Abstract

Beijing is a megalopolis with a serious water shortage that has been further exacerbated by an unreasonable industrial water structure. This article uses an input-output method to calculate the water use coefficients in each industrial sector in Beijing and analyses the water use characteristics of the various industrial sectors. Then, an industry association index that represents the influence and sensitivity of sectors is combined with the water use characteristics to readjust the industrial structure with the objectives of water conservation and sustainable economic development. The results indicate that the agricultural water use coefficient is the highest and that the coefficients are generally higher for sectors in secondary industry than for those in tertiary industry. In addition, all coefficients display a downward trend. The water use multipliers vary widely among sectors. In secondary industry during the study period, the number of high water use sectors remained stable, the number of potentially high water use sectors increased and the number of general water use sectors decreased. A comprehensive analysis of the water use characteristics and industrial structure correlations could provide a reference for the optimal allocation of water resources.

INTRODUCTION

Located in the Haihe Basin, Beijing is a megalopolis with serious water shortages. In 2017, the water resources in Beijing totalled 2.98 billion m3 (BWAA 2017), and according to the resident population of 21.71 million at the end of this year, the per capita water resource total is 137 m3 (domestic water, industrial water, agricultural water, and ecological water accounts for about 46%, 9%, 13%, and 32% of the total, respectively), which is far from the global average value. The water demand will remain high given continued population growth and the spread of urbanization. Optimizing the water use structure to maximize the economic output given the limited water inputs has become an important topic (Fu et al. 2017; Li et al. 2018). An in-depth analysis of the industrial water use characteristics associated with industrial structure optimization has great practical significance.

Studies have shown that a large portion of water resources are consumed by agriculture and the thermal power industry in China (Okadera et al. 2015; Zhuo et al. 2016). However, when considering the relationships among industrial sectors, the economic system contains many sectors, and sectors in the same industrial chain can be divided into upstream and downstream sectors according to their relative positions. A sector consumes water resources in its production process, and when products are reused as raw materials for production in the downstream sector, the water resources hidden in the upstream sector will flow to the downstream sector, which may significantly affect the water use of the downstream sector. For example, direct water use by the food processing industry and gaming industry is small (Li & Chen 2014), but their direct water use is significantly lower than their indirect water use because the food processing industry is an important downstream sector of agriculture, and food production requires a large number of water-intensive products provided by agricultural sectors. For the gaming industry, direct water use is low, and its large indirect water use is mainly reflected in the accommodation and catering activities associated with the sector, labour employment and the purchase of game-related souvenirs. Current water resource management policies focus mainly on sectors with large direct water use but neglect sectors with large indirect use; that is, the water consumption of the entire economic chain is neglected (Liu et al. 2017). Therefore, these policies are not conducive to the control of water use or to the formulation of targeted water-saving policies. Therefore, a thorough understanding of the connection between economic development and water use, as gained by exploring sectors with high indirect water use by means of industry association, will have great practical significance for controlling the water use of the entire economic system.

Scholars have conducted research on the water use efficiency of industry sectors, and the research methods mainly include the input-output method, water footprint, LMDI factor analysis and data envelopment analysis methods. Compared with other methods, the input-output method has a simple calculation process and low data demand. This approach is also the most advantageous in tracking indirect water use because it can link producers and consumers through their economic activities based on the input-output table and effectively avoids the truncation error caused by artificial boundary selection. For water resources, this method can account for indirect water use by tracing the upstream components of the studied sector and identifying the closest real industrial water use characteristics. The Leontief inverse matrix used in the input-output calculation process also plays a crucial role in describing the relationship between environmental resources and the economy (Peters & Hertwich 2008; Li & Han 2017). In relevant studies of industrial water accounting, the input-output method has been widely used (Gray & McKean 1975; Zhao et al. 2010; Zhang & Song 2011; Zhi et al. 2015; Li & Han 2017). The input-output method has been further integrated with ecological network analysis and other methods to build socioeconomic networks and explore water flow characteristics among sectors (Wang & Chen 2016; Wang et al. 2019). All these studies revealed that economic development relies strongly on water resources. Existing research on industrial water use based on the input-output method has mainly focused on water use coefficient analysis; however, when making industrial structure adjustment plans based on water use intensity, the complex interrelationships among sectors are not taken into consideration in this method. Moreover, the number of years that can be analysed is limited, making it difficult to see trends over time. Therefore, using the input-output method, this article further considers the relationship between sectors by means of an industry association index, and explores water use status in the various industrial sectors of Beijing.

The main contributions of this article include the following three aspects. First, the water use coefficients of the various sectors are calculated using the input-output method, and the variation characteristics of these coefficients are compared across years, which helps to reveal the complex water connections between sectors. Second, based on the indexes of the water use characteristics, the sectors are divided into three categories: high, potentially high, and general water use sectors. Identifying key sectors in this manner will help to determine the overall goal for industrial water savings. Third, an industry association index, which represents the influence and sensitivity of the sectors, is combined with the water use characteristics; the purpose of this step is to complete industry restructuring under the dual premise of ensuring stable economic development and saving water, thereby achieving the bidirectional goal of saving water and developing the economy.

Study area

Beijing is located in the Haihe River Basin, which covers an area of 16,410 km2 between 39°26′N and 41°03′N and between 115°25′E and 117°30′E. Beijing belongs to the continental monsoon climate zone; it is hot and rainy in summer and cold and dry in winter. The multiyear average precipitation is 595 mm/y (1950–2017), presenting a significant decreasing trend. Related research shows that future precipitation in Beijing will show an upward trend (Zhu et al. 2012); however, because Beijing belongs to the continental monsoon climate zone, precipitation will not increase significantly. The change in total water resources is greatly affected by precipitation, and the annual average water resources total 2.61 billion m3 (2003–2017). In terms of the water supply, groundwater and reclaimed water were the main components of Beijing's water supply in 2017, totalling 1.66 billion m3 and 1.05 billion m3 and accounting for 42% and 27% of the total supply, respectively. The amount of water transferred from the South-to-North Water Diversion Project and surface water accounted for relatively small proportions of the total supply (22% and 9%, respectively). The weighted average terminal water price of various water sources is 6.76 yuan/m3 (Tian et al. 2014). The contradiction between the water supply and demand will remain prominent, and this difference is manifested as follows: (1) water supply capacity is very limited, the basic water demand must be met by transferred water; (2) the long-term over-exploitation of groundwater can lead to a decline in the groundwater level; (3) water pollution has exacerbated the water crisis; (4) the large population and rapid urbanization have increased the domestic water demand; and (5) the coordinated development of Beijing-Tianjin-Hebei has led to high water use in the tertiary industry.

RESEARCH METHODS AND DATA SOURCES

Water use coefficient calculation

In industrial water use, the direct water use coefficient refers to the ratio of the total amount of water used in the production process to the total output of a sector. The greater this value is, the more water is required per unit of output product (Gu et al. 2014); the calculation formula is as follows: 
formula
(1)
where is the direct water use coefficient of sector is the total water use, and is the total output (compared with other economic indicators, taking the total output as the denominator can reflect the resource input of the sector in the entire economic production stage). The direct water use coefficients for all sectors can be expressed in matrix form as .
In addition to direct water use, sector production also requires indirect water from other upstream sectors. Indirect and direct water collectively constitute total water use and are characterized by the total water use coefficient (Gu et al. 2014), which can be calculated by multiplying the direct water use coefficient matrix with the Leontief inverse matrix; the following formula is then obtained (Gu et al. 2014): 
formula
(2)
where is the total water use coefficient matrix and is the Leontief inverse matrix, represents the identity matrix; is the direct consumption coefficient matrix, which is obtained by , is the intermediate input matrix of the input-output table, is the total input matrix constituted by .
The water use multiplier refers to the ratio of a sector's total water use coefficient to its direct water use coefficient. A large index value implies that the total water use is higher than the direct water use, and the indirect water consumption, which is the difference between these two terms, is correspondingly high. In this case, the given sector has relatively close economic and technical contact with other sectors. The corresponding formula is as follows (Tian & Tang 2010): 
formula
(3)

It is worth noting that the monetary unit in the input-output table is 10,000 yuan (1 yuan ≈ 0.132, 0.147, and 0.159 USD in 2007, 2010, and 2012, respectively), which is a common unit used in Chinese statistics. In this article, unless otherwise specified, the monetary unit is 10,000 yuan.

Classification of industrial water use characteristics

The extent of industrial water use is relative and is measured by the overall level of water use by the local economy. The measurement standard involves a comparison of relative water use indexes across various sectors. Sectors are divided into three categories according to their water use characteristics (Kong 2013), as shown in Table 1.

Table 1

Classification of industrial water use characteristics

ClassificationClassification foundation
High water use sector Rsj ≥ 1 or Rqj ≥ 1 
Potentially high water use sector Rsj < 1 and Rqj < 1, R∂j ≥ 1 
General water use sector Rsj < 1 and Rqj < 1, R∂j < 1 
ClassificationClassification foundation
High water use sector Rsj ≥ 1 or Rqj ≥ 1 
Potentially high water use sector Rsj < 1 and Rqj < 1, R∂j ≥ 1 
General water use sector Rsj < 1 and Rqj < 1, R∂j < 1 

In Table 1, sectors are divided into high, potentially high, and general water use sectors, and three new indexes are included in the calculation; that is, the relative water use structure coefficient (), relative water use coefficient (), and relative water use multiplier (). The detailed meanings of the indexes are as follows.

refers to the extent of the influence of the water use of one sector on the water use of all sectors; a value greater than 1 indicates that the degree of influence of water use in this sector exceeds the average level. The computational formula is as follows: 
formula
(4)
where is the total number of industrial sectors and the other coefficients are the same as above.
refers to the ratio of the direct water use coefficient for a sector relative to the average direct water use coefficient for all sectors. If the value is greater than 1, the water use in that sector is higher than the average level. The corresponding formula is as follows: 
formula
(5)
where is the average direct water use coefficient for all sectors. is the sum of direct water use for all sectors, and is the sum of the total inputs of all sectors.
refers to the ratio of the water use multiplier for a sector to the average water use multiplier for all sectors. This variable reflects the extent to which a change in water use in a sector influences the water use in all sectors and is calculated as follows: 
formula
(6)
where is the average relative water use multiplier for all sectors and is the water use multiplier of sector .

Industry association index

In terms of water conservation, the national economic system constantly aims to reduce water use; in terms of economic output, the goal is to obtain the largest possible total output from the economic system. To compare and balance these two objectives, the diffusion coefficient and inducing coefficient are introduced for further analysis.

is the degree of diffusion relative to production in other sectors when a given sector increases its per unit output. A high value indicates a strong effect of a given sector on other sectors, and the computational formula is as follows (Liu 2002): 
formula
(7)
where is the total demand in a given sector relative to that in other sectors, is the corresponding element in matrix , and is the weight, which can be expressed as follows: 
formula
(8)
where is the ratio of the output in sector to the total output of all sectors.
represents the sensitivity of a sector when all other sectors increase their per unit output. This variable reflects the inducing capability of a given sector caused by productivity changes in the entire economic system and can be calculated as follows (Liu 2002): 
formula
(9)
where is the corresponding element in the Leontief inverse matrix and is the ratio of the input from sector relative to the total input.

Data sources

Direct water use data were obtained from the Beijing Water Resources Bulletin (BWAA 2007, 2010, 2012), Beijing Statistical Yearbook (NBS 2008, 2011, 2013), and relevant statistical reports. Industry output data were obtained from the Beijing input-output table in 2007, 2010 and 2012. In addition, some sectors were merged to match the industrial water use data; with agriculture included, there were a total of 39 sectors, and the codes and names of sectors are shown in Table 2. Among them, agriculture is represented by A1. Secondary industry consists of 24 sectors, including mining, manufacturing, energy processing, and other industrial sectors, represented by S1–S24. Tertiary industry consists of 14 sectors, including various circulation sectors and service sectors, represented by T1–T14.

Table 2

Sector code and name

CodeNameCodeName
A1 Agriculture S19 Instruments & cultural office machinery manufacturing industry 
S1 Coal mining & washing industry S20 Craft & other manufacturing industry 
S2 Petroleum & gas extracting industry S21 Scrap & waste industry 
S3 Metallic ore mining & selection industry S22 Electricity, heat production & supply industry 
S4 Non-metallic ore, other ore mining & selection industry S23 Gas production & supply industry 
S5 Food manufacturing & tobacco processing industry S24 Water production & supply industry 
S6 Textile industry T1 Transportation, storage, & postal service industry 
S7 Textile, clothing, shoes, leather, & down product industry T2 Information transmission, computer service, & software industry 
S8 Wood processing & furniture manufacturing industry T3 Wholesale & retail trade industry 
S9 Paper printing, educational, & sports goods manufacturing industry T4 Accommodation & catering industry 
S10 Petroleum processing, coking, & nuclear fuel processing industry T5 Financial industry 
S11 Chemical industry T6 Real estate industry 
S12 Non-metallic mineral products industry T7 Leasing & business service industry 
S13 Metal smelting & calendaring processing industry T8 Scientific research, technical service & geological prospecting industry 
S14 Metal product industry T9 Water conservancy, environment & public facilities management industry 
S15 General & special equipment manufacturing industry T10 Resident service & other service industry 
S16 Transportation equipment manufacturing industry T11 Education 
S17 Electrical machinery & equipment manufacturing industry T12 Health, social security & social welfare industry 
S18 Communications equipment, computers & other electronic equipment manufacturing industry T13 Culture, sports & entertainment industry 
T14 Public administration & social organization industry 
CodeNameCodeName
A1 Agriculture S19 Instruments & cultural office machinery manufacturing industry 
S1 Coal mining & washing industry S20 Craft & other manufacturing industry 
S2 Petroleum & gas extracting industry S21 Scrap & waste industry 
S3 Metallic ore mining & selection industry S22 Electricity, heat production & supply industry 
S4 Non-metallic ore, other ore mining & selection industry S23 Gas production & supply industry 
S5 Food manufacturing & tobacco processing industry S24 Water production & supply industry 
S6 Textile industry T1 Transportation, storage, & postal service industry 
S7 Textile, clothing, shoes, leather, & down product industry T2 Information transmission, computer service, & software industry 
S8 Wood processing & furniture manufacturing industry T3 Wholesale & retail trade industry 
S9 Paper printing, educational, & sports goods manufacturing industry T4 Accommodation & catering industry 
S10 Petroleum processing, coking, & nuclear fuel processing industry T5 Financial industry 
S11 Chemical industry T6 Real estate industry 
S12 Non-metallic mineral products industry T7 Leasing & business service industry 
S13 Metal smelting & calendaring processing industry T8 Scientific research, technical service & geological prospecting industry 
S14 Metal product industry T9 Water conservancy, environment & public facilities management industry 
S15 General & special equipment manufacturing industry T10 Resident service & other service industry 
S16 Transportation equipment manufacturing industry T11 Education 
S17 Electrical machinery & equipment manufacturing industry T12 Health, social security & social welfare industry 
S18 Communications equipment, computers & other electronic equipment manufacturing industry T13 Culture, sports & entertainment industry 
T14 Public administration & social organization industry 

RESULTS AND DISCUSSION

Analysis of the direct water use coefficient and total water use coefficient

The direct water use coefficient and total water use coefficient for agriculture (A1) both declined, but they were ranked first during the study period (Figure 1). Both of these coefficients in the secondary industrial sectors were generally higher than those in the tertiary industrial sectors, and they showed a downward trend in most sectors. The extent of the decrease in the secondary industrial sectors was greater than that in the tertiary industrial sectors; these characteristics are closely related to Beijing's industry structural adjustment and enhancements in water use efficiency in recent years. Sectors in the secondary and tertiary industries in 2012 are presented as study samples.

Figure 1

Water use coefficients for each sector: (a), (c), and (e) represent the water use coefficients of the secondary industrial sectors in 2007, 2010, and 2012, respectively; (b), (d), and (f) represent the water use coefficients of the tertiary industrial sectors in 2007, 2010, and 2012, respectively.

Figure 1

Water use coefficients for each sector: (a), (c), and (e) represent the water use coefficients of the secondary industrial sectors in 2007, 2010, and 2012, respectively; (b), (d), and (f) represent the water use coefficients of the tertiary industrial sectors in 2007, 2010, and 2012, respectively.

As observed in Figure 1(e), in secondary industry, the energy processing and supply sectors, such as petroleum processing, coking & nuclear fuel processing (S10) (134.32 m3/10,000 yuan) and electricity, heat production & supply (S22) (89.23 m3/10,000 yuan), had relatively high direct water use coefficients, which indicates that these sectors directly consumed a large amount of water. The petroleum & gas extracting industry (S2) (0.42 m3/10,000 yuan) and coal mining & washing industry (S1) (0.35 m3/10,000 yuan) had relatively low direct water use coefficients, which indicates that these sectors had relatively low dependence on water. For the total water use, the metal smelting & calendaring processing industry (S13) (214.9 m3/10,000 yuan) and S22 (208.48 m3/10,000 yuan) had relatively high coefficients. To ensure an increase in the per unit output, the entire economic system must thus increase water use by a large amount. Expanding the production scale in these sectors would place great pressure on water resources in Beijing. Sectors with low total water use coefficients, such as S19, S6, and S8, use less water than other sectors, and their development is suitable for the current water shortage situation in Beijing. Compared with the fundamental sectors, the light sectors had lower direct water use coefficients, but the gap between the direct and total water use coefficients was greater than that in the fundamental sectors. High-tech sectors had low water use coefficients, and the gap between the two types of coefficients was small.

As illustrated in Figure 1(f), in the tertiary industrial sectors, water conservancy, environment & public facilities management (T9) (22.33 m3/10,000 yuan) and education (T11) (6.42 m3/10,000 yuan) had relatively high direct water use coefficients, and these sectors had high water use intensities in their production processes. The financial industry (T5) (0.16 m3/10,000 yuan) and information transmission, computer service & software industry (T2) (0.24 m3/10,000 yuan) had relatively low values and are less dependent on water resources during the production process than are other sectors. With respect to total water use, T9 (22.46 m3/10,000 yuan) and T11 (7.58 m3/10,000 yuan) had relatively high total water use coefficients and require the economic system to provide a large amount of water to ensure their development. T5 (1.30 m3/10,000 yuan) and T2 (1.06 m3/10,000 yuan) had relatively low values; therefore, expanding production in these sectors will not increase the pressure on regional water resources.

Analysis of the water use multiplier

The indirect pulling function of industrial water use can be reflected by the water use multiplier. The water use multipliers of 39 sectors in Beijing in 2012 are listed in Table S1 in the Supporting Information. Among the secondary industrial sectors, S2 and the electrical machinery & equipment manufacturing industry (S17) had relatively high water use multipliers. Of the tertiary industrial sectors, T5 and the transportation, storage and postal service industry (T1) had high water use multipliers. Sectors with high water use multipliers had relatively close economic and technical contact with other sectors; they also had high total water use coefficients but low direct water use coefficients, and their indirect water demand exceeded their direct demand. Taking S2 as an example, the corresponding water use multiplier was 82.05, meaning that when this sector consumed 1 m3 of water, 81.05 m3 of water was indirectly consumed; thus, the sector had a strong effect on water utilization. The average water use multiplier was 21.44 for secondary industrial sectors, 2.75 for tertiary industrial sectors, and 14.55 overall. Indirect water use in secondary industrial sectors was greater than that in tertiary industrial sectors. In secondary industrial sectors, the average water use multiplier was 26.65 for high-tech industries, 22.98 for fundamental industries, and 11.03 for light industries; these values reflect the large difference in indirect water use among sectors. The water use multiplier for A1 was only 1.09, meaning that agriculture consumed little indirect water, and its production relied mainly on direct use.

Two sectors are taken as examples to provide more detailed information. The water use multiplier for S17 was 65.90, the direct water use coefficient was 0.98 m3/10,000 yuan, and the total water use coefficient was 64.58 m3/10,000 yuan, suggesting that 10,000 yuan in output from this sector required the direct consumption of only 0.98 m3 of water, and the remaining 63.6 m3 of water was consumed by goods and services from other sectors. The production in this industry had strong stimulating effects on other sectors. S19 had low direct and total water use coefficients, and its water use multiplier was low, which indicates that sector production was not highly dependent on water resources. This is a water-saving sector, which is generally beneficial for sustainable development in Beijing.

Analysis of direct water use and total water use

In 2007, 2010 and 2012, the direct water use of various sectors was 3.48 billion m3, 3.52 billion m3, and 3.59 billion m3, respectively, and the total water use was 9.71 billion m3, 9.01 billion m3, and 9.66 billion m3, respectively, 2.79, 2.56, and 2.69 times the direct use. The difference between these two values is the indirect water use. As the capital of China, Beijing has frequent economic trade with other provinces in China and other countries in the world; therefore, this part of indirect water use also includes water embodied in imports from other provinces and other countries. The analysis above indicates that the water use coefficients show a downward trend; if the water use coefficients in 2010 and 2012 were maintained at the 2007 level, the direct water use would increase by 2.29 billion m3 and 3.95 billion m3, respectively, and the total water use would increase by 6.75 billion m3 and 8.57 billion m3, respectively. Therefore, in the context of rapid economic development, technological progress plays a crucial role in maintaining the relative stability of water use.

Cluster analysis based on industry associations and water use characteristics

Currently, Beijing is facing severe industrial water supply and demand pressure. Adjusting the industry structure and developing a water-saving industry pattern could have important practical significance for Beijing. In an economic system, improper industrial structure adjustment may have a series of negative effects on the entire economic system. The inducing coefficient and diffusion coefficient were introduced to measure the degree of industry association from an economic perspective. By introducing the diffusion coefficient and inducing coefficient, we can identify the influence of a sector on the economic system and its sensitivity to economic system changes to identify the sectors with high influences and determine the sensitivity of the economic system. This approach allows us to effectively analyse the intrinsic characteristics of the industrial structure and provide suggestions for water-saving industry patterns.

Sectors were divided into high, potentially high, and general water use sectors. The dichotomy method was selected, with the diffusion coefficient used as the abscissa, the inducing coefficient as the ordinate, and 1 as the boundary. The entire coordinate system was divided into four quadrants, and the quadrant classification map of industrial water use characteristics and industry associations is presented in Figure 2.

Figure 2

Water use characteristic classification map for sectors in Beijing. Please refer to the online version of this paper to see this figure in colour: http://dx.doi.org/10.2166/ws.2019.152.

Figure 2

Water use characteristic classification map for sectors in Beijing. Please refer to the online version of this paper to see this figure in colour: http://dx.doi.org/10.2166/ws.2019.152.

Figure 2(a)–2(f) represent the water use characteristics of the secondary industrial sectors and tertiary industrial sectors in 2007, 2010, and 2012, respectively. The red diamonds, blue squares, and green triangles represent the high, potentially high, and general water use sectors, respectively.

In Figure 2, with respect to secondary industrial sectors during the study period, the number of high water use sectors remained stable, whereas the number of potentially high water use sectors increased and that of general water use sectors decreased. Sectors in the secondary industries were distributed among the four quadrants, and the general water use sectors had a tendency to concentrate in the third quadrant. The inducing and diffusion coefficients of these sectors tended to decrease. The distribution of high water use sectors was relatively stable, and they were mainly distributed in the first and second quadrants. The corresponding inducing coefficients were also large. Potentially high water use sectors were mainly distributed in the third and fourth quadrants, and their inducing coefficients were small. Compared with the secondary industrial sectors, sectors in tertiary industry were mainly concentrated in the first and third quadrants. The inducing and diffusion coefficients of these sectors were both large or both small, and they remained stable over the three study periods. The water use characteristics of secondary industrial sectors in 2012 (Figure 2(e)) were investigated in detail to explain the connections between industrial water use characteristics and industry associations.

Sectors in the first quadrant had intertwining relationships with other sectors, indicating that they could greatly affect the development of other sectors. There were six sectors located in this quadrant, all of which were secondary industrial sectors. Both the inducing and diffusion coefficients of these sectors were larger than 1. Moreover, these sectors were closely associated with and affected the development of other sectors. Among these six sectors, S1 had potentially high water use, and its was small, which meant that it has room to improve with respect to total water use control. The corresponding value was also small, suggesting that the water use efficiency of the sector was higher than the average level. The government should implement appropriate policy measures to ensure the continued development of this type of sector. This sector had a high value, which indicated that a large amount of hidden water was contained in its products, and it had close technical contact with other sectors. S16 was a general water use sector. Both the and values of this sector were less than 1, which meant that it had a high water use efficiency and the potential to improve with respect to total water use control. The development of this sector can affect the development of other sectors. Therefore, this sector should be advanced.

Sectors in the second quadrant had high inducing coefficients and low diffusion coefficients. These sectors had a strong stimulatory effect on the development of other sectors; additionally, they had a restrictive effect on the development of some sectors. Among them, S10 and S11 were high water use sectors, and S6 was a potentially high water use sector. With respect to the high water use sectors in this quadrant, relevant departments should focus on increasing the water use efficiency. Then, industrial restructuring should be considered, and ultimately, sector transfers can be implemented if necessary. These sectors had high inducing coefficients; therefore, a virtual water trade method could be used in these sectors to replace the current method, thereby satisfying the demands of other sectors.

Sectors in the third quadrant had weak inducing and diffusion coefficients, and the association degree was poor. There were eight sectors located in this quadrant; among them, S23, S17, and S15 were potentially high water use sectors, and the others were general water use sectors. The influence of potentially high water use sectors in this quadrant was low with respect to the entire economic system, and they did not significantly impact the development of other sectors; however, they had high water use. Therefore, relevant departments should implement policies to limit the development of these sectors and related transfers, and these sectors should eventually be phased out of the industry structure.

Sectors in the fourth quadrant had weak inducing coefficients and strong diffusion coefficients. These sectors could greatly affect the entire economic system but were not easily affected by the system. Six sectors appeared in this quadrant. S24 was a high water use sector; S8, S14, S21, and S20 were potentially high water use sectors; and S12 was a general water use sector. The development of high water use sectors in this quadrant should rely on the corresponding administrative departments, and each industry department should strictly enforce the industrial water quota and consider implementing a restructuring strategy.

Policy guidelines

Some suggestions are proposed to increase the water savings by industry in Beijing.

  • (1)

    To decrease industrial water use in Beijing, the government must strengthen research into water-saving technologies in the various sectors and promote the application of new water-saving devices. Technical renovations to water equipment should be implemented in high water use sectors, such as S10, S22, S3, T9, T4, T11 and T6. Outdated high water-use equipment and processes should be eliminated, and the water prices in these sectors should appropriately increase. By improving wastewater reuse and recycled water technologies and perfecting the ladder water price system, the water use by industry in Beijing should substantially decrease. These measures need to be included in the first wave of industrial restructuring. If the water savings results are still not significant, relevant departments should implement industrial transfers.

  • (2)

    Efforts need to coincide with the ‘Water Pollution Control Action Plan’ by strictly controlling total industrial water use and imposing stringent water resource management policies. A water use efficiency assessment system needs to be implemented. Water conservation objectives and tasks will need to be included in local government performance evaluations. A reclaimed water reuse management system needs to be established and continually improved. For high water use industries, the related departments should enforce the use of water-saving equipment that meets the market access standards and make efforts to perform water conservation work under legal protection.

  • (3)

    The system will need to use capital advantages and construct coordinated development mechanisms that couple water use and economic development. Water prices must be considered an important limiting factor in industrial water utilization. Development should focus on high-tech industries, such as T2, because they have low water use and strong economic driving effects. Moreover, under the premise of guaranteeing a high water savings efficiency, modern service industries that are closely related to traditional Chinese cultural activities should be actively pursued.

CONCLUSION AND PROSPECTS

This research used the input-output method and 2007, 2010, and 2012 as reporting periods to calculate the direct water use coefficient, total water use coefficient, and water use multiplier index for each sector in Beijing. Then, the industry associations were combined with water use characteristics to adjust the industrial structure. The goal of this paper was to provide a basis for management decisions to support water conservation and steady economic growth. The results indicate that agriculture had the highest water use coefficient. The coefficients were generally higher for sectors in secondary industry than in tertiary industry, and they all showed a downward trend. Agriculture had a small water use multiplier, and this index varied widely among sectors in secondary and tertiary industries. The number of high water use sectors in secondary industry remained stable, the number of potentially high water use sectors increased, and the number of general water use sectors decreased. With respect to tertiary industrial sectors, the numbers of all categories of sectors remained stable. A comprehensive analysis of the water use characteristics and industry structure correlation can provide a reference for industrial restructuring.

For future studies, three corresponding prospects are listed below. First, water resources are different from nonrenewable energy and are not completely consumed in the utilization process. Therefore, a study of the relationship between water and the economy from the perspective of net water use might be reasonable. Second, industrial restructuring in this article is considered only from the perspective of water resources. If other resources, especially energy and land use, which are also closely related to water and the economy, are taken into account in future research, a more reasonable industrial structure optimization strategy might be obtained. Finally, the input-output table for 2012 in Beijing is the most recent version, leading to a notable time delay between the year the data were obtained and the publication of this article. Future research can update the time series and propose more appropriate policy suggestions.

SUPPLEMENTARY MATERIAL

The Supplementary Material for this paper is available online at https://dx.doi.org/10.2166/ws.2019.152.

REFERENCES

REFERENCES
BWAA (Beijing Water affair Authority)
2007
Beijing Water Resources Bulletin
.
China Water & Power Press
,
Beijing
(in Chinese).
BWAA (Beijing Water affair Authority)
2010
Beijing Water Resources Bulletin
.
China Water & Power Press
,
Beijing
(in Chinese).
BWAA (Beijing Water affair Authority)
2012
Beijing Water Resources Bulletin
.
China Water & Power Press
,
Beijing
(in Chinese).
BWAA (Beijing Water affair Authority)
2017
Beijing Water Resources Bulletin
.
China Water & Power Press
,
Beijing
(in Chinese).
Fu
Q.
Xiao
Y.
Cui
S.
Liu
D.
Li
T.
2017
Optimization of water use structure and plantation benefit of unit water consumption using fractional programming and conditional value-at-risk model
.
Int. J. Agric. Biol. Eng.
10
(
2
),
152
162
.
Gray
S. L.
McKean
J. R.
1975
An Economic Analysis of Water Use in Colorado's Economy: An Input-Output Analysis
.
Colorado State University
,
Fort Collins, CO, USA
.
Kong
W.
2013
The Study of Industry Water Constraints and Industry Development Strategy of Beijing
.
Beijing University of Technology
,
Beijing, China
(in Chinese).
Liu
Q.
2002
Research on structural analysis method of input-output coefficient
.
Stat. Res.
2
,
40
42
(in Chinese)
.
Liu
S.
Wu
X.
Han
M.
Zhang
J.
Chen
B.
Wu
X.
Wei
W.
Li
Z.
2017
A three-scale input-output analysis of water use in a regional economy: Hebei province in China
.
J. Cleaner Prod.
156
,
962
974
.
NBS (National Bureau of Statistics of China)
2008
China Statistical Yearbook
.
China Statistics Press
,
Beijing
(in Chinese).
NBS (National Bureau of Statistics of China)
2011
China Statistical Yearbook
.
China Statistics Press
,
Beijing
(in Chinese).
NBS (National Bureau of Statistics of China)
2013
China Statistical Yearbook
.
China Statistics Press
,
Beijing
(in Chinese).
Okadera
T.
Geng
Y.
Fujita
T.
Dong
H.
Liu
Z.
Yoshida
N.
Kanazawa
T.
2015
Evaluating the water footprint of the energy supply of Liaoning Province, China: a regional input–output analysis approach
.
Energ. Policy
78
,
148
157
.
Peters
G. P.
Hertwich
E. G.
2008
CO2 embodied in international trade with implications for global climate policy
.
Environ. Sci. Technol.
42
(
5
),
1401
1407
.
Tian
G. L.
Tang
P.
2010
Research on Measurement of Water Using of Each Sector Based on Water Input-output Model
. In:
International Conference on Information Management, Innovation Management and Industrial Engineering
,
577
582
.
Tian
G. L.
Wu
Q.
Xu
C.
2014
Compensation standard of water price in benefited areas of south-to-north water diversion based on visual water theory
.
China Water Resour.
2
,
14
16
(in Chinese)
.
Zhang
R.
Song
H.
2011
Optimal Allocation of Water Resources Research Based on Input-Output Method of Hebei Province
. In:
Fourth International Joint Conference on Computational Sciences and Optimization
, pp.
570
573
.
Zhu
L.
Chen
Y.
Yan
R.
Shen
T.
Jiang
L.
Wang
Y.
2012
Characteristics of precipitation and temperature changes in Beijing City during 1951–2010
.
Res. Sci.
34
,
1287
1297
(in Chinese)
.

Supplementary data