1. Introduction
Since 1978, rapid industrial progress has contributed irreplaceably to China’s economic development [
1,
2]. However, due to the blind pursuit of economic development speed, rough industrial production leads to resource problems, and environmental management becomes the bottleneck restricting economic development. China’s industry has yet to shake off the development pattern of high consumption and emissions [
3], particularly evident in the industrial water use (IWU) system. In 2020, China’s IWU reached 121.80 million tons, and the corresponding industrial wastewater discharge reached 2.07 million tons [
4]. Among the direct pollution sources, the wastewater discharge from industrial pollution sources is second only to that of comprehensive sewage outlets [
5], which is far behind the wastewater discharge standard of developed countries.
Since 2016, China has vigorously advocated the industrial green development (IGD) to promote high-quality industrial sustainable development in the 13th and 14th Five-Year Plans while protecting the environment and conserving resources. Made in China 2025 and Industrial Green Development Plan (2016–2020) pointed out that China’s industrial development has entered the green transformation period. Besides, as a typical water-deficient country, improving the IWU efficiency is particularly important in promoting IGD, which needs a series of policy guidelines and legal guarantees as the basis. The formulation of these policies and laws must be based on the scientific evaluation of IWU and IGD systems. Such assessment can also provide policymakers with quantitative information to help them conduct policy analysis effectively [
6].
Water is an unrecognized but essential connecting factor to achieving sustainable development goals, so the practical analysis of the relationship between IWU and IGD systems is of great significance to industrial green and sustainable development in China. However, there are few types of research on the relationship between IWU and IGD systems, mainly investigating the relationship between IGD system and carbon emissions or technological innovation system. Feng and Chen [
7] adopted the spatial Dubin model to explore the spatial impact of environmental regulation and technological innovation in the IGD system. Wang et al. [
8] adopted the entropy weight method and grey relation analysis to construct the IGD and regional non-waste systems, respectively. They explored the relationship between these two systems through the improved coupling coordination model. It is also found that the technology innovation system significantly positively affects the IGD system [
3]. Wang et al. [
9] discussed the impact of carbon abatement and technological innovation systems on the IGD system. Dong et al. [
10] discussed how industrial convergence affects the IGD system. Mennicken et al. [
11] believed that adequate investment in carbon emissions reduction and technological innovation systems could effectively promote IGD. However, only a few previous researchers have discussed the relationship between the IGD and the IWU systems. Most research involving the IWU system only discussed the relationships between industrial economic growth and total consumption or the intensity of IWU [
12,
13,
14,
15], which need to be more. Furthermore, the spatial–temporal evolution characteristics of the degree have yet to be discussed in depth. This is the gap that our research is trying to fill.
IWU efficiency is the main index for reflecting the performance of the IWU system. At present, the measurement models of water use efficiency can be mainly divided into the stochastic frontier analysis (SFA) [
16,
17,
18] and data envelopment analysis (DEA) [
19,
20,
21]. However, the SFA model has a few limitations. It is only suitable for one output scenario, and whether the result is mainly correct depends on the predicted forms of the production functions. The DEA model does not need to set the functional forms and can overcome these shortages of subjective indicator selection and weighting. Then, this paper adopts the DEA model to measure the IWU efficiency. Compared to the traditional DEA model, the slack-based measure (SBM) model could directly deal with the output excess and output insufficiency of relevant decision-making units and is not affected by the whole statistical data set, so the results are more effective [
22,
23].
Since green development was put forward [
24], the cores of relevant definitions emphasize that economic and environmental benefits can be improved through environmentally friendly financial behaviors to achieve sustainable development goals [
25,
26,
27,
28]. As a component of green development, IGD has gradually become the focus of academic research with the development of new industrialization. The Global Reporting Initiative presented a global sustainability index considering the economic, environmental and social dimensions to reveal the IGD [
29]. OECD [
30] established a complete indicator system of green development composed of economy, environment, and human well-being. Liu et al. [
31] measured the IGD degree in China based on three aspects: environment, economy, and technology. Wang et al. [
27] constructed the evaluation index system of the IGD based on three aspects: Environment, economy, and society. Wang et al. [
32] revealed the differences of the IGD system in China’s three major urban agglomerations from the aspects of the environment, resources, and society by adopting entropy-weighted TOPSIS. Wang et al. [
33] measured the performance of the IGD system in China from the perspective of resource consumption and environmental pollution. Economy, environment and technology are the main aspects for selecting assessing indicators to evaluate the performance of the IGD System.
Due to the unbalanced distribution of industrial sectors and further technological development, the quantitative relationships between IWU and IGD systems of provinces present spatial differences. Traditional exploratory spatial data analysis (ESDA) and mathematical statistics can effectively reflect the spatial differences, spatial distribution, and spatial correlation [
34,
35,
36,
37,
38]. With the rapid development of China’s industry, it is of great significance to analyze the quantitative relationship between IWU and IGD systems in the temporal dimension for industrial adjustment and technological upgrading. However, traditional ESDA and mathematical statistics are difficult to represent either the characteristics spatial correlation or temporal dimension [
39]. Exploratory spatial–temporal data analysis (ESTDA) can integrate temporal dimension analysis into spatial analysis more effectively and has been widely used in the analysis of land use, resource utilization, pollution control and other fields. Dong et al. [
10] adopted ESTDA to reveal the spatial–temporal dynamics of the multidimensional poverty index in China from 2007 to 2017. By applying ESTDA, Zhang et al. [
40] analyzed China’s carbon footprint’s spatial pattern and dynamic spatial–temporal interaction from 2001 to 2013. Liu et al. [
41] used ESTDA and the spatial econometric model to analyze the spatial–temporal evolution characteristics and influencing factors of PM2.5 in the Yangtze River Economic Belt. Li et al. [
42] analyzed the spatial–temporal dynamic evolution of GDP per capita for 2303 counties in China from 1998 to 2015 based on the ESTDA framework.
Based on the above analysis, we could find that existing research focuses more on the interaction between the IGD system and technological innovation or energy system, but few study the relationship between IWU and IGD systems. Besides, the method involved is not able to evaluate the dynamic evolution between systems effectively. In the coupling framework between resource and economic systems, this research focuses on the coupling coordination relationship between IWU and IGD systems and discusses its temporal and spatial variations to provide policy suggestions for achieving sustainable development goals. The objectives of this study are as follows: (1) Construct an index system to evaluate the performance of IWU and IGD systems. (2) Propose a new coupling coordination framework to evaluate the coupling coordination degree between IWU and IGD systems. (3) Reveal the characteristics of spatial–temporal variations in coupling coordination degree by adopting ESTDA.
The remaining structure of the article is organized as follows.
Section 2 introduces the construction of an index system, methods, and data acquisition.
Section 3 presents the evaluation results and discussion of this research. In
Section 4, the conclusions, limitations, and future research directions are provided.
3. Results
3.1. Analysis of Industrial Water Use System
The IWU efficiency of each province during 2009–2018 is presented in
Table 4. The IWU efficiency of most provinces was less than one, indicating that the IWU of these provinces needs to be optimized. On the whole, the IWU efficiency of each province presented an upward trend. From 2009 to 2018, the average IWU efficiency increased from 0.401 to 0.703, with an increase of about 75.19%. The improvement of IWU efficiency is closely related to the promulgation of water-related policies, including establishing a strict water resource management system, strictly controlling water for production, and effective wastewater treatment.
During 2009–2018, Chongqing, Jiangsu, Beijing, Shanghai, and Guangdong ranked in the top five. The mean values of their IWU efficiency were 0.906, 0.897, 0.889, 0.872, and 0.846, respectively. Chongqing is located in the core of the upper reaches of the Yangtze River Economic Belt. It is a critical area of the Chinese government’s western development plan and a vital pilot area of IGD. The resource advantage, policy guarantee, and industrial adjustment ensure that the IWU efficiency of Chongqing consistently ranked first in the western region. The remaining four provinces are all in the eastern region, which have advanced industrial structure and economic foundation and are in the first echelon of the national industrial transfer and industrial restructuring. They have banned industries with high water use consumption, adopted advanced water-saving technologies, and formulated strict water policies.
Ningxia, Gansu, Guizhou, Fujian, and Shaanxi were the five provinces with lower IWU efficiency. The mean values of their IWU efficiency were 0.409, 0.437, 0.441, 0.455, and 0.493, respectively. In these provinces, the problems, such as unbalanced economic development, low awareness of water conservation, and the widespread existence of industries with high water use consumption, lead to the waste of water resources in industrial production and low water use efficiency. However, there is no denying that the IWU in all the provinces mentioned above has improved since 2011 (the beginning of the 12th Five-Year Plan), which is due to the Chinese government’s increasing emphasis on IGD, developing the recycling industry and optimizing the comprehensive utilization of water resources.
3.2. Analysis of Industrial Green Development System
Table 5 presents the index weights of China’s IGD system based on the entropy-weighted TOPSIS model. The four primary indexes indicators are ranked as follows: Industrial economy (27.23%) > Industrial innovation (26.51%) > Industrial resources (24.98%) > Industrial environment (21.28%). Economic development is still the primary goal of IGD, which is a necessary foundation to support the implementation of China’s 14th Five-Year Plan. Industrial innovation is also a prerequisite required to improve industrial production efficiency, reduce resource consumption and reduce environmental pollution, and it plays the leading role in the IGD systems. The Chinese government has been increasing investment in industrial innovation in recent years. Industrial resource consumption and industrial pollution are important indexes to measure the sustainability and environmental friendliness of the IGD system.
The top five secondary indexes are as follows: Number of industrial R&D employees (8.70%) > value-added growth rate (8.32%) > industrial energy consumption (7.33%) > labor productivity (7.30%) > industrial water use consumption (7.21%). Number of industrial R&D employees is the industrial innovation index. China’s industrial structure is transforming; energy-intensive and inefficient heavy industries still occupy a large proportion of the industry, which is a significant cause of high energy and resource consumption and heavy pollution. Increasing investment in industrial innovation can not only support the adjustment of industrial structure but also drive the IGD [
51]. Value-added growth rate and Labor productivity are the important indexes to measure the performance of the industrial economy. As the supporting industry of the national economy, industry bears the critical responsibility of promoting social and economic development. The other two indexes are industrial resources indexes. Resource consumption and environmental pollution is the bottleneck restricting the IGD. Since the 12th Five-Year Plan period, the Chinese government has made great efforts to introduce advanced industrial production technology and increase investment in environmental pollution control to ensure the green transformation of industrial enterprises.
Table 6 shows the performance of the IGD system of 30 provinces in China. From 2009 to 2018, IGD has been improved in most provinces. Zhejiang, Jiangsu, Shanghai, Tianjin, Beijing, and Guangdong were the six provinces with outstanding performance in the IGD system, and the mean values of their performance of the IGD system exceeded 0.4. They are all the most advanced industrial provinces in China and are all located in the eastern region. Compared with other provinces, they have a complete and efficient industrial structure and are the earliest provinces to promote industrial adjustment. Based on a strong economy, they have vigorously developed high-tech industries, increased investment in environmental pollution control, and eliminated industries with high resource consumption and pollution. Excellent transportation and geographical conditions also ensure their access to advanced production technology, foreign investment, and policy preferences.
The provinces in the central region, such as Anhui, Shanxi, and Jilin, were in the second group. Although their performance was not as good as that of the provinces in the eastern region, it should not be ignored that their IGD has maintained an annual growth rate of more than 4%. In recent years, the provinces in the central region have increased cooperation with provinces in the eastern region to introduce advanced energy-saving and water-saving technologies. Since the implementation of the Plan on the Rise of the Central Region [
52], the Chinese government has increased the policy preference for the provinces in the central region, especially allocating funds to recruit R&D employees and attract investment for these provinces, to build them into modern equipment manufacturing and high-tech industrial bases. At the same time, the government strictly controlled environmental pollution to ensure the sustainable development of the industrial economy.
The provinces in the western region were at the bottom in terms of the performance of the IGD system during 2009–2018. As the only municipality directly under the central government in the western region, the industrial development of Chongqing was very rapid. Except for Chongqing, the IGD degree of other provinces in the western region increased slowly, and even Guangxi, Ningxia and Qinghai showed a downward trend. Compared with other provinces in the western region, Chongqing is more attractive to foreign investment and has easier access to high technology, supporting its IGD. Other provinces in the western region, in order to ensure economic growth, undertake the backward industries eliminated from the central and eastern regions more, resulting in the intensification of resource consumption and environmental pollution.
As shown in
Table 6 and
Figure 2, there were only two provinces (Jiangsu and Zhejiang) whose performance of the IGD system was greater than 0.4 in 2009, while in 2018, there were six provinces (Jiangsu, Zhejiang, Shanghai, Beijing, Tianjin, Shandong and Chongqing) whose IGD degree was greater than 0.4.
Figure 2 also presents the composition of the IGD degree in 2009 and 2018. Obviously, except for Shaanxi, the industrial economic index score ratio to the IGD degree has decreased in the other 29 provinces. The slowdown in economic growth is one reason for this phenomenon. More importantly, the Chinese government is paying more and more attention to water consumption and industrial environment control instead of focusing solely on the industrial economy. Compared to 2009, the scores of industrial resources and industrial environment increased significantly in most provinces in 2018, especially the provinces in the eastern and central regions. A strong economy can support enough investment in industrial innovation. Jiangsu, Zhejiang, Shandong, Chongqing, Beijing, and Tianjin, all of which are economically developed regions, experienced the most noticeable increase in industrial innovation.
3.3. Analysis of Coupling Coordination
The coupling coordination model is used to calculate the coupling coordination degree between IWU and IGD systems during 2009–2018, and the results are shown in
Figure 3 and
Figure 4. By 2018, the five provinces with the highest coupling degree are Zhejiang (0.980), Shanghai (0.969), Jiangsu (0.950), Tianjin (0.948), and Henan (0.908). The five provinces with the lowest coupling degree were Xinjiang (0.190), Qinghai (0.210), Jiangsu (0.255), Guangxi (0.294), and Hainan (0.301). During the research period, the coupling degree of the provinces in the eastern region remained high, and the fluctuation range was small. Their coupling degree was mainly in the high coupling stage and run-in stage. The coupling degree of provinces in the central region decreased and then increased. The coupling development between IWU and IGD systems in these provinces experienced a process from deterioration to improvement. The coupling degree of provinces in the western region was generally low. Except for Chongqing, the coupling degree of other provinces was mainly in low couple stage and irrelevant stage. As shown in
Figure 4a, from 2009 to 2013, the number of provinces in low couple stage and irrelevant stage generally showed an increasing trend. After a decline in 2014, they showed an increasing trend until reaching a peak in 2016 and falling back from 2016 to 2018. The number of provinces in the high coupling stage and run-in stage generally showed a downward trend. After a sudden increase in 2014, they showed a downward trend until reaching the bottom in 2016, and showed an upward trend from 2016 to 2018.
In
Figure 3, the coupling coordination degree of the 23 provinces showed an upward trend during 2009–2018. In 2009 (
Figure 3a), Zhejiang (0.674), Jiangsu (0.668), Shanghai (0.665), Chongqing (0.616), and Beijing (0.591) ranked among the top five. In 2018 (
Figure 3j), Jiangsu (0.859), Zhejiang (0.836), Shanghai (0.817), Beijing (0.796), and Tianjin (0.759) ranked the top five. In these provinces, the construction of efficient IWU system and well-considered IGD system ensure that their coupling coordination degree has been maintained at a high value. On the contrary, the coupling coordination degree of the other 7 provinces (Guangxi, Hainan, Liaoning, Gansu, Qinghai, Ningxia, and Xinjiang) all decreased during the research period. Among these seven provinces, the IWU efficiency of Liaoning and Hainan has been effectively improved. Nonetheless, their IGD degree was always low, which affected the coupling and coordinated development between IWU and IGD systems. Guangxi, Gansu, Ningxia, Xinjiang, and Qinghai performed poorly in the IWU and IGD systems, ranking at the bottom of the 30 provinces. These provinces scored low in industrial resources, industrial environment, and innovation, leading to the decline of their coupling coordination degree between the IWU and IGD systems.
Obviously, as shown in
Figure 4, the research period can be divided into two stages: 2009 to 2013 (
Figure 4a) and 2014 to 2018 (
Figure 4b). During 2009–2013, 30 provinces have developed just four kinds of coupling coordination stages. The numbers of provinces in the high dysregulation stage were 0, 0, 5, 5, and 3, respectively. The numbers of provinces in the basic dysregulation stage were 11, 13, 12, 11, and 14, respectively. The numbers of provinces in the run-in stage were 15, 13, 7, 6, and 6, respectively. The numbers of provinces in the basic coordination were 4, 4, 6, 8, and 7, respectively. During this period, the coupling and coordinated development between IWU and IGD systems deteriorated. There were no provinces in the high coordination stage, the number of provinces in the high dysregulation stage and basic dysregulation stage showed an upward trend. Provinces in the run-in stage were gradually changed to basic dysregulation stage and high dysregulation stage. During 2014–2018, 30 provinces developed five kinds of coupling coordination stages. The numbers of provinces in the high dysregulation stage were 0, 3, 1, 3, and 0, respectively. The number of provinces in the basic dysregulation stage was 9, 11, 13, 10, and 8, respectively. The number of provinces in the run-in stage was 11, 7, 7, 8, and 10, respectively. The provinces in the basic coordination stage were 7, 6, 5, 8, and 9, respectively. The provinces in the high coordination stage were 7, 6, 5, 8, and 9, respectively. Compared with the previous period, the provincial coupling coordinated development has been improved. Number of provinces in the high dysregulation stage and basic dysregulation stage showed a downward trend, the number of provinces in the high coordination stage increased gradually. Provinces in the run-in stage and basic dysregulation stage gradually changed to the basic coordination stage.
3.4. Analysis of Spatial-Temporal Characteristics of Coupling Coordination
3.4.1. LISA Time Path Analysis
Based on the evolution characteristics of the coupling coordination degree shown in
Section 3.3 (
Figure 3 and
Figure 4), the research period was divided into two development stages: 2009–2013 and 2014–2018. The relative length, tortuosity, and movement direction of coupling coordination degree between IWU and IGD systems in China were calculated based on the LISA time path, as shown in
Figure 5.
Based on the natural break point method in ArcGIS and manual classifications, the relative lengths of the LISA time path were classified into a low relative length (Γ
i < 0.33), medium relative length (0.33 < Γ
i < 0.9), higher relative length (0.9 < Γ
i < 1.3), and high relative length (Γ
i > 1.3) (
Figure 5a,b). During 2009–2013, there were 11 provinces with a relative length greater than one. There were nine provinces with higher relative length and six provinces with high relative length, these provinces included Xinjiang, Yunnan, Jilin, Shanxi, Henan, and Anhui, indicating that the local spatial structure was highly dynamic. In contrast, there were 14 provinces with medium relative length, and only Shanghai was in low relative length. These provinces, such as Shanghai, Jiangsu, Shandong, Fujian, Anhui, Beijing, Tianjin, Chongqing, and Sichuan, are mostly located in the eastern and central region, with stable local spatial structures. During 2014–2018, there were 14 provinces with a relative length greater than 1. Although the proportion of provinces with relative length greater than the national average increased, only Henan was in high relative length. Moreover, the number of provinces with low relative length increased to seven. Generally, during 2009–2018, the spatial structure of the coupling coordination degree between IWU and IGD systems was gradually stable and the spatial structure of the eastern region is more robust than that of the central and western regions.
Similarly, the tortuosity of the LISA time path was classified into a low tortuosity (0.45 <
fi < 3.24), medium tortuosity (3.24 <
fi < 8.22), higher tortuosity (8.22 <
fi < 19.58), and high tortuosity (19.58 <
fi < 67.32) (
Figure 5c,d). During 2009–2013, In all provinces except Shanghai, the tortuosity was more significant than 1, indicating that most provinces were affected by the spatial dependence of local structures to varying degrees. Gansu and Henan were in high tortuosity, and Jiangsu and Qinghai were in higher tortuosity. It reflected the solid dynamic characteristics of their spatial transition process. A total of 18 provinces were in low tortuosity, accounting for 60%, and these provinces presented a continuous patchy distribution. During 2014–2018, the tortuosity of all provinces was more significant than 1, indicating that all provinces were affected by the spatial dependence of local structures. Jiangsu and Zhejiang were in high tortuosity, Guangdong, Qinghai, Gansu, Inner Mongolia, Hebei, Shandong, and Henan were in higher tortuosity, thus reflecting the intense volatility of the spatial dependence of coupling coordination degree in these provinces and a dynamic change process among these provinces and their neighborhoods. The provinces with low tortuosity decreased to only seven, accounting for 23.33%. In contrast, the proportion of provinces with high and higher tortuosity increased to 30%, indicating that the coupling coordination degree between IWU and IGD systems had a strong spatial dependence. Its evolution volatility has been enhanced during the research period.
The LISA movement direction was used to characterize the spatial integration of the spatial variation in the coupling coordination between IWU and IGD systems in China. Four types were delineated by referring to the average level. As shown in
Figure 5e,f, during 2009–2013, Jiangsu, Zhejiang, Shanghai, Beijing, Tianjin, Shandong, Fujian, Henan, and Hubei showed a synergistic high growth trend, while Chongqing, Guizhou, Guangdong, and Shanxi showed a synergistic negative growth trend. Provinces with high synergistic growth are mainly located in the eastern region, while those with low synergistic growth are primarily located in the western region. During 2014–2018, the number of provinces with high synergistic growth decreased from eight to seven, and the number of provinces with low synergistic growth increased from three to five, indicating that the spatial integration of the spatial evolution in coupling coordination was slightly enhanced in this period compared to the previous period. The spatial variation of the coupling coordination degree between IWU and IGD systems in China performed differently in the east, central, and western regions. This situation remains consistent in the short term.
3.4.2. LISA Spatial-Temporal Transition
This paper constructs the spatial–temporal transition matrix to explore the transfer characteristics of the local correlation types of coupling coordination degree between IWU and IGD systems. In
Table 7, from 2009 to 2018, LL/LL had the highest transfer probability (0.333) with 91 samples. In terms of the transition type, the most common transition was Type I, which had a ratio of 82%, thus indicating that the spatial agglomeration of coupling coordination degree between IWU and IGD systems had a relatively higher path dependence and locked spatial features. The reason was that, in the process of rapid industrialization, the low IWU efficiency and low IGD degree caused by the resource-intensive industrial development mode could not be changed in the short term. The ratios of transition types II, III, and IV were 0.078, 0.067, and 0.033, respectively, thus reflecting that the internal factors of the unit had a more significant impact on their concentration of coupling coordination than the neighborhood overflow. The
SF was only 0.144, and
SC was 0.841 from 2009 to 2018. It indicates that the distribution of coupling coordination degree between IWU and IGD systems in China has strong transfer sluggishness, and the coupling coordination degree was relatively stable.