[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Next Article in Journal
Multi-Criterion Analysis of Cyclone Risk along the Coast of Tamil Nadu, India—A Geospatial Approach
Next Article in Special Issue
Spatio-Temporal Relevance Classification from Geographic Texts Using Deep Learning
Previous Article in Journal
Using a Cost-Distance Time-Geographic Approach to Identify Red Deer Habitat Use in Banff National Park, Alberta, Canada
Previous Article in Special Issue
PMGCN: Progressive Multi-Graph Convolutional Network for Traffic Forecasting
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
IJGI
Volume 12
Issue 8
10.3390/ijgi12080340
ijgi-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Analysis of PM2.5 Synergistic Governance Path from a Socio-Economic Perspective: A Case Study of Guangdong Province

by
Kunkun Fan
1,2,
Daichao Li
1,2,*,
Cong Li
1,2,
Xinlei Jin
1,2,
Fei Ding
1,2 and
Zhan Zeng
3
1
The Academy of Digital China (Fujian), Fuzhou University, Fuzhou 350108, China
2
Key Laboratory of Spatial Data Mining and Information Sharing of Ministry of Education, Fuzhou University, Fuzhou 350002, China
3
Hunan Cartographic Publishing House, Changsha 410007, China
*
Author to whom correspondence should be addressed.
ISPRS Int. J. Geo-Inf. 2023, 12(8), 340; https://doi.org/10.3390/ijgi12080340
Submission received: 14 July 2023 / Revised: 11 August 2023 / Accepted: 15 August 2023 / Published: 16 August 2023
(This article belongs to the Special Issue Innovative GIS Models and Approaches for Large Environmental and Urban Applications in the Age of AI)
Browse Figures
Figure 1
<p>Location of the study area.</p> "> Figure 2
<p>Technical roadmap.</p> "> Figure 3
<p>Flowchart of scenario construction methodology.</p> "> Figure 4
<p>Correlation between PM<sub>2.5</sub> and influencing factors in various cities.</p> "> Figure 5
<p>The predicted annual average concentration of PM<sub>2.5</sub> in Guangdong Province from 2020 to 2025.</p> "> Figure 6
<p>Coupling and coordination relationship of PM<sub>2.5</sub>-pollution-influencing factors in Guangdong Province from 2020 to 2025.</p> ">
Versions Notes

Abstract

:
Analyzing the influencing factors of PM2.5 concentration, scenario simulations, and countermeasure research to address the problem of PM2.5 pollution in Guangdong Province is of great significance for governments at all levels for formulating relevant policies. In this study, the ChinaHighPM2.5 dataset and economic and social statistics for Guangdong Province from 2010 to 2019 were selected, and a PM2.5 pollution management compliance path formulation method based on the multi-scenario simulation was proposed by combining the differences in city types and PM2.5 concentration prediction. Based on the prediction model of PM2.5 concentration constructed by the Ridge and SVM models and facing the PM2.5 pollution control target in 2025, the urban PM2.5 pollution control scenario considering the characteristics of urban development was constructed. According to the scenario simulation results of the PM2.5 prediction model, the PM2.5 pollution control path suitable for Guangdong Province during the 14th Five-Year Plan period was explored. The coupling coordination model was used to explore the spatial and temporal pattern evolution of PM2.5 pollution collaborative governance in various prefecture-level cities under the standard path, and the policy recommendations for PM2.5 pollution control during the 14th Five-Year Plan period are proposed. The results showed the following: ① in the case of small samples, the model can provide effective simulation predictions for the study of urban pollutant management compliance pathways. ② Under the scenario of PM2.5 management meeting the standard, in 2025, the annual average mass concentration of PM2.5 in all prefecture-level cities in Guangdong Province will be lower than 22 μg/m3, and the annual average concentration of PM2.5 in the whole province will drop from 25.91 μg/m3 to 21.04 μg/m3, which will fulfil the goal of reducing the annual average concentration of PM2.5 in the whole province to below 22 μg/m3, as set out in the 14th Five-Year Plan for the Ecological Environmental Protection of Guangdong Province. ③ Under the path of PM2.5 control and attainment, the regional coordination relationship among prefecture-level cities in Guangdong Province is gradually optimized, the number of intermediate-level coordinated cities will increase, and the overall spatial distribution pattern will be low in the middle and high in the surrounding area. Based on the characteristics of the four city types, it is recommended that a staggered development strategy be implemented to achieve synergy between economic development and environmental quality. Urban type I should focus on restructuring freight transportation to reduce urban pollutant emissions. City type II should focus on urban transportation and greening. For city type III, the focus should be on optimizing the industrial structure, adjusting the freight structure, and increasing the greening rate of the city. For city type IV, industrial upgrading, energy efficiency, freight structure, and management of industrial pollutant emissions should be strengthened.

1. Introduction

PM2.5 pollution has always been one of the major factors of atmospheric environmental problems in Guangdong Province [1,2,3]. The 2019 Bulletin on the State of the Environment in Guangdong Province showed that the annual average concentration of PM2.5 in the province was 27 μg/m3, which was still higher than the World Health Organization’s PM2.5 transition level 2 standard (25 μg/m3). With the successive introduction of relevant policies during the 14th Five-Year Plan period, the new policies have put forward more stringent requirements for the treatment of PM2.5 pollution in Guangdong Province. The 14th Five-Year Plan for Ecological Environment Protection of Guangdong Province proposes reducing the annual average concentration of PM2.5 in the province to below 22 μg/m3 by 2025. Therefore, finding a path to achieve PM2.5 pollution control standards is of great significance for prefectural municipalities in Guangdong Province in formulating policies on environmental pollution control.
Currently, research on PM2.5 pollution management pathways is mainly carried out in three aspects: analysis of PM2.5 influencing factors [4,5,6], prediction model construction [7,8,9], and PM2.5 pollution management [10,11,12,13,14,15,16]. In terms of PM2.5 influencing factors, the existing research has only established indicator characteristics from the relationship between meteorological conditions [17,18,19] and PM2.5 concentration to reveal the role of meteorological parameters in influencing the daily and seasonal changes in PM2.5 concentration as well as its spatial distribution, ignoring the comprehensive impacts of economic development and policy adjustments on the management of PM2.5 levels, such as industrial structure, energy consumption, and environmental protection policies. Therefore, this study integrated relevant statistical indicators on the basis of existing research and constructed an indicator system for PM2.5 influencing factors from the perspective of socio-economic factors. In terms of PM2.5 prediction models, traditional prediction models have certain limitations. Although the classical statistical model is easy to use, its prediction effect needs to be improved [20], the mechanism of the air quality model is complex, and the prediction performance is average [21]. With the maturity of machine learning methods, deep learning models based on neural networks are beginning to be used in the study of air pollution [22], and based on regression methods [23], neural networks [24], RF [25], LightGBM [26], and LSTM [27] have been widely used to study the nonlinear relationship between economic development and PM2.5 concentration. Although machine learning-based models have superior predictive performance, they require a large number of samples to support them [28,29,30]. The coupling of multiple single models maximizes the integration of model strengths and improves prediction accuracy [31,32,33,34]. Therefore, this study explored hybrid models coupling classical statistical and machine learning approaches for more accurate PM2.5 concentration predictions. In terms of PM2.5 pollution management, the existing studies mainly focused on the construction of the same management scenarios for multiple cities within the study area, ignoring the differences in socio-economic characteristics between cities [35,36,37,38]. Therefore, this study constructed PM2.5 management scenarios that take into account the socio-economic characteristics of the cities, and simulated the prediction and analysis of these scenarios. These predictions were combining with the coupled degree of coordination model to explore the evolution characteristics of the coupled coordination relationship of PM2.5 pollution influencing factors in different scenarios, and we then analyzed the spatial and temporal patterns of PM2.5 pollution management for the purpose of proposing paths of synergistic management of PM2.5 pollution according to the local conditions, with a view to realizing the synergistic management of Guangdong Province and the sustainable development of the city.
This study used Guangdong Province as the study area; we combined air pollution control policies such as the Three-Year Action Plan for Winning the Battle for the Blue Sky with socio-economic indicators such as the influencing factors of PM2.5 pollution, and analyzes the rationality of the indicators. Based on the historical data of PM2.5 and influencing factors in each prefecture-level city, a hybrid model coupling Ridge and SVM was constructed to realize PM2.5 concentration predictions. Based on the current development status of each prefecture-level city, a scenario construction method that takes into account the characteristics of urban development was proposed, and the PM2.5 management scenarios for each prefecture-level city in the 14th Five-Year Plan period were constructed using this method. Finally, based on the PM2.5 prediction model, the PM2.5 management scenario was simulated and predicted, and based on the results, the path of PM2.5 pollution management in Guangdong Province during the 14th Five-Year Plan period was determined. The evolution of the spatial and temporal patterns of PM2.5 pollution management processed and coupling coordination relationships were analyze by using the coupling coordination degree model to reveal the influence relationship among socio-economic development, air pollution management, and regional coordinated development. This research aimed to explore the interactions between economic and social development, environmental protection, and coordinated management through the prediction of PM2.5 concentrations, simulation of management scenarios, and a comprehensive study of the coupling and coordination relationships, so as to provide scientifically sound policy recommendations for the PM2.5 management targets of the 14th Five-Year Plan for Guangdong Province.

2. Materials and Methods

2.1. Study Area

Guangdong Province is located in the southeastern coastal region of mainland China, with 21 prefecture-level cities under its jurisdiction, mainly comprising the four regions of Eastern Guangdong, Western Guangdong, Northern Guangdong, and the Pearl River Delta (Figure 1), covering a total area of about 180,000 square kilometers and a population of more than 113 million. As one of the most populous and economically developed provinces in China, Guangdong Province brings together 8% of the country’s population in less than 2% of the country’s land area, with a well-developed socio-economy, and a Gross Regional Product (GRP) of approximately RMB 129,118,858,000,000 in 2022, which has ranked first in the country for the 31st consecutive year, and with manufacturing, trade, and services being the main drivers of the socio-economy in Guangdong Province. However, due to the problems brought about by its rapid economic development, the province’s atmospheric environmental pollution problem has also become more and more serious. With the “Outline of the Fourteenth Five-Year Plan and Vision 2035 for the National Economic and Social Development of Guangdong Province”, “Fourteenth Five-Year Plan for the Ecological Civilization Construction of Guangdong Province”, and “Fourteenth Five-Year Plan for the Ecological Environmental Protection of Guangdong Province”, etc., the management of PM2.5 in Guangdong Province has entered into a new stage, and there are new objectives for the management of PM2.5 in the new era and situation. In 2025, the average annual concentration of PM2.5 in Guangdong Province will reach the target concentration of the 14th Five-Year Plan, and as the largest province in China in terms of economic output (GDP), with a complex and diverse industrial structure and a large population, Guangdong Province is facing the double pressure of economic development and environmental governance. In this context, exploring a path for PM2.5 pollution management that minimizes the impact of urban socio-economic development has a profound significance for the governments of Guangdong Province at all levels in formulating and implementing the relevant management programs.

2.2. Variable Selection and Data Sources

Due to the late start of PM2.5 concentration monitoring in the prefecture-level cities of Guangdong Province, which began in 2013, there are fewer samples of PM2.5 annual average concentration data; however, the use of remote sensing imagery to estimate PM2.5 concentrations can yield longer time series of PM2.5 data. Wei et al. [39] generated the ChinaHighPM2.5 dataset at a 1 km resolution from 2000 to 2020 based on ground-based measurements, satellite remote sensing products, and other data. Considering the great impact of the novel coronavirus epidemic in 2020 on the socio-economy of Guangdong Province and its significant difference from normal years [40], this research did not take it as a valid research year to ensure that the accuracy and reliability of the data and that the results were not affected by bias. In this study, we selected the 2010–2019 annual average PM2.5 concentration data from this dataset and extracted the PM2.5 data for Guangdong Province using ArcGIS.
This research combined the Three-Year Action Plan for Winning the Battle for Blue Sky, the 14th Five-Year Plan, and related studies and to select the relevant indicators of influencing factors, which belong to six aspects: industrial structure, energy consumption, road traffic, cargo transportation, urban greening, and pollution emissions. In terms of industrial structure, the higher the proportion of secondary industries in GDP, the higher the PM2.5 concentration, but with the increase in the proportion of tertiary industries, the tertiary industries catching up and overtaking the secondary industries can curb air pollution to a certain extent [41]. Therefore, the proportion of secondary industries in GDP (SGDP) and the proportion of tertiary industries in GDP (TGDP) were chosen to measure the city’s industrial structure. In terms of energy consumption, reducing energy consumption per unit of GDP (ECPG) can help to improve economic efficiency and air quality [42], and this indicator was chosen to represent the intensity of urban energy consumption. In regard to road transportation, controlling the number of automobiles can help alleviate urban traffic congestion and tailpipe emissions, and therefore the growth rate of civilian vehicle ownership (GROC) was chosen to characterize urban road transportation. In terms of cargo transportation structure, since the total amount of road freight and waterway freight in Guangdong Province accounts for more than 95% of the total freight, the implementation of “public-to-water” measures was mainly taken into account, and the proportion of highway freight to total freight (HFTF) and the proportion of waterway freight to total freight (WFTF) were chosen to characterize the freight transportation structure. Regarding urban greening, this study selected urban green area per capita (GAPC) to measure the level of urban greening. Finally, urban industrial sulfur dioxide emissions (ISDE) was selected as an indicator to measure the level of urban pollution emissions.
In this study, SGDP, TGDP, GROC, and GAPC data were obtained from the Guangdong Provincial Statistical Yearbook; HFTF and WFTF data were obtained from the China Urban Statistical Yearbook; and ECPG and ISDE data were obtained from the Guangdong Provincial Government Information Disclosure System upon request. The period of the data spanned 2010–2019.

2.3. Research Methods

In this study, the RR-SVM hybrid model coupled with classical statistics and machine learning methods was used to predict PM2.5 concentration. Based on this model, the research on PM2.5 pollution control paths was realized by constructing and simulating governance scenarios. The main processes were as follows: first, we determined the reasonableness and validity of PM2.5-influencing factors using grey correlation analysis. Second, we constructed the RR-SVM hybrid model based on the 2010–2019 PM2.5 and variable data for each prefectural city in Guangdong Province, coupling Ridge regression and Support Vector Machine (SVM) methods, and used the Particle Swarm Optimization (PSO) algorithm to find the optimal parameters of the model. Then, based on the air pollution management policy and planning, the historical change pattern of variables, and the socioeconomic status of the study area, a basic library of PM2.5 management scenarios was constructed. Then, the development status of each prefecture-level city was categorized according to its current development status; based on the categorization results, the PM2.5 pollution management scenarios of each type of city were constructed in combination with the base scenario library. Finally, the scenario simulation prediction was carried out by using the hybrid model, and according to the results, the standard path of PM2.5 pollution control in Guangdong Province during the 14th Five-Year Plan period was determined. Then, the coupling coordination degree model of the urban PM2.5-influencing factors was constructed, and the spatial and temporal evolution laws of the coupling coordination relationship between urban PM2.5-influencing factors in different prefecture-level cities are further explored based on six aspects: STIR, ECPG, WHFR, GROC, GAPC, and ISDE. The technical roadmap is shown in Figure 2.

2.3.1. Grey Correlation Model

According to the related study by He Xiang et al. [43], the steps for calculating the correlation between each influencing factor and the PM2.5 concentration are as follows:
(1) First identify the sequence of features to be used as a reference and the sequence of factors to be compared:
The annual average PM2.5 concentration values of each city for a total of 10 years from 2010 to 2019 were selected as the characteristic sequence x 0 t , i.e., x 0 t = x 0 1 ,   x 0 2 , ⋯ ,   x 0 10 . The data for 8 influencing factors such as SGDP, TGDP, GROC, and GAPC of each city are selected as the factor series, i.e., x i t = x i 1 ,   x i 2 , ⋯ ,   x i 10 ,   i = 1 ,   2 ,   ⋯ ,   8 .
(2) If the factors have different scales and the data are not directly comparable, they are standardized.
(3) Calculate the correlation coefficient:
ξ x 0 t , x i t = m i n i m i n t x 0 t − x i t + ρ m a x i m a x t x 0 t − x i t x 0 t − x i t + ρ m a x i m a x t x 0 t − x i t
In the formula, x 0 t is the characteristic sequence and x i t is the factor sequence; m i n i m i n t x 0 t − x i t and m a x i m a x t x 0 t − x i t are the minimum value and maximum value of extreme deviation, respectively; and ρ is the resolution, which generally takes a value of 0.5.
(4) Calculate the grey correlation R:
R i = 1 10 ∑ t = 1 10 ξ x 0 t ,   x i t
When R ∈ 0 , 0.3 , the correlation is slight; when R ∈ 0.3 , 0.6 , the correlation is medium; and when R ∈ 0.6 , 1.0 , the correlation is strong [43].

2.3.2. PM2.5 Prediction Model Approach

(1)
Ridge Regression
When socioeconomic factors are used for PM2.5 concentration prediction, the high correlation between variables can lead to extreme instability and overfitting of traditional linear regression models. Meanwhile, the ridge regression model retains all the variables while resolving the multicollinearity among the variables [44]. The ridge regression model is formulated [45] as follows:
y = β 0 + β 1 x 1 + β 2 x 2 + ⋯ + β p x p + λ ∑ j = 1 p β j 2
where y is the PM2.5 concentration; x 1 ,   x 2 ,   ⋯ ,   x p represent the eight influencing factors such as SGDP and TGDP; β 1 ,   β 2 ,   ⋯ ,   β p represent the regression coefficients; and λ is the regularization parameter.
(2)
SVM Model
The SVM model can solve the problems of non-linearity, small samples, and high dimensions, and effectively avoid over-fitting [46], and is thus often used in prediction or classification research. This study used this method to establish a nonlinear relationship between PM2.5 and the influencing factors to achieve the prediction of PM2.5 concentrations. The specific formula is as follows.
Let there be a total of n samples. Each sample x i contains p influences such as SGDP ( x i , 1 ,   x i , 2 ,   ⋯ ,   x i , p ) and the target variable PM2.5 concentration y. The SVM model equation [47] is
y = ∑ i = 1 n α i K x i , x j + b − λ ∑ i = 1 n α i 2
where α i is the Lagrange multiplier, K x i , x j is the kernel function, b is a constant term, and λ is the regularization parameter.
In SVM modeling, the kernel function is used to map data from the original space to a higher dimensional space, thus enabling the model to deal with nonlinear relationships. The radial basis kernel (RBF kernel) is one of the most commonly used kernel functions, and is used in this study with the formula:
K x i , x j = e − γ ‖ x i − x j ‖ 2
where γ is the parameter of the kernel function, which is used to control the distance between the samples in the high-dimensional space. To enable the model to achieve an ideal effect, the PSO algorithm was used to optimize the regularization parameter λ and the kernel function parameter γ .
(3)
R-SVM Model Construction Process
In this study, the ridge regression model was first constructed using the ridge module in the Sklearn package of the Python language based on the data for annual average PM2.5 concentrations and socio-economic factors in each prefecture-level city from 2010 to 2019. Then, the variable data from 2010 to 2019 were used as input data and predicted using the ridge model to obtain the linear prediction result of PM2.5 from 2010 to 2019, which was used as one of the input variables in the SVM model to explain the linear relationship between PM2.5 and this variable.
In constructing the SVM model, the ridge model prediction results and the original independent variables were set as the characteristic variables of the model, and PM2.5 was set as the model predictor variable. Although the SVM model is suitable for small sample datasets, there are only ten years of historical data for each city, which is still a small amount of data for modeling. For this reason, this study adopted the method used by Wang et al. [10], using the data of 21 cities as the training data of the model at the same time, and this method effectively improved the sample size and stability of the model and avoids overfitting.
In this study, the root mean square error (RMSE), mean absolute error (MAE), and goodness of fit (R2) were used as model accuracy evaluation indexes.

2.3.3. Scenario Construction Method

Due to certain differences between the socio-economic characteristics of different prefecture-level cities, using the traditional method to construct the same scenario simulations for all prefecture-level cities will cause the loss of the socio-economic development characteristics of the city. To solve this problem, this study utilized the similarities and differences between the socio-economic characteristics of different prefecture-level cities to classify the cities, and based on the characteristics of each type of city, we constructed a PM2.5 governance scenario intending to achieve the PM2.5 governance standard in 2025. The flowchart of the method is shown in Figure 3.
The specific steps are as follows:
(1) According to the relevant policies and planning objectives of the study area, the historical pattern of change and the current development of PM2.5-influencing factors, and relevant studies, we constructed a base scenario library with three levels of low, medium, and high numerical magnitude for each influencing factor indicator in terms of the direction of regulation for industrial structure, energy consumption, cargo transportation, road traffic, greening of cities, and pollutant emissions, and set three levels of regulation intensity for each level (loose, intensive, and strict regulation) to match the degree of socio-economic development of different cities. The main role of this library was to provide a reference when constructing PM2.5 governance scenarios for subsequent types of municipalities, and to select from the base scenario library the indicator regulation levels that match the socio-economic development of that type of municipality.
(2) The natural breakpoint method was used to grade the social and economic index data of each prefecture-level city in 2019 to establish the category characteristics of the social and economic variables of each prefecture-level city. We used the K-means algorithm to cluster the grading results and classified the cities according to the clustering results to obtain the city classification results.
(3) Based on the degree of development of socio-economic indicators in each type of city, we selected the matching indicator regulation intensity from the base scenario library and constructed the urban PM2.5 pollution management scenarios for 2020–2025.

2.3.4. Coupling Coordination Degree Model

In this study, due to the differences in the unit magnitude of the data of each indicator, the efficacy of different indicators was found to have both positive and negative effects. To ensure the reasonableness of the evaluation results, it was necessary to standardize the indicators before each calculation to eliminate the differences in magnitude and direction between the original data, and the formulas [48] are as follows:
n i j = o i j − m i n ( o j ) m a x o j − m i n ( o j ) ,   n i j   is   a   positive   indicator
n i j = m a x ( o j ) − o i j m a x ( o j ) − m i n ( o j ) ,   n i j   is   a   negative   indicator
where n i j is the standardized value of the indicator n i j ; o i j is the original value; and m a x ( o j )   and   m i n ( o j )   are the maximum and minimum values of indicator i , respectively. In this way, GAPC was found to be a positive indicator, while STIR, ECPG, WHFR, GROC, and ISDE were found to be negative indicators.
The coupling coordination degree model [49,50,51] further draws on the coordination degree model for the comprehensive evaluation and study of the whole system based on the coupling degree. The functional relationship equation of the coupling coordination degree measurement model of multi-systems is as follows:
C = ∏ i = 1 n U i 1 n ∑ i = 1 n U i n 1 n
where C is the value of the coupling degree, n is the number of subsystems, and U i   is the value of each subsystem.
The coupling coordination degree D is calculated as
T = ∑ i = 1 n w i U i
D = C T
where U i is the normalized value of the ith subsystem and w i is the weight of the ith subsystem. In most studies, it is assumed that the importance of each subsystem is the same, so w i was set to be the same value and ∑ i = 1 n w i = 1 . The classification criteria of the coordination level and coordination development degree are shown in Table 1 below.

3. Results

3.1. PM2.5 Concentration Prediction and Analysis of Influencing Factors

3.1.1. Correlation Analysis of PM2.5-Influencing Factors

In this study, the correlation between PM2.5 concentration values and indicators of influencing factors in all prefectural-level cities of Guangdong Province was calculated using grey correlation analysis based on the data for annual average PM2.5 concentrations and indicators of the influencing factors from 2010 to 2019. In the existing studies, the correlation is considered mild when the grey correlation R ∈ (0, 0.3], moderate when R ∈ (0.3, 0.6], and strong when R ∈ (0.6, 1.0] [43]. According to Figure 4, among the eight indicators, SGPD had the highest correlation with PM2.5 concentration, with an average correlation of 0.85 for each prefecture-level city. The correlation between Dongguan City and Shanwei City reached 0.97 and 0.94, respectively, and the rest of the prefectures and cities were strongly correlated, except for Zhaoqing City; the average correlation between TGDP and PM2.5 concentration of the prefectures and cities was 0.73, and the correlation was strong. Dongguan and Shanwei had the highest correlation between this indicator and PM2.5 concentration, 0.97 and 0.91, respectively. The average correlation between ECPG and PM2.5 concentration for each prefecture-level city in Guangdong Province was 0.84, and the correlation was strong, while Zhanjiang City had the lowest correlation of 0.62. The average correlation between WHFR and PM2.5 concentration in each prefecture-level city was 0.77 and 0.74, and all cities had correlations greater than 0.6, which is a strong correlation. Except for the moderate correlation between GROC and PM2.5 concentration in Shenzhen, the correlation of all other cities was strong. The correlations between GAPC and PM2.5 concentration of Shantou City, Huizhou City, Zhongshan City, Zhuhai City, Jiangmen City, and Guangzhou City were lower than 0.6, which is a moderate correlation, and the remaining 15 cities all had a strong correlation. The average correlation between ISDE and PM2.5 concentration in Guangdong Province was 0.59, which is a moderate correlation.
In terms of each indicator, the average correlation between ISDE and PM2.5 in Guangdong Province was medium, and the average correlation between the rest of the indicators and PM2.5 concentration was strong. This indicates that industrial structure, energy consumption, transportation, and urban greening in Guangdong Province have strong correlations with PM2.5 concentrations, and have influenced PM2.5 concentrations to a large extent in the past decade. Regionally, the grey correlations between PM2.5 concentrations and socio-economic factors in each city were large, and their correlations were moderate or strong, indicating that the indicators played an important role in the changes in PM2.5 concentrations, and the selection of indicators is reasonable.

3.1.2. PM2.5 Concentration Prediction

Issues can arise due to the small sample size of the data for each prefecture-level city, while too many feature variables will lead to the overfitting phenomenon and reduce the generalization ability of the model. Linear transformation is one of the most common solutions for this issue, reducing the number of indicators and retaining the information of the original indicators when two indicators are highly correlated [52]. Therefore, in this model, the ratio of SGDP and TGDP (STIR) was used to represent SGDP and TGDP, and the ratio of HFTF and WFTF (WHFR) was used to represent HFTF and WFTF. Through this indicator compression, the original indicators can be preserved and the multicollinearity can be reduced. The results of the ridge regression model are shown in Table 1, and the results showed that the predicted values are in good agreement with the actual values, and the highest and lowest R2 values in prefecture-level cities were 0.96 and 0.87, respectively, which better explains the PM2.5 having a linear relationship with some of the influencing factors.
The ridge model prediction results with the original independent variables were set as the feature variables of the RR-SVM model, while PM2.5 was set as the predictor variable, and the RR-SVM prediction model was obtained after training. The original independent variables and PM2.5 were set as the feature variables and predictor variables of the single ridge and SVM models, and the single SVM prediction model obtained after training was used as the control group. The model results are shown in Table 2.
The results in Table 2 show that the RR-SVM model demonstrated better prediction performance than the single model. The RR-SVM model’s R2 ranged between 0.93 and 0.99, the MAE ranged between 0.32 and 1.25 μg/m3, and the RMSE ranged between 0.365 and 1.635 μg/m3. Compared with the single model, the RR-SVM model improved the R2 by 0.041 on average, reduced the MAE by 0.446 μg/m3 on average, and reduced the RMSE by 0.475 μg/m3 on average; therefore, the RR-SVM model coupled with classical statistics and machine learning outperformed the single model. Therefore, the RR-SVM model coupling classical statistics and machine learning performs better than a single model. The RR-SVM prediction model can improve the prediction accuracy of PM2.5 under small sample conditions and provide effective support for PM2.5 prediction.

3.2. PM2.5 Governance Scenario Construction and Simulation Analysis

To identify the PM2.5 management path that minimizes the impact on urban socio-economic development, this study combined the expected targets of each indicator during the 14th Five-Year Plan period, the historical pattern of change, the current state of development, and related research, and is based on six modes of regulation, including industrial structure, energy consumption, freight transport structure, road traffic, urban greening, and pollution emissions. The base scenario library shown in Table 3 was constructed. This base scenario library provides a reference basis for the construction of subsequent urban PM2.5 governance scenarios.
In the case that certain socio-economic indicator data are more dispersed, direct cluster analysis may be affected by extreme values. Therefore, in this study, based on the consideration of the distribution of the values of each indicator and the average level for the province or the country, the data for the socio-economic indicators of municipalities in 2019 were graded using the natural breakpoint method. Among them, STIR was categorized into five levels; ECPG, WHFR, GROC, and ISDE were categorized into four levels; and GAPC was categorized into three levels. The grading results are shown in Table 4.
The K-means algorithm was used to cluster the results of the grading of urban socio-economic indicators, and according to the clustering results, 21 cities were classified into different city categories based on similarities in socio-economic characteristics. Table 5 shows the clustering results of the graded data and city classifications of each socio-economic indicator.
Combined with Table 4 and Table 5, it can be observed that there were large differences in the socio-economic characteristics of various types of cities, and various types of cities had different development patterns and characteristics. The first category of cities was dominated by tertiary industry, with limited space for optimization of the industrial structure, lower energy consumption intensity but still with space for optimization, a higher carrying capacity of waterway freight transport, moderate automobile ownership, large differences in the green space area, and higher industrial sulfur dioxide emissions. In the second category of cities, secondary industry accounted for a lower proportion of GDP than tertiary industry, which can be appropriately transferred to the industrial structure, with a slightly higher intensity of energy consumption, a weaker carrying capacity of waterway freight transportation, fast growth in automobile ownership, the lowest per capita green space area, and higher industrial sulfur dioxide emissions. In the third category of cities, secondary industry accounted for a higher proportion of GDP, and these cities should promote the development of tertiary industry. In these cities, energy consumption intensity was the lowest, the waterway freight carrying capacity was moderate, the growth of automobile ownership was slightly lower, the per capita green space area was close to the average level of the province, and the industrial sulfur dioxide emissions were slightly higher. The fourth category of cities, demonstrating a similar proportion of GDP for the secondary and tertiary industries, should promote the development of tertiary industries. In these cities, the intensity of energy consumption was much higher than the average level of the province, there was a strong carrying capacity of waterway freight transportation, the growth of automobile ownership was relatively low, the per capita green space area was the highest, and the industrial sulfur dioxide emissions were higher.
Therefore, when developing urban PM2.5 management scenarios, the differences between the socio-economic characteristics of various types of cities should be taken into account, and a targeted focus should be placed on the mode of regulation. To this end, this study categorized the indicators into high, medium, and low classes based on the results of their grading in Table 3, according to the space in which each indicator can be adjusted. For example, the STIR was generally low in the first category of cities, indicating that there is less room for industrial structure optimization in this category of cities, so the STIR regulation needs were classified as low-grade, while the STIR was generally high in the third category of cities, so it was classified as high-grade. The specific results are shown in Table 6.
According to the index control demand levels of each type of city in Table 6, the corresponding “low”, “medium”, or “high” control levels were selected from the base scenario library in Table 2, and three PM2.5 pollution control scenarios were constructed for each type of city, namely loose, intensive, and strict, as shown in Table 7.
Based on the governance scenarios in Table 7, the socioeconomic data for 2020–2025 were calculated and used as input data for the PM2.5 prediction model to predict the PM2.5 concentration in each city. The results showed that among the first group of cities, Qingyuan City can reduce PM2.5 concentration to below the target under the loose control scenario, while Meizhou City can achieve PM2.5 concentration control by using only the Intensive scenario. In contrast, Guangzhou, Shaoguan, Shenzhen, and Yunfu would need to adopt the most stringent control intensity to keep the concentration below the target. Among the second group of cities, Heyuan and Shanwei were able to reduce PM2.5 concentrations to below the target under the loose scenario, while Maoming and Yangjiang needed to adopt the intensive scenario to achieve the control target, and only Jieyang needed the strict scenario. Among the Category 3 cities, Chaozhou, Shantou, and Zhongshan can accomplish the control target by adopting the loose scenario, while Dongguan and Foshan need to adopt the strict scenario. In the fourth category of cities, Zhuhai City can adopt the loose scenario, while Huizhou City, Jiangmen City, Zhanjiang City, and Zhaoqing City need to adopt the intensive scenario to control the PM2.5 concentration below the target. Therefore, all 21 prefecture-level cities in Guangdong Province can achieve the target of reducing PM2.5 concentrations to below 22 μg/m3 by 2025 with minimal impact on the socioeconomic development of the cities.
Therefore, under the premise of ensuring minimal impact on urban socio-economic development, and based on the PM2.5 prediction results under different scenarios for each prefecture-level city, each prefecture-level city in Guangdong Province can adopt the PM2.5 pollution management pathways shown in Table 8.
Figure 5 demonstrates the predicted 2020–2025 average annual PM2.5 concentrations in Guangdong Province under the PM2.5 pollution management pathways presented in Table 8. As can be seen from Figure 5, all cities in Guangdong Province will show a significant decrease in the annual average PM2.5 concentration. Among them, Zhaoqing, Maoming, Jieyang, and Shantou will be the first to meet the 2025 PM2.5 annual average concentration control requirement in 2023. By 2024, all cities will be able to achieve the PM2.5 concentration control target, except for six cities, namely Yangjiang, Jiangmen, Yunfu, Foshan, Guangzhou, and Meizhou. By 2025, all 21 prefectural-level cities in Guangdong Province will be able to fulfil the PM2.5 control targets during the 14th Five-Year Plan period, and the annual average concentration of PM2.5 in Guangdong Province will drop from 25.91 μg/m3 in 2019 to 21.04 μg/m3, thus completing the task of controlling the concentration of PM2.5 pollution.

3.3. Analysis of PM2.5 Pollution Collaborative Governance Pattern

Figure 6 demonstrates the changes in the coupled coordination relationship of PM2.5 pollution management for each prefecture-level city in Guangdong Province in 2020–2025 under the PM2.5 pollution management compliance pathway. Combining Figure 4 and Figure 5, it can be seen that under the scenario simulation of the attainment pathway, with the reduction of PM2.5 concentration, the overall level of coupled coordination of factors influencing the management of PM2.5 pollution in Guangdong Province in 2020–2025 shows an increasing trend. From 2000 to 2022, due to higher PM2.5 concentrations, coupling coordination was dominated by moderate dissonance, mild dissonance, and near dissonance, with a lesser degree of reluctant coordination. After 2023, the environmental quality will be improved, the level of coordination will gradually rise, and some of the areas on the verge of dysfunction will be upgraded to barely coordinated and primary coordinated, which indicates that with the management of PM2.5 pollution and the concentration of the standard, the regional coordination relationship will be gradually optimized, and the number of intermediate coordinated cities will increase, while the level of some barely coordinated cities will be upgraded to primarily coordinated.
From the evolution of the spatial patterns of PM2.5 pollution management in Guangdong Province, the coupling and coordination of the relationship between the spatial differences are significant, demonstrating the distribution characteristics of being high in the center of the city and low in the surrounding areas. The eastern cities demonstrate a relatively good coupling and coordination degree, while cities in the central region generally display lower values, and in the same type of city, the same spatial accumulation characteristics are evident. The dissonant region is dominated by cities in central China, mainly Jiangmen City, Foshan City, Guangzhou City, Dongguan City, Huizhou City, and other regions with high PM2.5 concentrations. The rapid economic development of such cities is accompanied by an increase in industrialization, dense traffic, and construction activities, leading to a rapid accumulation of PM2.5 and a decrease in air quality in the region, with a poorer degree of coordination. In barely coordinated regions, the regional distribution is more scattered, mainly relating to the lower PM2.5 concentration of Zhaoqing City, Maoming City, Yangjiang City, and other cities. The level of socio-economic development in such cities is basically at the same level, socio-economic development tends to be stabilized, the differences between the various influencing factors are relatively small, and the city’s economic development and environmental protection are relatively balanced, to ensure the stability of the city’s air quality, and the level of its coupling and coordination is in the barely coordinated stage. High-level coordination areas are mainly distributed in the eastern part of the coastal economic belt, and from 2020 to 2022, they were mainly dominated by reluctant coordination and primary coordination, and there were fewer intermediate coordination areas. With the coordinated advancement of pollution reduction and economic development, Shantou Shanwei and other cities will start to progress from reluctant coordination to primary coordination and show a grouping trend by 2023 through measures such as strengthening environmental management and promoting industrial upgrading. By 2025, the overall spatial distribution pattern of PM2.5 pollution control coupling and coordination relationships in Guangdong Province will be low in the center and high in the surrounding areas.

4. Conclusions

(1) In the case of small samples, the hybrid model coupling classical statistics and machine learning methods can achieve more accurate fitting, and the results are more stable and reliable compared with a single model, which can provide effective simulation prediction for the identification of urban pollutant management pathways to reach pollution targets.
(2) The results of the PM2.5 scenario simulation for 2020–2025 show that the PM2.5 synergistic management pathway proposed in this study can effectively reduce the average annual concentration of PM2.5 in Guangdong Province and improve the coordination of urban economic and social development. Under the PM2.5 synergistic management scenario, Guangdong Province will accomplish the expected goal of controlling the annual average PM2.5 concentration in the “14th Five-Year Plan for Ecological Environmental Protection of Guangdong Province” by 2025, and realize the synergistic development of the economy and the environment.
(3) Considering the challenges of air pollution and the imbalances in economic development across the 21 prefecture-level cities in Guangdong Province, this study proposes PM2.5 collaborative governance policy recommendations tailored to the social and economic characteristics of each city. These recommendations, focused on six key aspects including industrial structure, freight distribution, road traffic, energy consumption intensity, urban greenery, and urban pollutant discharge, offer valuable insights to enhance urban governance capacity, drive economic advancement, and facilitate pollution prevention and control during the 14th Five-Year Plan period.

5. Discussion

This study took Guangdong Province as the study area, and used the grey correlation, ridge, and SVM models to simulate and predict the PM2.5 management scenarios in various cities. It analyzed the coupling and coordination relationship between PM2.5-influencing factors, and identified the paths of PM2.5 pollution control in various cities, which provides an important reference for the completion of PM2.5 pollution control in Guangdong Province during the 14th Five-Year Plan period. For different prefecture-level cities, this paper proposes corresponding synergistic management measures.
(1) For Guangzhou City, Meizhou City, Shenzhen City, Shaoguan City, Yunfu City, and Qingyuan City, during the 14th Five-Year Plan period, priority should be given to adjusting the freight transport structure and reducing urban pollutant emissions to collaborate in combating PM2.5 pollution. In terms of freight structure, Guangzhou, Shenzhen, Shaoguan, Yunfu, and Qingyuan are all located in the Pearl River water system area, so they can adjust the freight structure through joint planning and resource integration, developing waterways, improving shipping capacity, promoting multimodal transport, adopting waterway transport, etc., to achieve the efficient flow of goods and optimal transport routes, and to reduce the reliance on road transport, thus reducing the emission of related pollutants. Meizhou, on the other hand, can actively promote road transport and reduce the emission of related pollutants, while it can also actively develop public water intermodal transportation and coordinate the promotion of the construction of Meizhou Port. In terms of reducing pollutant emissions from urban industries, more stringent corporate pollution emission standards can be formulated, and measures such as increasing environmental protection inspections and upgrading equipment can be taken to reduce the direct emissions of pollutants. By optimizing the freight transport structure and reducing urban pollutant emissions, it not only helps to improve the efficiency of the flow of goods, but it also promotes the sustainable development of the economy in the region and reduces the impact of air pollution on people’s health, thus realizing the synergistic development of PM2.5 environmental management and economic and social development.
(2) For Heyuan City, Jieyang City, Maoming City, Shanwei City, and Yangjiang City, the main focus should be on urban road traffic and urban greening, supplemented by measures such as adjusting the industrial structure, lowering the intensity of energy consumption, and reducing pollutant emissions, to synergistically control PM2.5 pollution and realize a virtuous cycle of economic development and environmental protection. To control urban road traffic pollution, new energy vehicles can be used to replace traditional fuel vehicles, and public transportation infrastructure can be improved to reduce vehicle emissions. In terms of urban greening, green belts can be constructed on both sides and in the center of the road to increase the adsorption rate of green plants and reduce the pollution caused by road traffic. At the same time, PM2.5 pollution can also be treated by adjusting the industrial structure, reducing the intensity of energy consumption, and implementing and reducing pollutant emissions, among other measures. These initiatives will help to reduce the level of air pollution, improve the environmental quality of cities, and at the same time promote the sustainable development of cities.
(3) For Chaozhou City, Dongguan City, Foshan City, Shantou City, and Zhongshan City, the main direction of collaborative governance should be to optimize the industrial structure, adjust the structure of cargo transportation, and improve the urban greening rate, thus improving environmental quality, reducing pollutant emissions, and reducing the concentration of atmospheric pollutants to achieve a reduction in PM2.5 concentrations and urban economic development through collaborative governance. In terms of optimizing the industrial structure, these five prefecture-level cities are dominated by secondary industries, and tertiary industries are developing slowly. It is necessary to encourage the upgrading and transformation of traditional industries, promote the development of high-end service industries through tax incentives, subsidies, and preferential policies, promote economic transformation and sustainable development, and improve the quality of urban development and people’s quality of life. In terms of adjusting the structure of cargo transportation, Dongguan City, Foshan City, and Zhongshan City are located at the estuary of the Pearl River Delta, and so they can be integrated into the Pearl River transportation network to promote the transfer of road transportation to the waterway transportation. Chaozhou City and Shantou City can promote the transfer of bulk cargo from “road to water” through marine transportation. In terms of improving the urban greening rate, we can reduce the concentration of urban air pollutants by building urban green corridors or parks and building green space networks. By optimizing industrial institutions, adjusting the transportation structure, and increasing the greening rate of cities, environmental quality can be effectively improved, pollutant emissions can be reduced, and the concentration of air pollutants can be lowered, thus realizing synergistic management of PM2.5 and the development of the city’s economy, and thereby injecting a new impetus into the city’s sustainable development.
(4) For Zhanjiang City, Huizhou City, Jiangmen City, Zhaoqing City, and Zhuhai City, in addition to the small improvement effect of urban greening on air quality, from the perspective of PM2.5 environmental governance and coordinated economic and social development, there is still much room for improvement in the industrial structure, energy consumption intensity, freight structure, and industrial pollutant emissions. In terms of industrial structure, Zhuhai City, Zhanjiang City, Huizhou City, Jiangmen City, and Zhaoqing City all have heavy industries as their leading industries. We should promote the development of high-tech industries, encourage the transformation and upgrading of traditional industries, improve production technology, develop clean energy, high-tech industries, green manufacturing, and other environmentally friendly industries, and improve the sustainability of economic development. In terms of energy consumption intensity, cities should implement energy conservation and emission reduction policies, improve energy management, identify energy consumption problems, and implement corrective measures to reduce energy consumption intensity. At the same time, governments should promote the replacement of traditional energy with clean energy, increase the reform of the energy supply structure, and promote the application of renewable energy. In terms of freight structure, Zhanjiang City and Zhuhai City are coastal cities, which can promote the transfer of freight transportation from road to sea and rail transportation by developing marine logistics and expanding port functions. Huizhou, Jiangmen, and Zhaoqing can adjust the freight structure by strengthening the construction of logistic centers and developing multimodal transport. In terms of pollutant emissions, cross-regional collaboration and cooperation should be strengthened, an effective information-sharing mechanism should be established, more stringent corporate emission standards should be formulated, environmental protection inspections should be strengthened, and enterprises should be encouraged to adopt cleaner production technologies and advanced equipment to reduce industrial pollutant emissions. Inter-regional consultations, joint actions, and the joint formulation of policy measures should be implemented to achieve the goal of collaborative governance. These measures will produce synergistic effects in the areas of industry, energy, transportation, and environmental protection, thus realizing the synergistic development of the economy and the environment.
However, this study still has some limitations. First, due to the difficulty in obtaining energy consumption data for each prefecture-level city, it failed to further consider the energy structure related to PM2.5 concentration, such as the proportion of new energy sources like wind energy and electric energy and the use of new energy vehicles in transportation. In addition, this study only explored the coupled and coordinated relationship between PM2.5-influencing factors under different scenarios of governance and failed to explore the coupled and coordinated relationship of PM2.5 synergistic governance in terms of the three systems of economic, social, and ecological benefits. Future research can conduct an in-depth analysis on the evolution process and coupled coordination relationship of PM2.5 synergistic governance by establishing a comprehensive PM2.5 governance index system and combining it with a coupled coordination degree model. By comprehensively considering multiple aspects such as economic, social, and ecological benefits, we can gain a more comprehensive understanding of the effects and mutual influences of PM2.5 governance, promote the effective synergistic utilization of resources in all aspects, and provide a scientific basis for decision-making on future PM2.5 governance strategies.

Author Contributions

Conceptualization, Daichao Li, Kunkun Fan and Cong Li; methodology, Daichao Li, Kunkun Fan and Cong Li; validation, Kunkun Fan and Cong Li; formal analysis, Kunkun Fan, Cong Li and Xinlei Jin; data curation, Xinlei Jin, Fei Ding and Zhan Zeng; writing—original draft preparation, Daichao Li and Kunkun Fan; writing—review and editing, Daichao Li and Kunkun Fan; supervision, Daichao Li and Zhan Zeng; project administration, Daichao Li. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by the Strategic Priority Research Program of the Chinese Academy of Sciences, grant number XDA23100502.

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study.

Data Availability Statement

The data of this work can be shared to the readers depending on the request.

Acknowledgments

We thank the editors of the ISPRS International Journal of Geo-Information reviewers for their valuable suggestions.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Zhang, L.; Wilson, J.P.; Zhao, N.; Zhang, W.; Wu, Y. The dynamics of cardiovascular and respiratory deaths attributed to long-term PM2.5 exposures in global megacities. Sci. Total Environ. 2022, 842, 156951. [Google Scholar] [CrossRef] [PubMed]
  2. Zhang, L.; Zhao, N.; Zhang, W.; Wilson, J.P. Changes in Long-Term PM2.5 Pollution in the Urban and Suburban Areas of China’s Three Largest Urban Agglomerations from 2000 to 2020. Remote Sens. 2022, 14, 1716. [Google Scholar] [CrossRef]
  3. Ma, X.; Jia, H. Particulate matter and gaseous pollutions in three megacities over China: Situation and implication. Atmos. Environ. 2016, 140, 476–494. [Google Scholar] [CrossRef]
  4. Zhang, L.; Wilson, J.P.; MacDonald, B.; Zhang, W.; Yu, T. The changing PM2.5 dynamics of global megacities based on long-term remotely sensed observations. Environ. Int. 2020, 142, 105862. [Google Scholar] [CrossRef] [PubMed]
  5. Chen, Z.; Chen, D.; Zhao, C.; Kwan, M.; Cai, J.; Zhuang, Y.; Zhao, B.; Wang, X.; Chen, B.; Yang, J. Influence of meteorological conditions on PM2.5 concentrations across China: A review of methodology and mechanism. Environ. Int. 2020, 139, 105558. [Google Scholar] [CrossRef] [PubMed]
  6. Zhu, M.; Guo, J.; Zhou, Y.; Cheng, X. Exploring the spatiotemporal evolution and socioeconomic determinants of PM2.5 distribution and its hierarchical management policies in 366 Chinese cities. Front. Public Health 2022, 10, 843862. [Google Scholar] [CrossRef] [PubMed]
  7. Zamani Joharestani, M.; Cao, C.; Ni, X.; Bashir, B.; Talebiesfandarani, S. PM2.5 prediction based on random forest, XGBoost, and deep learning using multisource remote sensing data. Atmosphere 2019, 10, 373. [Google Scholar] [CrossRef]
  8. Qi, Y.; Li, Q.; Karimian, H.; Liu, D. A hybrid model for spatiotemporal forecasting of PM2.5 based on graph convolutional neural network and long short-term memory. Sci. Total Environ. 2019, 664, 1–10. [Google Scholar] [CrossRef]
  9. Ge, L.; Wu, K.; Zeng, Y.; Chang, F.; Wang, Y.; Li, S. Multi-scale spatiotemporal graph convolution network for air quality prediction. Appl. Intell. 2021, 51, 3491–3505. [Google Scholar] [CrossRef]
  10. Wang, P.; Feng, H.; Bi, X.; Fu, Y.; He, X.; Zhang, G.; Niu, J. Phase objectives analysis for PM2.5 reduction using dynamics forecasting approach under different scenarios of PGDP decline. Ecol. Indic. 2021, 129, 108003. [Google Scholar] [CrossRef]
  11. Ceng, P.; Wei, X.; Duan, Z. Study on spatial and temporal differentiation and optimization path of collaborative pollution control in urban agglomerations of China. Resour. Environ. Yangtze Basin 2023, 32, 1317–1333. [Google Scholar]
  12. Cheng, N.; Zhang, D.; Li, Y.; Xie, X.; Chen, Z.; Meng, F.; Gao, B.; He, B. Spatio-temporal variations of PM2.5 concentrations and the evaluation of emission reduction measures during two red air pollution alerts in Beijing. Sci. Rep. 2017, 7, 8220. [Google Scholar] [CrossRef] [PubMed]
  13. Zhang, H.; Wang, S.; Hao, J.; Wang, X.; Wang, S.; Chai, F.; Li, M. Air pollution and control action in Beijing. J. Clean. Prod. 2016, 112, 1519–1527. [Google Scholar] [CrossRef]
  14. Liu, J.; Kiesewetter, G.; Klimont, Z.; Cofala, J.; Heyes, C.; Schöpp, W.; Zhu, T.; Cao, G.; Sanabria, A.G.; Sander, R. Mitigation pathways of air pollution from residential emissions in the Beijing-Tianjin-Hebei region in China. Environ. Int. 2019, 125, 236–244. [Google Scholar] [CrossRef] [PubMed]
  15. Ye, C.; Chen, R.; Chen, M.; Ye, X. A new framework of regional collaborative governance for PM2.5. Stoch. Environ. Res. Risk Assess. 2019, 33, 1109–1116. [Google Scholar] [CrossRef]
  16. Wang, Y.; Liu, Z.; Huang, L.; Lu, G.; Gong, Y.; Yaluk, E.; Li, H.; Yi, X.; Yang, L.; Feng, J. Development and evaluation of a scheme system of joint prevention and control of PM2.5 pollution in the Yangtze River Delta region, China. J. Clean. Prod. 2020, 275, 122756. [Google Scholar] [CrossRef]
  17. Li, W.; Yang, G.; Qian, X. The socioeconomic factors influencing the PM2.5 levels of 160 cities in China. Sustain. Cities Soc. 2022, 84, 104023. [Google Scholar] [CrossRef]
  18. Yu, Q. Analysis of PM2.5 Influencing Factors Based on Various Statistical Methods—A Case Study of Beijing in 2021; Atlantis Press: Paris, France, 2022; pp. 195–204. [Google Scholar]
  19. Zhao, H.; Liu, Y.; Gu, T.; Zheng, H.; Wang, Z.; Yang, D. Identifying spatiotemporal heterogeneity of PM2.5 concentrations and the key influencing factors in the middle and lower reaches of the yellow river. Remote Sens. 2022, 14, 2643. [Google Scholar] [CrossRef]
  20. Shahriar, S.A.; Kayes, I.; Hasan, K.; Hasan, M.; Islam, R.; Awang, N.R.; Hamzah, Z.; Rak, A.E.; Salam, M.A. Potential of ARIMA-ANN, ARIMA-SVM, DT and CatBoost for atmospheric PM2.5 forecasting in Bangladesh. Atmosphere 2021, 12, 100. [Google Scholar] [CrossRef]
  21. Liu, B.; Jin, Y.; Xu, D.; Wang, Y.; Li, C. A data calibration method for micro air quality detectors based on a LASSO regression and NARX neural network combined model. Sci. Rep. 2021, 11, 21173. [Google Scholar] [CrossRef]
  22. Zhang, G.; Rui, X.; Fan, Y. Critical review of methods to estimate PM2.5 concentrations within specified research region. ISPRS Int. J. Geo-Inf. 2018, 7, 368. [Google Scholar] [CrossRef]
  23. Yue, H.; Duan, L.; Lu, M.; Huang, H.; Zhang, X.; Liu, H. Modeling the Determinants of PM2.5 in China Considering the Localized Spatiotemporal Effects: A Multiscale Geographically Weighted Regression Method. Atmosphere 2022, 13, 627. [Google Scholar] [CrossRef]
  24. Bai, Y.; Zeng, B.; Li, C.; Zhang, J. An ensemble long short-term memory neural network for hourly PM2.5 concentration forecasting. Chemosphere 2019, 222, 286–294. [Google Scholar] [CrossRef] [PubMed]
  25. Stafoggia, M.; Bellander, T.; Bucci, S.; Davoli, M.; De Hoogh, K.; De’Donato, F.; Gariazzo, C.; Lyapustin, A.; Michelozzi, P.; Renzi, M. Estimation of daily PM10 and PM2.5 concentrations in Italy, 2013–2015, using a spatiotemporal land-use random-forest model. Environ. Int. 2019, 124, 170–179. [Google Scholar] [CrossRef] [PubMed]
  26. Zhong, J.; Zhang, X.; Gui, K.; Wang, Y.; Che, H.; Shen, X.; Zhang, L.; Zhang, Y.; Sun, J.; Zhang, W. Robust prediction of hourly PM2.5 from meteorological data using LightGBM. Nat. Sci. Rev. 2021, 8, nwaa307. [Google Scholar] [CrossRef] [PubMed]
  27. Wang, Z.; Zhou, Y.; Zhao, R.; Wang, N.; Biswas, A.; Shi, Z. High-resolution prediction of the spatial distribution of PM2.5 concentrations in China using a long short-term memory model. J. Clean. Prod. 2021, 297, 126493. [Google Scholar] [CrossRef]
  28. Pan, B. Application of XGBoost Algorithm in Hourly PM2.5 Concentration Prediction; IOP Publishing: Wales, UK, 2018; p. 12127. [Google Scholar]
  29. Du, P.; Wang, J.; Hao, Y.; Niu, T.; Yang, W. A novel hybrid model based on multi-objective Harris hawks optimization algorithm for daily PM2.5 and PM10 forecasting. Appl. Soft Comput. 2020, 96, 106620. [Google Scholar] [CrossRef]
  30. Moisan, S.; Herrera, R.; Clements, A. A dynamic multiple equation approach for forecasting PM2.5 pollution in Santiago, Chile. Int. J. Forecast. 2018, 34, 566–581. [Google Scholar] [CrossRef]
  31. Wang, P.; Zhang, H.; Qin, Z.; Zhang, G. A novel hybrid-Garch model based on ARIMA and SVM for PM2.5 concentrations forecasting. Atmos. Pollut. Res. 2017, 8, 850–860. [Google Scholar] [CrossRef]
  32. Zhou, Y.; Chang, F.; Chang, L.; Kao, I.; Wang, Y.; Kang, C. Multi-output support vector machine for regional multi-step-ahead PM2.5 forecasting. Sci. Total Environ. 2019, 651, 230–240. [Google Scholar] [CrossRef]
  33. Babu, C.N.; Reddy, B.E. A moving-average filter based hybrid ARIMA–ANN model for forecasting time series data. Appl. Soft Comput. 2014, 23, 27–38. [Google Scholar] [CrossRef]
  34. Gairaa, K.; Khellaf, A.; Messlem, Y.; Chellali, F. Estimation of the daily global solar radiation based on Box–Jenkins and ANN models: A combined approach. Renew. Sustain. Energy Rev. 2016, 57, 238–249. [Google Scholar] [CrossRef]
  35. Zhang, X.; Fung, J.C.; Zhang, Y.; Lau, A.K.; Leung, K.K.; Huang, W.W. Assessing PM2.5 emissions in 2020: The impacts of integrated emission control policies in China. Environ. Pollut. 2020, 263, 114575. [Google Scholar] [CrossRef] [PubMed]
  36. Wu, R.; Dai, H.; Geng, Y.; Xie, Y.; Masui, T.; Liu, Z.; Qian, Y. Economic impacts from PM2.5 pollution-related health effects: A case study in Shanghai. Environ. Sci. Technol. 2017, 51, 5035–5042. [Google Scholar] [CrossRef] [PubMed]
  37. Shi, Y.; Zhu, Y.; Gong, S.; Pan, J.; Zang, S.; Wang, W.; Li, Z.; Matsunaga, T.; Yamaguchi, Y.; Bai, Y. PM2.5-related premature deaths and potential health benefits of controlled air quality in 34 provincial cities of China during 2001–2017. Environ. Impact Assess. 2022, 97, 106883. [Google Scholar] [CrossRef]
  38. Ju, H.; Zhang, S.; Yan, Y. Spatial pattern changes of urban expansion and multi-dimensional analysis of driving forces in the Guangdong-Hong Kong-Macao Greater Bay Area in 1980–2020. Acta Geogr. Sin. 2022, 77, 1086–1101. [Google Scholar]
  39. Wei, J.; Li, Z.; Lyapustin, A.; Sun, L.; Peng, Y.; Xue, W.; Su, T.; Cribb, M. Reconstructing 1-km-resolution high-quality PM2.5 data records from 2000 to 2018 in China: Spatiotemporal variations and policy implications. Remote Sens. Environ. 2021, 252, 112136. [Google Scholar] [CrossRef]
  40. Lai, W.D.; Deng, Z.X. The influence and countermeasures of the “COVID-19” on the economic development of coastal economic zone of guangdong province. Int. J. Econ. Behav. Organ. 2020, 8, 44–48. [Google Scholar] [CrossRef]
  41. Zhu, L.; Li, T.; Ma, L.; Liu, Z. The influence of industrial structure adjustment on Haze Pollution: An empirical study of Jing-jin-ji region. Ecol. Econ. 2018, 34, 141–148. [Google Scholar]
  42. Li, G.; Fang, C.; Wang, S.; Sun, S. The effect of economic growth, urbanization, and industrialization on fine particulate matter (PM2.5) concentrations in China. Environ. Sci. Technol. 2016, 50, 11452–11459. [Google Scholar] [CrossRef]
  43. He, X.; Lin, Z.; Liu, H.; Qi, X. Analysis of the driving factors of PM2.5 in Jiangsu province based on grey correlation model. Acta Geogr. Sin. 2016, 71, 1119–1129. [Google Scholar]
  44. Hoerl, A.E.; Kennard, R.W. Ridge regression: Applications to nonorthogonal problems. Technometrics 1970, 12, 69–82. [Google Scholar] [CrossRef]
  45. Hoerl, A.E.; Kennard, R.W. Ridge regression: Biased estimation for nonorthogonal problems. Technometrics 1970, 12, 55–67. [Google Scholar] [CrossRef]
  46. Vapnik, V. The Nature of Statistical Learning Theory; Springer Science & Business Media: Berlin/Heidelberg, Germany, 1999. [Google Scholar]
  47. Joachims, T. Making Large-Scale SVM Learning Practical; Technical Report; Universität Dortmund: Dortmund, Germany, 1998. [Google Scholar]
  48. Li, Y.; Li, Y.; Zhou, Y.; Shi, Y.; Zhu, X. Investigation of a coupling model of coordination between urbanization and the environment. J. Environ. Manag. 2012, 98, 127–133. [Google Scholar] [CrossRef] [PubMed]
  49. Liu, Y.B.; Li, R.D.; Song, X.F. Grey associative analysis of regional urbanization and eco-environment coupling in China. Acta Geogr. Sin. 2005, 60, 237–247. [Google Scholar]
  50. Liu, Y.; Li, R.; Li, C. Scenarios simulation of coupling system between urbanization and eco-environment in Jiangsu province based on system dynamics model. Chin. Geogr. Sci. 2005, 15, 219–226. [Google Scholar] [CrossRef]
  51. Chen, M.; Lu, D.; Zhang, H. Comprehensive evaluation and the driving factors of China’s urbanization. Acta Geogr. Sin. 2009, 64, 387–398. [Google Scholar]
  52. Xu, Y.; Fu, H. Descending dimension method in linear algrbra. Coll. Math. 2020, 36, 111–117. [Google Scholar]
Ijgi 12 00340 g001
Figure 1. Location of the study area.
Figure 1. Location of the study area.
Ijgi 12 00340 g001
Ijgi 12 00340 g002
Figure 2. Technical roadmap.
Figure 2. Technical roadmap.
Ijgi 12 00340 g002
Ijgi 12 00340 g003
Figure 3. Flowchart of scenario construction methodology.
Figure 3. Flowchart of scenario construction methodology.
Ijgi 12 00340 g003
Ijgi 12 00340 g004
Figure 4. Correlation between PM2.5 and influencing factors in various cities.
Figure 4. Correlation between PM2.5 and influencing factors in various cities.
Ijgi 12 00340 g004
Ijgi 12 00340 g005
Figure 5. The predicted annual average concentration of PM2.5 in Guangdong Province from 2020 to 2025.
Figure 5. The predicted annual average concentration of PM2.5 in Guangdong Province from 2020 to 2025.
Ijgi 12 00340 g005
Ijgi 12 00340 g006
Figure 6. Coupling and coordination relationship of PM2.5-pollution-influencing factors in Guangdong Province from 2020 to 2025.
Figure 6. Coupling and coordination relationship of PM2.5-pollution-influencing factors in Guangdong Province from 2020 to 2025.
Ijgi 12 00340 g006
Table 1. Standards for division of coordination level and coordination development degree.
Table 1. Standards for division of coordination level and coordination development degree.
Coordination DegreeCoordination LevelCoordination DegreeCoordination Level
[0, 0.1)Extreme imbalance[0.5, 0.6)Reluctant coordination
[0.1, 0.2)Serious imbalance[0.6, 0.7)Primary coordination
[0.2, 0.3)Moderate disorder[0.7, 0.8)Intermediate coordination
[0.2, 0.3)Mild disorder[0.8, 0.9)Good coordination
[0.4, 0.5)Endangered disorder[0.9, 1)Quality coordination
Table 2. Comparison of the accuracy of PM2.5 prediction results for three models by city.
Table 2. Comparison of the accuracy of PM2.5 prediction results for three models by city.
CityRidgeSVMRR-SVM
MAE/(μg·m−3)RMSE/(μg·m−3)R2MAE/(μg·m−3)RMSE/(μg·m−3)R2MAE/(μg·m−3)RMSE/(μg·m−3)R2
Chaozhou0.75860.97770.94931.29421.55360.87190.52000.62750.9791
Dongguan1.01531.49540.92691.19741.44230.93200.48590.71840.9831
Foshan1.57271.88790.90801.39621.83000.91360.44780.49590.9937
Guangzhou1.30841.66780.92930.90401.13500.96730.61961.03970.9725
Heyuan1.03931.31870.93842.25872.92730.69660.78321.09250.9577
Huizhou0.90731.05300.95691.37602.07460.83260.32460.36540.9948
Jiangmen1.04441.35880.90951.11141.37010.90800.39260.47500.9889
Jieyang0.77010.86540.95641.21821.43420.88030.62500.77320.9652
Maoming0.79260.90790.95892.43162.96500.66880.54610.84460.9731
Meizhou1.65252.03680.89343.79074.49740.48041.25311.63520.9313
Qingyuan0.83051.22000.95042.09222.33810.81800.54010.70010.9837
Shantou0.83450.95890.95952.68833.11390.57300.61090.75440.9749
Shanwei1.33671.74190.92582.20402.75850.81390.82901.25820.9613
Shaoguan1.14241.23430.94461.42792.12500.83580.65690.99240.9642
Shenzhen1.10221.45840.90641.47241.79320.85850.84231.05940.9506
Yangjiang0.71860.89810.94021.34651.75690.77130.39070.47370.9834
Yunfu1.40121.53570.93013.08423.99020.52790.97171.26630.9524
Zhanjiang1.58751.91210.86712.45112.94790.68400.68620.96040.9665
Zhaoqing1.30981.43480.91362.10752.64620.70620.58710.90060.9660
Zhongshan1.08971.40740.92092.25362.46440.75760.72460.94550.9643
Zhuhai0.98811.25810.92011.65432.15690.76530.55320.83120.9651
Table 3. Base scenario library.
Table 3. Base scenario library.
FactorLowerMiddleHigh
Low WidthLow StrongLow SeverityMiddle WidthMiddle StrongMiddle SeverityHigh WidthHigh StrongHigh Severity
SGDPDown 0.1%Down 0.2%Down 0.3%Down 0.4%Down 0.6%Down 0.8%Down 1.0%Down 1.5%Down 2.0%
TGDPUp 0.1%Up 0.2%Up 0.3%Up 0.4%Up 0.6%Up 0.8%Up 1.0%Up 1.5%Up 2.0%
ECPGDown [11%]Down [12%]Down [13%]Down [14%]Down [15%]Down [16%]Down [17%]Down [18%]Down [19%]
HFTFDown 0.1%Down 0.2%Down 0.3%Down 0.5%Down 0.7%Down 0.9%Down 1.0%Down 1.5%Down 2.0%
WFTFUp 0.1%Up 0.2%Up 0.3%Up 0.5%Up 0.7%Up 0.9%Up 1.0%Up 1.5%Up 2.0%
GROCDown 2%Down 3%Down 4%Down 5%Down 6%Down 7%Down 8%Down 10%Down 12%
GAPCUp [5%]Up [10%]Up [15%]Up [20%]Up [25%]Up [30%]Up [35%]Up [40%]Up [45%]
ISDEDown [10%]Down [15%]Down [20%]Down [25%]Down [30%]Down [35%]Down [40%]Down [45%]Down [50%]
Note: Values with “[]” indicate the total amount of change for the period 2020–2025, while those without “[]” indicate the amount of change per year.
Table 4. Classification results of indicators.
Table 4. Classification results of indicators.
RankSTIRECPG/(tce·10kCHY−1)WFHRGROC/%GAPC/m2ISDE/10 kt
10.00–0.380.31–0.430.00–0.050.00–0.0311.5–14.90.07–0.21
20.38–0.630.43–0.630.05–0.140.03–0.1214.9–19.10.21–0.41
30.63–0.830.63–1.630.14–0.340.12–0.1819.1–23.80.41–0.80
40.83–1.001.63–1.740.34–1.230.18–0.22 1.80–1.63
51.00–1.33
Note: STIR is the ratio between the share of the secondary sector in GDP (SGPD) and the share of the tertiary sector in GDP (TGDP); ECPG is the energy consumption per unit of GDP; WHFR is the ratio between the share of waterborne freight traffic in total freight traffic (WHTF) and the share of highway freight traffic in total freight traffic (HFTF); GROC is the rate of growth of civilian car ownership; GAPC is the per capita area of green space in the city; and ISDE is the sulfur dioxide emissions of the urban industry.
Table 5. Clustering results of each indicator, classification data, and city classification results.
Table 5. Clustering results of each indicator, classification data, and city classification results.
CategoryCitySTIRECPGWHFRGROCGAPCISDE
Category 1Guangzhou114212
Meizhou211313
Qingyuan224234
Shaogua223323
Shenzhen213131
Yunfu223423
Category 2Heyuan221331
Jieyang331333
Maoming322322
Shanwei321431
Yangjiang331323
Category 3Chaozhou513232
Dongguan514213
Foshan512223
Shantou412232
Zhongshan422221
Category 4Huizhou544224
Jiangmen423213
Zhangjiang343322
Zhaoqing433214
Zhuhai333211
Note: The numbers in the table are the classification levels of the corresponding indicators.
Table 6. Regulation of demand in each regulation mode.
Table 6. Regulation of demand in each regulation mode.
CategorySTIRECPGWHFRGROCGAPCISDE
Category 1Lower ↓Lower ↓High ↓Middle ↓Middle ↑High ↓
Category 2Middle ↓Middle ↓Lower ↓High ↓High ↑Middle ↓
Category 3High ↓Lower ↓Middle ↓Lower ↓Middle ↑Middle ↓
Category 4High ↓High ↓High ↓Lower ↓Lower ↑High ↓
Note: “↑” and “↓” in the table represent the increase or decrease in the index, respectively.
Table 7. PM2.5 governance scenarios for each city category.
Table 7. PM2.5 governance scenarios for each city category.
FactorCategory 1Category 2Category 3Category 4
Loose TypeIntensive TypeStrict TypeLoose TypeIntensive TypeStrict TypeLoose TypeIntensive TypeStrict TypeLoose TypeIntensive TypeStrict Type
STIRlow widthlow strengthlow
strictness		medium widthmedium strengthmedium strictnesshigh widthhigh strengthhigh
strictness		high widthhigh strengthhigh strictness
ECPGlow widthlow strengthlow
strictness		medium widthmedium strengthmedium strictnesslow widthlow strengthlow
strictness		high widthhigh strengthhigh strictness
WHFRhigh widthhigh strengthhigh
strictness		low widthlow strengthlow
strictness		medium widthmedium strengthmedium strictnesshigh widthhigh strengthhigh strictness
GROCmedium widthmedium strengthmedium strictnesshigh widthhigh strengthhigh
strictness		low widthlow strengthlow
strictness		low widthlow strengthlow strictness
GAPCmedium widthmedium strengthmedium strictnesshigh widthhigh strengthhigh
strictness		medium widthmedium strengthmedium strictnesslow widthlow strengthlow strictness
ISDEhigh widthhigh strengthhigh
strictness		medium widthmedium strengthmedium strictnessmedium widthmedium strengthmedium strictnesshigh widthhigh strengthhigh strictness
Note: The fields in this table correspond to those in Table 6. For example, “low width” in energy intensity corresponds to “low width” in energy intensity in Table 4, that is, [11%].
Table 8. PM2.5 pollution control paths for each city.
Table 8. PM2.5 pollution control paths for each city.
CitySGDPTGDPECPGHFTFWFTFGROCGAPCISDE
Chaozhou down 1.0% up 1.0% down 2.48% down 0.5% up 0.5% down 2.0% up 3.08% down 4.68%
Dongguan down 2.0% up 2.0% down 2.86% down 0.9% up 0.9% down 4.0% up 4.47% down 6.92%
Foshan down 2.0% up 2.0% down 2.86% down 0.9% up 0.9% down 4.0% up 4.47% down 6.92%
Guangzhou down 0.3% up 0.3% down 2.29% down 2.0% up 2.0% down 7.0% up 4.47% down 10.91%
Heyuandown 0.4% up 0.4% down 2.48% down 0.1% up 0.1% down 8.0% up 5.13% down 4.68%
Huizhou down 1.5% up 1.5% down 3.25% down 1.5% up 1.5% down 3.0% up 1.60% down 9.48%
Jiangmen down 1.5% up 1.5% down 3.25% down 1.5% up 1.5% down 3.0% up 1.60% down 9.48%
Jieyang down 0.8% up 0.8% down 2.86% down 0.3% up 0.3% down 12.0% up 6.38% down 6.92%
Maoming down 0.6% up 0.6% down 2.67% down 0.2% up 0.2% down 10.0% up 5.77% up 5.77%
Meizhou down 0.2% up 0.2% down 2.11% down 1.5% up 1.5% down 6.0% up 3.78% down 9.48%
Qingyuandown 0.1% up 0.1% down 1.92% down 1.0% up 1.0% down 5.0% up 3.08% down 8.16%
Shantou down 1.0% up 1.0% down 2.48% down 0.5% up 0.5% down 2.0% up 3.08% down 4.68%
Shanweidown 0.4% up 0.4% down 2.48% down 0.1% up 0.1% down 8.0% up 5.13% down 4.68%
Shaoguan down 0.3% up 0.3% down 2.29% down 2.0% up 2.0% down 7.0% up 4.47% down 10.91%
Shenzhen down 0.3% up 0.3% down 2.29% down 2.0% up 2.0% down 7.0% up 4.47% down 10.91%
Yangjiang down 0.6% up 0.6% down 2.67% down 0.2% up 0.2% down 10.0% up 5.77% up 5.77%
Yunfu down 0.3% up 0.3% down 2.29% down 2.0% up 2.0% down 7.0% up 4.47% down 10.91%
Zhanjiang down 1.5% up 1.5% down 3.25% down 1.5% up 1.5% down 3.0% up 1.60% down 9.48%
Zhaoqing down 1.5% up 1.5% down 3.25% down 1.5% up 1.5% down 3.0% up 1.60% down 9.48%
Zhongshan down 1.0% up 1.0% down 2.48% down 0.5% up 0.5% down 2.0% up 3.08% down 4.68%
Zhuhai down 1.0% up 1.0% down 3.05% down 1.0% up 1.0% down 2.0% up 0.82% down 8.16%
Note: The data in the table are year-to-year changes.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Fan, K.; Li, D.; Li, C.; Jin, X.; Ding, F.; Zeng, Z. Analysis of PM2.5 Synergistic Governance Path from a Socio-Economic Perspective: A Case Study of Guangdong Province. ISPRS Int. J. Geo-Inf. 2023, 12, 340. https://doi.org/10.3390/ijgi12080340

AMA Style

Fan K, Li D, Li C, Jin X, Ding F, Zeng Z. Analysis of PM2.5 Synergistic Governance Path from a Socio-Economic Perspective: A Case Study of Guangdong Province. ISPRS International Journal of Geo-Information. 2023; 12(8):340. https://doi.org/10.3390/ijgi12080340

Chicago/Turabian Style

Fan, Kunkun, Daichao Li, Cong Li, Xinlei Jin, Fei Ding, and Zhan Zeng. 2023. "Analysis of PM2.5 Synergistic Governance Path from a Socio-Economic Perspective: A Case Study of Guangdong Province" ISPRS International Journal of Geo-Information 12, no. 8: 340. https://doi.org/10.3390/ijgi12080340

APA Style

Fan, K., Li, D., Li, C., Jin, X., Ding, F., & Zeng, Z. (2023). Analysis of PM2.5 Synergistic Governance Path from a Socio-Economic Perspective: A Case Study of Guangdong Province. ISPRS International Journal of Geo-Information, 12(8), 340. https://doi.org/10.3390/ijgi12080340

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop