Study on an Evaluation Model for Regional Water Resource Stress Based on Water Scarcity Footprint

Abstract

Under the multiple pressures of intensifying global climate change disruption and rapid economic growth, China has become one of the countries facing the most serious water scarcity problems. Based on the ISO 14046 standard and the framework of water scarcity footprint theory, this study will break through the static limitations and lack of dimensions of traditional characteristic factors (i.e., water stress) and construct a water stress evaluation index system that combines nature, economy, and society. The results indicate that in recent years, regional water stress in China has exhibited significant spatiotemporal variations and spatial clustering, primarily driven by composite factors, with an overall decreasing trend. Among them, Shanghai is the highest-pressure area and Shaanxi is the lowest-pressure area, which is mainly due to the spatial projection of the coupling effect of multi-dimensional factors. In addition, the obstacle degree analysis method shows that indicators such as the utilization rate of water resource development constitute cross-regional constraints. To this end, all regions should make efforts to regulate and control the water use structure, introduce water-saving technologies, and strengthen water-saving publicity according to their needs. Therefore, this study not only provides a scientific basis for in-depth understanding of the distribution law and influencing mechanism of water stress but also provides an important reference for the rational allocation and sustainable use of water resources by upgrading the characteristic factors to system control signals.

1. Introduction

China, as the world’s sixth-largest water resource country, boasts a total volume of approximately 3.1 trillion cubic meters. However, its per capita water availability reaches only 35% of the global average. Nearly two-thirds of cities nationwide face varying degrees of water scarcity, exhibiting a distinct spatiotemporal heterogeneity characterized by “abundance in the south versus scarcity in the north, and plentiful supply in summer versus shortages in winter.” Spatially, the Yangtze and Pearl River basins and areas south of them account for over 80% of the country’s total water resources, while the Yellow River basin and inland northwestern regions consistently face water scarcity challenges. Temporally, precipitation is primarily concentrated during June to September, triggering the simultaneous occurrence of seasonal floods and droughts. This spatial imbalance also leads to a severe mismatch with the distribution of population and economic activities. For instance, in key economic zones such as the Beijing–Tianjin–Hebei region and the Huang–Huai–Hai Plain, the utilization rate of water resources has generally exceeded 100%, resulting in groundwater over-extraction across an area of 267,600 square kilometers. This has triggered ecological degradation and deterioration of water quality [1,2]. Under the dual stress of increasing extreme precipitation/drought events and the demands of high-quality development, water scarcity has become a rigid constraint on regional sustainable development. Therefore, it is essential to create a scientific and rational assessment method for water scarcity to solve the problem of water scarcity. This approach enables accurate identification of water scarcity levels across diverse regions, thereby providing a robust basis for rational water allocation, efficient resource utilization, and the development of targeted water conservation policies. This, in turn, can effectively mitigate the conflict between water scarcity and regional development, promoting sustainable socioeconomic progress [3,4].

The water footprint theory, pioneered by Hoekstra, has been extensively applied in water resources management. This framework enables comprehensive assessment of regional water resources by quantifying both water consumption during human production/consumption processes and water requirements for pollutant dilution and degradation [5,6,7,8,9,10,11]. Currently, numerous countries and organizations worldwide have established water footprint assessment tools or standards, with prominent examples including the Water Footprint Network (WFN) and the International Organization for Standardization (ISO) [12,13]. The Water Footprint Network (WFN) has developed a series of standardized methodologies, guidelines, and specifications encompassing water footprint accounting, impact assessment, and the mitigation/offsetting of negative impacts. The International Organization for Standardization (ISO) has established the ISO 14046 international standard for water footprint assessment [14]. However, these two frameworks differ in their definitions of water footprint: WFN primarily focuses on volumetric assessment, whereas ISO emphasizes impact assessment. It should be noted that both methods have a solid scientific basis and are suitable for different research and use scenarios. Among them, the volumetric approach provides an important basis for the total accounting of water consumption, while the impact-oriented approach further evaluates the impact of water consumption on the environment by introducing impact factors. Since this study focuses on regional water stress effects, an impact-oriented approach is more in line with the research objectives. Specifically, the ISO 14046 standard further specifies that water footprint impact assessment primarily encompasses three key components: (1) water availability footprint, (2) water scarcity footprint, and (3) water degradation footprint [13]. The purpose of water availability footprint assessment is to determine the potential environmental impacts of products, processes, or organizations related to water availability. Its evaluation scope can also cover freshwater, i.e., other types of water resources (e.g., groundwater, rainwater, etc.). If the evaluation only focuses on water quantity, it should be called a water scarcity footprint, and a characterization model should be constructed based on the regional differences in water scarcity. The characterization factors in the model should be used to calculate the water scarcity footprint. If the water scarcity footprint evaluates only one type of water resource, a qualifier should be used to specify the specific type (e.g., freshwater scarcity footprint). Therefore, a broad water availability footprint is an overarching concept that includes a water scarcity footprint focusing more on the water quantity dimension. It is important to note that water degradation footprints belong to another separate type of evaluation that aims to assess the potential environmental impacts of products, processes, or organizations related to water quality, such as eutrophication, ecotoxicity, etc. In the current phase, this study focuses on the quantitative assessment of the water scarcity footprint, aiming to systematically analyze the impact of water consumption on regional water resource pressure and provide a foundational evaluation of the magnitude and spatial distribution of water scarcity.

The quantification model for water scarcity footprint fundamentally transforms physical water consumption into scarcity impact values through characterization factors. These factors represent regional scarcity weights per unit water consumption, typically derived from Google Earth Engine datasets or demand-to-supply ratio methodologies [15,16]. However, the aforementioned methodologies exhibit certain limitations to varying degrees. For instance, although Google Earth Engine data can provide rich geographical information, the real-time nature and accuracy of its data may be limited by the resolution and update frequency of satellite images, and it is difficult to accurately reflect the dynamic changes in water resources in local areas. Although the supply and demand ratio method can reflect the supply and demand situation of water resources, it only starts from the perspective of the ratio of water consumption and the available water resources, and it relies on natural factors in the transition, ignoring the coupling between economic water consumption intensity and social agglomeration effect, resulting in insufficient comprehensiveness and accuracy in evaluating regional water resource scarcity [17,18]. Notably, technological advancements during socioeconomic development may enhance water use efficiency, thereby alleviating water scarcity. Similarly, sociocultural factors could significantly influence water allocation and utilization patterns. Consequently, these multidimensional drivers should be comprehensively incorporated into water stress indices.

In order to break through the above limitations, this study intends to build a multi-dimensional water resource pressure evaluation model based on the “natural–economic–social” factor and use the obtained regional water resource pressure index as a characteristic factor of the water scarcity footprint quantitative model. This is done to improve the construction of the water scarcity footprint quantitative model and provide a theoretical basis and decision support for the sustainable use of water resources.

Table 1. Model comparison.

Model	WAVE	AWARE	ReCiPe	Multi-Dimensional Comprehensive Pressure Index of Water Resources
Source	Berger et al. [19]	Boulay et al. [20,21]	Ridoutt et al. [22]	This work
Considerations	The amount of water consumed due to evaporation rate is taken into account	Available water remaining	Consolidate consumable and degradable water into a single indicator	Physical scarcity + socio-economic adaptation
Spatial resolution	High (Basin/Country)	High (Basin/Country)	Medium (Country)	Medium (Country/Province)
Indicator dimensions	Single physical dimension	Single physical dimension	Two dimensions	Multiple dimensions
Advantages	High standardization, helpful interpretation of virtual water research	Scientific rigor, high resolution, clear physical significance	Covering the impact of water quantity and water quality in a single indicator	Comprehensive evaluation, identifies socio-economic vulnerability, supports differentiated policies
Disadvantages	The implementation is complex	The concept is slightly more complex and mainly reflects physical scarcity	Complex model, limited accuracy	High data demand, not yet standardized

systematically compares the existing mainstream models with the models proposed by this research institute. Specifically, we will conduct objective empowerment based on the entropy weight method, combine cluster analysis to identify spatial distinction types, systematically evaluate the space–time pattern of water resource pressure in 31 provinces in China in recent years, and explore the main influencing factors affecting water resource pressure based on obstacle analysis methods. The research framework is illustrated in Figure 1. This study aims to address three core research aims: (1) Analyzing the spatiotemporal heterogeneity and spatial clustering characteristics of water stress; (2) Deciphering the interactive feedback mechanisms among the three subsystems of natural, economic, and social factors; (3) Designing region-specific regulatory pathways to mitigate water stress.

2. Research Methods and Data Sources

2.1. Construction of Evaluation Index System

We selected a water resource stress evaluation index system (

Table 2. Water resource stress assessment indicator system.

Influencing Factor	Specific Indicator	Unit	Calculation Method	Significance	Attribute
Natural Factors	Total water resources per unit area (X1)	104 m3/km2	Total water resources/Regional area	Measures renewable freshwater availability	-
	Irrigation water use per cultivated mu (X2)	m3	Irrigation water use/Actually irrigated area	Reflects water-saving technology effectiveness	+
	Water resources utilization rate (X3)	%	(Total water supply/Total water resources) × 100%	Indicates water development intensity	+
Economic Factors	Water use per 104 yuan GDP (X4)	m3/104 yuan	Total water use/GDP	Measures water use efficiency	+
	Tertiary industry proportion (X5)	%	(Tertiary industry GDP/Total GDP) × 100%	Reflects economic structure optimization	-
	Industrial water use per 104 yuan value-added (X6)	m3/104 yuan	Industrial water use/Industrial value-added	Indicates industrial water efficiency	+
Social Factors	Urbanization rate (X7)	%	(Urban population/Total population) × 100%	Measures urban development level	+
	Population density (X8)	persons/km2	Total population/Total area	Reflects population stress	+
	Per capita domestic water use (X9)	m3/day	Residential water use/Population	Indicates household water consumption intensity	+

Note: “Mu” is a unit of land area in China, with 1 mu = 666.67 m².

) based on three dimensions: natural factors, economic factors, and social factors. We conducted separate analysis and assessment for each dimension, reflecting the hierarchical nature of water resource stress analysis. Further, specific indicators were selected based on the principle of comprehensiveness and hierarchically categorized to enhance the pertinence of the regional water resource stress evaluation system in China. The indicator selection in this study was based on a synthesis of existing literature, with systematic consolidation of relevant indicators to enhance the scientific rigor and representativeness of the research [23,24,25]. Remarkably, in determining the indicators of natural factors, we mainly selected variables that are driven by physical geography and climatic conditions or those that directly reflect the innate endowment and constraints of regional water resources. For example, X₁ is the most direct natural endowment indicator, and its high value directly determines the regional water supply potential. X₂ is strongly dependent on climatic conditions (evaporation, precipitation) and can largely reflect the degree of drought in different regions. X₃ represents the degree of exploitation of the total water resources in the region, and if the X₃ of a region is high, it is largely due to the insufficient background of its natural water resources to meet the basic survival and development needs. This indicator is also a passive response and direct reflection of human activities to fragile natural conditions. Similarly, when determining the indicators of economic factors, we focus on selecting parameters that can reflect the operational efficiency, technical level, and industrial structure characteristics of the regional economic system. For example, X₄ and X₆ are the core economic indicators, which directly measure the efficiency relationship between economic output and water consumption. A corresponding increase means that the economic value of this unit of water consumption increases and the pressure on water resources increases. X₅ is a key economic and social structure indicator, which macroscopically determines the upper limit of the overall economic water efficiency. This is because the service-based economic structure itself has the inherent attribute of low water resource dependence. As for social factors, we mainly selected a series of indicators that reflect the characteristics of population distribution, social spatial structure, and the pressure of residents’ living activities on the water resources system. Among them, X₇ and X₈ are the core indicators driving water demand and social pressure, which determine the distribution pattern and concentration of water use activities in space and time. X₉ is a concrete embodiment of social structural pressure on the consumption side of residents, and the continuous growth of domestic water demand will further exacerbate the pressure on water resources. Additionally, the selection of indicators was rigorously constrained by data availability across different regions to ensure practical applicability. In addition, principal component analysis (PCA, Python 3.12.10 code in Supplementary Materials) was further used to verify the rationality of the index, and the results showed that the cumulative variance contribution rate of the first two principal components (PC1 + PC2) in all dimensions significantly exceeded the general standard of 70% (natural factors 86.88%, economic factors 80.19%, and social factors 92.29%). This indicates that the variables within each dimension have a high degree of intrinsic correlation and are suitable for constructing comprehensive indicators.

2.2. Water Stress Evaluation Model

2.2.1. Entropy Weight Method

The entropy weight method is an objective weighting approach that determines indicator weights based on the dispersion degree of dataset values, effectively eliminating subjective human influences in the weighting process [26,27,28]. Rooted in information theory, entropy serves as a comprehensive measure of system disorder. A smaller entropy value indicates greater variability in the indicator’s data range, implying that the indicator exerts a stronger influence in comprehensive evaluations and should therefore be assigned a higher weight. The specific calculation process is detailed as follows:

(1)

Data Standardization

The positive and negative indicators are standardized according to the following formula, and the values of each index are compressed into the interval of [0, 1] to eliminate the dimensional differences of each index:

$$\mathrm{positive}\mathrm{indicators}:\hspace{1em}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}^1={\displaystyle \frac{{\mathrm{X}}_{\mathrm{i}\mathrm{j}}-\mathrm{m}\mathrm{i}\mathrm{n}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}}{\mathrm{m}\mathrm{a}\mathrm{x}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}-\mathrm{m}\mathrm{i}\mathrm{n}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}}}$$

$$\mathrm{negative}\mathrm{indicators}:\hspace{1em}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}^1={\displaystyle \frac{\mathrm{m}\mathrm{a}\mathrm{x}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}{-\mathrm{X}}_{\mathrm{i}\mathrm{j}}}{\mathrm{m}\mathrm{a}\mathrm{x}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}-\mathrm{m}\mathrm{i}\mathrm{n}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}}}$$

where ${\mathrm{X}}_{\mathrm{i}\mathrm{j}}$ is the standard value of the sample data, and $\mathrm{m}\mathrm{i}\mathrm{n}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}$ and $\mathrm{m}\mathrm{a}\mathrm{x}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}$ are the minimum and maximum values of the sample data, respectively.

(2)

Calculation of Numerical Proportion

The proportional value (${\mathrm{P}}_{\mathrm{i}\mathrm{j}}$) of the i-th item under the j-th indicator is calculated as follows:

$${\mathrm{P}}_{\mathrm{i}\mathrm{j}}={\displaystyle \frac{{\mathrm{X}}_{\mathrm{i}\mathrm{j}}^1}{{\sum }_{\mathrm{i}=1}^{\mathrm{m}}{\mathrm{X}}_{\mathrm{i}\mathrm{j}}^1}}$$

(3)

Entropy Value Calculation

We calculate the entropy of the k-value, the j-th indicator, as follows:

$$\mathrm{K}={\displaystyle \frac{1}{\mathrm{l}\mathrm{n}\left(\mathrm{m}\right)}}$$

$${\mathrm{e}}_{\mathrm{j}}=-\mathrm{K}\times {\sum }_{\mathrm{i}=1}^{\mathrm{m}}{\mathrm{P}}_{\mathrm{i}\mathrm{j}}\times \mathrm{l}\mathrm{n}\left({\mathrm{P}}_{\mathrm{i}\mathrm{j}}\right);\mathrm{k}>0;\hspace{1em}{\mathrm{e}}_{\mathrm{j}}\ge 0$$

(4)

Coefficient of variation calculation

The coefficient of variation for the j-th indicator is calculated as follows:

$${\mathrm{d}}_{\mathrm{j}}=1-{\mathrm{e}}_{\mathrm{j}}$$

(5)

Entropy weight calculation

The weight of the j-th indicator is as follows:

$${\mathrm{w}}_{\mathrm{j}}={\displaystyle \frac{{\mathrm{d}}_{\mathrm{j}}}{{\sum }_{\mathrm{j}=1}^{\mathrm{m}}{\mathrm{d}}_{\mathrm{j}}}}$$

2.2.2. Integrated Water Resource Stress Index

According to the standardization and weighting of the indicators, the comprehensive water stress index of the three major factors of nature, economy, and society can be calculated. The specific formula is as follows:

$${\mathrm{Z}}_{\mathrm{k}}={\sum }_{\mathrm{j}=1}^{\mathrm{n}}{\mathrm{w}}_{\mathrm{j}}\times {\mathrm{X}}_{\mathrm{i}\mathrm{j}}^1$$

$$\mathrm{W}\mathrm{S}\mathrm{I}=\mathsf{\alpha }{\mathrm{Z}}_1+{\mathsf{\beta }\mathrm{Z}}_2+\mathsf{\gamma }{\mathrm{Z}}_3$$

where Z represents the three major factors, and k = 1, 2, and 3 represent the natural, economic, and social factors, respectively. n is the number of evaluation indicators in each factor. ${\mathrm{Z}}_{\mathrm{k}}$ is the comprehensive evaluation index of the sub-factors. Among them, ${\mathrm{w}}_{\mathrm{j}}$ and ${\mathrm{X}}_{\mathrm{i}\mathrm{j}}^1$ could be obtained via the calculation in Section 2.2.1. Moreover, WSI is a comprehensive evaluation index of composite factors. ${\mathrm{w}}_{\mathrm{j}}$ indicates the weight of the indicator. $\mathsf{\alpha },\mathsf{\beta }$, and $\mathsf{\gamma }$ are the relative importance of natural, economic, and social factors, respectively. This paper considers the three sub-factors to be equally important, so $\mathsf{\alpha }=\mathsf{\beta }=\mathsf{\gamma }=1/3$. This assignment is mainly based on the following considerations. First of all, there is a lack of unified and quantifiable theoretical support for the relative importance of these three factors in the academic community. If other weighting schemes are assigned, it will inevitably bring greater subjective judgment, and the use of the equal weight method could minimize subjective preferences. Secondly, the water resource system is a complex system in which these three elements affect each other, and it is difficult to judge the absolute dominance of a certain element in different scenarios. Therefore, it is prudent and reasonable to consider them as having equal potential importance in the absence of targeted assumptions. Meanwhile, this equal-weight treatment method has a precedent in the construction of composite indicators of resources and the environment [29,30,31].

In conclusion, the core purpose of this study is to integrate multi-dimensional indicators into a comprehensive index to achieve overall evaluation and comparison of complex systems. It is worth noting that our weight determination process is not based on subjective value judgments but chooses a hybrid model of the entropy weight method (within subsystems) and equal weight method (between subsystems). First of all, the entropy weight method is a data-driven objective weighting method that is derived from the variation of the index data itself, which effectively avoids human subjectivity and is very suitable for the comparative ranking goal of this study. Secondly, the use of the equal-weight method at a higher level is due to the lack of scientific consensus to distinguish between economic, social, and environmental absolute importance, which is a transparent and conservative strategy that minimizes subjective value judgments and ensures the robustness and interpretability of results. Similarly, Na et al. used the objectively empowered entropy weight method and the subjectively empowered chromatography method to solve the problem of insufficient evaluation indicators and methods for global low-carbon campuses [32]. Zhao et al. used entropy weight and analytic hierarchy process to weigh the systematic index of the public health practice teaching evaluation system to bridge the gap between traditional qualitative assessment and objective, data-driven methods [31,33].

2.2.3. Obstacle Analysis

The obstacle degree model can quantify the impact of each index on regional water stress in China by analyzing the contribution of each index [31,34,35] It is calculated as follows:

$${\mathrm{O}}_{\mathrm{i}\mathrm{j}}={\displaystyle \frac{{\mathrm{w}}_{\mathrm{j}}(1-{\mathrm{X}}_{\mathrm{i}\mathrm{j}}^1)}{{\sum }_{\mathrm{j}=1}^{\mathrm{n}}(1-{\mathrm{X}}_{\mathrm{i}\mathrm{j}}^1)}}$$

where ${\mathrm{O}}_{\mathrm{i}\mathrm{j}}$ is the obstacle degree of indicator j in region i; ${\mathrm{w}}_{\mathrm{j}}$ indicates the weight of the j-th indicator; ${\mathrm{X}}_{\mathrm{i}\mathrm{j}}^1$ is the normalized value of indicator j in region i.

2.2.4. Data Source

The research subjects of this study were 31 provinces in China (except Hong Kong, Macao, and Taiwan). The data sources used were the China Statistical Yearbook, Provincial and Municipal Statistical Yearbook, China Water Resources Bulletin, Provincial and Municipal Water Resources Bulletin, etc. Due to the limitation of the completeness of the data series, the data series used were 2019–2023.

2.2.5. Model Validation

To clarify the sensitivity of the entropy weight method to data distribution, we take the Z_k, dominated by natural factors in Beijing in 2019, as an example. In this case, the initial weights of indicators X₁, X₂, and X₃ are 0.12, 0.16, and 0.72, respectively, and the initial Z_k is 0.2674. We first replace the normalized initial data of high-weight X₃ with the national average, typical value (national maximum), and outlier value (national minimum), and Z_k becomes 0.2671, 0.2634, and 0.2638 respectively. Similarly, we replace the initial data after the normalization of low-weight X₁ with the national average, typical value (national maximum), and outlier value (national minimum), and Z_k becomes 0.2672, 0.2680, and 0.2664. The results show that the absolute change in high-weight indicators is significantly greater than that of low-weight indicators, indicating that the numerical changes in high-weight indicators will have a more significant impact on the comprehensive evaluation result Z_k, exposing its robustness limitations.

In addition, PCA is used to further explore the rationality of weight allocation. The PCA results of the three subsystems showed that the variance contribution rate of the first principal component (PC1) of the natural subsystem was 58.80%, of which the X1 and X₂ loads were higher and the signs were opposite (0.79 and −0.58), showing significant synergistic change characteristics. The contribution rate of PC1 of the economic subsystem was 45.81%, and both X4 and X6 showed high positive loads (0.62 and 0.65), reflecting the core dimension of their joint characterization of economic water efficiency. The contribution rate of PC1 of the social subsystem was 51.06%. X₉ dominates this dimensional information with an extremely high load of 0.99, while X₇ and X₈ contribute weakly. These results explain another limitation of the entropy weight method, which cannot distinguish whether the high weights arise from the theoretical importance of indicator independence or from the overlap of information caused by correlations between other indicators.

Moreover, the combination of different indicators may indeed lead to the same Z_k value (such as the high and low index scores canceling each other off), which further reflects that although the composite index is convenient for comparison, it cannot reflect the difference in system content structure.

Therefore, the comprehensive evaluation model constructed in this study should be regarded as an efficient data-driven diagnostic tool. In practice, it can effectively identify the key indicators that contribute the most to system differences, and it is recommended that decision-makers combine the knowledge of specific fields and the internal relationship between indicators to deeply analyze the causes of weights to formulate more targeted water resources management strategies.

3. Results and Discussion

3.1. Spatiotemporal Analysis of Water Stress Index

3.1.1. Comprehensive Evaluation Index of Natural Factors

The distribution of natural factors could reflect the use of water resources in various aspects such as water resource endowment and water-saving technology effectiveness in each region. Herein, the total water resources per unit area, the actual water consumption per mu of cultivated land, and the development and utilization rate of water resources were selected from the perspectives of total amount, land per capita, and exploitation, respectively. The entropy weight comprehensive index method was used to obtain the comprehensive evaluation index of natural factors in various provinces in China (except Hong Kong, Macao, and Taiwan) to measure the water resources stress of each region from the perspective of natural factors. The results are shown in Figure S1.

From the perspective of time distribution (Figure S1), the evaluation index caused by natural factors in each region has a small change range with the year, and there is no obvious change trend. The evaluation index increases in some years and decreases in some years. This is because the natural factors are mainly related to the actual water consumption per mu of cultivated land, the per capita water resources, and the development and utilization rate of water resources, and these indicators will fluctuate to a certain extent every year. Overall, however, the relative change in the index was not substantial, and no pronounced color transitions were observed. From the perspective of the average evaluation index over the years, Ningxia has the highest comprehensive evaluation index of natural factors of 0.8306, which is mainly due to the precipitous increase in the utilization rate of water resource development in Ningxia, with the lowest in Zhejiang (0.0790). It can be seen that the spatial differences in the distribution and use of water resources in different regions of China are large, which further indicates that there are large differences in water stress between regions. Secondly, from the perspective of the annual average index, the evaluation index of regions with close geographical location is relatively close; for example, Jiangxi, Fujian, Hunan, and Hubei provinces are geographically adjacent to each other, and the evaluation index is close. Liaoning, Jilin, and Heilongjiang provinces are geographically adjacent to each other, and the evaluation index is close. These results indicate that there may be spatial aggregation in the evaluation index of natural factors in different regions of China. Therefore, it can be seen that there are overall differences and local aggregation distribution characteristics of the natural factor evaluation index caused by the difference in the distribution of water resources in China. Herein, the annual average natural factor evaluation index was further plotted into an evaluation index heat map using Excel software, as shown in Figure 2.

As can be seen from Figure 2, the comprehensive evaluation index of China’s regional natural factors is divided into five categories from light to dark, with darker colors indicating greater the water stress caused by natural factors and lighter colors indicating less water stress. The range of the lightest color was 0~0.12, the range of the darkest color was more than 0.58, and the difference was significant. If the areas with the strongest water stress are regarded as being in the first category, Ningxia is included in the first category, and the third category includes Shanghai, Jiangsu, and Tianjin. The third category includes eight regions: Xinjiang, Hebei, Hainan, Beijing, Guangxi, Gansu, Guangdong, and Inner Mongolia. The fourth category includes 12 regions: Fujian, Heilongjiang, Shanxi, Jiangxi, Liaoning, Tibet, Qinghai, Henan, Jilin, Shandong, Hubei, and Hunan. The fifth category includes seven regions: Yunnan, Shaanxi, Anhui, Sichuan, Guizhou, Chongqing, and Zhejiang. Overall, the evaluation index is high in the north and low in the south, and the difference is obvious, which is mainly due to the fact that the distribution of water resources in China is more in the south and less in the north. This is consistent with the distribution trend of water stress in China reported in existing studies [18,36]. Locally, the geographical adjacent or close areas have the same or close color; that is, the comprehensive evaluation index of natural factors is close and the agglomeration characteristics are obvious, which is in line with the first principle of geography.

3.1.2. Comprehensive Evaluation Index of Economic Factors

The assessment of economic factors can reflect the correlation between the overall water use efficiency and sustainability of each region, such as water consumption per 10,000 yuan of GDP and industrial structure, which can quantify the water cost and transformation potential of economic growth. In this study, we will select three indicators of water consumption per 10,000 yuan of GDP, the proportion of the tertiary industry, and the water consumption of 10,000 yuan of industrial added value, respectively, to measure the degree of water resource consumption in each region from three aspects: macro resource utilization intensity, structural transformation direction, and industrial technology efficiency. The comprehensive evaluation index of economic factors is shown in Figure S2.

From the perspective of the change trend of the comprehensive evaluation index of economic factors in each region, the change range of each region is not large, but there is a more or less increasing or decreasing trend. From the perspective of the average annual comprehensive evaluation index, Xinjiang has the largest evaluation index of 0.6017, and Beijing has the smallest evaluation index of 0.0075. It can be seen that the spatial difference is large. The above distribution of the evaluation index in each region may be attributed to the following aspects: First, water stress depends on water use efficiency, such as water consumption per 10,000 yuan of GDP. In the case of creating the same GDP, the larger the index, the smaller the water consumption, and the smaller the water stress. Similarly, there is water consumption per 10,000 yuan of industrial added value. Focusing on technological progress in the industrial field can effectively improve the efficiency of industrial technology and reduce dependence on water resources. Second, water stress is closely related to industrial structure. Generally speaking, the water use structure (agriculture, industry, life, ecology, etc.) in each region is different, among which the primary industry uses the most water, and the tertiary industry uses less water. Quantifying the proportion of the tertiary industry can reveal the potential of industrial upgrading to alleviate water resource stress. The higher the proportion of tertiary industry, the smaller the water resource stress. As shown in Figure S2, the comprehensive evaluation index of economic factors is relatively low in several developed regions such as Beijing, Tianjin, and Zhejiang, which is largely due to the improvement of water use efficiency and the rational transformation of economic structure. In addition, the comprehensive evaluation index of economic factors also shows the characteristics of local spatial agglomeration, which may be attributed to the similarity of industrial structure and economic development in regions with similar geographical locations. In order to more intuitively show the spatial distribution of China’s regional economic factor comprehensive evaluation index, we used Excel to draw the relevant heat map again; see Figure 3 for details.

Figure 3 shows that China’s regional evaluation index can be divided into five parts according to the comprehensive evaluation index of economic factors. From the perspective of grouping intervals, the smallest interval of the evaluation index is 0~0.16, and the largest interval is more than 0.53. Xinjiang, Tibet, Heilongjiang, and Guangxi are the four regions with the darkest colors, indicating that water stress is the greatest. Among them, Xinjiang has the greatest water resource stress, which may be mainly due to its location in the northwest, relatively underdeveloped economy, largest water consumption index per 10,000 yuan of GDP, low water use efficiency, and greater water resource stress. In addition, Beijing, Tianjin, Zhejiang, Shandong, and other regions have the lightest color, which may be due to the abundant water supply and high water use efficiency in these regions, so the water stress is less. Taking Beijing as an example, Beijing’s water consumption per 10,000 yuan of GDP and 10,000 yuan of industrial added value are relatively the smallest among all regions in China, and the proportion of tertiary industry is the largest among all regions, so Beijing’s comprehensive evaluation index of economic factors is the lowest and the water stress is the smallest. Similarly, geographically adjacent or close areas are the same or close in color, and the agglomeration characteristics are obvious. Therefore, the comprehensive evaluation index of regional economic factors in China exhibits pronounced spatial heterogeneity and significant agglomeration characteristics.

3.1.3. Comprehensive Evaluation Index of Social Factors

The assessment of social factors could reveal the structural mechanisms and systemic vulnerabilities of water stress. In this study, three indicators of urbanization rate, population density, and per capita daily domestic water consumption were selected to quantify the stress degree of human activities on the water system. Among them, the urbanization rate reflects the direction of economic structural transformation and can characterize the effect of industrial and population agglomeration. Population density determines the geographical concentration of water demand, which can reflect the intensity of spatial stress distribution. The per capita daily domestic water consumption of residents can directly quantify the water cost of lifestyle and characterize the intensity of consumption behavior. The comprehensive evaluation index of social factors is shown in Figure S3.

As can be seen from Figure S3, the comprehensive evaluation index of social factors in Shanghai is the largest, which is 0.8917, and the evaluation index of Gansu is the smallest, which is only 0.0524, with a large spatial difference range. It is worth noting that the larger the comprehensive evaluation index of social factors, the greater the water stress caused by social factors. Among them, Shanghai, Beijing, and Tianjin are significantly ahead in the comprehensive evaluation index, which is closely related to their population density overload and consumption upgrading to promote per capita water use. Secondly, the spatial agglomeration is obvious. The composite evaluation index of geographically close or adjacent regions is also similar. The geographical proximity also means that there is a high degree of similarity between physical geographical features and population distribution patterns. For example, the evaluation indices of Anhui, Henan, Hubei, Hunan, and Jiangxi in the central region are similar. Similarly, we use a heat map (Figure 4) to visualize the spatial distribution characteristics of the comprehensive evaluation index of regional social factors in China.

Figure 4 divides China’s 31 provinces (excluding Hong Kong, Macao, and Taiwan) into five categories, the first of which includes Shanghai. The second category includes Beijing. The third category includes Tianjin, Jiangsu, Guangdong, Zhejiang. Chongqing, Hainan, Chongqing, Fujian, Shandong, Hubei, Anhui, Henan, Hunan, Guangxi, Jiangxi, Liaoning, and Sichuan are in the fourth category. Xinjiang, Shaanxi, Guizhou, Shanxi, Jilin, Heilongjiang, Yunnan, Ningxia, Inner Mongolia, Qinghai, Tibet, and Gansu belong to the fifth category. From the minimum interval of 0~0.12 to the maximum interval of 0.65, it can be seen that the comprehensive evaluation index of social factors has significant regional differences. From the perspective of the overall geographical distribution of the region, it belongs to the geographical distribution of high in the east and low in the west. Secondly, the color of the comprehensive evaluation index of regions adjacent or close to each other is similar, which indicates that the comprehensive evaluation index of regional social factors in China also has significant spatial agglomeration characteristics.

3.1.4. Comprehensive Evaluation Index of Composite Factors

The above studies used the entropy method to calculate the comprehensive evaluation index of natural resources (reflecting the background conditions of water resources), the comprehensive evaluation index of economy (representing industrial water intensity), and the comprehensive evaluation index of social factors (quantifying population agglomeration and consumption behavior) in each region of China, qualitatively analyzing the geographical distribution of these indices in each region of China. The results show that there are significant differences and spatial agglomeration of water stress caused by natural, economic, and social factors in different regions. However, there are inevitable limitations of single-element analysis, and it is necessary to start from the perspective of systems theory, regarding natural, economic, and social development and water resources as a composite system and judging the magnitude of water stress through the size of the comprehensive index. In this study, we believe that the three factors of nature, economy, and society are equally important, so when calculating the composite factor comprehensive evaluation index, the weight of these three indicators is taken as 1/3. The calculation results are shown in Figure S4.

Figure S4 shows that the composite factor comprehensive evaluation index of Shanghai is the largest, which is 0.4825, and the evaluation index of Shaanxi is the smallest, which is only 0.1074, with a large spatial difference range. In order to further illustrate the trend of China’s regional composite factor comprehensive evaluation index from 2019 to 2023, we draw a related trend change map (Figure 5). The results show that in the past five years, the evaluation index of each region has shown a downward trend year by year, advancing towards a healthy and sustainable development state. Combined with the analysis results in Section 3.1.1, Section 3.1.2 and Section 3.1.3, Shanghai’s high position is mainly due to its comprehensive evaluation index of natural factors and social factors. However, the comprehensive evaluation index of high composite factors in Ningxia and Xinjiang is attributed to the higher values of the two comprehensive evaluation indexes of natural factors and economic factors. In addition, Shaanxi’s comprehensive evaluation index is low, mainly because the water stress caused by natural, economic, and social factors in the region is relatively small in the past five years, and the comprehensive ranking of each index is low, resulting in a low ranking of the final composite evaluation index.

Figure 6 shows the spatial distribution characteristics of the regional comprehensive water stress index in China. From the perspective of group intervals, the smallest interval of the evaluation index was 0~0.16, the largest interval was more than 0.38, and the difference was more significant. From the perspective of spatial distribution, Xinjiang, Tibet, and Ningxia in Northwest China showed darker colors, indicating that the water stress in these regions was high, which may be due to the scarcity of water resources and the low water resource utilization efficiency. Southeast regions, such as Jiangsu, Shanghai, Guangdong, and Guangxi, all show high water stress, which may be related to their high population density and frequent economic activities. In addition, the water stress in the central part of the country and most of the southwest and northeast regions is relatively low, which may be related to factors such as relatively low population density, effective policy guidance, and the promotion of feasible water-saving technologies.

3.2. Analysis of Influencing Factors of Water Stress Index

In order to further explore the obstacle factors of the water stress index in China and improve countermeasures, we studied the obstacle factors of water stress from the index level. According to

Table 3. The main obstacle factors and degree regarding the water stress index in China.

Ranking	1		2		3		4
	Factor	Obstacle %	Factor	Obstacle %	Factor	Obstacle %	Factor	Obstacle %
Beijing	X8	0.27	X3	0.26	X4	0.19	X6	0.12
Tianjin	X3	0.26	X8	0.26	X4	0.20	X6	0.13
Hebei	X3	0.27	X8	0.26	X4	0.20	X6	0.13
Shanxi	X3	0.27	X8	0.26	X4	0.20	X6	0.13
Inner Mongolia	X3	0.29	X8	0.27	X4	0.19	X6	0.14
Liaoning	X3	0.28	X8	0.27	X4	0.20	X6	0.13
Jilin	X3	0.28	X8	0.27	X4	0.20	X6	0.13
Heilongjiang	X3	0.28	X8	0.27	X4	0.18	X6	0.14
Shanghai	X9	0.32	X3	0.27	X4	0.23	X6	0.08
Jiangsu	X3	0.27	X8	0.27	X4	0.24	X6	0.09
Zhejiang	X8	0.27	X3	0.27	X4	0.20	X6	0.12
Anhui	X3	0.27	X8	0.26	X4	0.24	X6	0.08
Fujian	X3	0.29	X8	0.28	X4	0.21	X6	0.12
Jiangxi	X3	0.29	X8	0.27	X4	0.22	X6	0.11
Shangdong	X3	0.26	X8	0.25	X4	0.20	X6	0.13
Henan	X3	0.26	X8	0.25	X4	0.20	X6	0.13
Hubei	X8	0.27	X3	0.27	X4	0.23	X6	0.10
Hunan	X3	0.28	X8	0.27	X4	0.22	X6	0.10
Guangdong	X3	0.29	X8	0.28	X4	0.20	X6	0.12
Guangxi	X3	0.30	X8	0.28	X4	0.23	X6	0.09
Hainan	X3	0.30	X8	0.28	X4	0.20	X6	0.12
Chongqing	X8	0.28	X3	0.27	X4	0.21	X6	0.11
Sichuan	X3	0.28	X8	0.27	X4	0.20	X6	0.13
Guizhou	X3	0.28	X8	0.26	X4	0.21	X6	0.11
Yunnan	X3	0.28	X8	0.26	X4	0.20	X6	0.12
Tibet	X3	0.29	X8	0.26	X4	0.24	X6	0.08
Shannxi	X3	0.27	X8	0.26	X4	0.20	X6	0.13
Gansu	X3	0.29	X8	0.26	X4	0.19	X6	0.13
Qinghai	X3	0.29	X8	0.26	X4	0.20	X6	0.13
Ningxia	X8	0.26	X4	0.18	X2	0.17	X3	0.16
Xinjiang	X3	0.30	X8	0.28	X6	0.26	X4	0.06

, it can be concluded that the water resource development utilization rate (X₃), water consumption per 10,000 yuan of GDP (X₄), water consumption per 10,000 yuan of industrial added value (X₆), and population density (X₈) are the main influencing factors common to all regions, and the other factors are slightly different in different regions. In conclusion, the main obstacle factors belong to the three subsystems of nature, economy, and society, which indicates that the interaction and synergy between these three subsystems should be fully considered when formulating water resource management strategies. Comprehensive measures should be taken to deal with the problem of water stress to achieve sustainable use of water resources and coordinated economic and social development.

4. Conclusions

In this study, we constructed an evaluation index system for water resource stress, and we used the entropy weight method and obstacle degree analysis to analyze the temporal and spatial distribution characteristics of water stress in China in recent years. We discussed the influencing factors and drew the following conclusions.

There are significant spatial and temporal differences and spatial agglomeration of regional water stress in China under the leading role of natural, economic, and social factors. The comprehensive evaluation index of natural factors shows the phenomenon of high in the north and low in the south, the comprehensive evaluation index of economic factors shows a weakening trend from the outside to the inside, and the comprehensive evaluation index of social factors shows the distribution of high in the east and low in the west. The comprehensive water resources stress index also has obvious regional differences and spatial agglomeration characteristics, which is in line with the first principle of geography. In view of this, cooperation between neighboring regions should be strengthened, and the efficiency of water resources utilization should be improved through the cooperation mechanism of water resource allocation and the sharing of water-saving technologies to achieve coordinated development of the region. For example, in the Beijing–Tianjin–Hebei region where water stress is high due to natural factors, in addition to strengthening water-saving measures, investment in the research, development, and application of water recycling technologies should also be increased. In Xinjiang, Tibet, and other places where economic factors have a great impact, attention should be paid to the transformation of water-saving technologies, appropriate development of industries with low water consumption and high added value, and the establishment of a strict system for the paid use of water resources. In addition, attention should also be paid to the coordinated development of social development and water resources.

(1)

The comprehensive water stress index shows that the water resource stress in Shanghai is the highest (most indicators are ahead of other regions), and the stress in Shaanxi is the lowest (most indicators are ranked behind). The water stress index of each region has shown a downward trend year over year in the past five years, indicating that China has achieved certain results in water resource management and utilization. However, the water resource stress in some areas is still high, and it is necessary to further strengthen the protection and management of regional water resources. For regions such as Shanghai, where water stress is greater, in addition to continuing to strengthen water-saving measures, it is also necessary to increase the research, development, and application of seawater desalination technology and increase the proportion of unconventional water sources. Meanwhile, it is necessary to strengthen the supervision of high-water-consuming enterprises, strictly control new high-water-consuming projects, and guide enterprises to change to water-saving production methods. For regions with relatively low stress, such as Shaanxi, it is necessary to continue to strengthen the protection of water resources to prevent the increase in water stress caused by factors such as economic development and population growth.

(2)

The main obstacle factors of water stress come from the three subsystems of nature, economy, and society, among which the development and utilization rate of water resources, the water consumption per 10,000 yuan of GDP, the water consumption of 10,000 yuan of industrial added value, and the population density are the main influencing factors common to all regions. These results indicate that natural conditions, economic development level, and population factors play an important role in the formation of water stress, and it is necessary to comprehensively consider these factors and formulate targeted water resources management and protection measures. In view of the main obstacle to the development and utilization of water resources, it is necessary to strengthen scientific planning and strict examination and approval of water resource development and utilization projects to avoid over-exploitation of water resources. For areas with high utilization rates of developed water, it is necessary to carry out post-development and utilization assessment of water resources and formulate corresponding water resource protection and restoration measures according to the assessment results. In view of the high water consumption per 10,000 yuan of GDP and the high water consumption of 10,000 yuan of industrial added value, it is necessary to establish an evaluation system for enterprise water use efficiency, carry out key supervision and rectification within a time limit for enterprises with low water use efficiency, and guide enterprises to adopt advanced water-saving technologies and production processes to improve water use efficiency. For areas with high population density, it is necessary to strengthen urban planning and population management, reasonably control the scale of cities and the rate of population growth, and at the same time strengthen publicity and education on water conservation, raise public awareness of water conservation, and reduce per capita water consumption.

In conclusion, this study enriches the scope of characteristic factors in the quantification model of water scarcity footprints, makes the assessment of water scarcity footprints more comprehensive, and provides a scientific basis for a deep understanding of the distribution laws and impact mechanisms of water resource pressure in China, providing a reference for the rational allocation and sustainable use of water resources. However, the study of water stress assessment will involve a wide range of factors due to the lack of reliable data sources, comprehensive data statistical methods, etc. In the future, we will continue to integrate factors such as climate, ecology, policy, system, and culture into the existing indicator system to fully reflect the complexity and systematicity of water resource pressure and further improve the quantitative model of water scarcity footprint. Meanwhile, we will also try to integrate other empowerment methods to reduce the calculation deviation of traditional entropy weight methods when processing data, further deepen the evaluation of water resources pressure, achieve a leap from diagnosis to governance, and provide more scientific and practical systematic solutions for the sustainable development of global resource-based cities. Furthermore, the current study has solely focused on water quantity-induced scarcity pressure. Given that regional water scarcity is inextricably linked to water quality impacts, future research will aim to develop a water degradation model and systematically analyze its coupling effects with water quantity pressure, thereby providing a more comprehensive understanding of the integrated impacts of human activities on water resources.