Spatial-Temporal Evolution Characteristics of Industrial Carbon Emissions in China’s Most Developed Provinces from 1998–2013: The Case of Guangdong

Abstract

Industry is widely valued as an important contributor to carbon emissions. Therefore, it is of great significance to analyze the industrial carbon emissions (ICE) in Guangdong, the strongest industrial province in China. We have adopted the carbon emission accounting model and standard deviational ellipse analysis model to analyze the temporal and spatial characteristics and evolution trends of the industry carbon emission amount and intensity in Guangdong from 1998 to 2013. The study results include: (1) Due to the rapid development of industry, Guangdong’s ICE showed a steady growth trend; (2) The distribution characteristics of ICE were characterized by the trend of taking the Pearl River Delta (PRD) region as the center and gradually spreading to the surrounding areas. From the perspective of industrial sectors, it can be divided into steady growth type, fluctuant growth type, basically stable type, and decrease type; (3) The spatial pattern of the ICE in Guangdong is basically the same as that of the total industrial output value, that is, the southwest-northeast pattern. This work is helpful for China’s carbon peak, especially for the formulation of industrial carbon peak policy and the sustainable development of the environment.

1. Introduction

As per the statistics of the International Energy Agency (IEA), 1/3 of the primary energy has been consumed by the industrial department, with carbon dioxide (CO₂) emissions accounting for nearly 1/4 of the amount in the world [1]. As an important CO₂ emission source, the industry has caused a great influence on climate warming, which has attracted extensive attention [2,3,4]. The industry as the largest carbon emissions industry has attracted a great deal of scientific research. Many scholars have analyzed spatio-temporal distribution features and variation trends of China’s industrial carbon emissions (ICE) from the macro perspective. For example, Wang et al. analyzed the regional carbon flow characteristics and carbon emission changes in China’s different industries by building the model with the input-output data between 2007 and 2012 and believed that the carbon emissions decreased from east to west [5]; Feng et al. evaluated the regional characteristics of China’s carbon emissions during 2005–2010 and believed developed eastern coastal provinces were responsible for 80% of the emissions, but there was also a trend of transfer to western regions [6], and in particular, the increment in the central region was large [7]. For example, the ICE in Anhui grew by an average annual growth rate of 10.50% from 26.2284 million tons to 117.2113 million tons during 2000–2015 [8]. Zhang et al. analyzed the spatio-temporal evolution of ICE in China’s 282 cities, believing that the trend of the ICE from 2003 to 2016 first increased and then slowly decreased. [9]; Wang et al. investigated spatio-temporal patterns and agglomeration features of carbon emissions by using exploratory spatial data analysis (ESDA) methods of China’s 30 provinces during 2006–2019 and believed that China’s CO₂ emissions were on the rise, and the highest average CO₂ emissions were in the eastern region [10].

Due to China’s vast territory, unbalanced social and economic development, energy consumption and carbon emissions of different regions vary widely [11]. While some provinces still seriously rely on industry, others have already entered the stage of post-industrialization [12,13,14]; in some provinces, the utilization level of clean energy is gradually improved, but in some other provinces, coal consumption is still heavily dependent on [15]. For example, Gao et al. investigated the carbon emission efficiency(CEE) of 28 industrial sectors during 2005–2017 by using a super efficiency slack-based measure model and distinguishing between direct and embodied carbon emissions [16]; Zhu et al. calculated the CEE of energy-intensive industries in China from provincial scale by using three-stage data envelopment analysis (DEA) model, believing that most of the eastern provinces had higher CEE than other provinces [17]. Peng et al. estimated the carbon emissions, energy consumption, and carbon emission intensity(CEI) for each sector in three northeastern provinces, and believed that these three provinces were more dependent on fossil energy than China was on fossil energy [18]; Wang et al. analyzed the efficiency and performance of industrial total factor carbon emissions of 37 subsectors in Liaoning Province during 2003–2012 by using DEA and Malmquist-Luenberger productivity index, believing that Liaoning Province industry had difficulty in reducing carbon emissions [19].

On the other hand, several researchers have investigated and evaluated the energy consumption and carbon emissions of China’s various industrial sectors from an industrial department viewpoint [20]. For instance, Tian et al. evaluated the cyclical changes in carbon emissions of the manufacturing industry in China from 1992 to 2012, and found that the industries with the highest CO₂ emissions were metal smelting and rolling, production and supply of electric power and steam, chemical industry, nonmetal mineral products and metal smelting and rolling [21]; recognizing the wide variation in carbon emissions across industries in China [22], some scholars analyzed the carbon emissions of individual industrial departments, such as the manufacture of machinery [23], iron and steel industry [24,25], electric power [26], cement industry [27,28], manufacture of food [29], textile industry and papermaking industry [30,31]. Dong et al. believed that the Production and supply of electric power, steam, and hot water, among Henan’s all major emission sectors during 2006–2018, had the largest increase in carbon emissions [32]. Wen et al. calculated the carbon emissions which was emitted by 38 industrial subsectors across China, believing that four subsectors accounted for 97.2% of the growth in total ICE during 2000–2017 [33].

At present, the influencing elements of ICE in China and the carbon reduction approaches have become another hot spot for research, which attempts to achieve energy conservation and emission reduction without affecting economic development. For example, Yuan and Zhao investigated the drivers of CO₂ emissions in energy-intensive industries from China during 2005 to 2010, believing that external inputs have mainly contributed to increasing the carbon emitted, while the demand change was the key to reducing carbon [34]; Zhang et al. analyzed the drivers contributing to CO₂ emissions from electric power industry in Jiangsu Province during 2002–2017 by building extended Stochastic Impacts by Regression on Population, Affluence, and Technology (STIRPAT) model, and projected carbon emissions from 2018 to 2030 through the Monte Carlo method [35]; by using the STIRPAT model and building a multi-constraint input/output optimization model, Wang et al. explored how to update the industrial structure and achieve the goal of low-carbon transformation for China’s industry [36]; Zhang et al. investigated the drivers of CO₂ emissions of Liaoning’s 41 industrial sub-industries by using the LMDI model and believed that energy efficiency dominates in reducing carbon emissions [37]; Zhang et al. analyzed the impact elements of ICE of Nanjing through STIRPAT model and believed that the most influencing factor of Nanjing’s ICE was the industrial energy structure, followed by the total population [38]; Kong et al. investigated various impact elements of CO₂ emissions in China by LMDI model and provided five ideas for reducing carbon emissions [39].

Guangdong is both China’s most industrially advanced area and its most developed province, and the manufacturing output value and the import and export output values take up 15% and 25% of the total values of China respectively [40,41]. With the rapid development of the economy especially in the Pearl River Delta (PRD). The literature now in existence mostly focuses on the connection between carbon emission and economic growth [42], dynamic change characteristics of the carbon emission and sink [43,44], influence factors of the carbon emission [41,45], quota allocation of the carbon emission among industries [46], etc., and in most studies, the three elements influencing the ICE in Guangdong are the industrialization process, energy efficiency, and energy structure [40]. For instance, by using a multi-resolution emission inventory model, Xu et al. evaluated the spatio-temporal variations of industrial CO₂ emissions in PRD during 2008–2012 and concluded that the feature of “one belt, one center” was presented by the carbon emissions in PRD [47].

Since 2000, the local governments have steadily become the primary executor of energy conservation and emission reduction policies, replacing national ministries and commissions [48]. Since 1989, Guangdong Province has been the most populous and economically developed province in China, PRD urban agglomeration is one of the largest in the world. The key indicator for understanding the evolution of China’s ICE pattern is studying the spatio-temporal evolution pattern of Guangdong’s ICE.

The threat of global warming is continuing to increase and has become a hot topic in the current study, thus attracting the common attention of scholars at home and abroad. For example, Voumik et al. analyze the influence of many different types of energy on carbon emissions from electricity and heat in the G7 countries and believed that using renewable energy and nuclear energy will effectively decrease carbon emissions [49]; Walsh et al. forecasted carbon emissions for a fossil fuels scenario by using the FeliX model and believed that anthropogenic emissions needed to peak in 10 years without the emergence of transformative technologies with the goal of the Conference of Parties in Paris [50].

This article proceeds as follows, we have searched the basic data of industrial enterprises in Guangdong from 1998 to 2013, and calculated ICE, in combination with the data from China Statistical Yearbook and Guangdong Statistical Yearbook, the spatial econometric models (ellipse analysis) containing the temporal and spatial influences are adopted, to investigate the characteristics of the ICE and existing problems in Guangdong from 1998 to 2013 from the point of view of industrial sectors, and the spatial change of the regional ICE can be analyzed more effectively. This study has important significance in revealing how carbon emissions are affected by multiple factors that are hidden by regional differences. Besides, it can also be used to provide the basis for making more focused carbon emission policies. This paper aims to solve the following problems: (1) Temporal and spatial characteristics of the carbon emission from Guangdong, industrial enterprises; (2) Evolution rule of the focus of the ICE in Guangdong. For the administrative department of the Guangdong government, it is expected that the study results can be of reference value to make the energy conservation and emission reduction policies, to actively respond to the “carbon peak and carbon neutrality” strategy.

2. Materials and Methods

2.1. Study Area

Guangdong Province is a region in southeastern China, with an area of 17.97 × 10⁴ km², a population of 106.44 million, and a GDP of RMB 62.16 billion yuan. Guangdong includes three parts: the PRD region, eastern Guangdong, western Guangdong, and northern Guangdong (Figure 1). The official statistical data on the carbon emissions in China has not been released currently, so the data on the ICE in Guangdong shall be calculated. Carbon emissions are typically calculated using the method recommended by the IPCC (IPCC, 2006). Since Guangdong Province’s statistical data only included three types of energy data: coal, oil, and natural gas, this paper adopted the fossil fuel carbon emission coefficient recommended by the National Development and Reform Commission: coal: 2.64 ton of CO₂/ton of Standard Coal Equivalent (SCE); petroleum: 2.08 ton of CO₂/ton of SCE; natural gas: 1.63 ton of CO₂/ton of SCE. The basic idea is that the main concept is that for each carbon emission source, the activity amount and emission coefficient are constructed, and the estimated value of the carbon emissions in the project of the emission is the product of the activity data and the emission coefficient [51].

(1)

Calculation Method for the Industry-Level Carbon Emissions

The standard coal consumption of fossil fuels in various industries has been provided in Guangdong Statistical Yearbook over the years, and the calculation formula for the standard coal consumption is shown as follows:

$$E_{jki}=e_{jki}\times F_k$$

(1)

In Formula (1), $E_{jki}$ means the standard coal consumption of k kind(s) of fossil fuel(s) of the industrial sector j in i (year); $e_{jki}$ means the consumption of k kind(s) of fossil fuel(s) of the industrial sector j in i (year); and $F_k$ means the standard coal conversion coefficient(s) of k kind(s) of fuel(s). The standard coal conversion coefficient in China Energy Statistical Yearbook is adopted.

The calculation formula for the carbon emission of the industrial sectors is shown as follows:

$$C_{ji}={\displaystyle \sum }_k{E}_{jki}\times M_k$$

(2)

In Formula (2), $C_{ji}$ means the carbon emissions of the industrial sector j in i (year); $E_{jki}$ means the standard coal consumption of k kind(s) of fossil fuel(s) of the industrial sector j in i (year); and $M_k$ means the respective carbon emission coefficient.

(2)

Calculation Method for the Total Industrial Carbon Emissions in Guangdong

The total ICE in Guangdong is the sum of the calculated values of all the sub-industries, $C_i$ means the carbon emissions of all the industrial sectors in the province in i (year), that is, the ICE amount in i (year) (Formula (3)).

$$C_i={\displaystyle \sum }_j{C}_{ji}$$

(3)

2.2. Data Source and Adjustment of the Industrial Sector Classification Method

The data on the names of industrial enterprises, total industrial output values of enterprises, industry category, and total industrial output values of sub-industries in all the cities, enterprise addresses, etc. in Guangdong is adopted from China Industrial Enterprise Database (from 1998 to 2013). China Industrial Enterprise Database is formed based on the information from the Statistics of Statistical Statement of Industrial Enterprises above the Designated Size by the National Bureau of Statistics, the statistical objects of the database are the industrial incorporated enterprises above the designated size. The social-economy statistical data and energy data in Guangdong from 1998 to 2013 are adopted from Guangdong Statistical Yearbook. From this, the industrial department in Guangzhou is finally divided into 36 industries for calculation and discussion (

Table 1. Adjusted Classification and Codes of Industrial Sectors in Guangdong.

Code	Industry Type	Code	Industry Type
7	Extraction of Petroleum and Natural Gas	27	Manufacture of Medicines
8	Mining and Dressing of Ferrous Metal Ores	28	Manufacture of Chemical Fibers
9	Mining and Dressing of Nonferrous Metal Ores	29	Rubber and Plastic Products
10	Mining and Dressing of Nonmetal Ores	30	Nonmetal Mineral Products
13	Processing of Farm and Sideline Food	31	Smelting and Pressing of Ferrous Metals
14	Manufacture of Food	32	Smelting and Pressing of Nonferrous Metals
15	Manufacture of Wine, Beverage, and Refined Tea	33	Metal Products
16	Tobacco Products	34	Manufacture of General-purpose Machinery
17	Textile Industry	35	Manufacture of Special-purpose Machinery
18	Manufacture of Textile Garments, Footwear, and Headgear	36	Manufacture of Automobile
19	Leather, Fur, Feather, Down, and Related Products	37	Manufacture of Railway, Ship, Aeronautics, and Other Transport equipment
20	Timber Processing, Bamboo, Cane, Palm Fiber & Straw Products	38	Manufacture of Electrical Machinery and Equipment
21	Manufacture of Furniture	39	Manufacture of Communication Equipment, Computers and
22	Papermaking and Paper Products	40	Other Electronic Equipment Manufacture of Instruments and Meters
23	Printing and Record Medium Reproduction	41	Other Manufactures
24	Manufacture of Cultural, Educational, Sports, and Entertainment Articles	44	Production and Supply of Electric Power and Heat Power
25	Petroleum Refining, Coking, and Nuclear Fuel Processing	45	Production and Supply of Gas
26	Manufacture of Raw Chemical Materials and Chemical Products	46	Production and Supply of Water

).

2.3. Model Calculation Methods

Standard Deviational Ellipse Analysis Method

The standard deviational ellipse (SDE) is a statistical analysis method of spatial pattern, which is expressed with shape feature, intensiveness, distribution, orientation, centrality, etc., and can objectively reflect the global feature of the geographic element in the spatial distribution [52]. For the standard deviational ellipse analysis graph, the parameter calculation can be carried out, and space visualization can be achieved through the ArcGIS spatial statistic module.

The calculation of the standard deviational ellipse is mainly determined through three variables, including rotation angle, the standard deviation in the long axis direction, and the standard deviation in the short axis direction. If the distribution of carbon emissions of enterprises (total industrial output value of enterprises) is anisotropic, then there must be a direction with the maximum dispersion, which is defined as the long axis, while the direction with the minimum dispersion perpendicular to it is defined as the short axis. They can be regarded as a result of rotating Axes X and Y in the Cartesian coordinate system by a certain angle. The rotation angle means the clockwise rotation angle of Axis Y of the standard deviational ellipse, the calculation formula of which is shown as follows:

$$\theta =\arctan \frac{\left[\sum {\left(x_i-\stackrel{-}{x}\right)}^2-\sum {\left(y_i-\stackrel{-}{y}\right)}^2\right]+\sqrt{\left\{{\left[\sum {\left(x_i-\stackrel{-}{x}\right)}^2-\sum {\left(y_i-\stackrel{-}{y}\right)}^2\right]}^2+4{\left[\left(x_i-\stackrel{-}{x}\right)\left(y_i-\stackrel{-}{y}\right)\right]}^2\right\}}}{2\sum \left(x_i-\stackrel{-}{x}\right)\left(y_i-\stackrel{-}{y}\right)}$$

(4)

$$e_1=2S_x=\sqrt{\frac{\sum {\left[\left(x_i-\stackrel{-}{x}\right)cos \theta -\left(y_i-\stackrel{-}{y}\right)sin \theta \right]}^2}{n-2}}$$

(5)

$$e_2=2S_y=\sqrt{\frac{\sum {\left[\left(x_i-\stackrel{-}{x}\right)sin \theta -\left(y_i-\stackrel{-}{y}\right)cos \theta \right]}^2}{n-2}}$$

(6)

In Formulas (4)–(6), $\left(x_i,y_i\right)$ means the coordinates of the ith calculated point; $\left(\stackrel{-}{x},\stackrel{-}{y}\right)$ means the mean center; $e_1$ means the long axis of the standard deviational ellipse; and $e_2$ means the short axis of the standard deviational ellipse.

3. Results

3.1. Distribution Change of Industrial Enterprises in Guangdong

As shown in Figure 2, since 1989, Guangdong has been the one with the fastest industrialization process and the highest industrial output value among all provinces in China, and the total industrial output value has increased continuously from 906.242 billion yuan in 1998 to 10,967.310 billion yuan in 2013, growing at 18.08% per year on average. However, the number of industrial enterprises in the whole province shows the characteristics of fluctuation change due to the impact of 2 successive economic crises and changes in the policies of the Chinese government: (1) Stage of decrease from 1998 to 2003. In 1997, the Asian financial crisis first broke out in Thailand and spread rapidly throughout Asia. As a member of the global economy and market, China was obviously affected by the Asian financial crisis, especially in Guangdong, where the impact was greatest and the export-oriented economy was most established. The exports in Guangdong dropped sharply, and a large number of enterprises closed down. Due to the influence, in only 1 year, the number of industrial enterprises was rapidly reduced from 17,966 in 1998 to 12,433 in 1999, and although the number subsequently rebounded, it still did not return to that level (15,756) in 1998 until 2003; (2) Stage of rapid growth from 2003 to 2008. With the progress of China’s reform and opening up and the improvement of the macroeconomic situation, at this stage, the industrial enterprises in Guangdong showed rapid growth. For example, in 2004, the number increased by 18,982 compared with that in 2003, and the highest peak of 51,070 was achieved in 2008; (3) Stage of adjustment since 2008. At this stage, the industrial enterprises in Guangdong have stepped into the optimization stage of “quantity reduction and quality improvement”. There are certain similarities and differences with the first stage. The same point is that the enterprises at this stage are also affected by the financial crisis (the global financial crisis started in 2008). But what is different is that at this stage, the central government of China and the Guangdong government have implemented the targeted policy adjustment. Firstly, it was encouraged to actively eliminate the outdated production capacity from the industry in Guangdong and improve the industrial structure. During this period, the number of textile, light industry, and resource processing industries was reduced continuously, but the agglomeration development momentum of advanced manufacturing industries such as machinery and electronics was strengthened [53]; what’s more, the “double transfer strategy” was implemented in the province to diffuse industrial enterprises within the province. The main industries of Guangdong were long concentrated in PRD, due to which the costs of environment, land, ecological environment, and labor force here were increased constantly, but the development in the northern, eastern, and western Guangdong was weak, and consequently, the regional imbalance of development in the province was further enlarged [54]. The “double transfer strategy” has been implemented in Guangdong since 2008, that is, systematic promotion of the industrial transfer from PRD to the northern, eastern, and western Guangdong and encouragement of the rural labor force in less developed areas to obtain employment in the cities in PRD and other regions. By 2011, in PRD, 4260 enterprises have withdrawn, 30,811 enterprises have been shut down, and 41,390 backward enterprises have been eliminated; 35 industrial parks have been built or upgraded in the less-developed area of Guangdong, and the number of project number and investments funds of these parks have reached a total of 2988 projects and 702.97 billion yuan [55]. It should be pointed out that with the quantity reduction caused due to industrial diffusion inside and outside Guangdong, the healthy development of the economy is promoted instead. For example, in 2013, the number of industries in the whole province was 112% of that in 1998 (20,171), but the total industrial output value in 2013 was not decreased, but increased significantly, being 9.77 times that in 1998, indicating that the industry of Guangdong has entered the stage of “quantity reduction and quality improvement”.

Considering the spatial distribution of enterprises (Figure 3), from 1998 to 2013, the quantity distribution of industrial enterprises in Guangdong showed a great difference in the spatial agglomeration level, and the agglomeration degree showed a trend of decentralization. In 1998, there was a significant concentration of industrial enterprises in the cities of PRD, especially in four cities including Guangzhou, Shenzhen, Foshan, and Dongguan, and in other cities, only northern Guangdong and Shantou and Zhanjiang in eastern and western Guangdong have more enterprises, but the concentration degree was far lower than that of the four cities. The industrial enterprises in the remaining 15 cities showed a less scattered state, forming a typical enterprise distribution model of “one center + eastern and western Guangdong”. Since then, the distribution of industrial enterprises in Guangdong has gradually formed a state of “decentralization”, which was mainly manifested as a state of gradual northeastward and southwestward diffusion from the four cities, but there were certain stages. For example, in 2003, the number of industrial enterprises mainly in Zhongshan and Zhuhai increased greatly, and there was a certain data amount increase in Zhaoqing and Jiangmen. On the contrary, the distribution density of the four cities of Guangzhou, etc. decreased to a certain extent, with the manifestation of uniform distribution of industrial enterprises within PRD. Since then, by means of the “double transfer” strategy, the distribution of industrial enterprises across the province further presented a trend of diffusion, and mainly expanded to eastern and western Guangdong. The expansion speed of the enterprises was obviously higher than that in other areas, with the pattern that the expansion speed in eastern Guangdong was greater than that in western Guangdong, and the number of enterprises in Chaozhou and Jieyang was increased greatly. It shall be noted that from the perspective of the overall situation of 15 years because the northern Guangdong region is mainly mountainous, the industrial enterprises are poor in terms of both number and expansion.

3.2. Temporal and Spatial Variation Characteristics of the Total Industrial Carbon Emissions in Guangdong

3.2.1. Time Change

Guangdong has the largest industrial output among Chinese provinces, especially since 1998, as industrialization developed rapidly, the number of industrial enterprises and the total industrial output value have increased rapidly, followed by the rapid increase of ICE, from 117.23 million tons to 371.47 million tons between 1998 and 2013, and the proportion of ICE in Guangdong increased from 14.08% to 22.41% of national carbon emissions (Figure 4). However, the rate of increase has slowed after 2000 indicating that the peak growth period of Guangdong’s ICE has passed. On the other hand, considering CEI, Guangdong’s industrial CEI is significantly below the national average and continues to decline, from 1.29 tons/ten thousand yuan to 0.34 tons/ten thousand yuanduring1998–2013, indicating that Guangdong’s industrial upgrading and energy conservation and emission reduction have produced good results.

3.2.2. Spatial Change

To evaluate the characteristics of total carbon emissions (TCE) and CEI in different regions of the province, we produced a regional distribution map of Guangdong’s carbon emissions (Figure 5) in four periods. In Guangdong, there is a significant regional difference in ICE. ICE is highly concentrated in the PRD region and lower in northern Guangdong. From the perspective of cities, PRD Region is the most developed area in Guangdong, where lots of industries are distributed intensively, and the value of industrial output and total ICE are far above those in the east, west, and north. For example, in 1998, Guangzhou, Foshan, and Shenzhen in PRD were the top three cities with a total ICE exceeding 10 million tons/year. Only Shanwei, Heyuan, and Yangjiang had total ICE of fewer than 1 million tons/year, and the remaining 15 cities had total ICE ranging from 1 million tons/year to 10 million tons/year. By 2013, the rankings of the total ICE in 21 cities have not changed much, and they still showed the feature that the total amount of PRD was greater than the total amount of other cities. The number of cities with a TCE greater than 10 million tons/year has increased to 11, mainly in PRD (7 cities), western Guangdong (2 cities—Maoming and Zhanjiang), and then eastern Guangdong (1 city—Jieyang). The total ICE in the remaining 11 cities was within the range of 1 million tons/year–10 million tons/year, and the TCE of all cities is more than 1 million tons/year. It should be noted that the distribution of the industrial CEI is not consistent with the ranking of the total ICE, and the industrial CEI in the developed area is relatively small on the contrary. For example, in 2013, Shenzhen ranked last in the province in terms of industrial CEI, at 0.26 ton/10 thousand yuan, although it ranked third in the province in terms of total ICE; in 2013, the total ICE in Chaozhou was only the thirteenth in the whole province, but the industrial CEI was 1.08 tons/10 thousand yuan, and Chaozhou has become the city with the highest emission intensity in the whole province. This phenomenon has been also shown in backward areas such as Yunfu, Shaoguan, Jieyang, etc. in the province.

3.3. Temporal and Spatial Evolution Characteristics at the Industry Level

3.3.1. Time Sequence Evolution of the Total Carbon Emissions of Different Industries

During 1998–2013, 36 industrial sectors’ total industrial output values in Guangdong presented the state of growth, which indicates that the industry is the main power source for Guangdong’s economic development. The TCE of different industries also increased. Except that the carbon emissions of six industries, including mining and dressing of nonferrous metal ores (9), processing of farm and sideline food (13), manufacture of food (14) and other manufactures (41), production and supply of gas (45), etc., were reduced, the carbon emissions of 33 industries presented the positive growth. Therein, the industries with the high average annual growth rate mainly included the manufacture of furniture (17.81%), manufacture of other electronic equipment (15.82%), printing and record medium reproduction industries (12.31%) and timber processing, bamboo, cane, palm fiber & straw products (12.21%). As per the variation trends of the TCE in different years, the following types can be obtained (the data of the coal mining and washing industry (6), the comprehensive utilization industry of waste resources (42), and metal product, machinery, and equipment repair industry (43) were not provided in some years, so these industries are excluded):

(1)

Steady growth type

16 industries with the codes 7, 15, 19, 20, 22, 24, 25, 26, 30, 31, 32, 36, 38, 39, 44, and 46 respectively are mainly included, which are labor- and capital-intensive industries. The main feature is that the TCE steadily rises along with the total industrial output value (Figure 6a). The ICE in these industries is characterized by steady growth with an increase in the industrial output values and is especially concentrated in the traditional high-energy consumption industries (Figure 6b). For example, in 2003, there were a total of 10 industries with a TCE greater than 10 million tons, therein, 8 industries’ carbon emissions showed stable growth. The steady growth type is mainly shown in the high energy consumption industries, and the energy conservation and emission reduction potential of these industries is limited, to which more attention shall be paid.

(2)

Fluctuant growth type

The main industry codes included are 8, 10, 17, 18, 21, 23, 27, 29, 33, 34, 35, 37, and 40 in 13 industries. The TCE of the above industries is significantly lower than those in the industries of steady growth type. In 2013, only the TCE in these two industries, including the manufacture of textile garments, footwear, and headgear as well as printing and record medium reproduction, were more than 10 million tons, which were less than 5 million tons in other industries. Simultaneously, it is clear from Figure 7 that the TCE in these industries has undergone two changes. From 1998 to 2010, a continuous growth process was shown, but the TCE in most industries reached the peak at the end of the Eleventh Five-year Plan (2010). While the total industrial output value continued to grow, TCE began to show a downward trend, which indicates that technological upgrading of these industries has caused a decline of the industrial CEI, and at the same time also shows that the relatively ideal results have been achieved through “promotion of the technical transformation of traditional industries and innovation-driven industrial upgrading” proposed in “the Twelfth Five-Year Plan” of Guangdong.

(3)

Basically-stable type

There are only a few industries of this type in general, including 4 industries with the codes 13, 14, 16, and 28 respectively (Figure 8). The main characteristics are that the carbon emissions present the fluctuation change but are stable in general and not greatly affected by the changes in GDP.

(4)

Decrease type

Similarly, the industries of this type are less in quantity, only including three industries, that are mining and dressing of nonferrous metal ores, other manufactures, and production and supply of gas (Figure 9). Generally, the TCE of these industries is not large, but the reduction amplitude is also very limited.

3.3.2. Time Sequence Evolution of the Carbon Emission Intensity of Different Industries

The CEE of different industrial sectors will be ignored if the evolution of the carbon emissions of various industries in Guangdong is just examined from the total ICE and average annual growth rate. Therefore, considering the industrial CEI of different industries, we have analyzed varying industries’ CEI. From 1998 to 2013, 36 industrial sectors’ CEI in the whole province trended downward. In 1998, the CEI in 20 industries below the average industrial level for the province as a whole (less than 1.29 tons/10 thousand yuan), and the number of such industries reached 23 in 2013 (less than 0.34 tons/10 thousand yuan). The nonferrous metal mining and processing industry, agricultural and sideline food processing industry, and food manufacturing industry were added. 36 industrial sectors with an average annual decline rate greater than 10% include 13 industrial sectors with the codes 8, 9, 13, 14, 16, 24, 25, 31, 32, 34, 44, 45, and 46 respectively, which are mainly labor- and capital-intensive industries, especially the primary product processing industries such as mining industry and smelting industry. The CEI at the initial stage is relatively high (in 1998), but the decline rate is large. When it drops to 2 tons/10 thousand yuan, the decline rate slows down and is gradually stabilized (Figure 10a). For instance, the industrial CEI of petroleum refining, coking, and nuclear fuel processing was as high as 13.60 tons/10 thousand yuan in 1998, and dropped to 0.76 ton/10 thousand yuan in 2013; the CEI of the smelting and pressing of ferrous metals has dropped from 7.00 tons/10 thousand yuan in 1998 to 0.90 ton/10 thousand yuan in 2013. This shows that remarkable results have been achieved in technical transformation and product upgrading of traditional high-energy consumption industries in Guangdong. The CEI of the labor- and technology-intensive industries, including the metal processing industry and manufacture of food, is obviously lower, usually not exceeding 5 tons/10 thousand yuan (in 1998), and has dropped to less than 1 ton/10 thousand yuan in 2013 (Figure 10b). While the decreased amplitude of the industrial CEI of the manufacturing industries such as the manufacture of automobiles and equipment is less than that of other industries on the whole, mainly because these industries are the manufacturers using advanced technology, with low dependence on energy and high industrial added value. So, the CEI was generally less than 1 ton/10 thousand yuan in 1998 and 0.1 ton/10 thousand yuan in 2013 (Figure 10c). From this point of view, this type of industry is featured in the high economic benefits and low CEI, which should be regarded as the key industries for energy conservation and emission reduction in Guangdong.

4. Discussion

4.1. The Relationship between Fossil Energy Consumption and Industrial Carbon Emissions in Guangdong Province

Although the share of fossil energy consumption has decreased from 84.70% in 1998 to 80% in 2013, in the long run, fossil energy is still the main energy type in Guangdong Province. In terms of the consumption structure of industrial fossil energy in Guangdong (Figure 11), from 1998 to 2013, the fossil energy in industrial consumption in Guangdong is still dominated by coal, and the share of coal consumption presents a steadily increasing trend. The coal consumption increased from 33.9688 million tons of standard coal in 1998 to 120.0933 million tons of standard coal in 2013, and during 1998–2013, the proportion of coal consumption in fossil energy increased from 71.87% to 81.47%. Except that in 2009, Guangdong’s industrial coal share was higher than the national industrial coal average (Figure 11a); though the total petroleum consumption increased from 13.0721 million tons of standard coal to 21.9815 million tons of standard coal during 1998–2013, the proportion presented the downward trend, and was decreased from 27.66% in 1998 to 14.91% in 2013, below the national average (Figure 11b); the importance was not attached to the natural gas consumption, the total consumption was increased from 0.2219 million tons of standard coal to 5.3388 million tons of standard coal during 2009–2013. Though its growth rate was the highest in the industrial consumption of fossil energy (with an average annual growth rate of 23.62%), the proportion in 2013 only reached 3.62% due to the smaller base, and similarly was lagged behind the national average level (Figure 11c).

However, it can be found that the carbon emission of coal is significantly higher than that of oil and natural gas through the ratio of fossil energy consumption to carbon emission. For example, in 2013, coal consumption accounted for 81.47% of the total fossil energy consumption, but the contribution rate of carbon emissions was 85.35%. The proportions of oil consumption and carbon emissions were 14.91% and 12.31% respectively, and the proportions of natural gas consumption and carbon emissions were 3.62% and 2.34% respectively. This shows that the obvious structural problem exists in the industrial energy consumption in Guangdong, that is, the proportions of cleaner petroleum and natural gas are too low. Especially the consumption proportion of natural gas has long been lower than 1%, which should be brought to the forefront. It is clear that reducing the proportion of fossil energy, especially coal in industrial consumption, and increasing the consumption of clean energy are important paths to reduce ICE in Guangdong Province.

4.2. Relationship between the Industrial Transfer and the Temporal and Spatial Evolution of the Carbon Emissions

It is clear from the relationship between the industrial concentration degree and carbon emissions in Guangdong that from 1998 to 2013, the industrial transfer in different regions of Guangdong caused a change in the relationship between both. Therefore, further analysis is necessary for the relationship between industrial transfer and carbon emission in the whole province. It should be pointed out that although Guangdong encourages enterprises in PRD to transfer to the north, east, and west, in fact, along with the industrial transfer within the province, the number of enterprises in PRD is still significantly increased at the same time, which reflects that international industries still transfer to PRD. For example, from 1998 to 2013, the industrial output values and carbon emissions of the manufacture of communication equipment, computers, and other electronic equipment rose rapidly, with the average annual growth rates reaching 18.56% and 15.82% respectively, but the industrial CEI was low, which was continuously decreased from only 0.09 ton/10 thousand yuan in 1998 to only 0.06 ton/10 thousand yuan in 2013, and the comprehensive benefits greatly exceeded those of other industries. Similar industries include technology-intensive industries of manufacture of special-purpose machinery (35), manufacture of automobiles (36), manufacture of electrical machinery and equipment (37), etc. With 6 representative advanced manufacturing industries (manufacture of general-purpose machinery, manufacture of special-purpose machinery, manufacture of automobiles, manufacture of electrical machinery and equipment, manufacture of communication equipment, computers, and other electronic equipment as well as the manufacture of instruments and meters) as examples, there were 7832 enterprises increased in PRD Region by 2013 in comparison with the situation in 1998, with the increased number accounting for 94.03% of that in the whole province. In these advanced manufacturing industries, the demands for fossil energy were relatively low, and at the same time, because of the advanced technology, the CEI was much less than that of the labor- and capital-intensive industries, and thus the CEE in PRD has been gradually improved.

On the other hand, subject to the factors of land cost, labor force, wage level, environmental management, etc., the labor- and capital-intensive industries in PRD mainly show a trend of transfer to the areas with low transport costs, low wages, loose environmental regulations, and low land costs. In these industries, many are characterized by high fossil energy consumption and high carbon emission. For instance, the carbon emissions of nonmetal mineral products, which represented 23.74% of the total ICE in 1998, reduced to 21.26% in 2013, ranking first in the industries, The carbon emission coefficient decreased from 6.23 tons/10 thousand yuan in 1998 to 1.83 tons/10 thousand yuan, ranking only second to the papermaking and paper products; the TCE from electricity and heat production and supply ranked only second to those of the nonmetal mineral products, the proportion was increased from 11.38% in 1998 to 17.57% in 2013. The carbon emission coefficient decreased from 10.94 tons/10 thousand yuan in 1998 (ranking second) to 1.04 tons/10 thousand yuan in 2013 (ranking third); the carbon emissions of the smelting and pressing of ferrous metals taking up 4.62% of the TCE in 1998 rose to 6.73% in 2013, the CEI was 7.00 tons/10 thousand yuan in 1998, ranking the third among 36 industries, and though it was decreased to 0.90 ton/10 thousand yuan in 2013, the rank remained unchanged; the papermaking and paper products had the fastest growing TCE which were increased from 3.94% in 1998 to 6.36% in 2013 but the smallest reduction amplitude of the CEI among all the industries which was reduced from 2.39 tons/10 thousand yuan in 1998 to 1.27 tons/10 thousand yuan in 2013, being the industry with the highest CEI; the proportion of the total ICE in the petroleum refining, coking, and nuclear fuel processing was rapidly decreased from 11.81% in 1998 to 7.98% in 2013, but the impact could not be neglected. Therefore, in this paper, 11 labor- and capital-intensive industries, including mining and dressing of nonmetal ores, processing of farm and sideline food, textile industry, manufacture of textile garments, footwear and headgear, leather, fur, feather, down and related products, timber processing, bamboo, cane, palm fiber & straw products, manufacture of cultural, educational, sports and entertainment articles, petroleum refining, coking and nuclear fuel processing, nonmetal mineral products, smelting and pressing of ferrous metals as well as smelting and pressing of nonferrous metals, are selected to form the change chart of the number of enterprises in the eastern, northern and western Guangdong in 1998 and 2013 (Figure 12), from which it can be found that the enterprises in these industries were transferred to these areas, and a total of 2856 enterprises were added in 15 years. The main transfer centers were concentrated in Jieyang, Shantou, and Chaozhou in eastern Guangdong, Zhanjiang, and Maoming in western Guangdong as well as Qingyuan and Shaoguan in northern Guangdong. From this, it is clear that the industrial development in Guangdong shows the overall trend of the shift of advanced manufacturing industries to PRD, textile and light industry to eastern Guangdong, and the resource processing industry to more advantageous locations in western Guangdong, which is in line with the current results of the research [53]. Through the increase of these industries, the development of the regional economy can be promoted on the one hand; however, these industries are basically featured in the high total amount and intensity of carbon emission, causing the pollution effect which cannot be underestimated, and the TCE and CEI of these cities have rapidly increased. For example, the rank of the CEI in Chaozhou rapidly changed from fifteenth in 1998 to first in the whole province in 2013, and in the same period, the rank of the intensity in Yunfu City rose from twelfth to second in the whole province.

As a result of industrial upgrading and strong technological innovation, PRD is extremely different from other regions in Guangdong in terms of the number and quality of industrial enterprises. Although the total amount of carbon emissions in PRD is large, the overall CEI is low and declines continuously, so the CEE is high. At the same time, due to the weak foundation and short development time, the gap between the eastern, western, and northern Guangdong and PRD is still large, and together with the low level of industrial transfer and lack of transfer of high-end industries, the quality of industrial development is obviously lower than that in PRD. With the transfer of the low-end industries, it is inevitable that carbon emission pollution will be transferred quickly, to which sufficient attention should be paid, otherwise the “pollution haven” effect will be caused easily.

4.3. Change and Trend of Spatial Pattern of Industrial Carbon Emissions in Guangdong

In the method of the standard deviational ellipse (SDE), the spatial distribution ellipse with the azimuth angle, short axis, center, and long axis as the basic parameters is adopted to quantitatively reflect the characteristics of the study object in the spatial layout, to describe the position and trend direction of the geographic element in the spatial distribution and the discretization degree in the main and secondary directions [56]. We have adopted the method of weighted standard deviation ellipse in ArcGIS software. According to the spatial location of provincial, municipal, and county administrative units, the standard deviation ellipse of the spatial distribution of each variable is determined by the corresponding carbon emission and industrial variable indicators, and the parameter calculation and spatial visualization are carried out through the ArcGIS spatial statistic module. Thus the standard deviation ellipse parameters of total industrial output values and ICE in Guangdong from 1998 to 2013 are obtained, and the SDE graph of the total industrial output value and ICE in Guangdong is drawn, to reflect the spatial distribution pattern of the carbon emissions with the change of the standard deviation of the long and short axes, position of the center point, specific movement, etc.

In the SDE analysis, the long axis represents the distribution direction of the geographic element in the main trend direction, while the short axis indicates the discretization degree (distribution range) of geographic elements in the secondary direction. As shown in Figure 13, the spatial distribution of the ICE and total industrial output value in Guangdong have highly similar features. The spatial shape evolution of the ICE has the following several characteristics: (1) The spatial pattern of the ICE matches that of the total industrial output value, which shows a southwest-northeast pattern; (2) The spatial rotation angle presents the trend of stable northward rotation, which has been increased from 65.88° in 1998 to 71.12° in 2013, slightly falling behind the change of the spatial rotation angle of the total industrial output value during that time, but the increments are basically the same; (3) Considering shape index, the spatial discretization degree of ICE is significantly greater than that of total industrial output value. The main trend direction of the ICE was expanded rapidly from the central point to the southwest and northeast, and its extension range and speed are much greater than the diffusion of the total industrial output value, while diffusion of the two on the short axis is basically consistent. In the examination of the changes in 15 years, it can be found that both of them have experienced a process from shrinking (from 1998 to 2007) to expanding (from 2008 to 2013). It shows that after 2008, with the outward transfer of industries in PRD, the carbon emission has been transferred, the impact range of which is more obvious than that of the industrial diffusion. Thus the following judgment can be made to the spatial pattern trend of the ICE in Guangdong: (1) The spatial shape of the ICE in Guangdong is basically consistent, that is, the southwest-northeast pattern (western and eastern Guangdong) will still be the main extension direction of the ICE, at the same time, it will slowly transfer to the northern Guangdong, and the trend of continuous increase and deterioration of the industrial carbon emissions in the western and eastern Guangdong in the future will be greater than that in the northern Guangdong; (2) Although the main direction of the ICE in Guangdong has a trendency to rotate northward, and the rotation angle decrease year by year and tends to be stable, the spatial distribution pattern of the ICE in Guangdong will remain basically stable; (3) Considering the moving tracks of the carbon emission ellipse graphs in different periods, carbon emission growth in eastern Guangdong is significantly higher than in western Guangdong, resulting in the continuous movement of the ellipse to the northeast, indicating that the pressure of industrial energy conservation and emission reduction in eastern Guangdong is greater.

5. Conclusions

The secondary industry is the major consumer of fossil energy, so it is an important contributor to carbon emissions. In comparison with previous research, the main contribution of the study is deepening of the following aspects: we adopt the enterprise database to analyze the spatio-temporal evolution of the ICE in Guangdong, which overcomes the problem that only macro-level analysis (provincial, prefecture-level or county-level) is provided in previous research, and can reflect the temporal and spatial variation characteristics of the carbon emissions from industrial enterprises in Guangdong more accurately; secondly, we apply the spatial bivariate analysis to discuss the spatial correlation between the distribution of the industrial enterprises and carbon emissions in Guangdong; thirdly, we analyze the relationship between the quantity, carbon emissions and industrial output values of all the industrial enterprises in Guangdong from 1998 to 2013 through the ellipse analysis, to reflect the characteristics of transfer of the carbon emission center, shape change of the carbon emission, etc. in Guangdong over the past 15 years more accurately. Some specific conclusions therefore obtained from this study can be summarized as follows:

(1) Based on Guangdong’s statistics from 1998 to 2013 and the data from the enterprise database, the changes in the industrial enterprise number, total industrial output value, and total ICE in Guangdong are analyzed. Compared with that, in 1998, the number of industrial enterprises in Guangdong increased by only 12% in 2013, and the total industrial output value and the total ICE increased steadily, with an average annual growth rate of 18.08% and 7.99% respectively. There has been a continuous reduction in the industrial CEI from 1.29 tons/10 thousand yuan in 1998 to 0.34 tons/10 thousand yuan in 2013, indicating that industrial upgrading and energy saving, and emission reduction in Guangdong have played a good role. The changes in the spatial distribution of the industrial enterprises and ICE have been analyzed. The study found that both of them have a highly positive correlation, showing a trend of expansion mainly to eastern and western Guangdong with PRD as the center.

(2) Coal still forms the main structure of Guangdong’s industrial fossil energy consumption, and the consumption proportion has a rising trend. The proportion of coal consumption increased from 71.87% in 1998 to 81.47% in 2013, being higher than the average level of the national industrial sector. Correspondingly, the proportion of relatively clean petroleum and natural gas is generally low, especially since the consumption proportion of natural gas is lower than 1% in the long term, clean energy transition in the industrial sectors of Guangdong Province has a long way to go.

(3) By analyzing the relationship between carbon emissions and industrial development from the perspective of industry, it can be found that the total output value of 36 industries in the province from 1998 to 2013 showed an increasing trend, the TCE of most industries except 6 industries of mining and dressing of nonferrous metal ores, etc., also rise therewith, which can be divided into 4 types as per the growth rate, including steady growth type, fluctuant growth type, basically stable type and decrease type. The TCE and industrial CEI growth rates in technology-intensive industries are lower than in labor- and capital-intensive industries.

(4) By analyzing the relationship between the regional industrial transfer and carbon emissions, we believe that the technology-intensive industries are mainly aggregated in PRD. The demand for fossil energy in these industries is relatively small. At the same time, their CEI is far lower than that of labor- and capital-intensive industries thanks to the advanced technology, so the CEE of PRD has gradually increased. In the same period, the labor- and capital-intensive industries in PRD mainly transferred toward the northern, eastern, and western Guangdong, causing a sharp increase in the amount and intensity of carbon emission in these areas. The change and trend of the spatial pattern of ICE in Guangdong are discussed with the standard deviational ellipse analysis method. It is found that the spatial pattern of the ICE in Guangdong is basically in line with that of the total industrial output value, which shows a southwest-northeast pattern, and the spatial rotation angle has a trend of gradually northward rotation.

Based on the study results, it is of great significance to analyze the temporal and spatial evolution rules of industrial carbon emissions in Guangdong, to achieve the goals of China‘s low-carbon economy and carbon emission reduction. There are two policy implications for policymakers:

(1) From the perspective of the analysis results in this paper, the labor- and capital-intensive industries in the industrial sectors of Guangdong show a “three-high” trend of total carbon emissions, carbon emission growth rate, and industrial carbon emission intensity. While the backward industries also transfer to the northern, eastern, and western Guangdong at the same time, due to which continuous increase of the economic gap with that of the Pearl River Delta and the “pollution haven effect” may be caused, and as a result, the carbon emissions and carbon emission situation may deteriorate. Therefore, the local governments should issue corresponding policies to reduce energy intensity and carbon intensity through reasonable industrial structure adjustment, as well as formulate reasonable industrial transfer policies to prevent the transfer of carbon emissions from backward industries in the Pearl River Delta to other regions in the province, and transform traditional industries through technological progress and energy efficiency to reduce carbon emissions.

(2) As a developing country in the process of rapid industrialization, it is unpractical to require China to noticeably slow down its economic development for a rapid decrease in carbon emissions in a short time, but as the most developed province in China, Guangdong should play an exemplary role in terms of the industrial upgrading and energy conservation and emission reduction. China should formulate policies to encourage the developed provinces to take the lead in implementing the strategic plan for industrial carbon reduction and take Guangdong as the pilot province. Other methods should be considered through further research, such as innovation of the clean technology and the implementation of the clean energy price subsidy. Of course, our current study has some uncertainties. This is mainly due to the limitation of the industrial enterprise data in Guangdong. As only the data by 2015 was released in the enterprise database, this paper only covers the data from 1998 to 2013. In the future, with more available data, the latest analysis of the industrial carbon emission in Guangdong can be achieved; on the other hand, under the space limitation, the study on the influence factors/mechanisms of industrial carbon emissions in Guangdong is not involved in this paper.