Analysis of the Spatial and Temporal Evolution of Energy-Related CO₂ Emissions in China’s Coastal Areas and the Drivers of Industrial Enterprises above Designated Size—The Case of 82 Cities

Abstract

The energy consumption by industrial enterprises above designated size in China’s coastal region is the main source of CO₂ emissions. This study analyzes the spatial and temporal evolution patterns and driving factors of CO₂ emissions due to the energy consumption by industrial enterprises above designated size. Enterprises in 82 cities in China’s coastal regions were studied from 2005 to 2020 based on their CO₂ emissions and socio-economic data. The Exploring Spatial Data Analysis (ESDA) methodology and Logarithmic mean Divisia Index decomposition (LMDI model) were used. The results show that, during the study period, energy-related CO₂ emissions from industrial enterprises above designated size in China’s coastal areas generally show a fluctuating upward trend. However, a few cities showed a trend from steady growth to a peak and then a slow decline, which may realize the “double carbon” target in advance. The spatial correlation of CO₂ emission intensity showed a decreasing and then increasing trend, and there were spatial aggregation characteristics in some cities. Among the driving factors, the pull effect is higher than the inhibition effect; the output scale contributes the most to the pull effect, and labor productivity contributes the most to the inhibition effect. The results of this study have a certain reference value for the realization of the “double carbon” target in China’s coastal regions.

1. Introduction

Jinping Xi, General Secretary of the CPC (Communist Party of China) Central Committee, during his inspection in Jiangsu, emphasized that cities are important carriers of modernization, but they are also places with the highest population density and concentrated pollution emissions. To achieve harmonious coexistence between humans and nature in the process of modernization, protecting the ecological environment of cities must be given a more prominent position [1]. The report of the 20th Party Congress further points out that promoting green and low-carbon economic and social development is a key link to achieving high-quality development [2]. The 2023 Report on the Work of the Government clearly puts the promotion of economic development to green transformation and the prevention and control of environmental pollution as the top priorities of this year’s government work report [3]. As an important agglomeration area for China’s industrial development, China’s coastal region has become responsible for high energy consumption and CO₂ emissions due to its high population density, high economic density and high energy intensity. Most of the coastal cities are cities with fast economic development in China, and some of them are also rich in coal resources, and the rapid development of their industrial enterprises above designated size economy is accompanied by an increase in CO₂ emissions caused by energy consumption, which has a greater impact on the urban environment. Therefore, the study of the spatial and temporal evolution of energy-related CO₂ emissions from cities in China’s coastal regions is key to promoting the prevention and control of environmental pollution in China’s coastal regions. It is also the most important task regarding the green transformation of the development mode of cities in China’s coastal regions and the construction of low-carbon cities. So, what has been the development trend in CO₂ emissions from industrial enterprises above designated size in China’s coastal regions in recent years? What are the main driving factors? What are the different characteristics between regions? Answering these questions will help us to understand the basic situation concerning energy-related CO₂ emissions from industrial enterprises above designated size in China’s coastal regions, the regional differences and evolutionary trends, as well as the main driving factors. It will also provide theoretical references for the realization of the “dual carbon goal”.

In recent years, with the introduction of carbon peaking and carbon neutrality targets, research on how to reduce CO₂ emissions has once again become a hot topic of academic interest. For example, Wang et al. [3,4,5] accounted for CO₂ emissions in Chinese cities and studied their drivers and found that different urban types in China have a different focus on CO₂ emission drivers, and that there are obvious spatial differences. Tian et al. [6] studied the CO₂ emissions of end-use energy consumption in Jiangsu Province and added secondary energy indicators to the total CO₂ emissions. Gu et al. [7] studied carbon emissions in metropolitan areas of Zhejiang Province and concluded that studying the relationship between metropolitan area construction and carbon emissions would be beneficial to urban emission reduction. In addition, scholars have also paid more attention to the CO₂ emissions of urban agglomerations and key cities; for example, Lin et al. [8] analyzed the spatial and temporal evolution characteristics and influencing factors of the industrial carbon emission efficiency of the Beijing Tianjin Hebei urban agglomeration and found that the industrial carbon emission efficiency of the Beijing Tianjin Hebei urban agglomeration is increasing; Cai et al. [9] concluded that the spatial characteristics of CO₂ emissions of the Yangtze River Delta urban agglomeration are driven and influenced by typical urban areas; Wu et al. [10,11,12,13] studied the carbon emission factors of energy consumption in Wuhan city and found that the main factors promoting the growth of its carbon emissions came from population expansion and economic growth.

Scholars at home and abroad have conducted a large amount of research on the drivers affecting CO₂ emissions, and the research methods have mainly included the STIRPAT model [14,15], the LMDI decomposition model, geographically weighted regression [16,17],and the environmental Kuznets curve [18,19]. Ang [20] found that the LMDI decomposition model requires easy access to data, can reasonably decompose factors and has no residual values in the results, and the method has been widely used in recent years. In the study of CO₂ emission drivers, economic development, population size, energy structure, industrial structure and energy intensity are generally considered as the five indicators [21,22,23,24,25,26,27,28,29,30,31]; however, in recent years, different scholars have expanded the drivers of CO₂ emission change into multiple indicators according to the research needs. Yang et al. [32] added transportation carbon intensity and transportation structure factors to the above five indicators in order to study the change in transportation carbon emissions in the Yangtze River economic zone; Liu et al. [33] concluded that electricity intensity and investment efficiency were the main factors inhibiting the growth of electricity consumption; Liu et al. [34] added technology state, labor input and labor force to their study and found that capital input is the key factor driving carbon emissions.

A review of the literature showed that there is room for expanding the existing studies: In terms of research regions, most studies have focused on China as a whole [3,4,5], provinces [35], urban agglomerations [36,37,38,39,40,41] and key cities [7,29]. Among these, urban agglomerations are mainly coastal urban agglomerations with high levels of economic development such as Beijing Tianjin Hebei, the Yangtze River Delta, Guangdong, Hong Kong, Macao and Chengdu Chongqing, as well as the Yellow River Basin, Liaoning and the Central South urban agglomerations and other important Chinese areas with high energy usage. The key cities are mainly provincial capitals and cities with high economic development. It can be seen that most of the research areas are located in China’s coastal provinces, and most of the research is based on comparative studies of isolated points. We have not found any study that took all the cities in China’s coastal region as the research sample. China’s coastal region brings together China’s economically developed cities, resource cities, and large cities with large carbon emissions, so integrating the research will be more conducive to grasping the realization of the “double carbon” goal of the region from an overall perspective. In addition, there have been fewer regional studies that analyze the spatial heterogeneity of CO₂ emissions from a geographic perspective, and China’s coastal region includes most of the cities in the south and the north of China. Therefore, this study spans a larger scale, and the area can be divided into the three major coastal regions of the north, the middle and the south for detailed study. From the perspective of the research methods, in terms of accounting for and estimating total CO₂ emissions, scholars have focused on accounting for small-scale studies and estimating for large-scale studies due to the availability of data and large workload. In addition, most of the unified caliber of energy consumption data in the city-level statistical yearbook is mainly in industrial enterprises above designated size. In 2022, various government departments jointly released the “Implementation Plan for Carbon Peaking in Industry”, which mentioned focusing on key industries and formulating the implementation plan for carbon peaking in iron and steel, building materials, petrochemicals and chemicals, non-ferrous metals, etc. Among them, the key industries are mainly industrial enterprises above designated size, while the research specifically concerning industrial enterprises above designated size is scarce. Scholars’ discussions regarding carbon emission drivers provide a good basis for this study, but, unfortunately, most of the literature has used a relatively small number of energy types, which makes it difficult to reflect the actual carbon emissions and carbon intensity changes in regions. There has been less analysis carried out on the regional correlation of carbon emissions.

Compared with previous studies, this study aims to explore and address the following questions: ① In terms of regional selection, this study specifies the research object as above a designated size in China’s coastal regions in order to address the problem of insufficient attention having been paid to this topic of China’s coastal region’s industrial enterprises above designated size in the existing research; ② In terms of research content, this study accounts for the CO₂ emissions and intensity triggered by the energy consumption by large-scale industrial enterprises above designated size in China’s coastal region over the past 15 years. It uses exploratory spatial data analysis to explore the spatial and temporal characteristics of the CO₂ emissions of China’s coastal region as a whole and the cities above each level on the basis of accounting. On the other hand, this study draws on the information from previous expert studies and selects four drivers of CO₂ emissions, namely output scale, energy structure, industrial structure and per capita energy consumption intensity. Additionally, it considers the element of labor force. In order to level the constant equation, the indicator of labor productivity is introduced into the LDMI model, which breaks the traditional formula of the drivers of CO₂ emissions. Moreover, it further clarifies the drivers of CO₂ emissions of industrialized industries along the coastal regions of China. In addition, labor productivity is introduced into the LMDI model to break the traditional drivers of CO₂ emissions and to further clarify the drivers of CO₂ emissions from industrial enterprises above designated size in China’s coastal regions; ③ In terms of research methodology, a cross-disciplinary integration is realized. Furthermore, the regional color of geography is introduced into the field of economics to analyze the spatial and temporal heterogeneity of CO₂ emissions within the region by each driver, to analyze the causes of CO₂ emissions and to identify the emission reduction potentials of the cities within the coastal region of China. This is with a view to providing theoretical references for the control of CO₂ emissions in the region.

2. Methods

2.1. Overview of the Study Area and Data Sources

2.1.1. Overview of the Study Area

A total of 114 cities above the prefecture level in 11 provinces (municipalities directly under the central government), including Tianjin, Hebei, Liaoning, Shandong, Jiangsu, Shanghai, Zhejiang, Fujian, Guangdong, Guangxi and Hainan provinces in the coastal region of China, were included in this study. Due to the problem of city representation and data acquisition (Section 2.1.2. Data sources), 28 cities in the northern coastal provinces, 25 cities in the central coastal provinces and 29 cities in the southern coastal provinces were selected as the study sample area in this study. Laiwu City was transferred to Jinan City in January 2019, and Hong Kong, Macao and Taiwan were temporarily excluded from this study.

2.1.2. Data Sources

Since some city statistical yearbooks have not been updated for the latest year, and the difference between the CO₂ emissions of cities above the city scale in China’s coastal region prior to 2006 was relatively small, this study selected 2005 to 2020 as the study period in order to ensure the uniformity and scientificity of the data. The authors reviewed the statistical yearbooks and other materials of 114 cities in China’s coastal region. Based on the requirement for data continuity and the objective fact that some of the cities’ statistical bureaus had not compiled statistics or stipulated that they had not disclosed the relevant data, this study selected 31 types of energy consumption data of 82 cities above the prefecture level in China’s coastal region. The city-level energy consumption data of Chinese cities were mainly obtained from the China Energy Statistical Yearbook 2006–2021 [42] and the statistical yearbooks of cities at or above the prefecture level. Some additional data were obtained from the China Environment Yearbook 2006–2014 [43], the statistical yearbooks of their provinces and the statistical departments of prefecture-level cities. Other city-related data on the number of all employees in industrial enterprises above designated size were obtained from the China City Statistical Yearbook 2006–2021 [44]. In addition, data on the industrial value added in the secondary industry and the industrial value added above the scale were obtained from the statistical bulletins on the national economic and social development of cities at all levels for all years from 2005 to 2020. Data were also taken from the yearbooks of cities at all levels from 2006 to 2021. The above basic data were calculated on the basis of the above data, and the missing data of individual years were handled by the differential interpolation method. In order to eliminate the influence of price fluctuations, the value added by industry and the value added by industrial enterprises above designated size were converted into constant prices, with 2005 as the base period.

2.2. Research Methodology

2.2.1. City-Scale CO₂ Emission Accounting

In this study, the carbon emission coefficient method is used to calculate the CO₂ emissions induced by the energy consumption by industrial enterprises above designated size in the coastal regions of China. The formula for the CO₂ emission factor method is as follows:

$${{\displaystyle C}}_{}={\displaystyle \sum {{\displaystyle E}}_i}\times {{\displaystyle K}}_i$$

(1)

where C is the total CO₂ emission (million tons) and E_i and K_i denote the consumption quantity of the ith energy source (10 ktce) and the corresponding carbon emission coefficient, respectively. The carbon emission coefficients of various energy sources are accounted for with reference to the China Energy Statistical Yearbook and the IPCC Guidelines for National Greenhouse Gas Inventories (2006 edition) [45]. The reference coefficients and CO₂ emission coefficients of each energy source converted into standard coal are shown in

Table 1. Energy species standard coal coefficient and CO₂ emission factor.

Energy Varieties	Standard Coal Coefficient (kgce/kg m3)	CO2 Emission Factor (104 t/104 t)	Energy Varieties	Standard Coal Coefficient (kgce/kg m3)	CO2 Emission Factor (104 t/104 t)
Raw coal	0.7143	2.492	Diesel	1.4571	2.167
Washed refined coal	0.9000	3.212	Fuel oil	1.4286	2.219
Other washed coal	0.2850	2.492	Liquefied petroleum gas	1.7143	1.828
Coal products	0.6000	2.631	Refinery dry gas	1.5714	2.162
Coke	0.9714	2.977	Petroleum brain	1.5000	2.126
Other coking products	1.3000	2.341	Lubricants	1.4143	2.126
Coke oven gas	6.1430	1.288	Solvent oil	1.4671	2.126
Blast furnace gas	1.2860	7.523	Paraffin wax	1.3643	2.126
Converter gas	2.0700	1.288	Petroleum coke	1.0914	2.126
Generator gas	1.7860	1.288	petroleum asphalt	1.4000	2.126
Other gas	3.5710	1.288	Other oil products	1.2000	2.126
Natural gas	13.300	2.162	Heat	0.03412	3.212
Liquefied natural gas	1.1757	2.660	Electric power	1.2290	6.113
Crude oil	1.4286	2.104	Residual heat and pressure	0.03412	3.212
Gasoline	1.4714	1.988	Other fuels	1.01000	2.4567

. From these, the CO₂ emissions of industrial enterprises above designated size in China’s coastal region from 2005 to 2020 were obtained.

2.2.2. Exploring Spatial Data Analysis (ESDA)

ESDA is measured by spatial relationships, and the properties of spatial data are identified by spatial autocorrelation. Global spatial autocorrelation, which is used in this study to analyze the overall degree of correlation in the spatial distribution of urban CO₂ emissions in China’s coastal provinces, assesses the significance of the index by calculating the Moran’s I value, Z score and p value using the following expression:

$$I=\frac{{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n}(y_i-y)(y_j-y)}}{{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n}{W}_{ij}{(y_i-y)}^2}}$$

(2)

where n is the total number of cities in China’s coastal regions; y_i and y_j are the CO₂ emissions of region i and region j, respectively; y is the average value of carbon emissions of all regions; and W_ij is the spatial weight matrix, obtained based on the Rook adjacency relationship. i denotes the global Moran’s I, with a range of values of [−1, 1]. If Moran’s I > 0, then carbon emissions are positively correlated, i.e., counties with higher (or lower) carbon emissions are spatially clustered. If Moran’s I < 0, then carbon emissions are negatively correlated, i.e., there are significant spatial differences between neighboring regions. The larger the Moran’s I, the greater the correlation. When Moran’s I = 0, it indicates a random spatial distribution of carbon emissions.

Local spatial autocorrelation analyzes whether a local spatial correlation exists among cities in China’s coastal provinces. This is often reflected by the local Moran’s I. The local spatial autocorrelation is calculated as follows:

$$I_i=Z_i{\displaystyle {\sum }_j^{}}{W}_{ij}{Z}_j$$

(3)

where I_i is the local Moran’s I; Z_i and Z_j are spatial units and standardized carbon emissions; and W_ij is the spatial weight matrix. The carbon emissions in the study area were categorized into the following types according to the local Moran’s I: high–high (H-H), low–low (L-L), low–high (L-H), high–low (H-L) and insignificant. Among them, H-H and L-L indicate that neighboring regions have a positive spatial correlation: H-H means that all neighboring regions are high-carbon emitting regions, and L-L means that all neighboring regions are low-carbon emitting regions. L-H and H-L indicate that a low-carbon emitting region is surrounded by a high-carbon emitting region and a high-carbon emitting region is surrounded by a low-carbon emitting region, respectively.

2.2.3. LMDI Model

This study adopts the LMDI method to decompose the driving factors affecting the CO₂ emissions of industrial enterprises above designated size in China’s coastal region into five factors: carbon emission factor, per capita energy consumption intensity, labor productivity, industrial structure, and output scale. This is according to the carbon emission equation of energy consumption:

$$C={\displaystyle \sum _i{\displaystyle \sum _j\frac{{{\displaystyle C}}_{ij}}{{{\displaystyle E}}_{ij}}\times \frac{{{\displaystyle E}}_{ij}}{{{\displaystyle P}}_{}}\times \frac{{{\displaystyle P}}_{}}{{{\displaystyle G}}_i}\times \frac{{{\displaystyle G}}_i}{{{\displaystyle S}}_{}}\times {{\displaystyle S}}_{}}}$$

(4)

where C is the total amount of CO₂ emissions (million tons); i is the number of industrial enterprises above designated size in each city of China’s coastal region; j represents the different energy types; E_ij and C_ij denote the amount of activity of the ith China’s coastal region’s industrial enterprises above designated size that consumes the jth type of energy (million tons) and its resulting carbon emissions (million tons), respectively; G_i denotes the value added by the industry of the ith type of industry (million Yuan); P denotes the number of all the employed workers in industrial enterprises above designated size (million people); and S denotes the value added by the secondary industry (million Yuan).

Let $A=\frac{{{\displaystyle C}}_{ij}}{{{\displaystyle E}}_{ij}}$, $B=\frac{{{\displaystyle E}}_{ij}}{{{\displaystyle P}}_{}}$, $D=\frac{{{\displaystyle P}}_{}}{{{\displaystyle G}}_i}$, $F=\frac{{{\displaystyle G}}_i}{{{\displaystyle S}}_{}}$. Then, Equation (4) is simplified to Equation (5):

$$C={\displaystyle \sum _i{\displaystyle \sum _j(A\times B\times D\times F\times S)}}$$

(5)

In Equation (5), A is the carbon emission factor of the jth type of energy consumed by industrial enterprises above designated size in the ith Chinese coastal region, which is also used to reflect the energy structure; B denotes the energy consumption intensity per capita; D denotes the inverse of GDP per capita, which is used to reflect the opposite trend in labor productivity; F denotes the industrial structure; and S denotes the value added by the secondary industry. According to the LMDI model, the effect of each driver of the change in CO₂ emissions from period 0 to period t is expressed as the following equation:

$$\mathsf{\Delta }C={{\displaystyle C}}^t-{{\displaystyle C}}^0=\mathsf{\Delta }A+\mathsf{\Delta }B+\mathsf{\Delta }D+\mathsf{\Delta }F+\mathsf{\Delta }S$$

(6)

$$\mathsf{\Delta }A={\displaystyle \sum _i{\displaystyle \sum _j\frac{{{\displaystyle C}}_{ij}^t-{{\displaystyle C}}_{ij}^0}{\ln {{\displaystyle C}}_{ij}^t-\ln {{\displaystyle C}}_{ij}^0}}}\ln (\frac{{{\displaystyle A}}_{ij}^t}{{{\displaystyle A}}_{ij}^0})$$

(7)

$$\mathsf{\Delta }B={\displaystyle \sum _i{\displaystyle \sum _j\frac{{{\displaystyle C}}_{ij}^t-{{\displaystyle C}}_{ij}^0}{\ln {{\displaystyle C}}_{ij}^t-\ln {{\displaystyle C}}_{ij}^0}}}\ln (\frac{{{\displaystyle B}}_{ij}^t}{{{\displaystyle B}}_{ij}^0})$$

(8)

$$\mathsf{\Delta }D={\displaystyle \sum _i{\displaystyle \sum _j\frac{{{\displaystyle C}}_{ij}^t-{{\displaystyle C}}_{ij}^0}{\ln {{\displaystyle C}}_{ij}^t-\ln {{\displaystyle C}}_{ij}^0}}}\ln (\frac{{{\displaystyle D}}_{ij}^t}{{{\displaystyle D}}_{ij}^0})$$

(9)

$$\mathsf{\Delta }F={\displaystyle \sum _i{\displaystyle \sum _j\frac{{{\displaystyle C}}_{ij}^t-{{\displaystyle C}}_{ij}^0}{\ln {{\displaystyle C}}_{ij}^t-\ln {{\displaystyle C}}_{ij}^0}}}\ln (\frac{{{\displaystyle F}}_{ij}^t}{{{\displaystyle F}}_{ij}^0})$$

(10)

$$\mathsf{\Delta }S={\displaystyle \sum _i{\displaystyle \sum _j\frac{{{\displaystyle C}}_{ij}^t-{{\displaystyle C}}_{ij}^0}{\ln {{\displaystyle C}}_{ij}^t-\ln {{\displaystyle C}}_{ij}^0}}}\ln (\frac{{{\displaystyle S}}_{ij}^t}{{{\displaystyle S}}_{ij}^0})$$

(11)

Equations (7)–(11) represent the amount of change in CO₂ emissions in industrial enterprises above designated size. The five factors of energy consumption structure, per capita energy consumption intensity, labor productivity, industrial structure and output to $\mathsf{\Delta }A$, $\mathsf{\Delta }B$, $\mathsf{\Delta }D$, $\mathsf{\Delta }F$ and $\mathsf{\Delta }S$ represent the contribution to CO₂ emissions from industrial enterprises above designated size in the coastal region of China.

3. Results and Analysis

3.1. Characterization of the Temporal Evolution of CO₂ Emissions

Most of the cities in the coastal region of China in this study are provincial, sub-provincial, populous and economic cities, and they are representative of the cities in each province. In order to better characterize the spatial and temporal evolution of energy-related CO₂ emissions from cities in the coastal region of China, the large-scale coastal region of China was divided into three meso-scale regions, namely the northern coastal region, the central coastal region and the southern coastal region, to which the cities in each coastal province belong. From 2005 to 2020, the energy-related CO₂ emissions of China’s coastal region as a whole, the northern coastal region, the central coastal region and the southern coastal region all showed a fluctuating upward trend (Figure 1). The energy-related CO₂ emissions of the coastal regions were characterized by the northern coastal region > the central coastal region > the southern coastal region, increasing from 2114.42 million tons, 1064.76 million tons, 702.33 million tons and 385.13 million tons, respectively, in 2005 to 5247.77 million tons, 2838.71 million tons, 1483.78 million tons and 925.27 million tons, respectively, in 2020.

In order to more intuitively reflect the regional city-by-city variability, the CO₂ emissions from industrial enterprises above designated size were divided into six classes: very low carbon emissions (below 10), low carbon emissions (10–25), medium carbon emissions (25.01–50), high carbon emissions (50.01–100), very high carbon emissions (100.01–150) and ultra-high carbon emissions (above 150). This was in order to explore the evolutionary characteristics of CO₂ emissions in China’s coastal regions and cities. According to Figure 1, specifically in 2005, the energy-related CO₂ emissions of the overall industrial enterprises above designated size in China’s coastal regions were dominated by very low carbon emissions, low carbon emissions and medium carbon emissions. Among them, Tangshan, Shanghai, Nanjing and other cities had the highest emissions and the largest number of very low carbon emission cities, totaling 27. This accounted for 33.73% of the total number of sample cities, which was due to the relatively slow development of the cities and mainly to light industry. This industry was mainly clustered in the central coast of northern Jiangsu (e.g., Nantong, Lianyungang, Suqian) southern Zhejiang (e.g., Quzhou, Jinhua) and the southern coast of northern Guangdong (e.g., Yunfu, Meizhou, Qingyuan). Tangshan, as a heavy industrial city in China, has a developed iron and steel industry, dense iron and steel enterprises and high coal consumption. Coal accounts for the largest share of total energy consumption in Tangshan. Shanghai and Nanjing were characterized by the rapid development of the secondary industry and increased energy consumption during this period, resulting in the rise in CO₂ emissions. In 2010, Tangshan in China’s coastal area had become an ultra-high-carbon-emission city. The number of high-carbon-emission cities had increased to 11, accounting for 9.63% of the total number of cities, while the number of low-carbon-emission cities and very-low-carbon-emission cities had shrunk to 22 and 13, respectively. The percentage of very-low-carbon-emission cities shrank by 18.07%, and the spatial scope shrank in the central coastal areas such as Suqian, Zhoushan and Lishui and in the southern coastal areas such as Haikou and Guilin. In 2015, the number of ultra-high-carbon-emission cities in China’s coastal areas increased to six, and Suzhou, Nanjing, Ningbo and other cities in the Yangtze River Delta along the central coast became the most important cities in China due to the continued rapid development of industrialization. Consequently, they became cities that consumed a huge amount of energy resources; very-high-carbon-emission cities increased their share by 8.44% compared with 2010, and these were mainly clustered in the northern coastal region. Meanwhile, the number of medium-carbon-emission cities, low-carbon-emission cities and very-low-carbon-emission cities showed a decreasing trend, being most obvious i very-low-carbon-emission cities. In 2020, the number of ultra-high carbon emission cities in China’s coastal area increased to eight, and the number of high carbon emission cities decreased to six. Among these, the energy-related CO₂ emission of Tangshan industrial enterprises above designated size reached up to 5385.191 million tons and the carbon emission of Dongying and Dalian rose from a high carbon emission to an ultra-high carbon emission level. The continuous increase in ultra-high-carbon-emission cities reflects the rapid development trend in industrial enterprises above designated size in China’s coastal area. At the same time, this also indicates that the environment is being damaged to a certain extent while the cities are developing. The number of very-high-carbon-emission cities had increased to 21 cities, and their spatial distribution was still dominant in the northern coast, with sporadic distribution in the central coast and the southern coast; medium- and low-carbon-emission cities were in balance with those in 2015, and the number of very-low-carbon-emission cities showed a continuous downward trend, narrowing from 28 in 2005 to 5 in 2020. These five cities have had a lower level of economic development in recent years and are mainly traditional cities with a low level of industrial development. Thus, this has led to a low proportion of CO₂ emissions related to industrial energy consumption above the scale for this type of coastal cities compared with other resource-based cities. This shows that the energy-related CO₂ emissions from industrial enterprises above designated size in China’s coastal areas, in terms of rank, have overall formed a situation in which economic- and resource-based cities are dominated by high carbon emissions, and other cities are dominated by medium carbon emissions.

3.2. Characteristics of the Spatial Evolution of CO₂ Emission Intensity

Since the cities within the Chinese coastal region included in this study are not independent units from each other, the spatial correlation and variability of energy-related CO₂ emissions among cities must be taken into account. Therefore, this study analyzed the spatial autocorrelation of the energy-related CO₂ emission intensity in the Chinese coastal region.

Table 2. Urban CO₂ emission intensity Moran’s I index and Z-value in China’s coastal region (2005–2020).

Year	Moran’s Index	Z Score	p-Value
2005	0.3669	4.3004	0.0000
2006	0.2664	3.1221	0.0017
2007	0.1740	2.0779	0.0377
2008	0.2397	2.8100	0.0050
2009	0.2032	2.3967	0.0165
2010	0.2664	3.1221	0.0018
2011	0.1590	1.9087	0.0563
2012	0.1535	1.8481	0.0646
2013	0.1667	1.9964	0.0459
2014	0.1775	2.1229	0.0338
2015	0.1566	1.8882	0.0590
2016	0.1368	2.7974	0.0051
2017	0.2382	2.8364	0.0046
2018	0.2793	3.2679	0.0011
2019	0.2878	3.3352	0.0008
2020	0.2972	3.4389	0.0005

reports the Moran’s I and Z-value of the energy-related CO₂ emission intensity of China’s coastal region year by year from 2005 to 2020. This showed that the Moran’s I value was greater than 0 during the study period, which suggests that the energy-related CO₂ emission intensity of China’s coastal region is characterized by obvious spatial positive correlation and spatial agglomeration. From a temporal perspective, the Moran’s I fluctuated greatly from year to year, and, except for some years, the Moran’s I of the energy-related CO₂ emission intensity in China’s coastal regions showed a tendency of decreasing and then increasing. This indicates that the spatial correlation, on the whole, shows a tendency of weakening and then strengthening.

In order to determine the specific location of the concentration of energy-related CO₂ emission intensity in each city within China’s coastal region, a local spatial autocorrelation test was also carried out. The data of energy-related CO₂ emission intensity in China’s coastal region in 2005, 2010, 2015 and 2020 were selected to more intuitively portray the concentration relationship of CO₂ emission intensity within each Chinese coastal region. According to the results of the local spatial autocorrelation analysis (In Figure 2), the overall energy-related CO₂ emission intensity showed a gradual shrinking trend, with 12 cities showing significant positive spatial correlations in 2005. This included one high–low and low–high agglomeration type each, four high–high agglomeration types and six low–low agglomeration types. In 2010, the number of high–high agglomeration types shrank to two, the low–low agglomeration type shrank to four and the low–high and high–low agglomeration types were 3 and 1, respectively. In 2015, it shrank to seven cities, where there was one each of both the high–high and high–low agglomeration types and three and two low–high and low–low agglomeration types, respectively. In 2020, the number of high–low agglomeration types remained unchanged, while the number of high–high agglomeration types increased to two, and there were two and three low-high and low-low agglomeration types, respectively, in 2015. During the study period 2005–2020, high–high agglomeration types were mainly in cities with development potential in terms of transportation and industry, such as Fushun, Benxi, Maoming, Zhanjiang and Qinhuangdao in the northern and southern coastal areas. In contrast, low–low agglomeration types were mainly in Ningde, Quanzhou, Yangzhou, Foshan and Zhongshan in the central and southern coastal areas, where the service industry is developing rapidly but the industrial development is relatively weak. Cities such as Shenyang, Liaocheng, Jining and Tianjin have relatively low carbon emission levels, while the surrounding high-carbon-emission areas have relatively high carbon emission levels, thus forming a low–high agglomeration pattern with the surrounding high-carbon-emission areas. The high–low agglomeration phenomenon only occurs in Quzhou because this region implements the strategy of “industry is king and industry is strong” and industrialization is developing rapidly.

3.3. Factor Decomposition of Urban Energy-Related CO₂ Emissions in China’s Coastal Region

The LMDI method was used to decompose the driving factors influencing the energy-related CO₂ emissions of industrial enterprises above designated size in China’s coastal region into five factors: carbon emission coefficient, per capita energy consumption intensity, labor productivity, industrial structure and output scale. The results of the decomposition of the effect of each factor on the incremental energy-related CO₂ emissions each year are shown in

Table 3. Results of decomposition of factors influencing energy-related CO₂ emissions from 2006 to 2020.

Year	$\mathbf{\Delta }\mathit{A}$	$\mathbf{\Delta }\mathit{B}$	$\mathbf{\Delta }\mathit{D}$	$\mathbf{\Delta }\mathit{F}$	$\mathbf{\Delta }\mathit{S}$	Total Effect
2006	64.1375	238.8040	−289.7121	56.6335	379.6763	449.5392
2007	27.5501	156.6668	−304.8799	57.7436	419.5108	356.5913
2008	0.2046	−95.6445	−265.5315	48.3928	361.4045	48.8259
2009	18.8341	79.4592	−203.5542	26.4812	337.0562	258.2764
2010	−98.7024	275.4876	−278.1084	34.5775	477.0447	410.2991
2011	129.8239	433.9620	−676.5293	37.5617	476.5764	401.3947
2012	122.2263	38.7568	−411.4503	33.5273	453.6521	236.7123
2013	8.9945	88.1427	−426.1942	27.8706	440.4791	139.2928
2014	509.7842	−444.8522	−395.3330	14.7173	360.9694	45.2856
2015	−0.2155	128.8691	−398.8168	−20.2976	284.0955	−6.3653
2016	−171.2638	17.0535	−51.7426	−6.9380	234.8575	21.9666
2017	−19.7668	931.3033	−1056.5616	26.4942	267.7226	149.1918
2018	−4.8924	751.9006	−757.6303	14.2103	251.9699	255.5581
2019	76.6828	498.1247	−597.8619	5.3032	246.2182	228.4669
2020	−138.0305	292.8969	−303.0942	56.8661	191.8655	100.5037
Accumulation	525.3667	3390.9305	−6417.0004	413.1436	5183.0988	3095.5392

The calculation results are based on the 2005 year and CO₂ emissions units are in million tons.

and Figure 3. In general, the pull effect is higher than the inhibitory effect among the five factors that effect change in the energy-related CO₂ emissions of energy consumption in China’s coastal region; among them, the largest pull effect is the scale of the output of industrial enterprises above designated size economy, and the largest inhibitory effect is the labor productivity effect. In addition, from the spatial dimension, the contribution of energy-related CO₂ emissions is the largest in the northern coastal cities, the second largest in the central coastal cities and the smallest in the southern coastal cities.

3.3.1. Output Scale Effect

As the largest contributor to CO₂ emissions, the expansion of the economic scale and income growth of industrial enterprises is the most important driving force for industrial enterprises to increase energy consumption. The output scale effect is mainly dominated by the pull effect. During the period 2005–2020, the cumulative contribution of the industrial enterprises above designated size economy on a regular basis was 5183.1 million tons. From the point of view of the time series change in output scale effect, during the period 2005–2022, the output scale effect was generally in a fluctuating downward trend. There were four turning points in 2007, 2009, 2011 and 2016, which are attributed to the following reasons: ① After the phenomenon of the industrial market disorder in the early stage of the Eleventh Five-Year Plan, China’s coastal areas carried out industrial restructuring, and industrial enterprises above designated size gradually recovered by 2007, but by 2008, industrial enterprises above designated size were in a downward trend. Due to the impact of the international financial crisis and the shrinking demand in the international market, the development of the industrial economy was once again affected, and the growth rate declined. Subsequently, the state promoted the steady growth of the industrial economy by expanding domestic demand and implementing a loose fiscal policy. For this reason, with the encouragement and support of the state policy, the industrial economy regained the momentum of rapid development in the period 2009–2011. Emissions also grew along with the growth, reaching a peak in 2011. ② During the “12th Five-Year Plan” period, the market side of the service industry began to gradually develop. In addition to the impact of national and regional development policies, due to the shrinking demand in foreign markets, China’s coastal industrial economic development above scale decelerated, and by 2016, the industrial economic growth rate reached the lowest value in the study period. Consequently, the output scale effect on CO₂ emissions from the energy consumption by industrial enterprises above designated size is minimized. From the perspective of spatial dimension, in terms of the contribution rate of different regions, the industrial scale effect was the largest in the northern coastal region, followed by the central coastal region, while the southern coastal region had the smallest contribution rate. As an important agglomeration of China’s heavy industry development, it can be seen that the CO₂ emissions from industrial energy consumption above scale in China’s coastal areas were mainly increased by the growth of the output scale of moderately skilled labor-intensive industries, such as the equipment manufacturing industry and the electromechanical manufacturing industry; traditional labor-intensive industries, such as textile, garment and food manufacturing; and moderately skilled capital-intensive industries, such as the petrochemical industry, metal smelting, calendaring, and so on. Therefore, in order to effectively control CO₂ emissions from industrial energy consumption above scale in China’s coastal areas, the economic scale of moderately skilled labor-intensive industries, traditional labor-intensive industries and moderately skilled capital-intensive industries should be reduced in each region.

From the specific point of view of each coastal city, the output scale contribution value of Tangshan’s industrial enterprises above designated size economy in 15 years is 394.75 million tons, which is the city with the largest contribution rate of the output scale effect; in addition to this, Guangzhou, Dalian, Handan, Nanjing, Shijiazhuang and Tianjin had a larger contribution to the output scale in the period 2005–2010, and the contribution rate to the output scale effect increased in the central coastal cities such as Xuzhou, Suzhou, Ningbo, and Zhoushan in the period 2011–2020, which is the most significant contribution rate of the output scale effect contribution increases.

3.3.2. Labor Productivity Effect

Labor productivity, which is the ratio of the fruits of the labor created by a worker over a certain period of time to his or her corresponding labor consumption, is an important influencing factor in promoting high economic growth and coordinating balanced regional development. Improvement in labor productivity helps to reduce energy intensity, which, in turn, helps to reduce carbon dioxide emissions from China’s industrial enterprises above designated size. In this study, the inverse of GDP per capita was used to reflect the opposite trend in labor CO₂ productivity, and the results showed that the cumulative contribution of the inverse of GDP per capita was −6417 million tons. The average contribution during the study period was −427.8 million tons, which was the biggest constraint affecting emissions by industrial enterprises above designated size in China’s coastal region. The labor productivity effect was the most important inhibitory effect in energy-related CO₂ emissions in Chinese coastal province cities. From an overall point of view, during the period 2005–2020, labor productivity is dominated by the pull effect, which presented the alternating characteristics of rising and falling overall, i.e., the overall balance of the period 2005–2010 was relatively stable, and a small upward trend began to appear in 2009; a sharp rise followed by a downward trend appeared in the period 2011–2015; in the period 2016–2020, the turning point in labor productivity was in 2017, when there was a substantial rise followed by a decline, maintained as a pull effect. The northern coastal region contributed the most to the labor productivity effect, accounting for approximately 56.2% of the total amount of the effect; from the specific point of view of the cities in the coastal provinces, Tangshan, Handan, Dongying, Zaozhuang, Benxi, Zibo, Shijiazhuang, Tianjin and other cities in the northern coastal provinces were the main cities inhibiting the labor productivity effect of the cities of the northern, central and southern coastal provinces. Linyi, Dongying, Benxi, Dalian and Zibo in the north, Suzhou, Suqian, Yangzhou and Nanjing in the center and Liuzhou, Maoming, Guilin, Nanning and Chaozhou in the south of the coastal provinces were the main influential cities.

3.3.3. Per Capita Energy Consumption Intensity Effect

Energy intensity reflects the economic efficiency of energy utilization. The factor decomposition results show that the per capita energy intensity effect was the main pull effect in energy-related CO₂ emissions from industrial enterprises above designated size, with a cumulative contribution of 3390.93 million tons. As a whole, the per capita energy intensity was dominated by the pull effect during the period 2005–2020, and the overall stage is characterized by a rise and then a fall in the period 2005–2010 and a suppression effect in 2008, indicating that there was a positive effect in emission reduction in this stage. In contrast, 2011–2015 showed a sharp decrease followed by a rising trend, with 2014 as the turning point. This may mean the energy intensity of high-energy-consuming industries became larger during this period, leading to an increase in energy-related CO₂ emissions of the whole industry. In 2016–2020, the per capita energy intensity appeared to rise sharply followed by a declining trend, with 2017 as the turning point, and this stage was maintained as a pull effect. Specifically, during the period 2005–2020, Tangshan, Zibo, Heze, Dongying and other cities in the northern coastal provinces and Quanzhou, Guangzhou, Zhanjiang and other cities in the southern coastal provinces were the main cities affected by the pulling effect of the effect, while the effect of per capita energy intensity in cities in the central coastal provinces, mainly in Nanjing, Shaoxing, Ningbo and Xuzhou, appeared to be inhibited.

3.3.4. Other Effects

The advanced industrial structure and diversified energy structure are conducive to energy conservation and emission reduction in China’s coastal region’s industrial enterprises above designated size. The industrial structure effect and the energy structure effect have been in a relatively steady decline, and the adjustment of both has promoted the reduction in energy-related CO₂ emissions, although the overall impact on energy-related CO₂ emissions is relatively small. From the specific point of view of each city, the energy consumption structure is typified by the pulling effect of the capital cities and sub-capital cities of various coastal areas, such as Shijiazhuang, Guangzhou, Tianjin, Shanghai, Shenzhen, Dalian, Guilin, Suzhou, Nanjing, Fuzhou, Ningbo, Jinan, Shenyang, Hangzhou, Qingdao and so on. The inhibitory effect is typified by resource-based cities such as Handan, Tangshan and so on, and is mainly dominated by the cities in the northern coastal provinces. However, the industrial structure is slightly different, being dominated by the more industrially developed cities such as Tangshan, Nanjing, Handan, Dalian, Shenyang, Shijiazhuang, Guangzhou, Shenzhen, Benxi, Quanzhou, Lianyungang etc. The inhibitory effect of the industrial structure is mainly dominated by the cities in the central and southern coastal provinces, such as Zhongshan, Ningbo, Nanjing, Tai’an, Foshan, Haikou and other cities in the coastal provinces.

4. Discussion

This study focused on the spatial and temporal evolution of CO₂ emissions from industrial enterprises above designated size in Chinese coastal cities and their driving factors. It had the aim of analyzing the reasons for the changes in CO₂ emissions triggered by the energy consumption by cities above the scale of Chinese coastal cities and providing scientific references for the realization of the “dual carbon” goal in Chinese coastal cities. Compared with previous studies, this study further subdivided China’s coastal region at the city level, and the energy consumption data of each city in this study were adopted from the statistical part of each city, which is more accurate than estimation methods such as nighttime light inversion.

The results of the study show that: ① The CO₂ emissions from industrial enterprises above designated size in the cities of China’s coastal region have obvious regional characteristics and are unique to the region. The reason for this is that in recent years, the energy structure of these coastal cities has been optimized, the proportion of low-carbon energy has increased and the industrial structure has been optimized. The closure and transfer of high-energy-consuming enterprises have led to a reduction in the number of above-scale industries at the source and an emphasis on environmental protection. Therefore, this type of city is expected to realize the goals of “carbon peak” and “carbon neutrality” ahead of schedule. However, the vast majority of cities, such as Tangshan, Tianjin, Dalian and other resource-based cities, are still experiencing a continuous rise in energy-related CO₂ emissions, and it will be difficult for them to realize the goal of “double carbon” in terms of energy use, energy consumption and energy efficiency. To achieve the “double carbon” goal, additional efforts are needed in terms of energy structure and industrial structure. ② During the study period, the Moran’s I index values of CO₂ emission intensity in China’s coastal regions were all greater than 0, indicating that the CO₂ emission intensity in China’s coastal regions has obvious spatial positive correlation and spatial agglomeration, and that the Moran’s I fluctuates greatly from year to year. The global Moran’s I value of CO₂ emission intensity in China’s coastal regions showed that, except for some years, the global Moran’s I value of the CO₂ emission intensity in China’s coastal regions shows the trend of decreasing and then increasing, indicating that the degree of spatial correlation tended to weaken and then strengthen. The high–high agglomeration type is mainly based on cities with development potential in terms of transportation and industry, such as Fushun, Benxi, Maoming, Zhanjiang, Qinhuangdao, etc., in the northern and southern coastal areas. In contrast, the low–low agglomeration type is mainly distributed in cities with a fast development of the service industry but relatively weak industrial development, such as Ningde, Quanzhou, Yangzhou, Foshan, Zhongshan, etc., in the central and southern coastal areas. Cities such as Shenyang, Liaocheng, Jining and Tianjin have relatively low levels of carbon emissions, while neighboring high-carbon-emission areas have relatively high levels of carbon emissions, thus forming a low–high agglomeration pattern with neighboring high-carbon-emission areas. The phenomenon of high–low agglomeration only occurs in Quzhou because this region has implemented the strategy of “industry is king and industry is strong”, and industrialization has developed rapidly. ③ Overall, among the five factors influencing the change in CO₂ emissions from energy consumption by the on-site industries in China’s coastal regions, the pull effect is higher than the inhibitory effect; the largest pull effect is the scale of economic output of the on-site industries, and the largest inhibitory effect is the effect of labor productivity. From the specific point of view of each driver, Tangshan City is the city with the largest contribution rate to the output scale effect, in addition to Guangzhou, Dalian, Handan, Nanjing, Shijiazhuang and Tianjin, during the period 2005–2010. The contribution rate of this effect rose in Xuzhou, Suzhou, Ningbo, Zhoushan and other cities in the central coastal region represented in 2011–2020; Tangshan, Handan, Dongying, Zaozhuang, Benxi, Zibo, Shijiazhuang, Tianjin and other cities in the northern coastal region were the main contributing cities to the labor productivity effect. In addition, the labor productivity effect of the cities in the northern, central and southern coastal region appeared to have a pulling effect from Linyi, Dongying, Benxi, Dalian and Zibo in the north, and Suzhou, Suqian, Yangzhou, Nanjing and the southern part of Liuzhou, Maoming, Guilin, Nanning, Chaozhou and other cities in the coastal region as the main impact cities. Coastal region cities were the main impact cities: Tangshan, Zibo, Heze, Dongying and other northern coastal region cities, in addition to Quanzhou, Guangzhou, Zhanjiang and other southern coastal region cities, demonstrated the per capita energy consumption intensity effect of the main cities playing a pulling role. Nanjing, Shaoxing, Ningbo and Xuzhou, mainly in the central coastal provinces, demonstrated the per capita energy consumption intensity effect of the inhibition of the energy consumption structure of Shijiazhuang, Guangzhou, Qingdao and other provincial capitals and sub-provincial capitals of various coastal regions, where the pulling effect is typical. The inhibition effect is typical of Handan, Tangshan and other resource-based cities. The industrial structure effect is slightly different, with Tangshan, Nanjing, Handan, Dalian, Lianyungang and other cities with more developed industries being the main focus. Its inhibition is mainly based on the cities in the central and southern coastal regions, such as Zhongshan, Ningbo, Nanjing, Tai’an, Foshan, Haikou and other coastal cities.

Based on previous impact studies of Chinese coastal city data, this study also had some limitations in the research process: first, the sample size of the study was not comprehensive enough, and some of the coastal cities were missing data, resulting in a sample size of 82 coastal regional cities; second, carbon emission drivers should generally also be considered in the context of the environmental regulations and at the level of industrial agglomeration. However, in order to ensure a constant equation and as we did not seek to find the relevant indicators of the environmental regulations, these were not selected for the time being. We will consider these kinds of factors more in future studies.

5. Conclusions

This study accounted for the energy-related CO₂ emissions of industrial enterprises above designated size in China’s coastal regions during the period 2005–2020. Furthermore, this study analyzed the spatial and temporal characteristics of energy-related CO₂ emissions of industrial enterprises above designated size in China’s coastal regions during the period 2005–2020. Based on the LMDI model, we focused on revealing the driving factors of CO₂ emission triggered by energy consumption during the development of industrial enterprises above designated size in China’s coastal regions during 2005–2020. The main conclusions of this study are as follows:

(1)

CO₂ emissions in China’s coastal region are growing rapidly, from 2152.23 million tons in 2005 to 52,477.77 million tons in 2020. This represents an increase of approximately 2.5-fold, and it has an overall fluctuating upward trend.

(2)

The energy-related CO₂ emissions of a small number of cities in China’s coastal region show a trend of steady growth to reach the peak and then a slow decline. Consequently, this type of city will be expected to realize the goals of “carbon peak” and “carbon neutrality” ahead of schedule, while resource cities will have to realize the goals of “dual carbon” and “carbon neutrality”. In order to achieve the “double carbon” goal, it is necessary to consider adjusting the energy structure and optimizing the industrial structure.

(3)

During the study period, the CO₂ emission intensity of China’s coastal regions had obvious spatial positive correlation and spatial agglomeration characteristics, and its spatial correlation tended to weaken and then strengthen.

(4)

Overall, among the five factors affecting the change in CO₂ emissions from energy consumption by the industrial sector in China’s coastal regions, the pull effect is higher than the inhibitory effect, the scale of economic output of the industrial sector is the biggest pull factor and the labor productivity effect is the biggest inhibitory factor.

The policy implications of the above findings include the following: ① Cities in China’s coastal provinces should increase their cooperation in energy conservation and emission reduction. First, according to the functional positioning of cities in coastal provinces, inter-regional cooperation in emission reduction should be strengthened by establishing a regional emission reduction responsibility-sharing mechanism and an emission reduction compensation system as soon as possible and by achieving carbon emission reduction targets through inter-city cooperation. ② The gap between the scale effect of industrial economic output and the labor productivity effect is reflected in the fact that the more developed cities have a leading base and speed relative to the other less-developed cities. Therefore, the future coordination of the economic development of China’s cities in the coastal provinces is crucial to the development of the economy. The key to coordinating the economic development of cities in coastal provinces in the future is to promote the labor productivity of the more developed cities to grow at a faster rate in order to drive the development of other cities. Talent attraction policies should be introduced to reduce brain drain and to attract more human resources to the region, thus promoting sustained growth in labor productivity through the continuous optimization of labor allocation. ③ Per capita energy consumption intensity is one of the main factors affecting CO₂ emissions. In order to reduce CO₂ emissions, industrial enterprises above designated size need to increase the development and implementation of new energy-saving technologies, including the conversion of industrial production capacity, the elimination of high-energy-consuming equipment, and the optimization and improvement of the energy structure and industrial structure.