Measuring the Gains and Losses of Virtual Water Flows in China’s Coastal Areas

Abstract

The virtual water flow behind product trade provides researchers with ideas to alleviate water problems in China’s coastal areas, with accompanying resource and economic implications. This paper adopts a multi-region input–output model to calculate virtual water flows in coastal areas and analyses resource benefits and economic benefits, by combining the water stress index and shadow price, to measure gains and losses of virtual water flow. This study shows that (1) China’s coastal areas depends on external water supplies; (2) virtual water flows between coastal and inland areas generated 38.26 billion m³ of net scarce water savings and CNY 31,751 billion of net economic benefits, indicating that coastal areas benefited from virtual water flows from both resource and economic perspectives; (3) virtual water flows among coastal provinces and cities caused 11.10 billion m³ of net scarce water losses, but generated CNY 9784 billion of net economic gains, indicating that a further intensification of water stress in coastal areas, but such a virtual water flow pattern was beneficial from an economic perspective. This paper reveals the resource and economic impacts of virtual water flow at the national scale and among coastal provinces and cities and further proposes suggestions for promoting the beneficial flow of virtual water in coastal areas.

1. Introduction

China’s coastal areas, with 14% of the country’s total water resources, contain 70% of the country’s population and 60% of the country’s economic output. The adjustment of the economy and the influx of the floating population put forward higher requirements for water resources, which lead to the contradiction between supply and demand being further intensified [1]. The outline of the 14th Five-Year Plan and 2035 Vision for the Development of the National Maritime Economy state that it is important to strengthen the management and protection of water resources in coastal areas and to improve the efficiency of water resource utilization in order to meet the challenges of water shortages, water pollution and the destruction of water ecosystems. There is a complex relationship between water resources’ allocation and economic growth [2], and water issues have become a bottleneck, constraining the realization of sustainable economic development in China’s coastal areas. Local water shortages will not only affect the local economy, but also affect China’s economy as a whole through the industrial chain, and there is a “Cascading” impact. Economic development requires higher water use efficiency to meet demands in order to alleviate water scarcity. In the context of highly marketized and complex commodity trade, a large amount of virtual water implicit in water-intensive commodities is accompanied by trade for inter-regional water redistribution, and the impact on China’s regional water security is becoming increasingly unnoticeable [3]. Water resources are transferred through regional trade in the form of virtual water; on the resource side, virtual water flows from water-rich areas to water-scarce areas, generally easing their water stress and generating scarce water savings. On the economic side, virtual water transfers from economically underdeveloped areas to developed areas, increasing the marginal economic value of water resources and generating economic gains. Therefore, it is of great practical significance to analyze the pattern of virtual water flows in China’s coastal areas to assess their gains and losses in terms of both resources and the economy and to grasp the impacts of virtual water flows on trade linkages on coastal areas.

The concept of virtual water was first proposed by Professor Allan [4], and Hoekstra [5] further extended this concept to the amount of water required in the production of goods and services. Commodity trade can lead to inter-regional virtual water flow (VWF) [6], which indicates that the virtual water flow is caused by a difference in the virtual water content in commodities in inter-regional trade. In recent years, scholars at home and abroad have carried out virtual water flow studies on different spatial scales, such as global [7,8,9], national [10,11,12], regional and watershed [13,14,15,16], as well as provincial and municipal [17,18,19]. Some scholars have found that, due to the spontaneity of trade, there are situations in which some water-rich areas import products from water-poor areas, in reality [20,21]. This virtual water flow direction goes against the original intention of the virtual water strategy. The hidden virtual water flows behind product trade not only have an impact on regional water resources, but also on the economy; however, the impact of China’s inter-regional virtual water flow on alleviating water resource stress and developing the economy is still unclear, and this deserves further exploration.

The same virtual water flow has different impacts on regions with different water demands and water resource endowment. Domestic and foreign studies have pointed out that exploring the impacts of virtual water flow requires further measurements of regional water stress. Oki and Kanae [22] first proposed that virtual water trade can serve as a “conservation mechanism” for global water resources, but when the exporting place is a water-poor area and the importing place is a water-rich area, water savings may come at the cost of exacerbating water stress in the exporting region, and the effect on alleviating water stress in the importing region is limited, while the effect on alleviating global water stress is also unclear. Many studies have considered the water used for production under water stress to reflect the consumption of scarce water by commodity production. Unlike water saving benefits, scarce water saving benefits indicate an overall improvement in the water stress situation among trading partners, and the amount of scarce water saved represents the degree to which water stress is alleviated through trade.

Currently, the indicators for assessing water stress are constantly evolving and improving. The Falkenmark indicator [23], defined as the ratio of available water resources to the population, is limited by the fact that it treats the per capita available water resources as the per capita water demand [24], while Pfister et al. [25] proposed the measurement of water stress by the share of water used of the total available water resources, which are the two earliest indicators proposed. Zhong et al. [26] applied the Falkenmark indicator and improved the trade-related one to study the impact of international trade on water scarcity; Feng et al. [10] introduced the ratio of water withdrawal to water availability to assess the inter-provincial flow of scarce virtual water; Zhao et al. [27] found that areas with water stress export more than areas without stress by studying the physical water transfer in China’s water diversion projects and the virtual water flow between regions; Tian et al. [28] analyzed the direction and trend of inter-provincial water stress transfer in China in 2007 and 2012; Zhao et al. [29] constructed a scarce water saving accounting framework to test whether virtual water flows alleviate or exacerbate regional water stress; and Wu et al. [30] compared resource- and environment-oriented virtual water transfers.

The sustainability and efficiency of virtual water flows can be reflected from both resource and economic perspectives [31]. Analyzing virtual water flow from an economic perspective, it also has the same rule as the resource effect; when virtual water flows from areas with a high potential value of water resources to areas with low potential value, it will lead to economic losses. Han et al. [32] not only evaluated the scarce water savings and losses generated by inter-provincial virtual water flow in 2015, but also assessed the economic benefits and losses based on the shadow price of water resources. An et al. [33] used the Yellow River Basin as an example to investigate the inequality of virtual water consumption and economic benefits implicit in trade based on the production side and the consumption side. Zheng et al. [34] further explored the unfair exchange of virtual water consumption and the added value caused by China’s exports in 2002–2017, and all of these scholars analyzed virtual water flows from an economic perspective.

Previous studies have balanced the gains and losses of virtual water flow from a resource perspective, but when measuring the water stress in Chinese provinces and cities, the seawater resources in the specific geographical location of coastal areas have not been considered. The total available water resources in inland areas and coastal areas are different. From an economic perspective, existing studies have given less consideration to the economic impact of virtual water flows, and the indicators for evaluating their economic benefits are vague, without providing policy recommendations from an economic perspective. There are significant differences in the amount of water used for production under water stress and the level of economic development in different regions of China, and virtual water flow generates total gains or losses from both resource and economic perspectives. Therefore, the effectiveness of the practical application of virtual water flow for alleviating water resource stress and developing the economy is currently unclear and deserves further exploration. The unique value of this study is that it not only analyzes the virtual water flow pattern between coastal areas and inland areas among coastal provinces and cities, but it also evaluates the gains and losses of virtual water flow from the dimensions of resource and economic benefits by considering the water stress index and shadow price of each region.

The structure of this study presents a clear and logical progression. The first section of this paper introduces the research background, reviews the progress of research and shortcomings of virtual water flow benefit analyses, and illustrates the unique value and research purpose of this study; the second section provides detailed explanations of the methodologies and data sources used; the third section not only clarifies the theoretical basis for improving the water stress index of coastal provinces and cities and conducting an economic benefit analysis based on shadow prices, but also defines the concepts of the terms involved in the benefit analysis in detail. In the fourth part, on the basis of describing the virtual water flow pattern in coastal areas nationwide, and among coastal provinces and cities, the resource and economic benefits of virtual water flow in the above two areas are evaluated by considering the differences in the water stress index (WSI) and shadow price (SP) of each region. The fifth section discusses the applicability of the benefit analysis method used in this paper, puts forward detailed policy suggestions for coastal areas according to the research results and further elaborates future research directions, while the sixth section summarizes the research findings of this paper. This paper aims to promote the beneficial flow of virtual water in China’s coastal areas, so as to provide reference for alleviating water stress in coastal areas and optimizing virtual water trade relationships.

2. Materials and Methods

2.1. Study Area and Data Sources

The study area covers 11 coastal provinces, cities and autonomous regions of China, which have the advantage of being located near the sea and convenient economic links. Therefore, this paper considers that coastal areas can directly utilize seawater and that the amount of seawater directly utilized is usually not included in total water resources. Seawater desalination is a technology that converts seawater into freshwater resources. Therefore, this paper incorporates the amount of direct utilization of seawater and seawater desalination into the total available water resources (Q*) of coastal provinces and cities to improve the water stress index of coastal provinces and cities. The situation of additional available water resources in coastal provinces and cities is shown in Figure 1.

The multi-regional input–output model (MRIO) is the main method for accounting for virtual water flow (VWF) and was first proposed by Isard [35] to reflect the product flow relationship among various regions and sectors in an economic system. China’s MRIO table of 2017, with 42 sectors and 31 provinces (cities and autonomous regions), was obtained from the Carbon Emission Accounts & Datasets (CEADs) [36]. The water consumption data for agriculture, forestry, animal husbandry and fisheries come from the agricultural water consumption listed in the 2018 China Statistical Yearbook [37]. Since there are no specific water consumption data for industrial and tertiary sub-sectors, according to the treatment method of Zhang and Anadon [38], this paper calculates industrial water consumption and tertiary industry water consumption based on the input ratio of “water production and supply industry” in different sectors of different provinces for the multi-regional input–output table.

2.2. Virtual Water Flow (VWF)

The MRIO model contains m regions, and each region has n sectors, which can be represented by the following equation:

$$x_i^r={\displaystyle \sum _{s=1}^m{\displaystyle \sum _{j=1}^n{x}_{ij}^{rs}}}+{\displaystyle \sum _{s=1}^m{f}_i^{rs}}$$

(1)

where $x_i^r$ is the total output of sector i in region r, $x_{ij}^{rs}$ is the intermediate input of sector i in region r to sector j in region s and $f_i^{rs}$ is the input of sector i in region r to meet the final demand of region s.

$$a_{ij}^{rs}=\frac{x_{ij}^{rs}}{x_j^s}$$

(2)

where A is the technical coefficient matrix, $A=a_{ij}^{rs}$, which indicates the direct input of sector i of region r when sector j of region s produces a unit product; $x_j^s$ is the total output of sector j in region s.

This is expressed in a matrix as

$$X={\left(I-A\right)}^{-1}F$$

(3)

where X indicates the vector of the total output, I is an identity matrix, ${\left(I-A\right)}^{-1}$ is the Leontief inverse matrix and F refers to the vector of the total final demand.

$V_i^r$ reflects the volume of direct and indirect water consumption driven by the final demand, that is, the amount of virtual water contained in sector i per unit of product.

$$V_i^r=k_i^r{\left(I-A\right)}^{-1}$$

(4)

where $k_i^r$ refers to the direct water consumption intensity of sector i in region r, which is equal to the ratio of $w_i^r$ to $x_i^r$; $w_i^r$ is the direct water consumption required for the production in sector i in region r; and $x_i^r$ is the total output of sector i in region r.

Based on Equation (4), the virtual water outflow ($VWO$) and virtual water inflow ($VWI$) can be obtained from the following equation:

$$VW{O}^{rs}=V^r{\displaystyle \sum _{s\ne r}{e}^{rs}}$$

(5)

$$VW{I}^{sr}={\displaystyle \sum _{s\ne r}{V}^s}{e}^{sr}$$

(6)

The net inflow $V{W}_{net}^{rs}$ from region r to region s (r ≠ s) is expressed as

$$V{W}_{net}^{rs}=VW{I}^{sr}-VW{O}^{rs}$$

(7)

where $e^{rs}$ is the trade volume from region r to region s.

2.3. Resource Benefits

This paper introduces the water stress index (WSI) to analyze the impact of the virtual water flow behind commodity trade on regional water resources. The calculation formula for coastal provinces and cities is as follows:

$$WSI=\frac{WW}{Q^{*}}$$

(8)

$$Q^{*}=Q+SWU+SWD$$

(9)

where $WW$ is water withdrawal, $Q$ is blue water resources, $SWU$ is the amount of the direct utilization of seawater and $SWD$ is seawater desalination. The calculation formula for inland provinces and cities is as follows:

$$WSI=\frac{WW}{Q}$$

(10)

In addition, this paper classifies the water stress index (WSI) into four classes: no stress (WSI ≤ 0.2), moderate stress (0.2–0.4), severe stress (0.4–1.0) and extreme stress (WSI > 1.0).

Based on the evaluation of the regional water stress index, this paper further evaluates the conservation or loss of scarce water resources.

1. Water savings generated by exporting product i from region r to s without considering water stress can be expressed by the following equation:

$$W{S}^{rs}=\left(V_i^s-V_i^r\right)\times e^{rs}$$

(11)

2. Scarce water savings, considering water stress in each region, are calculated as follows:

$$SW{S}^{rs}=\left(WS{I}^s\times V_i^s-WS{I}^r\times V_i^r\right)\times e^{rs}$$

(12)

where $e^{rs}$ is the trade volume of product i from region r to s, $V_i^s$ and $V_i^r$ are the virtual water content (VWC) of product i in regions s and r, respectively, and VWC refers to the amount of water required to produce a unit of product i. $WS{I}^s$ and $WS{I}^r$ are the water stress index of regions s and r, respectively. A positive value of $SW{S}^{rs}$ indicates that the export of product i from region r to s saves scarce water, while a negative or zero value of $SW{S}^{rs}$ indicates losses or zero savings. $WS{I}^s\times V_i^s$ can be defined as the virtual scarce water content (VSWC), which represents the amount of virtual scarce water required to produce a unit of product i in region s.

2.4. Economic Benefits

Data Envelopment Analysis (DEA) is a commonly used method to measure the efficiency of water use, which is then used to assess the shadow price (SP) of water. The shadow price of water resources can be represented by the following equation:

$$SP=\theta \times \left(2E/C\right)$$

(13)

where $\theta$ is the water use efficiency of each province, obtained by the DEA model; $E$ is the GDP of each province; and $C$ is the total water use of each province. In addition, this paper classifies the shadow price into 3 classes: low shadow prices (SP < 100), medium shadow prices (100–300) and high shadow prices (SP > 300).

The net economic benefit ($EV{W}^{sr}$) generated by the virtual water flow from region s to r can be calculated by the following formula:

$$EV{W}^{sr}=\left(S{P}^r-S{P}^s\right)\times E^{sr}$$

(14)

where $S{P}^r$ and $S{P}^r$ are the shadow prices of water resources in regions r and s, respectively, and $E^{sr}$ is the net virtual flow from region s to r.

3. Theory

3.1. Theoretical Basis for the Improvement of the Water Stress Index in Coastal Provinces and Cities

The total amount of water resources in coastal areas is relatively high, but per capita possession is insufficient. With the continuous advancement of industrialization and urbanization, the government has to rely on large-scale groundwater extraction and the implementation of long-distance artificial water transfer projects to alleviate the shortage of freshwater resources. However, such an approach to utilizing terrestrial freshwater resources is inappropriate. Opening up seawater as a water source is an important way to alleviate the shortage of water resources in China’s coastal areas.

The “seawater resource” includes two aspects: desalination and direct utilization of seawater. The overall scale of the national seawater desalination project has been growing year by year, and this article selects seawater desalination data from the research year 2017. In addition, seawater’s direct utilization mainly relies on seawater in cooling applications; therefore, the seawater cooling water in each coastal province and city was extracted, as shown in

Table 1. List of water resources in coastal provinces and cities.

Provinces and Cities	Seawater Cooling Water (Billion tons/Year)	Seawater Desalination (Million tons/Day)	Blue Water (Billion m3)	Total Available Water (Billion m3)	Percentage of Additional Water Resources in Total Available Water (%)
Liaoning	92.94	8.77	186.3	279.6	33.4
Tianjin	12.09	31.72	13.0	26.2	50.5
Hebei	38.72	17.35	138.3	177.7	22.2
Shandong	83.08	28.26	225.6	309.7	27.2
Shanghai	31.62	/	34.0	65.6	48.2
Jiangsu	42.40	0.51	392.9	435.3	9.7
Zhejiang	306.84	22.78	895.3	1203.0	25.6
Fujian	225.19	1.12	1055.6	1280.8	17.6
Guangdong	418.37	8.13	1786.6	2205.3	19.0
Guangxi	54.20	/	2388.0	2442.2	2.2
Hainan	39.40	0.27	383.9	423.3	9.3

Note: billion is 10⁹, i.e., 1000 million.

.

The available water resources in China refer to the available blue water (surface water and groundwater), so green water is not considered in our water stress analysis. Based on previous analyses, this paper tries to propose an improved water stress index based on Pfister’s ratio of water withdrawal to available water resources [25], which incorporates seawater’s direct utilization and seawater desalination into the total available water resources of coastal provinces and cities ($Q^{*}$) so as to improve the water stress index of coastal provinces and cities. The amount of the direct utilization of seawater ($SWU$) and seawater desalination ($SWD$) in coastal provinces and cities are derived from the National Seawater Utilisation Report 2017 (Ministry of Natural Resources, 2018).

For their water saving calculation, most studies only considered the difference in water consumption between the export and import regions to produce the same product; they did not take into account the regional water shortage situation. The same water consumption has different effects on water stress in water-rich and water-poor areas [29]. Therefore, this paper quantifies the amount of water used for production in terms of water stress and further evaluates the saving or loss of scarce water resources in virtual water trade. It is mainly based on the following two points: 1. Water savings generated by exporting product i from region r to s without considering water stress; 2. Scarce water savings, considering the water stress in each region.

3.2. Theoretical Basis for Economic Benefits Analysis Based on Shadow Prices

In the field of water resources, the shadow price assessment method considers the ecological, social and economic value of water resources, including as water supply, irrigation, ecological protection, its landscape value, etc. Shadow price is a widely used index to evaluate the relative utilization efficiency of resource and opportunity costs, which can reveal the regional heterogeneity of resources in an economic system [39]. A larger shadow price indicates a higher marginal value for the resource, while a shadow price of zero indicates that the resource is in surplus and makes no marginal contribution to economic development. It is worth noting that the shadow price is a theoretical value that allows for relative comparisons of the differences in the value of water resources between areas, indicating that, under the same water depletion scenario, areas with a high shadow price would suffer greater economic losses than areas with a low shadow price.

Referring to study [32], this paper uses shadow price to internalize the value of water resources into their market price and further reveals the transfer of the economic value of water resources in the process of virtual water flows, aiming to provide a scientific basis for policymakers.

3.3. Relevant Concepts in the Benefit Analysis

Heterogeneity exists in productive water use under water stress and the potential value of water resources in different regions; virtual water flows generate gains and losses in both resource and economic benefit dimensions. We define the concepts of the terms involved in the benefit analysis as follows:

Terms:

Water Stress Index (WSI): the ratio of water withdrawal to available water resources, measuring the conflict between supply and demand for water resources.

Virtual Water Content (VWC): the amount of water consumed in the production of a unit of product, including direct and indirect water consumption.

Scarce Virtual Water Content (SVWC): the water used for production under water stress, obtained by multiplying VWC by WSI, meaning the amount of virtual scarce water required for the production of a unit of product.

Resource benefits: the scarce water saving or losses accompanied with virtual water flow.

Water saving: indirect water saving occurs when goodss flow from areas with a small VWC to areas with a large VWC.

Scarce water saving: scarce water saving is generated when virtual water flows from areas with a smaller SVWC to areas with a larger SVWC, and the amount of scarce water saving represents the extent of water stress alleviated.

Shadow Price (SP): the potential economic value of water resources.

Economic benefits: The flow of water resources between regions with different economic conditions will generate economic gains or losses. Economic gains occur when virtual water flows from areas with a low shadow price to areas with a high shadow price, and vice versa, resulting in economic losses.

4. Results

4.1. Virtual Water Flow Pattern

4.1.1. Overall Virtual Water Flow Patterns in Coastal Areas

Based on Equations (5) and (6), virtual water inflows and outflows in coastal provinces and cities were calculated (

Table 2. Virtual water flow momentum in coastal provinces and cities (unit: billion m³).

Virtual Water Flow Relationships	Virtual Water Inflow			Virtual Water Outflow			Net Outflows
	From Coastal Provinces and Cities	From Inland Areas	Total Inflows	Flows to Coastal Provinces and Cities	Flows to Inland Areas	Total Outflows
Liaoning	10.3	25.2	35.5	13.5	12.9	26.4	−9.1
Tianjin	10.4	12.1	22.5	4.4	6.4	10.8	−11.7
Hebei	10.1	23.3	33.4	14.1	21.3	35.4	2.0
Shandong	7.9	22.6	30.5	3.7	5.7	9.3	−21.2
Shanghai	7.5	14.5	21.9	10.6	12.0	22.6	0.7
Jiangsu	9.9	38.4	48.4	28.6	37.9	66.5	18.2
Zhejiang	28.3	57.1	85.4	10.5	27.7	38.2	−47.2
Fujian	7.9	25.6	33.5	5.0	12.9	18.0	−15.6
Guangdong	50.3	34.7	85.0	17.9	39.2	57.1	−27.9
Guangxi	7.8	13.4	21.1	34.2	11.9	46.1	25.0
Hainan	2.8	10.0	12.8	10.5	6.1	16.6	3.8
Coastal areas	153.0	276.9	430.0	153.0	194.0	347.0	−83.0

). In virtual water flows nationwide, the total virtual water outflow in coastal areas was 347.0 billion m³ and the total inflow was 430.0 billion m³, showing a net virtual water inflow state, with a net inflow of 83.0 billion m³. The virtual water inflow from coastal provinces and cities was almost the same as the outflow to coastal provinces and cities, so the net inflow of virtual water in coastal areas mainly came from inland areas; that is, coastal areas import a large number of water-intensive products in the process of their development.

In terms of the virtual water flow in each coastal province and city, Liaoning, Tianjin, Shandong, Zhejiang, Fujian and Guangdong were net inflow provinces, while the rest of the coastal provinces and cities had a net outflow of virtual water. Zhejiang had the largest net inflow, at 47.2 billion m³, followed by Guangdong, with a net inflow of 27.9 billion m³ of virtual water. There was a large gap between the net outflow of virtual water in each province, and the net virtual water outflow of Guangxi was 25.0 billion m³, which was about 38 times that of Shanghai.

4.1.2. Patterns of Virtual Water Flows among Coastal Provinces and Cities

The net flow involved in the flow of virtual water among coastal provinces and cities was 153.0 billion m³ (

Table 3. Virtual water flow momentum among coastal provinces and cities (unit: billion m³).

	To	Liaoning	Tianjin	Hebei	Shandong	Shanghai	Jiangsu	Zhejiang	Fujian	Guangdong	Guangxi	Hainan	Outflow
From
Liaoning		/	3.4	0.2	0.3	0.1	0.4	0.6	0.1	7.9	0.1	0.3	13.5
Tianjin		0.4	/	0.8	0.2	0.2	0.5	0.9	0.4	0.6	0.2	0.1	4.4
Hebei		1.3	2.4	/	2.9	1.1	1.2	2.7	0.5	1.3	0.5	0.2	14.1
Shandong		0.5	0.2	0.3	/	0.2	0.5	1.2	0.1	0.5	0.1	0.1	3.7
Shanghai		0.6	0.4	3.5	0.8	/	1.5	1.4	0.5	1.1	0.7	0.2	10.6
Jiangsu		2.5	1.8	1.6	1.3	1.8	/	12.7	2.2	3.1	1.3	0.4	28.6
Zhejiang		0.9	0.3	1.0	1.0	1.5	2.0	/	1.9	1.5	0.3	0.2	10.5
Fujian		0.7	0.1	0.3	0.1	1.5	0.3	1.2	/	0.5	0.3	0.1	5.0
Guangdong		2.6	1.1	1.5	0.8	0.8	2.0	5.6	1.3	/	1.5	0.8	17.9
Guangxi		0.7	0.3	0.9	0.4	0.3	1.4	1.9	0.6	27.3	/	0.3	34.2
Hainan		0.1	0.34	0.1	0.1	0.1	0.2	0.2	0.2	6.4	2.8	/	10.5
Inflow		10.3	10.4	10.1	7.9	7.5	9.9	28.3	7.9	50.3	7.8	2.8	153.0
Net inflow		−3.2	6.0	−4.1	4.2	−3.2	−18.7	17.8	2.9	32.4	−26.4	−7.7	/

). Tianjin, Zhejiang, Fujian, Shandong and Guangdong had greater virtual water inflows than their outflows to other coastal provinces, among which Guangdong and Zhejiang had the largest net virtual water inflows, acting as the virtual water consumers in coastal areas. In addition to the above five provinces, the rest of the coastal provinces and cities played the role of suppliers, among which Guangxi and Jiangsu were the two main sources of virtual water in coastal areas. The two largest virtual water flows between coastal provinces and cities were Guangxi–Guangdong and Jiangsu–Zhejiang, at 25.8 billion m³ and 10.7 billion m³, respectively.

4.2. Resource Benefits

4.2.1. Resource Benefits of Coastal Provinces and Cities under Virtual Water Flows Nationwide

We constructed a scarce-water-saving accounting framework that simultaneously takes into account water stress and the water used for production in each region and calculates the WSI of 31 provinces in China; the results are shown in

Table 4. Water stress index (WSI) by province in China.

Province	WSI	Province	WSI	Province	WSI	Province	WSI
Ningxia	6.120	Shanxi	0.575	Hubei	0.232	Guangxi	0.117
Shanghai	1.597	Henan	0.553	Shaanxi	0.207	Sichuan	0.109
Jiangsu	1.358	Xinjiang	0.542	Guangdong	0.197	Hainan	0.108
Beijing	1.326	Gansu	0.49	Hunan	0.171	Guizhou	0.098
Tianjin	1.048	Heilongjiang	0.486	Fujian	0.150	Yunnan	0.071
Hebei	1.022	Liaoning	0.469	Jiangxi	0.150	Qinghai	0.033
Shandong	0.676	Anhui	0.370	Zhejiang	0.149	Tibet	0.007
Inner Mongolia	0.607	Jilin	0.321	Chongqing	0.118	/	/

, which shows that 18 provinces had different degrees of water stress, mainly in coastal and northern regions, of which 6 provinces faced extreme water stress, and 4 of them were coastal provinces and cities (Shanghai, Jiangsu, Tianjin and Hebei).

The composition of the resource benefits of virtual water flows between coastal and inland areas is shown in Figure 2; this study found that virtual water flows nationwide saved 112.66 billion m³ of scarce virtual water, of which the virtual water flows among coastal provinces and cities yielded a total of 15.36 billion m³ of scarce water savings. Inflows into coastal provinces and cities (Figure 2a) saved 40.73 billion m³ of scarce virtual water, and virtual water flows from inland areas generated 25.37 billion m³ of scarce water savings, which were the main contributors to scarce water saving, of which 13.97 billion m³ of scarce virtual water was saved by inland–Jiangsu, which accounted for 55% of the water saving generated by outflows from inland areas. Outflows from coastal provinces and cities (Figure 2b) saved 87.29 billion m³ of scarce virtual water, and the scarce water saving was mainly generated by virtual water flows to inland areas, accounting for about 82% of savings, especially Guangdong–inland, which generated 33.53 billion m³ of scarce water savings.

Irrational virtual water flows resulted in 74.40 billion m³ of scarce water losses, of which 26.46 billion m³ of scarce water losses was caused by virtual water flows among coastal provinces and cities. Inflows into coastal provinces and cities (Figure 2c) resulted in 52.79 billion m³ of scarce water losses, of which virtual water flows originating from inland areas resulted in 26.32 billion m³ of scarce water losses, accounting for about half of the losses. In terms of the inflow place, the virtual water flow to Zhejiang led to the largest scarce water losses, with the two largest connections being Jiangsu–Zhejiang and inland–Zhejiang, at 11.88 billion m³ and 11.20 billion m³, respectively. Outflows from coastal provinces and cities (Figure 2d) resulted in 48.07 billion m³ of scarce water losses, of which 21.61 billion m³ of the scarce virtual water lost was lost to inland areas, especially from the outflow of virtual water from Jiangsu, which accounted for 61%, while the virtual water flow to Zhejiang also led to a huge loss of scarce virtual water. Overall, virtual water flows in coastal and inland regions produced a total of 38.26 billion m³ of net scarce water savings, suggesting that coastal areas benefited from virtual water flows nationwide, from a resource perspective.

Specifically for inland provinces and cities, the virtual water flow connections that virtual water flows through nationwide to generate scarce water savings were mostly inflows into high-water-stress provinces and cities such as Ningxia, Xinjiang and Jiangsu and outflows from low-water-stress provinces and cities such as Guangdong and Hebei (Figure 3a). Guangdong contributed the largest scarce water savings through outflow virtual water, totaling 35.17 billion m³. Among its outflows, the flows from Guangdong to Ningxia and Xinjiang were the two virtual water flow connections with the greatest scarce water savings, 16.28 billion m³ and 12.70 billion m³ of scarce water, respectively. The underdeveloped province Ningxia was the largest receiver (inflow) of scarce water savings, with a total of 43.50 billion m³ saved through the import of virtual water, and its significant water saving connections were the virtual water flows from Guangdong and Hebei.

Most of the virtual water flow connections that led to scarce water losses were outflows from provinces with high water stress, such as Jiangsu and Xinjiang, and inflows into provinces with low water stress, such as Zhejiang (Figure 3b). Particularly, the outflow from Jiangsu contributed the largest water losses (29.09 billion m³). Jiangsu provided Zhejiang with a large amount of virtual water, which supported its economic development. However, this connection caused the largest scarce water losses (11.88 billion m³). Xinjiang was the second largest contributor to water losses (11.16 billion m³). In terms of inflow places, the virtual water flow to Zhejiang led to the largest scarce water losses, mainly because Zhejiang had almost no water stress (WSI = 0.149), and the inflow of a large amount of net virtual water led to 25.77 billion m³ of scarce water losses.

4.2.2. Resource Benefits of Virtual Water Flows among Coastal Provinces and Cities

As shown in

Table 5. Scarce virtual water savings and losses among coastal provinces and cities (unit: billion m³).

	To	Liaoning	Tianjin	Hebei	Shandong	Shanghai	Jiangsu	Zhejiang	Fujian	Guangdong	Guangxi	Hainan	Total
From
Liaoning		/	2.66	0.10	−0.02	0.05	0.45	−0.22	−0.04	−1.98	−0.04	−0.09	0.86
Tianjin		−0.18	/	0.18	−0.10	0.01	0.51	−0.81	−0.32	−0.46	−0.17	−0.11	−1.45
Hebei		−0.71	0.33	/	−1.79	0.29	0.88	−2.44	−0.46	−1.09	−0.39	−0.21	−5.58
Shandong		0.04	0.10	0.30	/	1.18	1.38	−0.60	−0.06	−0.16	−0.04	−0.06	2.06
Shanghai		−0.39	−0.14	−1.35	−0.76	/	2.42	−1.85	−0.65	−1.39	−0.89	−0.25	−5.25
Jiangsu		−2.53	−1.50	−1.12	−1.37	−0.13	/	−16.28	−2.76	−3.88	−1.59	−0.53	−31.70
Zhejiang		0.38	0.27	0.90	0.19	4.03	4.40	/	0.00	0.07	0.02	0.01	10.29
Fujian		0.36	0.12	0.42	0.07	2.76	0.52	−0.01	/	0.02	0.04	0.00	4.30
Guangdong		0.98	1.40	1.57	0.35	1.42	3.84	−0.21	−0.07	/	0.07	0.03	9.39
Guangxi		0.17	0.15	0.44	0.07	0.25	1.45	−0.08	−0.02	1.15	/	−0.01	3.58
Hainan		0.04	0.21	0.06	0.02	0.13	0.32	0.00	0.00	1.32	0.30	/	2.41
Total		−1.84	3.61	1.50	−3.34	10.00	16.18	−22.51	−4.39	−6.40	−2.70	−1.22	−11.10

, the net value of each virtual water connection that generated net savings adds up to the total amount of scarce water savings. Virtual water flows among coastal provinces and cities generated a total of 15.36 billion m³ of scarce water savings. Since the WSI of Tianjin (1.048) was larger than that of Liaoning (0.469), the flow from Liaoning to Tianjin generated 2.66 billion m³ of scarce water savings, while the flow from Tianjin to Liaoning led to 0.18 billion m³ of scarce water losses; that is, the flow from Liaoning to Tianjin generated 2.48 billion m³ of scarce water savings, which was the largest scarce water saving connection. Shanghai–Jiangsu, Zhejiang–Shanghai and Fujian–Shanghai likewise generated significant scarce water savings.

Virtual water flows among coastal provinces and cities lost 26.46 billion m³ of scarce virtual water, and the outflows from Jiangsu contributed the largest losses (17.83 billion m³), accounting for 67% of the total scarce water losses. Among them, Jiangsu–Zhejiang was the largest scarce water losses connection, which may be due to the extreme water stress in Jiangsu (WSI = 1.358) and the large net virtual water flow from Jiangsu to Zhejiang (about 11.88 m³). Overall, virtual water flows among coastal provinces and cities resulted in 11.10 billion m³ of net scarce water losses, indicating that virtual water flows among coastal provinces and cities exacerbated the water shortage in coastal areas. The above situation is mainly due to the fact that the resource benefit is determined by the WSI and the amount of water used for production under water stress.

4.3. Economic Benefits

4.3.1. Economic Benefits for Coastal Provinces and Cities under Virtual Water Flows Nationwide

Due to the heterogeneity of economic development levels, there are large differences in the shadow price of water resources among different provinces in China. As shown in

Table 6. Shadow prices (SP) by province in China (unit: CNY/m³).

Province	SP	Province	SP	Province	SP	Province	SP
Tianjin	905	Zhejiang	157	Hunan	93	Qinghai	54
Beijing	544	Guangdong	141	Hainan	91	Jiangxi	52
Shandong	450	Shaanxi	134	Inner Mongolia	91	Jilin	45
Chongqing	287	Fujian	132	Liaoning	88	Gansu	42
Guizhou	263	Henan	126	Shanxi	83	Guangxi	39
Hebei	206	Sichuan	114	Yunnan	79	Heilongjiang	24
Shanghai	191	Jiangsu	97	Ningxia	60	Xinjiang	19
Hubei	183	Anhui	95	Tibet	56	/	/

, the results show that Tianjin, Beijing and Shandong were the top three provinces and cities with the highest shadow prices (905 CNY/m³, 544 CNY/m³ and 450 CNY/m³, respectively), while Guangxi (39 CNY/m³), Heilongjiang (24 CNY/m³) and Xinjiang (19 CNY/m³) had the lowest shadow prices.

The composition of the economic benefits of virtual water flows between coastal and inland areas is shown in Figure 4. Virtual water flows generated economic gains of CNY 39,961 billion, of which a total of CNY 11,186 billion was generated by virtual water flows among coastal provinces and cities. Inflows into coastal provinces and cities (Figure 4a) generated economic gains of CNY 34,085 billion, of which the virtual flow originating from inland areas generated CNY 22,898 billion, accounting for 67% of the total economic gains. Inland areas contributed notable economic gains, particularly in the virtual water flow to coastal provinces and cities with high shadow prices, such as Tianjin, Shandong and Zhejiang. The outflow from coastal provinces and cities (Figure 4b) generated economic gains of CNY 17,062 billion, of which the virtual flow to inland areas generated CNY 5876 billion, mainly because it originated from provinces and cities with low shadow prices, such as Hebei and Jiangsu.

Irrational flows resulted in CNY 8210 billion of economic losses, of which a total of CNY 1402 billion was lost in the virtual water flows among coastal provinces and cities. Inflows into coastal provinces and cities (Figure 4c) resulted in CNY 3540 billion of economic losses, of which virtual water flows originating from inland areas accounted for 60%. In terms of the inflow place, the virtual water flow to Guangxi resulted in large economic losses, mainly because Guangxi was the coastal province with the lowest shadow price. The outflow from coastal provinces and cities (Figure 4d) resulted in economic losses of CNY 6072 billion and the virtual flow to inland areas lost CNY 4670 billion. Overall, virtual water flows between coastal and inland areas generated a total net economic gain of CNY 31,751 billion, indicating that the flow of virtual water nationwide played a positive role in promoting economic development. This means that, under the same water depletion scenario, coastal areas are more likely to suffer huge economic losses than inland areas.

Specifically for inland provinces and cities, the virtual water flow connections that generated economic gains were mostly inflows into high-shadow-price provinces and cities such as Tianjin and Shandong and outflows from low-shadow-price provinces and cities such as Xinjiang (Figure 5a). Tianjin and Shandong generated CNY 1381 billion and CNY 7070 billion of economic gains by importing virtual water, respectively. Among them, the virtual water flow connection with the largest economic gain was Inner Mongolia–Tianjin, which generated a potential economic value of nearly CNY 3294 billion. Underdeveloped provinces and cities (Xinjiang, Inner Mongolia and Guangxi) had lower shadow prices, and their large virtual water flows to high-shadow-price provinces and cities created significant economic gains. Xinjiang was the largest contributor to economic gains (CNY 5036 billion), mainly because Xinjiang was the largest net virtual water outflow province and had the lowest shadow price.

The irrational outflow from high shadow price provinces and cities (Tianjin and Hebei) and the irrational inflow into low shadow price provinces and cities (Shanxi and Guangxi) have caused economic losses (Figure 5b). A large amount of virtual water flowed from provinces and cities with high shadow prices (Tianjin, Chongqing, and Hebei) to provinces and cities with low shadow prices (Shanxi, Guangxi), causing huge economic losses. Specifically, due to Tianjin’s high shadow price, its export of virtual water led to the largest economic losses (CNY 1751 billion). In addition, Chongqing–Guangxi and Hebei–Shanxi were the two largest economic loss connections, loosing CNY 763 billion and CNY 521 billion, respectively.

4.3.2. Economic Benefits of Virtual Water Flows among Coastal Provinces and Cities

As shown in

Table 7. Economic gains and losses from virtual water flows among coastal provinces and cities (unit: billion CNY).

	To	Liaoning	Tianjin	Hebei	Shandong	Shanghai	Jiangsu	Zhejiang	Fujian	Guangdong	Guangxi	Hainan	Total
From
Liaoning		/	2783	24	111	10	4	42	6	420	−7	1	3394
Tianjin		−356	/	−569	−107	−157	−422	−673	−283	−434	−176	−96	−3273
Hebei		−157	1689	/	711	−16	−129	−132	−39	−87	−76	−26	1738
Shandong		−181	71	−69	/	−43	−175	−355	−43	−148	−49	−39	−1031
Shanghai		−65	264	54	206	/	−138	−47	−29	−54	−102	−19	70
Jiangsu		−23	1418	173	443	166	/	759	76	137	−77	−3	3069
Zhejiang		−60	249	48	280	50	−117	/	−49	−25	−40	−10	326
Fujian		−29	112	21	39	89	−9	29	/	4	−23	−5	228
Guangdong		−135	854	97	249	38	−88	90	−12	/	−151	−41	901
Guangxi		35	269	143	169	46	83	221	59	2798	/	18	3841
Hainan		0	288	9	28	6	1	15	8	318	−146	/	527
Total		−971	7997	−69	2129	189	−990	−51	−306	2929	−847	−220	9784

, the net value of each virtual water connection that generated net gains adds up to the total economic gains. The virtual water flows among coastal provinces and cities generated CNY 11,186 billion of economic gains, with Tianjin and Guangdong being the largest receivers of these economic gains, accounting for 48% and 29% of the total economic gains, respectively. The two virtual water connections with the largest economic gains were Guangxi–Guangdong and Liaoning–Tianjin, generating net economic gains of CNY 2647 billion and CNY 2427 billion, respectively. The virtual water flow from coastal provinces and cities with low shadow prices (Guangxi, Liaoning and Jiangsu) contributed significantly to the economic gains, accounting for 71% of the total economic gains.

Virtual water flows from coastal provinces and cities with a high shadow price to ones with a low shadow price caused economic losses of CNY 1402 billion. Tianjin–Zhejiang was the largest economic loss connection, mainly due to the fact that the net outflow of virtual water from Tianjin to Zhejiang, with the large SP gap between the two provinces of 748 CNY/m³, caused an economic loss of CNY 673 billion, while the flow from Zhejiang to Tianjin generated an economic gain of CNY 249 billion; that is, Tianjin to Zhejiang caused a net economic loss of CNY 424 billion. In addition, the virtual water flow of Tianjin–Fujian and Hebei–Liaoning also caused significant losses. Virtual water flows to Zhejiang and Liaoning led to the greatest economic losses, accounting for 64% of losses.

Overall, the virtual water flows among coastal provinces and cities generated CNY 9784 billion of net economic gains, meaning that the virtual water flow among coastal provinces and cities promoted their overall economic development. But due to the large spatial scope of China’s coastal areas, there are great differences in the resource endowment and economic development level of each coastal province and city; economies are affected differently by virtual water flows.

5. Discussion

5.1. Applicability of the Benefit Analysis Methodology

Due to the unique geographical location of coastal provinces and cities, their total amount of available water resources is relatively more than that of inland provinces and cities, which has been neglected in previous studies. In order to fill this gap, we incorporate the amount of the direct utilization of seawater and seawater desalination into the total amount of water available in coastal provinces and cities, aiming to more accurately evaluate the resource benefits of virtual water flows. In addition, virtual water flows also have an economic impact, and the depth of previous studies, from this perspective, is relatively shallow, so this paper quantifies the potential economic value of water resources in various provinces and cities based on shadow prices, and further evaluates the contribution of virtual water flows to the economy.

Water-rich areas transfer water stress to water-poor areas by importing virtual water [30]. Under this virtual water flow pattern, China experienced a net loss of 80 billion m³ of scarce water due to inter-provincial virtual water trades [29]. Our study found that virtual water flows between coastal provinces and cities and inland areas generated a total of 38.26 billion m³ of net scarce water savings, and the virtual water flow among coastal provinces and cities resulted in 11.10 billion m³ of net scarce water losses, indicating that the virtual water flows among coastal provinces and cities further exacerbated the overall water stress. From the perspective of resource benefits, the virtual water flows among coastal provinces and cities are developing in an unsustainable direction. In addition, the existing virtual water flow pattern in China’s coastal areas generated huge economic benefits. The virtual water flows between coastal provinces and inland areas generated a total of CNY 31,751 billion of net economic gains, and virtual water flows among coastal provinces and cities generated a total of CNY 9784 billion of net economic gains. From the perspective of economic benefits, the virtual water trade pattern in coastal areas is generally moves from underdeveloped to developed places, and this flow pattern increases the marginal economic value of water resources. This is also consistent with the findings of Han et al. [32].

On the whole, this virtual water flow pattern can alleviate the water stress of the inflow places to a certain extent, save their local water resources, and develop industries with higher added value and economic benefits; but for an outflow place with a water shortage and underdeveloped economy, the continuous outflow of virtual water will not only exacerbate the depletion of water resources in the outflow place but also widen the economic gap between the outflow place and the inflow place, which is not conducive to regional coordinated development and is unsustainable.

Specific to coastal provinces and cities, the current trade pattern increased the potential economic value of water resources but was not sustainable from a resource perspective. Our study shows that Zhejiang imported a large amount of virtual water from Jiangsu, and the water stress in Jiangsu was higher than that of Zhejiang, while its shadow price was lower than that of Zhejiang; that is, the virtual water flowed to areas with low water stress and a higher marginal value to their water resources, which improved the adaptability of Zhejiang’s economy to water depletion but exacerbated the water stress in Jiangsu. Therefore, based on the calculation of the virtual water flow in coastal areas, the research results of the benefits evaluation in this paper can provide an important reference for coastal provinces and cities to formulate policies for water stress alleviation and water trade relationship optimization.

5.2. Suggestions for China’s Coastal Areas

Optimizing virtual water trade relationships is likely to have a wider hydro-economic impact than direct water-saving measures. In summary, in order to promote the coordinated and sustainable development of coastal areas and enhance their resilience to the threat of water resource shortage, this article proposes the following suggestions for formulating policies and measures:

(1) Effectively allocate and utilize unconventional water sources.

Coastal areas should make full use of the ocean as a huge water source and expand the scope and scale of their unconventional water source utilization, promoting seawater as cooling water in key industries such as coastal thermal power and nuclear power. Desalinated seawater can be used as a water source for living and production in coastal water-poor areas. In addition, rainwater collection and utilization for agricultural irrigation and domestic use should be strengthened, especially by taking advantage of the high rainfall in southern coastal areas.

(2) Promote water-saving technologies and other measures.

First of all, the advancement of water-saving technology is the key to improving water use efficiency, such as vigorously developing water reuse technology and building a sewage resource utilization system. Secondly, a reasonable pricing mechanism for water resources should be established to maximize economic gains and provide practical incentives to encourage water-saving behavior. Finally, coastal areas should utilize abundant marine resources, such as developing aquaculture (seaweed, shellfish and fish farming) to reduce stress on local water resources and increase the export value of their marine products and the scale of their virtual water trade to increase their economic benefits.

(3) Scientifically formulate water resource compensation policies.

Political means are the key to reshaping virtual water trade relationships and addressing water shortages. We suggest that the government take the lead in formulating and implementing reasonable virtual water resource compensation policies (similar to ecological compensation policies). For example, provinces and cities with high shadow prices, such as Tianjin and Guangdong, have generated significant economic gains due to the net inflow of virtual water, and the government can subsidize other provinces and cities by imposing virtual water trade fees on them.

(4) Balancing economic beneficiaries and virtual water contributors.

Economic beneficiaries and contributors to the virtual water trade should develop a set of cooperation models to enhance the sustainability of this trade. For example, in the virtual water flow between Jiangsu and Zhejiang, Zhejiang should be encouraged to provide financial assistance and technology sharing to improve Jiangsu’s water use efficiency. The sustainability of Jiangsu’s water resources determines whether Zhejiang can continue to import sufficient virtual water to support its economic development. The calculation of the potential economic benefits of each trade flow in this study can be used as a reference for the amount of financial assistance.

5.3. Future Research Directions

This paper focuses on virtual water flows in China’s coastal areas, incorporates direct seawater use and seawater desalination into the total available water resources and improves the water stress index of coastal provinces and cities, making it closer to the actual water stress situation. In addition, this paper considers the economic perspective of virtual water flows and analyzes the economic benefits of virtual water flows through shadow prices, and this evaluation method considers the hidden cost of water resources and reveals the potential impact of water consumption on economic development. However, in terms of data processing, since there are no statistics on the water consumption of specific sectors in China, this paper refers to the data processing experience of existing studies and estimates the water consumption of the industrial and service sub-sectors with the help of the existing total water consumption, and these estimated results are different from the actual water use. In addition, except for resource and economic benefits, virtual water flows also have an impact on the social dimension. Future research could incorporate social benefits into the analysis and explore the impact of virtual water flows on social development in coastal areas, including employment, income distribution, quality of life and other factors. This could provide decision support for the development of appropriate social policies and equitable development.

6. Conclusions

Coastal areas had a net inflow of 83.0 billion m³ from the overall virtual water flow, which mainly came from external water supplies. The total virtual water net flow among coastal provinces and cities was 153.0 billion m³, and Guangdong and Guangxi were the provinces with the largest net virtual water inflow and outflow, respectively.

In terms of the resource benefits of virtual water flows, a total of 38.26 billion m³ of net scarce water savings were generated by the virtual water flows between coastal and inland areas, indicating that virtual water flows played an important role in alleviating water stress in coastal areas. The virtual water flows among coastal provinces and cities resulted in net scarce water losses of 11.10 billion m³, indicating that virtual water flows among coastal provinces and cities further exacerbated the overall water stress; coastal areas need to strengthen their local water resource management while making reasonable use of external virtual water.

In terms of the economic benefits, virtual water flows between coastal and inland areas generated net economic gains of CNY 31,751 billion; this means that under the same scenario of water resource depletion, coastal areas were more likely to suffer huge economic losses than inland areas. The virtual water flows among coastal provinces and cities generated net economic gains of CNY 9784 billion, indicating that virtual water flows have boosted the economic development of coastal areas in the above two areas.