A Study of the Impact of Industrial Land Development on PM_2.5 Concentrations in China

Abstract

To promote the sustainable use of land resources and improve air pollution control, this study investigates the spatiotemporal dynamics of industrial land development and the heterogeneity of PM_2.5 concentrations across regions. Based on national land transaction data and PM_2.5 raster datasets, the analysis employs Moran’s I, a hot and cold spot analysis, and multivariate linear regression to examine how the transaction frequency, transaction area, and total transaction price of industrial land influence PM_2.5 concentrations in 286 cities from 2010 to 2021. The study focuses on quantifying the impact of industrial land development on PM_2.5 concentrations. The main findings are as follows: (1) the frequency of industrial land transactions varies significantly across regions, with clear intra-regional differences. The transaction area and total transaction price decrease in the following order: “East-West-Central-North-East” and “East-Central-West-North-East”, respectively. (2) The spatial clustering of PM_2.5 concentrations has intensified, with hot spots concentrated in Eastern and Central cities. Cold spots are distributed in bands along the Southern coast and scattered patterns in Heilongjiang Province. (3) The influence of industrial land development on PM_2.5 concentrations has generally weakened nationwide, with the strongest effects observed in the Eastern region. Among the development indicators, the impact of the transaction area is increasing, while those of the transaction frequency and total price are declining, showing clear regional disparities. Therefore, integrating sustainable development principles into the adjustment of the industrial land market is essential for effective air pollution prevention.

1. Introduction

Fine particulate matter (PM_2.5) refers to airborne particles with an aerodynamic diameter of less than 2.5 μm. Long-term exposure poses serious health risks, particularly to the human respiratory and cardiovascular systems [1,2]. Elevated PM_2.5 concentrations significantly impact both socio-economic systems and global ecosystems. These effects include intensifying the urban heat island phenomenon—a process where urban areas experience higher temperatures than surrounding rural areas due to heat absorption by buildings and pavement [3], reducing the carbon sequestration capacity of agroecosystems [4], and disrupting the Earth–atmosphere radiation balance [5]. Consequently, PM_2.5 pollution has become a major focus in global air pollution control efforts. There is a growing consensus that industrial production, mining emissions, and energy consumption are the primary sources of PM_2.5 pollution [6,7,8,9]. With rapid industrialization and surging energy and resource demand in China, the challenge of air pollution control has become increasingly severe. According to the 2021 Bulletin of China’s Ecological and Environmental Conditions, 121 out of 339 cities failed to meet national air quality standards. Given the strong link between industrial development and GDP growth, local governments in China have leveraged their authority over land markets to allocate more land to industrial purposes [10]. While the government-led allocation of land resources promotes economic development and urbanization, it also intensifies pressure on environmental and resource systems. Industrial land serves as a key platform for production activities and industrial restructuring, involving substantial material and energy consumption [11,12]. Therefore, its development significantly affects the sustainable use of urban land and the quality of the atmospheric environment. Investigating the impact of industrial land development on the PM_2.5 concentration is practically significant for advancing market-oriented factor reforms, promoting sustainable urban land use, and optimizing government land allocation policies from a land economics perspective to address air pollution. Conventional studies often rely on land use classifications, such as comparing industrial versus residential zones, to assess their impact on PM_2.5 pollution. However, these methods frequently overlook the dynamic nature of land use changes and fail to capture market-driven industrial expansion. In contrast, this study introduces market-based indicators—the transaction frequency, area, and price—as novel metrics of industrial growth that more sensitively reflect the dynamic link between economic activity and environmental pollution.

Current research on PM_2.5 primarily focuses on five areas: sources [13,14], chemical composition [15,16], spatial–temporal characteristics and driving factors [17,18,19], model-based estimation and prediction [20,21,22], and governance strategies [23,24]. Few studies specifically examine the relationship between industrial land development and PM_2.5 concentrations. Existing research that touches on this topic often approaches it from the perspective of land use and land use types [25,26,27]. The existing literature can be broadly categorized into studies on direct and indirect impacts. Direct impact studies primarily explore factors such as the consumption structure [28,29], industrial agglomeration [30,31,32], and urbanization [33,34,35]. Indirect impacts are examined through land resource mismatches [10,36,37], the industrial structure [38,39], and fiscal decentralization [40,41,42]. However, most previous studies have relied on land use data to assess the scale and development of industrial areas. These studies often overlook the role of land-based economic activities in air pollution and focus solely on economic output, neglecting the input–output dynamics of industrial land within the market economy. In practice, industrial land functions not only as a key production factor but also as a strategic tool for local governments to attract investment [43]. Industrial land plays a vital role in economic activities, reflected in its frequency of circulation, transaction scale, and value realization. It directly influences the local industrial production efficiency, scale, and development quality and indirectly impacts the air quality. Despite increasing attention to urban air pollution and land use, existing studies have rarely provided a systematic analysis of how industrial land development affects PM_2.5 concentrations across both time and space. Most focus on static land classifications or general urbanization indicators, overlooking the role of dynamic land market activities.

To address this gap, this study investigates the spatiotemporal evolution of industrial land development and PM_2.5 concentrations in China from 2010 to 2021. It further explores how this relationship varies across Eastern, Central, Western, and Northeastern regions, identifying key influencing factors and their temporal dynamics. These insights aim to deepen our understanding of how market-oriented land development contributes to environmental outcomes and to offer targeted policy recommendations for more sustainable industrial planning. This study offers two main contributions. First, by utilizing national land transaction data, the study assesses the current state of industrial land circulation and comprehensively evaluates its development in terms of the transaction frequency, area, and price. Second, from a regional perspective, the study applies a multiple linear regression model to examine the impact of industrial land development on PM_2.5 concentrations. It explores the relationships among land supply behavior, air pollution, and regional heterogeneity, thereby offering a theoretical foundation for region-specific policymaking.

2. Materials and Methods

2.1. Study Area

In 2021, China’s gross domestic product (GDP) is expected to reach 114 trillion yuan, with 59 trillion yuan in the Eastern region, 25 trillion yuan in the Central region, 24 trillion yuan in the Western region, and 6 trillion yuan in the Northeastern region. This shows clear regional disparities in economic development. In terms of the land market supply, the total supply of state-owned construction land in China in 2021 will be 690,000 hectares, a 61.2% increase from 2010. Of this, 175,000 hectares (25.8%) will be allocated to industrial, mining, and warehousing land. As the industrial scale expands, China’s total energy consumption in 2021 will reach 5.24 billion standard tons of coal, an increase of 61% compared to 2010. Coal consumption will account for 56%, while clean energy consumption will make up 25.5%. Additionally, the PM_2.5 concentration will decrease by 9.1% year-on-year. This indicates that the government is gradually regulating the allocation of industrial land in energy-efficient and environmentally friendly industrial production areas to achieve air pollution control over time. The overview of China’s region is shown in Figure 1.

2.2. Research Methodology

The main process of this study includes data collection, spatial and temporal dynamic analysis, and an analysis of the impact of industrial land development on the PM_2.5 concentration and finally puts forward relevant policy and optimization recommendations; the workflow is shown in Figure 2.

2.2.1. Spatial Autocorrelation

Tobler’s first law of geography states that spatial correlation exists for anything and the closer things are to each other, the higher the degree of spatial correlation [44]. Spatial autocorrelation analysis is used to measure the spatial distribution characteristics of the PM_2.5 concentration, i.e., whether the PM_2.5 level in a certain area is affected by neighboring areas. In this paper, Global Moran’s I (Global Moran’s Index) is used to test the degree of spatial agglomeration of the PM_2.5 concentration, the value of which ranges from −1 to 1. Positive values indicate a positive spatial correlation and negative values indicate a negative spatial correlation, and the larger the absolute value is, the stronger the spatial correlation, and vice versa, the weaker it is [45]. The formula is as follows:

$$\mathrm{I}={\displaystyle \frac{\mathrm{n}}{{\displaystyle {\sum }_{\mathrm{i}=1}^{\mathrm{n}}{\displaystyle {\sum }_{\mathrm{j}=1}^{\mathrm{n}}{W}_{ij}}}}}\times {\displaystyle \frac{{\displaystyle {\sum }_{\mathrm{i}=1}^{\mathrm{n}}{\displaystyle {\sum }_{\mathrm{j}=1}^{\mathrm{n}}{W}_{ij}\left(x_i-\overline{x}\right)\left(x_j-\overline{x}\right)}}}{{\displaystyle {\sum }_{i=1}^N{\left(x_i-\overline{x}\right)}^2}}}$$

(1)

where I is the global spatial autocorrelation index, n is the number of prefectures, x_i, and x_j are the sample values of prefecture i and prefecture j, respectively, x is the mean of the sample values, and W_ij is the spatial weight.

2.2.2. Cold and Hot Spot Analysis

The coefficient, the highest proposed by Getis and Ord, is commonly used for cold hotspot analysis, which identifies the degree of aggregation of the spatial distribution of attribute values of things and the clustering characteristics of their location [46]. This study applies cold hotspot analysis to identify and analyze the spatial distribution of the transaction area (TA) and the total transaction price (TTP) of industrial land, which is also used to analyze the spatial distribution pattern of the PM_2.5 concentration and to explore whether there is any significant spatial clustering of the market activities of industrial land and PM_2.5. It is calculated as follows:

$${\mathrm{G}}_{\mathrm{i}}^{*}={\displaystyle \frac{{\displaystyle {\sum }_{j=1}^n{w}_{i,j}{x}_j-\frac{1}{n}}{\displaystyle {\sum }_{j=1}^n{x}_i}{\displaystyle {\sum }_{j=1}^n{w}_{i,j}}}{\sqrt{\frac{1}{n}{\displaystyle {\sum }_{j=1}^n{x}_j^2-\frac{1}{n^2}}\left({\displaystyle {\sum }_{j=1}^n{x}_j}\right)}\times \sqrt{\frac{n}{n-1}{\displaystyle {\sum }_{j=1}^n{w}_{i,j}^2-\frac{1}{n-1}}{\left({\displaystyle {\sum }_{j=1}^n{w}_j}\right)}^2}}}$$

(2)

where ${\mathrm{G}}_{\mathrm{i}}^{*}$ is the local autocorrelation index of sample i, w_ij denotes the spatial weight, n is the total number of samples, and x_i and x_j are the observed values of sample i and sample j, respectively. When ${\mathrm{G}}_{\mathrm{i}}^{*}$ = 0, it means that the sample is the result of random generation; when ${\mathrm{G}}_{\mathrm{i}}^{*}$ > 0, it means that the sample is a positively aggregated hotspot, indicating that the PM_2.5 concentration is high in the region and that the surrounding areas also have high concentrations, forming a PM_2.5 hotspot; if applied to the industrial land indicator, it means that industrial land development is highly concentrated in the region, forming an industrial land development hotspot. When ${\mathrm{G}}_{\mathrm{i}}^{*}$ < 0, it means that the sample is a negatively aggregated cold zone indicating that the PM_2.5 concentration in the region is low and the surrounding area is also low, forming a PM_2.5 cold zone; if applied to the industrial land indicator, it means that the level of industrial land development in the region is low, forming a cold zone for industrial land development. The significance test ${\mathrm{G}}_{\mathrm{i}}^{*}$ will be divided into 99%, 95%, and 90% cold and hot spot zones and no statistical significance zones.

2.2.3. Multiple Linear Regression Analysis

Multiple linear regression is a statistical analysis method that establishes a linear quantitative equation between multiple variables and analyzes it using sample data. This study utilizes this method to quantify the extent to which industrial land development affects PM_2.5 concentrations and to control for other factors that may affect PM_2.5 concentrations [47]. In this paper, the following model is constructed by combining existing studies [48,49,50,51]:

$$\mathrm{Y}=\mathrm{C}+a_1{x}_1+a_2{x}_2+a_3{x}_3+Control(b_1{x}_4+b_2{x}_5\dots +b_7{x}_{10})$$

(3)

where the explanatory variable Y is the value of the PM_2.5 concentration; the core explanatory variables x₁, x₂, and x₃ represent the transaction frequency (TF), transaction area (TA), and total transaction price (TTP) in the development of industrial land in turn; the control variables x₄–x₁₀ represent the regional gross domestic product (GDP), the urban industrial structure (IND), the amount of public fixed asset investment (FI), the length of road (Road), the built-up area (UA), the local financial gap (FG), and the number of permanent residents at the end of the year (Pop); C is a constant term, and the rest are impact coefficients. Before conducting the regression estimation, we performed multicollinearity diagnostics by calculating Variance Inflation Factors (VIFs) for all explanatory and control variables. All VIF values were found to be less than 5, indicating that multicollinearity is not a concern in our models. To normalize the distribution of variables and reduce heteroscedasticity, we applied log-transformation to all continuous variables. This includes the GDP, fixed asset investment, transaction frequency, transaction area, and total price. The transformation helped to stabilize variances and improve model linearity.

2.3. Data Sources

The industrial land data are sourced from all industrial land transactions on the China Land Market website (https://www.landchina.com/, accessed on 7 February 2025) from 2010 to 2021, totaling 469,071 transactions. These include information on the region, latitude and longitude, transaction area, price, date, and more. The transaction area ranges from 0.64 hm² to 3577.16 hm², with land categorized as either existing or new construction and the land use type being industrial. The data on the average annual PM_2.5 concentration are sourced from the National Earth System Science Data Centre (http://geodata.nnu.edu.cn/, accessed on 7 February 2025). Seven control variables, including the GDP and road length, were obtained from the 2010–2021 China Urban Statistical Yearbook and China Urban Construction Statistical Yearbook, supplemented by city-specific statistical yearbooks. Missing data were interpolated where necessary. In view of the availability of data, this paper does not collect statistics from Hong Kong, Macao, Taiwan Province of China, and Tibet, as shown in

Table 1. Data sources and calculations.

Variable	Name	Calculation Method	Attributes	Data Source
Explanatory variable	PM2.5 concentration	Arcgis 10.5 Extraction		National Earth System Science Data Center (http://geodata.nnu.edu.cn/, accessed on 7 February 2025)
Core explanatory variables	Transaction Frequency	Summarize and organize	+	China Land Market Network (https://www.landchina.com, accessed on 7 February 2025)
	Transaction Area	Summarize and organize	+
	Total Transaction Price	Summarize and organize	+
	Gross Regional Product	Direct access	+	China City Statistical Yearbook and China Urban Construction Statistical Yearbook (https://data.cnki.net/yearBook, accessed on 7 February 2025)
	City Industrial Structure	Value Added of Secondary Industry/Value Added of Tertiary Industry	+
	Public Fixed Asset Investment	Direct access	+
Control variable	Length of Road	Direct access	+
	Size of Built-up Area	Direct access	+
	Local Financial Gap	(Current year’s fiscal revenue—Current year’s fiscal expenditure)/Per capita GDP of the previous year	+
	Year-end Resident Population	Direct access	+

.

3. Results

3.1. Analysis of the Temporal and Spatial Dynamics of Industrial Land Development

3.1.1. Analysis of the Spatial and Temporal Dynamics of the Frequency of Industrial Land Transactions

As shown in Figure 3, from 2010 to 2021, the frequency of industrial land transactions first increased and then decreased, with significant fluctuations, peaking at about 54,000 transactions in 2013, and then declining to about 600,000 transactions in 2021. This decline was primarily due to the slowdown in secondary industry development caused by the impact of the COVID-19 pandemic, which directly affected the land market. Observing the four economic regions, the frequency of industrial land transactions varied significantly across regions during the study period, with clear differentiation among provinces. The frequency decreased in the order of “east-central-west-northeast.” At the provincial level, the top five transaction frequencies are in Jiangsu, Shandong, Zhejiang, Hebei, and Anhui provinces, while the bottom five are Ningxia, Shanghai, Qinghai, Beijing, and Hainan. This is mainly due to differences in regional development positioning, the coverage of economic regions, and the industrial structures of provinces. For example, the Eastern region covers the second-largest number of provinces, following the Western region, and was prioritized in the “Twelfth Five-Year Plan” for leading the country’s economic development. Jiangsu has the largest manufacturing cluster in the country, while Hainan’s economy is primarily based on tourism and service industries.

3.1.2. Analysis of the Temporal and Spatial Dynamics of the Area of Industrial Land Transactions

As shown in Figure 4, the national industrial land transaction area during the study period follows a trend similar to the transaction frequency, peaking at approximately 216,000 hectares in 2013 and declining to a minimum of about 25,000 hectares in 2021. Observing the four major economic regions, industrial land in each region shows a decreasing trend of “East-West-Central-Northeast” during the study period, which differs from the trend in the industrial land transaction frequency. At the provincial level, the top five in terms of the transaction area are Shandong, Jiangsu, Xinjiang, Hebei, and Hubei, while the bottom five are Qinghai, Tianjin, Shanghai, Beijing, and Hainan. This is primarily due to the vast size of the Western region, where, with the advancement of urbanization, the area of industrial land per transaction is higher than in the Central region. Additionally, land resource endowment varies greatly across provinces and cities, with different land-use patterns focusing on different areas.

Using industrial land transaction data and the ArcGIS 10.5 platform to generate spatial points, four time nodes—2010, 2014, 2018, and 2021—were selected for analysis. The spatial aggregation of industrial land transaction areas was then examined using the Getis-Ord Gi* tool. As shown in Figure 5, in 2010, point clusters were concentrated in the Eastern region, while hot spot clusters were concentrated in the Western and Central regions. By 2014, the number of hot spots had decreased significantly, while cold spots increased and gradually clustered in the West, with sporadic distribution in the Northeastern region. By 2018, the number of cold and hot spots continued to decrease, and spatial agglomeration weakened significantly, presenting a uniform point-like distribution across the four major economic regions. By 2021, the number of cold and hot spots reached the minimum during the study period. Hot spots were mainly concentrated in Guangdong, Guangxi, and Shaanxi provinces, while cold spots were predominantly in Jiangsu, Zhejiang, Hebei, and Shanghai.

3.1.3. Analysis of the Temporal and Spatial Dynamics of the Total Price of Industrial Land Transactions

As shown in Figure 6, compared to the trends in the transaction frequency and area, the total price of industrial land transactions in the country remained relatively flat during the study period, peaking at approximately 469 billion in 2020 before sharply declining to about 77 billion in 2021. The overall trend showed a decrease in the order of “East-Central-West-Northeast.” Observing by economic region, the total transaction price in the Eastern region is dominated by the cluster formed by Jiangsu, Shandong, Zhejiang, and Guangdong. The Central region shows a more balanced development in the total transaction price across its provinces. In the Western region, the core of development is centered around Sichuan, Chongqing, Guangxi, and Shaanxi, forming a sub-center. In the Northeastern region, the development is driven by Liaoning, Heilongjiang, and Jilin, forming the driving force of the pattern.

Similar to the previous section on transaction area, the Getis-Ord Gi* tool is used to analyze the spatial aggregation dynamics of the total price of industrial land transactions. As shown in Figure 7, in 2010, hot spot clusters were distributed in a band shape across Sichuan, Yunnan, Guizhou, Guangxi, and Chongqing Municipality, while cold spot clusters were distributed in a line shape across Zhejiang, Anhui, and Henan provinces. By 2014, the number of cold and hot spots had increased significantly. Hot spot clusters were mainly concentrated in the Northeast region, while cold spot clusters were spread across the coastal provinces and cities in the Eastern region in an expansive agglomeration. By 2018, the number of cold and hot spots had increased significantly, with an even distribution across the four major economic regions. Cold spots were primarily located in the Eastern and Western regions. Hot spot clusters were evenly distributed across the four major economic regions, while cold spot clusters were located in the Eastern and Central regions. By 2021, the number of cold and hot spot clusters decreased significantly. The spatial development of clusters weakened, and the degree of spatial agglomeration decreased notably.

3.2. Characteristics of Spatial and Temporal Heterogeneity of PM_2.5

3.2.1. Characteristics of Temporal Heterogeneity

As shown in Figure 8, the national PM_2.5 concentration increased until 2013, peaking at 57 µg/m³, and then steadily declined to 29 µg/m³ by 2021. Observing the four major economic regions, during the study period, average PM_2.5 concentrations in the Northeast and West regions were generally lower than the national average. The East region followed a similar trend to the national average, while the Central region was consistently higher. All regions peaked in 2013 at 48, 46, 58, and 68 µg/m³, respectively, and then declined to their lowest values in 2021, reaching 27, 27, 28, and 33 µg/m³, respectively. According to the National Environmental Quality Standard (GB3095-2012) [52], a PM_2.5 concentration below 35 µg/m³ is considered to meet the standard. By 2021, PM_2.5 concentrations in all regions of the country are expected to meet this standard, with the highest reduction being 48.5% compared to 2013. Combined with the analysis of industrial land development trends in terms of the transaction area and frequency, there is a correlation between the average PM_2.5 concentration and industrial land development over time. This reflects that the expansion of industrial urbanization, based on land expansion in the form of “spreading out a big cake,” is being effectively controlled with the deepening of ecological civilization construction.

3.2.2. Characteristics of Spatial Heterogeneity

The spatial autocorrelation analysis of PM_2.5 concentrations in 286 prefecture-level cities in China from 2010 to 2021, using the global Moran index tool in ArcGIS 10.5, reveals a positive Moran index that passes the 1% significance test and shows a growing trend. The Moran index increased from 0.873 in 2010 to 0.951 in 2021, indicating that the spatial clustering of PM_2.5 concentrations has strengthened. The results of the Moran index are shown in

Table 2. Moran’s index timescale.

	2010	2012	2014	2016	2018	2021
M	0.8733	0.9353	0.9017	0.9615	0.9802	0.9507
P	0.01	0.01	0.01	0.01	0.01	0.01
Z	47.9978	54.4117	49.5791	52.8516	53.8538	52.2478

.

Using the Getis-Ord Gi* tool, this study explores the heterogeneity of the PM_2.5 spatial concentration in 286 prefecture-level cities across China from 2010 to 2021. As shown in Figure 9, from 2010 to 2014, the hotspot area is clustered in a faceted manner in the Eastern and Central regions, while the cold spot area is belt-shaped, spanning Guangdong, Guangxi, and Yunnan, and clustered in Heilongjiang. Over time, the thermal effect of spatial agglomeration weakens, and the hotspot area gradually shrinks inward. From 2014 to 2018, the hotspot area further clusters inward, and the cold effect of spatial agglomeration in the Northeast region intensifies. From 2018 to 2021, the spatial pattern is established, with the hotspot area showing a face-like concentration around Shandong, Hebei, Shanxi, Shaanxi, Hubei, Anhui, and Jiangsu and the cold spot area exhibiting a belt-like distribution in Guangdong, Guangxi, and Yunnan, along with a slice-like distribution in Heilongjiang.

3.3. Analysis of the Impact of Industrial Land Development on PM_2.5 Concentrations

Considering the changes in administrative divisions of Chinese cities, this paper follows the principles of data accessibility and uniformity, selecting 286 cities from four major economic regions as the scope of the study. Six time points—2010, 2012, 2014, 2016, 2018, and 2021—are chosen to measure the development of industrial land based on the transaction frequency, transaction area, and total transaction price, while controlling for GDP and the industrial structure of the cities. Based on seven variables—GDP, fixed investment in urban utilities, road length, built-up area, local financial gap, and year-end resident population—a multiple linear regression model is used to examine the impact of industrial land development on PM_2.5 concentrations during 2010–2021. This is performed by comparing impacts horizontally across regions and vertically across differences within regions. To eliminate data bias and heteroscedasticity, the above variables are transformed into logarithms.

3.3.1. Northeast Region

As shown in

Table 3. Table of influence degree of industrial land development on PM_2.5 concentrations in Northeast China from 2010 to 2021.

Year	2010		2012		2014		2016		2018		2021
	B	t	B	t	B	t	B	t	B	t	B	t
Constant	−0.246	−0.108	−1.853	−0.705	0.19	0.1	2.039 ***	3.106	2.256 ***	5.328	2.921 ***	7.902
TF	0.172 **	2.167	0.416 ***	5.489	0.444 *	1.856	0.136 *	2.053	0.107 **	2.4	0.058 **	2.144
TA	1.168 ***	5.448	0.742 ***	3.146	1.143 **	2.137	0.268 **	2.447	0.184 **	2.574	0.415 ***	2.87
TTP	0.227 *	1.87	0.516 **	2.142	0.403 ***	2.807	0.243 ***	4.004	0.234 *	2.041	0.209 ***	3.382
GDP	−0.494 ***	−3.497	−0.992	−1.384	−0.35 **	−2.243	−0.234 ***	−4.036	−0.215 **	−2.648	−0.362 ***	−3.975
IND	−0.203	−1.134	−0.245	−0.691	−0.31	−1.379	−0.158 **	−2.287	−0.161 ***	−3.701	−0.274 ***	−3.576
FI	−0.118	−0.709	−0.122	−0.917	−0.289 ***	−3.188	−0.145 ***	−2.855	−0.135 **	−2.491	−0.071 ***	−3.571
Road	−0.008	−0.04	−0.019	−0.372	−0.064	−0.637	−0.033	−1.283	−0.023	−0.735	−0.067	−1.127
UA	0.024	0.188	0.052	0.348	−0.054	−0.354	−0.018	−0.663	−0.004	−0.22	0.011	0.381
FG	0.064	0.279	0.063	0.322	0.078	0.563	0.086 *	1.724	0.022	0.313	0.024	1.385
Pop	0.433	0.599	0.149	0.768	0.575	1.588	0.125	1.338	0.103 *	2.068	0.038 *	1.921
R2	0.933			0.873			0.846			0.841
R2adjust	0.903			0.818			0.779			0.772
F	31.826			15.852			12.632			12.166

Note: *, **, and *** indicate that the test of significance was passed at 10 percent, 5 percent, and 1 percent; B is a non-standard parameter, and t is a t-test. Specific explanation of variable abbreviations: the transaction frequency (TF), transaction area (TA), total transaction price (TTP), the regional gross domestic product (GDP), the urban industrial structure (IND), the amount of public fixed asset investment (FI), the length of road (Road), built-up area (UA), local financial gap (FG), and the number of permanent residents at the end of the year (Pop). The same as below.

, there is a significant positive correlation between industrial land development in the Northeast region and PM_2.5 concentrations, meaning that the larger the scale of industrial land development, the more severe the PM_2.5 pollution. In terms of sub-dimensions, the impact of the transaction frequency on PM_2.5 concentrations first increases and then decreases, peaking at 0.416 in 2012, and then decreasing to a minimum of 0.058 in 2021. The impact of the transaction area on PM_2.5 concentrations shows a continuous downward trend, with an impact coefficient of 1.168 in 2010, decreasing to a minimum of 0.184 in 2018, before rebounding to 0.415 in 2021. The impact of the total transaction price on PM_2.5 concentrations first increases and then decreases, peaking at 0.516 in 2012, before decreasing to a minimum of 0.209 in 2021. The transaction area of industrial land has the greatest impact on PM_2.5 concentrations, primarily because it is related to the industrial occupancy scale, which directly influences the amount of industrial production accommodated. Overall, the impact of industrial land development on PM_2.5 concentrations in the Northeast has decreased substantially.

The large number of old industrial bases in the Northeast impedes the transformation and upgrading of the heavy industry. As a result, the government indirectly controls PM_2.5 concentrations by renewing, expanding, and reserving industrial land, offering large amounts of land in the market to attract more efficient and green production enterprises, reducing high-polluting enterprises, and improving the regional ecological environment.

3.3.2. Central Region

As analyzed in

Table 4. Table of influence degree of industrial land development on PM_2.5 concentration in Central China from 2010 to 2021.

Year	2010		2012		2014		2016		2018		2021
	B	t	B	t	B	t	B	t	B	t	B	t
Constant	2.155 ***	4.453	2.786 ***	6.615	2.219 ***	4.641	1.496 **	2.202	0.721	1.239	1.971 ***	4.028
TF	0.183 ***	3.871	0.093 ***	3.242	0.129 **	2.442	0.141 **	2	0.282 ***	3.427	0.068 *	1.855
TA	0.191 ***	3.192	0.149 ***	3.555	0.81 ***	5.334	0.225 **	2.494	0.34 **	2.575	0.391 ***	3.629
TTP	0.284 ***	3.496	0.123 ***	2.724	0.236 ***	3.551	0.168 **	2.601	0.313 ***	4.617	0.262 ***	4.999
GDP	−0.054	−0.777	−0.121 ***	−5.37	−0.456 ***	−4.32	−0.107	−1.439	−0.244 **	−2.307	−0.02	−0.537
IND	−0.196 **	−2.61	−0.03	−0.767	−0.09 ***	−4.917	−0.046	−0.453	−0.157 **	−2.211	−0.24 ***	−3.91
FI	−0.022	−0.437	0.002	0.028	0.105 **	2.61	−0.125	−1.262	−0.098	−1.531	−0.216 **	−2.623
Road	0.096	1.478	0.027	0.437	0.004	0.103	−0.083 ***	−2.985	−0.064	−0.916	−0.095 **	−2.032
UA	0.012	0.409	0.035	0.756	−0.376 ***	−4.571	−0.038	−0.872	−0.042	−0.709	−0.029	−0.545
FG	−0.084 *	−1.993	0.129	1.577	−0.043	−0.775	0.061	1.064	0.019	0.847	−0.028	−1.224
Pop	−0.073 ***	−3.789	−0.086	−1.335	−0.061	−1.045	0.175 *	1.845	0.055	0.989	0.03	0.401
R2	0.518		0.566		0.565		0.369		0.493		0.569
R2adjust	0.448		0.503		0.502		0.277		0.419		0.506
F	7.42		9.01		8.974		4.032		6.702		9.107

Note: *, **, and *** indicate that the test of significance was passed at 10 percent, 5 percent, and 1 percent.

, there is a significant positive correlation between all industrial land development in the Central region and the PM_2.5 concentration. Departing from the sub-dimension, the influence of the transaction frequency on the PM_2.5 concentration fluctuates greatly, with an overall weakening trend, reaching a peak of 0.282 in 2018 and decreasing to a minimum value of 0.068 in 2021; the influence of the transaction area on the PM_2.5 concentration shows a fluctuating growth trend, and the influence coefficient grows from 0.191 in 2010 to a maximum value of 0.81 in 2014, and then, in 2021, it weakened to 0.391; the impact of the total transaction price on the PM_2.5 concentration is relatively smooth compared to the frequency of transactions and the transaction area, reaching a minimum value of 0.168 in 2016, peaking at 0.313 in 2018, and dropping back to 0.262 in 2021. It can be seen that the transaction area of industrial land still has the greatest impact on the PM_2.5 concentration, and in general, the development of industrial land in Central China has a decreasing trend in the degree of impact on the PM_2.5 concentration of industrial land development in the Central region and is on a downward trend. In 2016, the Central government issued the “Thirteenth Five-Year Plan for Promoting the Rise of the Central Region”, which locates the Central region in the advanced manufacturing center and the demonstration area for the construction of an ecological civilization. Provinces are intensively developing industrial parks, introducing advanced manufacturing enterprises, and getting rid of the traditional inefficient and energy-consuming manufacturing industry while taking into account the construction of an ecology. The region is at a critical stage of synergistic development of production and ecology, with constraints on the development of industrial urbanization.

3.3.3. Western Region

As analyzed in

Table 5. Table of influence degree of industrial land development on PM_2.5 concentration in Western China from 2010 to 2021.

Year	2010		2012		2014		2016		2018		2021
	B	t	B	t	B	t	B	t	B	t	B	t
Constant	−3.021 ***	−2.862	−0.257	−0.247	−1.395	−0.863	−0.16	−0.142	1.319	1.405	0.296	0.471
TF	0.317 *	1.681	0.168 ***	3.232	0.312 ***	2.916	0.167 ***	3.279	0.332 ***	2.662	0.223 *	1.867
TA	0.54 ***	7.207	0.426 *	1.811	0.615 ***	3.845	0.618 ***	2.96	0.768 ***	5.262	0.569 ***	6.35
TTP	0.406 ***	3.176	0.242 ***	2.733	0.577 *	1.969	0.415 *	1.681	0.515 **	2.248	0.295 ***	3.286
GDP	−0.559 **	−2.441	0.128	1.131	0.145	0.366	−0.186	−1.55	−0.214 *	−1.838	0.16 *	1.698
IND	−0.101	−1.303	−0.128	1.213	−0.118	−0.705	−0.186	−1.153	−0.178	−1.544	−0.074	−1.064
FI	0.061 *	1.748	−0.271 **	−2.504	−0.01	−0.052	−0.173	−0.602	−0.158	−1.439	0.028	0.412
Road	0.085 *	1.723	−0.22 *	−1.82	−0.003	−0.033	−0.061	−0.528	−0.035	−0.765	0.029	0.914
UA	0.089	0.818	−0.108	−0.511	0.154	0.414	0.029	0.241	0.015	0.131	0.041	0.937
FG	0.174	1.414	0.141 *	1.837	0.168	0.775	0.037	0.5	0.025	0.336	0.044	1.388
Pop	0.311	1.284	0.26	1.396	0.17	1.083	0.086	0.942	0.032	0.491	0.08	1.252
R2	0.783		0.788		0.789		0.783		0.774		0.846
R2adjust	0.753		0.759		0.76		0.754		0.743		0.825
F	26.37		27.195		27.276		26.401		24.978		41.083

Note: *, **, and *** indicate that the test of significance was passed at 10 percent, 5 percent, and 1 percent.

, there is a significant positive correlation between industrial land development in the Western region and PM_2.5 concentrations. In terms of sub-dimensions, the impact of the transaction frequency on PM_2.5 concentrations fluctuates and weakens, decreasing to a minimum of 0.168 in 2012, rising to a peak of 0.332 in 2018, and then declining to 0.223 by 2021. The impact of the transaction area on PM_2.5 concentrations shows a relatively flat growth trend, decreasing to a minimum of 0.426 in 2012, rising to a peak of 0.768 in 2018, and then falling back to 0.569 by 2021. The impact of the total transaction price on PM_2.5 concentrations fluctuates and weakens, decreasing to a minimum of 0.242 in 2012, reaching a peak of 0.577 in 2014, and then continuing to weaken to 0.295 in 2021. It can be observed that the overall impact of industrial land development on PM_2.5 concentrations in the Western region continues to decline.

With the continued advancement of the Western development strategy and the “Belt and Road” initiative, the Western region is promoting environmentally friendly, green, high-energy enterprises by regulating oil and natural gas production, fostering clean energy bases, and optimizing the energy supply structure, thereby significantly reducing pollutant emissions.

3.3.4. Eastern Region

As analyzed in

Table 6. Table of influence degree of industrial land development on PM_2.5 concentration in Eastern China from 2010 to 2021.

Year	2010		2012		2014		2016		2018		2021
	B	t	B	t	B	t	B	t	B	t	B	t
Constant	−1.767 *	−1.963	−0.361 **	−2.237	−0.308	−1.582	−2.191 *	−1.905	0.213	0.176	−2.116 **	−2.056
TF	0.393 ***	3.784	0.301 ***	3.53	0.177 ***	2.927	0.214 ***	3.327	0.114 **	2.339	0.108 **	2.044
TA	0.779 ***	5.056	0.437 ***	6.465	0.701 ***	4.307	0.686 ***	10.503	0.866 ***	6.528	0.689 ***	9.237
TTP	0.413 ***	3.221	0.321 **	2.403	0.442 **	2.614	0.378 **	2.197	0.382 **	2.173	0.435 ***	2.832
GDP	−0.083	−0.899	−0.745	−0.683	−0.232	−1.486	0.11	1.128	−0.054	−0.917	0.085	0.526
IND	−0.015	−0.581	−0.036	−0.373	−0.174	−1.196	−0.152	−0.962	−0.348 **	−2.072	−0.103	−0.552
FI	−0.012	−0.164	−0.208	−1.118	−0.139	−0.1	−0.024	−0.16	−0.28 **	−2.074	−0.057	−1.069
Road	−0.002	−0.059	0.019	0.609	0.006	0.144	−0.009	−0.248	−0.163	−1.043	−0.044	−0.369
UA	0.041	0.527	0.067	1.534	0.01	0.1	0.007	0.136	0.008	0.157	0.01	0.141
FG	0.047	0.468	0.15	1.219	0.035	0.299	0.028	0.202	0.03	0.144	0.08	0.412
Pop	0.21	1.54	0.2	1.275	0.154	0.795	0.173	1.45	0.221	1.419	0.29	1.428
R2	0.812			0.821			0.806			0.875
R2adjust	0.787			0.797			0.787			0.858
F	32.383			34.442			31.256			52.306

Note: *, **, and *** indicate that the test of significance was passed at 10 percent, 5 percent, and 1 percent.

, there is a significant positive correlation between industrial land development in the Eastern region and PM_2.5 concentrations. In terms of sub-dimensions, the impact of the transaction frequency on the PM_2.5 concentration continues to weaken, with the impact coefficient decreasing from 0.393 in 2010 to 0.108 in 2021. The impact of the transaction area on the PM_2.5 concentration shows an intermittent weakening trend, decreasing to a minimum of 0.437 in 2012, rising to a peak of 0.866 in 2018, and then decreasing to 0.689 in 2021. The impact of the total transaction price on the PM_2.5 concentration showed fluctuating growth, decreasing to a minimum of 0.321 in 2012, reaching a maximum of 0.442 in 2014, and then continuing to weaken to 0.435 in 2021, with the degree of impact increasing by 5% compared to 2010. It can be observed that the influence of industrial land development on PM_2.5 concentrations in the Eastern region shows a fluctuating downward trend.

This is mainly due to the “13th Five-Year Plan,” which supports the Eastern region in leading the development policy by encouraging provinces to build advanced manufacturing R&D bases and new industrialization hubs and relocate high-energy-consuming enterprises. This aims to establish a high-precision industrial structure and lay a solid foundation for the development of a green economy. At the same time, regional ecological and environmental monitoring networks are being built, and air pollution prevention and control projects are being implemented to improve the air quality management system.

3.3.5. Comparative Analysis

Temporal changes in the four economic regions at six time points are summarized by dimension. As shown in Figure 10, based on transaction frequency, the impact of each region on PM_2.5 concentrations shifts from “East-West-Central-Northeast” to “West-East-Central-Northeast” during the study period. The overall impact shows a decreasing trend, with the largest decrease occurring in the Northeast region, from 0.444 in 2014 to 0.058 in 2021. The overall impact is weakening, with the Northeast experiencing the largest decrease, from 0.444 in 2014 to 0.058 in 2021. This is due to the Northeast having a large number of old industrial bases compared to other regions. With the renewal of these bases and the expansion of new industrial parks, industrial upgrading has accelerated, resulting in improved industrial pollution emissions.

Observing the transaction area, the impact of various regions on PM_2.5 concentrations decreases in the order of “Northeast-East-West-Central” to “East-West-Northeast-Central.” The overall impact shows an increasing trend, with the Central region showing the largest increase. The impact coefficient rose from 0.149 in 2012 to 0.81 in 2014, before declining to 0.391 in 2021. The Central region has abundant energy and raw material resources, with many resource cities compared to other regions. As the concept of ecological civilization is increasingly integrated into the development strategy, Central provinces are accelerating the transformation and upgrading of these cities by improving land use efficiency and intensifying the development of advanced manufacturing industries. Observing the total transaction price, the influence of each region on PM_2.5 concentrations generally decreases in the order of “East-West-Central-Northeast.” The overall influence fluctuates and weakens, with the largest decline in the Northeast region, from 0.516 in 2012 to 0.209 in 2021. This further indicates that the improvement in air quality in the Northeast is due to the renewal of old industrial bases and industrial upgrading.

These regionally divergent trends can be linked to both policy and structural economic differences. For instance, the Eastern region—while heavily industrialized—benefited from early policy shifts during the 13th Five-Year Plan that promoted high-tech and clean manufacturing industries, along with stricter environmental regulations. In contrast, the Central region saw increased industrial land development due to policies aimed at economic rebalancing and industrial relocation from the coast to inland areas, which temporarily raised PM_2.5 levels. Meanwhile, the Western region’s relatively low pollution trend is associated with national strategies such as the Western Development Policy and the Belt and Road Initiative, which favored green infrastructure and energy-efficient industries. Finally, the significant improvement in the Northeast reflects state-led renewal programs for old industrial bases, aimed at phasing out heavy polluters and attracting environmentally friendly enterprises.

In summary, the impact of the transaction area and total transaction price on PM_2.5 concentrations in the Eastern region is the highest. This is primarily due to the intensive circulation of capital, population, and land elements, the high level of industrial intensification, and the development of industrial supporting facilities, which enhance the value of industrial land and promote its expansion. Although environmental protection measures have been implemented, they only limit the growth of industrial pollution emissions. The industrial land base has been improved through the renewal of old industrial bases and industrial upgrading. Although a series of environmental protection measures have been implemented, they have only curbed the growth of industrial pollution emissions, and the base remains at a high level.

4. Discussion

4.1. Characterization of the Spatial and Temporal Distribution of Industrial Land Development and Analysis of the Impact of PM_2.5

This study examines the spatial and temporal characteristics of industrial land development and its impact on PM_2.5 concentrations. It explores the relationship between industrial land development and PM_2.5 pollution across various regions and time periods using a spatial analysis of national industrial land transaction and PM_2.5 data. Unlike conventional land classification approaches, this study examines land market transaction behavior, which dynamically reflects the interplay among local policy implementation, market demand shifts, and industrial land expansion. This method goes beyond the limitations of static spatial zoning and is more aligned with current research into pollution drivers under China’s ongoing land market reform. This study reveals that market-driven industrial land transactions are positively correlated with the PM_2.5 concentration, highlighting how transaction-based indicators can uncover regional and temporal variations more effectively than traditional zoning studies [5,53].

First, the spatio-temporal dynamic analysis shows that the transaction frequency, area, and total price of industrial land align closely with the trend of PM_2.5 concentrations from 2010 to 2021. Before 2013, rapid economic growth and industrialization led to a significant increase in the industrial land transaction frequency and area, resulting in a peak PM_2.5 concentration of 57 µg/m³. The overdevelopment of industrial land, particularly in the Eastern and Central regions, worsened air pollution, as reflected by the strong correlation among the rising transaction frequency, area, and PM_2.5 concentrations [25]. This result highlights the negative impact of industrial emissions on air quality in areas with intensive industrial land development, especially heavy, energy-intensive industries that significantly affect PM_2.5 levels. Since 2013, strengthened national environmental protection policies have gradually reduced PM_2.5 concentrations to 29 µg/m³ by 2021. This trend aligns with the slowing pace of industrial land development, particularly in the Eastern and Central regions, where the transaction frequency and area have declined, suggesting that environmental protection policies and industrial upgrading have partially mitigated the impact of industrial land development on PM_2.5 concentrations. Moreover, the observed fluctuations in the coefficients of control variables across regions can be attributed to differences in regional economic structures, levels of industrial development, and the infrastructure distribution. For example, the impact of the GDP and fixed asset investment may vary depending on whether a region is dominated by heavy manufacturing and service industries or is undergoing an economic transition. Notably, while the overall trend has improved, the extent of change and mechanisms of influence vary by region, indicating that the strength of policy implementation, industrial transformation processes, and differences in local governments’ environmental awareness are key factors affecting PM_2.5 concentration changes [51,54]. Additionally, the variation in control variable coefficients across regions is expected, given the diverse economic bases and development levels among regions. For instance, the GDP and fixed investment may have stronger effects in industrial-heavy provinces than in service-oriented or economically transitioning areas. Such heterogeneity reflects the real-world differences in how economic factors shape pollution levels. In addition to macroeconomic drivers, several plausible mechanisms may explain how industrial land development leads to increased PM_2.5 concentrations. First, land transactions often precede large-scale construction activities, such as plant construction, road expansion, and supporting infrastructure. These activities generate dust, diesel exhaust, and construction waste emissions. Second, the influx of industrial projects typically increases the demand for energy and raw materials, stimulating emissions from both onsite fuel combustion and offsite supply chains. Third, the process of land development in many Chinese cities involves land leveling, demolition, and soil movement, which directly contributes to particulate matter release. These pathways highlight how land transactions serve as a proxy for high-emission activities in the early and mid-stages of industrialization.

Second, this study highlights regional differences in the impact of industrial land development on PM_2.5 concentrations in the Eastern, Central, Western, and Northeastern regions [48]. In the Eastern and Central regions, the high industrial land development frequency and area make PM_2.5 concentrations more sensitive, as reflected by larger fluctuations during peak industrial land transaction periods. For example, between 2012 and 2014, PM_2.5 concentrations in the Eastern region remained high due to numerous industrial land transactions and the concentration of energy-intensive industries. Although PM_2.5 concentrations have since declined with industrial structural transformation, they remain affected by intensive industrial activities, highlighting the ongoing industrial pollution issue in the Eastern region [26]. In the Northeast, the impact of industrial land development has lessened. As local governments intensify industrial transformation and environmental management, especially in renovating old industrial bases, PM_2.5 concentrations in the Northeast have shown a more significant decline. Although the industrial land transaction area and frequency have increased in the Northeast, government measures, including clean energy substitution and industrial upgrading, have effectively mitigated the impact of industrial emissions on PM_2.5 concentrations [30]. The Western region exhibits different characteristics compared to other regions. Industrial land development in the Western region is relatively low, and PM_2.5 concentrations remain low due to limited industrialization and the influx of green, energy-efficient industries driven by the Belt and Road Initiative. This suggests that the low-pollution, low-energy-consumption industrial structure in the Western region has helped mitigate the impact of industrial land development on PM_2.5 concentrations [43].

In summary, the results show that the impact of industrial land development on PM_2.5 concentrations is influenced by the combined effects of the regional economy, industrial structure, environmental policies, and complex temporal and spatial interactions. Although PM_2.5 concentrations have declined due to enhanced government environmental measures, regional variability remains, highlighting the need for greater attention to regional differences and industrial structure characteristics in future policymaking and governance.

4.2. Limitations and Future Research Perspectives

First, while the control variables selected in this study, such as the GDP, urban industrial structure, and road length, explain regional differences, they do not account for all possible influencing factors. For example, climate change, seasonal factors, and other socioeconomic influences (e.g., population movement, changes in consumption patterns) may significantly affect PM_2.5 concentrations. Future studies should consider these factors to more comprehensively analyze the relationship between industrial land development and air quality.

Second, this study focuses on China as a whole, using a rich sample of nationwide data and discussing urban area divisions. However, policies, economic development levels, and ecological conditions vary across regions, which may limit the applicability of the conclusions to specific areas. To improve the broad applicability of the study, future research should focus on specific regional cases and explore the effects of policies and their impact on PM_2.5 concentrations in different regions.

Furthermore, the findings of this study are situated within the context of China’s state-led land market, where local governments play a pivotal role in land supply, industrial planning, and environmental governance. In countries with predominantly private land markets, such as those in North America or parts of Europe, the mechanisms linking industrial land transactions to PM_2.5 emissions may differ considerably. The market behavior, transaction frequency, and price signals in these regions are driven more by private investment strategies than government-led planning. As such, caution should be exercised when applying these conclusions to contexts with different land governance systems. Future comparative studies are encouraged to explore how institutional differences in land markets mediate the relationship between industrial development and environmental outcomes.

4.3. Policy Implications

First, industrial land market entry standards should be gradually regulated, and trading rules for land and resources should be improved. Local governments should consider the strategic positioning of economic development and the industrial structure, clarify new industrialization requirements, regulate industrial land entry standards, monitor the regional transaction frequency, limit the transactions of small, non-contiguous, and scattered industrial land, improve market player identification rules, and raise the market entry threshold.

Secondly, the orderly vacating of old industrial parks and the centralized supply of industrial land in batches are on the same track. Conduct an inventory of old industrial parks, clearing inefficient, energy-consuming enterprises to increase the “stock” of industrial land. Simultaneously, optimize the industrial spatial layout by combining regional development corridors, supplying land in batches, and focusing on digesting the existing industrial land.

Finally, it is essential to guide regional industrial land prices and identify high-quality production capacity within parks. Local governments should consider their industrial development status, adjust the scope and depth of land policy coverage, guide regional land prices to support local enterprises, enhance foreign enterprise standards to prevent poor-quality production capacity, and attract high-quality production capacity to the park.

5. Conclusions

The innovation of this study lies in considering the flow frequency and circulation scale of industrial land from the perspective of land transactions, quantifying industrial land development through the three dimensions of frequency, area, and total transaction price, and analyzing its spatio-temporal dynamics, as well as the spatio-temporal heterogeneity of PM_2.5 concentrations. Additionally, this study investigates the mechanisms by which industrial land development impacts PM_2.5 concentrations in different regions. This detailed regional analysis clarifies the role of industrial land in air pollution, with a particular focus on the trade area, which has a more significant effect on PM_2.5 concentrations. Finally, the variability in the impact of industrial land development on PM_2.5 across regions is analyzed to provide a theoretical basis for the differentiated formulation of environmental governance policies. The main findings are summarized as follows:

First, from 2010 to 2021, industrial land development initially increased and then decreased, with similar trends in the transaction frequency and area and relatively stable total transaction prices. The transaction area and total price decreased in the following order: “East-West-Central-North-East”, “East-Central-West-North-East”. Spatially, the transaction area and total transaction price decreased in the order of “East-West-Central-Northeast” and “East-Central-West-Northeast”, respectively. The number of cold and hot spots for the transaction area and total price continued to decrease spatially. The spatial development trend of clusters weakened, and spatial agglomeration decreased significantly, with points primarily concentrated in the Eastern and Central regions.

Second, the national PM_2.5 concentration increased and then decreased during the study period, reaching its lowest point in 2021. PM_2.5 levels in all regions met the national standard. Spatially, the clustering effect of hot and cold spots continues to weaken. Hot spots are clustered around Shandong-Hebei-Shanxi-Hubei-Anhui-Jiangsu, while cold spots are distributed in belt-shaped patterns in Guangdong-Guangxi-Yunnan and piecemeal along the Heilongjiang River.

Third, a significant positive correlation exists between the transaction frequency, area, total price of industrial land, and PM_2.5 concentration in the four major economic regions. The overall impact tends to weaken, with the Eastern region showing a higher influence than the others. A comparative analysis reveals that the transaction area tends to increase the PM_2.5 concentration, while the influence of the transaction frequency and total price weakens, though total price changes show significant variation.