of cropland weakened, which is closely related to national policies, such as the ban on straw burning and the promotion of conservation tillage techniques, reducing PM_2.5 emissions associated with cropland. This led to the saturation of the vegetation’s adsorption effect, thereby decreasing the reduction effect.

Regarding the overall reduction effect, it shows a declining trend because PM_2.5 concentrations in the IHS radiation areas of the BTH region generally decreased from 2015 to 2020, resulting in lower PM_2.5 values on impervious surfaces. This trend is related to the active promotion of industrial green transformation and the application of environmental technologies, which have reduced pollution caused by industrial emissions to some extent, leading to a relative decrease in the reduction effect. In addition, the area of vegetation has decreased. Although some vegetation types show good reduction effects, the overall PM_2.5 concentration reduction rate by vegetation types is directly affected by the decrease in vegetation area. For example, the area of closed deciduous broadleaved forest was 23.69 km² in 2015, but decreased to 21.82 km² in 2020.

4. Discussion

4.1. Spatiotemporal Distribution of PM_2.5 Concentrations in Different Types of Industrial Plants in IHS Radiation Areas (2015–2020)

To comprehensively evaluate the spatiotemporal variations in PM_2.5 within the IHS radiation areas of the BTH region, this section categorizes IHSs into three types based on their operational status: continuously operating, discontinued, and newly established. The PM_2.5 distribution patterns of each category are analyzed separately. This classification allows for the identification of differences in PM_2.5 trends and vegetation-mediated reduction effects under varying levels of industrial activity, providing insights into the interactions between industrial emissions and ecological regulation. The analysis in this section is based on the description of the study area and data sources in Section 2, as well as the data processing procedures and PM_2.5 reduction-rate calculations presented in Section 3, offering empirical support for the detailed case studies in Section 4.1.1, Section 4.1.2 and Section 4.1.3 and establishing a scientific foundation for the recommendations on regional air quality improvement and vegetation planning presented in Section 5.

4.1.1. PM_2.5 Spatiotemporal Characteristics of Continuously Operating Industrial Plants

Yanjin Building Materials Group Co., Ltd., located in Langfang City, Hebei Province, China with geographic coordinates of 116°49′30″E, 40°1′0″N, is a cement plant (as shown in Figure 7a). The plant remained operational from 2015 to 2020, but the number of IHS fire points showed a continuous decline. As shown in Figure 7c,d, the PM_2.5 concentrations in the IHS radiation area fluctuated between 2015 and 2020. The average PM_2.5 concentration in 2015 was 79.900 µg/m³, and in 2020, it was 39.586 µg/m³. The PM_2.5 concentration in the region decreased by 50.46% in 2020 compared to 2015. As shown in Figure 7b, the land cover around the plant is primarily composed of impervious surfaces, with only small areas of rainfed cropland and irrigated cropland, which have relatively weak reduction capabilities for PM_2.5 and may even contribute to PM_2.5 generation. However, according to monitoring data, the PM_2.5 concentration in the region decreased significantly in 2020 compared to 2015. Although the vegetation in the area has limited reduction effects and may even contribute to the accumulation of PM_2.5, the reduction in the number of IHS fire points during the plant’s production process led to a significant decrease in PM_2.5 emissions, driving the reduction in local PM_2.5 concentrations. This demonstrates that IHSs are one of the important sources of PM_2.5 emissions. This change also resulted in Langfang City’s overall PM_2.5 concentration reduction rate by vegetation types being higher in 2020 than in 2015, making Langfang one of the cities in the BTH region with the most significant air quality improvement.

The Tangshan Baoliyuan Coking Co., Ltd., located in Tangshan, Hebei Province, China is a cement plant (as shown in Figure 8a). The plant was continuously in operation from 2015 to 2020, and the number of IHS fire points has steadily increased. As shown in Figure 8c,d, the PM_2.5 concentration in the IHS radiation area fluctuated between 2015 and 2020. The average PM_2.5 concentration in 2015 was 65.774 µg/m³, while in 2020, it was 40.191 µg/m³. The PM_2.5 concentration in 2020 decreased by 38.92% compared to 2015. As shown in Figure 8b, the land cover around the plant includes vegetation types such as open deciduous broadleaved forest (0.15 < fc < 0.4), closed evergreen needle-leaved forest (fc > 0.4), and closed deciduous needle-leaved forest (fc > 0.4), which have strong adsorption and settling effects on PM_2.5, effectively reducing the dispersion of PM_2.5 in the air. These efficient vegetation types largely offset the emissions pressure caused by increased IHS activities. Although it is one of the most industrially active cities in the BTH region, Tangshan’s vegetation consistently maintained a positive reduction rate for PM_2.5 in both 2015 and 2020. This has kept the overall vegetation-based PM_2.5 reduction rate in Tangshan at a high level among cities in the BTH region, contributing to generally good air quality in the city.

4.1.2. PM_2.5 Spatiotemporal Characteristics of Discontinued Industrial Plants

Hebei Tianyu Coal Chemical Power Co., Ltd., located in Handan City, Hebei Province, China has geographic coordinates of 114°2′30″E–114°6′30″E, 36°33′30″N–36°36′30″N. It is a coal mine/coal chemical plant (as shown in Figure 9a). The plant was operational in 2015 but ceased operations by 2020 (with zero IHS fire points). As shown in Figure 9c,d the PM_2.5 concentration in the IHS radiation area fluctuated between 2015 and 2020. The average PM_2.5 concentration in 2015 was 76.805 µg/m³, and in 2020, it was 50.386 µg/m³. The PM_2.5 concentration in the region decreased by 34.47% from 2015 to 2020. As shown in Figure 9b, the dominant land-cover classes in this area were rainfed cropland and herbaceous cover, both of which showed relatively weak PM_2.5 concentration reduction rates by vegetation types. This indicates that the shutdown of IHS emissions was likely a key factor contributing to the local decrease in PM_2.5 concentrations. The study suggests that even without relying on high-efficiency reduction vegetation, the shutdown of IHSs can still reduce PM_2.5 concentrations, and the overall PM_2.5 concentration reduction rate by vegetation types in Handan City can still be maintained.

4.1.3. PM_2.5 Spatiotemporal Characteristics of Newly Added Industrial Plants

Located in Qinhuangdao City, Hebei Province, China, Delong Foundry Development Co., Ltd., with geographic coordinates of 119°12′30″E–119°13′30″E, 40°28′30″N–40°29′30″N, is a steel plant (as shown in Figure 10a). The plant was inactive in 2015, with zero IHS fire points, and resumed production by 2020. As shown in Figure 10c,d, PM_2.5 concentrations within the IHS radiation area changed between 2015 and 2020. The average PM_2.5 concentration in 2015 was 49.675 µg/m³, and in 2020, it was 33.0637 µg/m³. The PM_2.5 concentration in the region decreased by 33.44% in 2020 compared to 2015. As shown in Figure 10b, the land-cover data of the region reveal that the vegetation types include closed deciduous broadleaved forest (fc > 0.4), closed evergreen needle-leaved forest (fc > 0.4), and closed deciduous needle-leaved forest (fc > 0.4). These types of vegetation have a strong reduction capacity for PM_2.5. In 2020, the PM_2.5 concentration in the region showed a decreasing trend compared to the 2015 concentration. Despite the presence of IHS emissions of PM_2.5 in 2020, local vegetation, especially broadleaved forest, significantly reduced PM_2.5 concentrations. This offset the impact of IHS emissions to a certain extent, such that the overall PM_2.5 reduction rate by vegetation in Qinhuangdao remained stable and consistently positive in 2020 relative to 2015, exerting a continuous positive purification effect.

4.2. Limitations of This Study

This study extracted the regional PM_2.5 concentration data of the BTH region in 2015 and 2020, and quantified the PM_2.5 concentration reduction by different vegetation types within the radiation areas of IHSs. The results indirectly demonstrate that the environmental quality in the IHS zones of the BTH region has been improved, providing a scientific basis for promoting the sustainable development of industry.

However, certain uncertainties still exist in this study. First, the PM_2.5 concentration data were not corrected for the interference of non-industrial factors such as meteorological conditions, temperature, straw burning, and agricultural activities, and thus cannot fully represent the PM_2.5 concentration solely contributed by IHSs. Second, in areas with overlapping complex IHSs, it is difficult to separate the specific PM_2.5 concentration generated by each individual heat source, which may lead to certain deviations in the calculated PM_2.5 concentration reduction rate by vegetation types. Third, this study directly adopted the published IHS radiation-area dataset developed by Xin Sui et al., in which the maximum influence distance was predefined as 10 km. To maintain consistency with the original dataset, this study did not modify this distance threshold. Therefore, no additional sensitivity analyses using other thresholds, such as 5 km and 15 km, were conducted, which may introduce uncertainty into the calculation of PM_2.5 concentration reduction rate by vegetation types. In addition, the CHAP dataset itself contains uncertainties, and the use of Kriging interpolation to fill missing values may further affect data accuracy.

Future research can be carried out on the following aspects. First, the PM_2.5 concentration data can be adjusted to exclude meteorological impacts, especially those from traffic emissions and other non-industrial sources, to more accurately quantify the contribution of IHSs to PM_2.5 pollution and improve the accuracy of reduction-rate calculations. Second, future studies should conduct multi-distance sensitivity analyses, such as using 5 km, 10 km, and 15 km influence thresholds, to further evaluate the effects of distance thresholds on IHS radiation-area delineation and PM_2.5 concentration reduction rates by vegetation types. Third, the study period can be extended by incorporating data after 2020 to analyze the long-term trends in PM_2.5 concentration reduction rates by vegetation types, so as to provide references for the optimization of environmental protection policies.

5. Conclusions

This study utilized multi-source data, including IHS information, PM_2.5 concentrations, land-cover data, and the IHS radiation-area dataset developed by Xin Sui et al., to analyze PM_2.5 concentration reduction rates by selected land-cover and vegetation classes within IHS radiation areas of the Beijing–Tianjin–Hebei region. A total of 12 selected land-cover and vegetation classes were included in the comparative analysis. The IHS radiation areas were used as spatial masks to extract PM_2.5 concentrations and corresponding land-cover classes. Quantitative analyses were then conducted to evaluate differences in PM_2.5 concentration reduction rates among these selected classes and to examine their spatiotemporal variation characteristics. Based on the above analysis, the main conclusions are as follows:

(1)

From 2015 to 2020, PM_2.5 concentrations in the IHS radiation areas of the Beijing–Tianjin–Hebei region exhibit an overall decreasing trend. Spatially, the pattern shifted from “higher PM_2.5 concentrations in the central–southern areas and lower concentrations in the northern areas” in 2015 to “marked improvement in the southern areas and steady improvement in the northern areas” in 2020, indicating reduced regional disparities and reflecting the effectiveness of industrial reform measures and regional environmental policies.

(2)

This study selected 12 land-cover and vegetation classes for comparative analysis; significant differences were observed in PM_2.5 concentration reduction rates among these selected classes. Open deciduous broadleaved forests achieve the highest reduction rate at 39.08%, whereas rainfed cropland and irrigated cropland show lower efficiencies, at −9.35% and −6.71%, respectively. These results indicate that vegetation structure and density are critical determinants of PM_2.5 mitigation, with high-density and structurally complex forest types demonstrating greater potential for particulate matter removal through processes such as particle adsorption, deposition, and microclimate regulation.

(3)

Case studies in industrial cities such as Handan and Tangshan demonstrate that efficient vegetation can partially mitigate PM_2.5 concentrations even under continuous industrial activity.

The results not only provide a novel perspective on the dynamics of regional air pollution and the environmental impacts of industrial emissions, but also demonstrate significant differences in PM_2.5 concentration reduction rate among selected land-cover and vegetation classes within IHS radiation areas, offering a scientific basis for regional air pollution control and vegetation greening planning. Future efforts should continue to prioritize the cultivation and protection of forest vegetation, with particular emphasis on high-efficiency types such as open deciduous broadleaved forests, to further enhance the ecological resilience of IHS radiation areas and promote sustainable and fundamental improvements in regional air quality.