Study on the Spatiotemporal Evolution and Driving Factors of Water Resource Carrying Capacity in Typical Arid Regions

Abstract

As an important indicator for assessing regional water resources, the study of the spatiotemporal evolution and driving factors of water resources carrying capacity (WRCC) is essential for achieving sustainable water resource utilization. This study focuses on Yulin City, a typical arid region located on the Loess Plateau in northwestern China. By constructing an evaluation index system for regional WRCC and combining an improved fuzzy comprehensive evaluation model with the TOPSIS evaluation model, a comprehensive WRCC evaluation model is established. Additionally, Geodetector is used to explore the main driving factors behind the evolution of regional WRCC. This multidimensional analytical framework aims to deeply analyze the dynamic evolution trends of WRCC and the driving mechanisms of different factors in its spatiotemporal changes. The results indicate that (1) from 2011 to 2020, the overall WRCC of Yulin City showed a trend of positive improvement, with Shenmu, Yuyang, and Fugu areas performing the best, and by 2020, more than half of the counties had achieved Grade 3 or above; (2) the spatial variability of WRCC in Yulin City was more significant than its temporal changes; and (3) in terms of driving mechanisms, the northern six counties gradually shifted from traditional economic-driven factors to ecological and environmental drivers, whereas the southern six counties remained constrained by economic factors. Overall, water resource factors remain the primary driving force for the socio-economic development and environmental sustainability of the entire Yulin City. The study provides valuable information for water resource allocation and differentiated management in arid regions.

1. Introduction

Water resources are fundamental to human survival and social development, serving as a critical guarantee for maintaining ecological health. Under the combined influences of climate change, accelerated urbanization, and rapid population growth, the conflict between water supply and demand is becoming increasingly pronounced, characterized by a decrease in supply and an increase in demand [1].

The impact of water scarcity is especially severe in arid regions, where it not only threatens public health and hampers regional economic growth [2] but also constrains food production, ecosystem functions, and urban development [3]. Meanwhile, the United Nations predicts that the global urbanization rate will reach 68% by 2025. While the accelerated pace of urbanization injects new vitality into urban economic development, it also poses threats to the ecological environment [4], further highlighting the urgency of water resource management [5]. Under the conditions of limited water resources, it has become a serious challenge for all countries to realize the sustainable development of society and the harmonious evolution of the ecological environment. Judging the capacity of regional water resources to support socio-economic development and environmental protection and planning urban water allocation and development within carrying capacity constitute a crucial strategy to address the water resource crisis [6]. Evaluating water resources carrying capacity (WRCC) in arid areas is particularly urgent as it is key to ensuring water security, promoting ecological civilization, and achieving sustainable social development.

Yulin City, located in northwestern China, faces multiple challenges in its current development. Its industrial structure is generally at a low to medium level, with uneven urban–rural development and weak industrial chain competitiveness. Yulin City is under dual pressure from resource exploitation and ecological protection, especially in water resource management, where there are significant challenges and potential risks of secondary desertification. Additionally, there are notable shortcomings in livelihood areas, with overall low resident income levels and substantial tasks in improving social welfare and construction. Given the complexity and sensitivity of water resources in Yulin City, its WRCC is more susceptible to external influences [7]. As a microcosm of water resources and ecological protection in Northwest China, the study of Yulin City’s water resources carrying capacity is typical and can provide a reference for similar regions. Meanwhile, to face the challenges of water shortage and ecological degradation brought about by global climate change and rapid economic development in the region, it is urgent and necessary to study the carrying capacity of its water resources. Therefore, evaluating the WRCC of Yulin City and accurately understanding its current state and potential will not only provide a scientific basis and direction for regional planning, spatial balance, and future development goals but also offer strong support for local water resource management and ecological protection.

WRCC is a crucial indicator for assessing whether regional water resources can sustain the population, economy, and environmental sustainability [8]. Currently, there are two main perspectives on the concept: the first emphasizes the capacity for water resource development, referring to the maximum scale of water resource utilization that can support economic development without harming the environment. The second focuses on the sustainable support capacity of water resources, defining WRCC as the maximum water resource capacity for achieving regional sustainable development.

Since water resources inherently support multiple systems, their carrying capacity evaluation is inevitably influenced by factors such as resources, the population, the economy, and the ecological environment. When constructing evaluation index systems and choosing evaluation methods, regional differences, data acquisition difficulty, and specific research objectives must be fully considered. Presently, there is no unified and standardized index system or evaluation method in this field, leading to significant uncertainty in indicator selection and subjective influence on evaluation results. Notably, many studies tend to focus excessively on the water resource system itself, neglecting the comprehensive consideration of other supporting systems [9]. This approach may result in WRCC assessment outcomes that deviate from the core concept, failing to accurately reflect the true carrying capacity of regional water resources. Therefore, to more accurately and reasonably evaluate the WRCC of a region, it is necessary to construct a comprehensive, multi-system index system and employ objective, scientific methods to accurately evaluate the regional WRCC.

Numerous scholars have explored WRCC research methods, conducting quantitative evaluations based on various assessment index systems by constructing different models. These models have evolved from single-indicator evaluations to multi-factor evaluations. Moreover, the evaluation perspectives have expanded from single-system analyses to coupled analyses of complex systems, attempting to delve into the internal mechanisms and understand how system dynamics drive WRCC. However, multi-factor evaluations of complex systems face several challenges in practice, such as the difficulty of obtaining data in remote areas, the uncertainty of key factors, and similar challenges. Consequently, many researchers currently use comprehensive indicators to evaluate WRCC and select evaluation models based on the actual conditions of the region. Presently, methods such as the PVAR model [10], entropy weight method [11], fuzzy comprehensive evaluation model [12], TOPSIS model [13], SD model [12,14], and principal component analysis (PCA) [15] are widely applied in constructing and evaluating WRCC indicators. Although these methods provide quantitative evaluation results for regional WRCC, factors such as regional differences and the complexity of evaluation indicators make it challenging to directly translate these model evaluation results into specific recommendations or feasible paths for regional development planning. This limitation reduces the reference value when formulating region-specific development strategies.

Therefore, it is crucial to further identify the driving factors behind the evolution of regional WRCC based on quantitative evaluations. Scholars have tried using various methods in different regions to explore the main driving factors and their mechanisms for WRCC, including the geographically and temporally weighted regression (GTWR) model [16], the obstacle degree model [17], and principal component analysis [18]. However, these methods have certain limitations in addressing spatial heterogeneity and indicator correlation. To overcome these limitations, the recently proposed Geodetector model has demonstrated unique advantages in WRCC driving force research. It can more comprehensively consider spatial heterogeneity and indicator correlation, complementing the deficiencies of other methods [19].

This study constructed a WRCC evaluation model that encompasses the population and society, economic development, and ecological environment systems. It investigated the internal mechanisms of the WRCC system by coupling an improved fuzzy comprehensive evaluation model based on the entropy weight method with the TOPSIS model to evaluate the WRCC of 12 counties and districts in Yulin City from 2011 to 2020. The affiliation function of the fuzzy comprehensive evaluation method was optimized to minimize the subjective influence, and the spatiotemporal evolution of regional WRCC patterns was explored. The Geodetector method was employed to analyze the intrinsic driving factors of WRCC spatiotemporal evolution in Yulin City from both single-factor and interactive-factor perspectives. The research results aim to provide a comprehensive and reasonable evaluation of the current WRCC in Yulin City and support regional water resource planning and management policies.

2. Study Area and Methods

To study the WRCC of Yulin City and explore its main driving factors, this research establishes a framework consisting of three main modules: (1) The Determination of Regional WRCC Evaluation Indicators; (2) Analysis of the Spatiotemporal Evolution of WRCC; and (3) Analysis of Main Driving Factors. Based on the fundamental characteristics of the region, relevant factors influencing WRCC will be investigated. Quantitative scoring and ranking of WRCC will be conducted to explore its spatiotemporal patterns and discuss development differences between different areas within the region. Finally, the main drivers of WRCC in the region are identified to analyze the regional development path, aiming to improve WRCC based on securing economic development.

2.1. Study Area

This study selects Yulin City, Shaanxi Province, China (36°57′~39°35′ N, 107°28′~111°15′ E) as the research area (Figure 1). Yulin City is located on the Loess Plateau in northwest China, with a total area of 42,920.20 km², making it the largest prefecture-level city in Shaanxi Province, encompassing 1 county-level city, 2 districts, and 9 counties. As of 2022, Yulin City has a permanent population of 3.6161 million and a regional GDP of 654.365 billion yuan, highlighting its significant regional resource advantages, particularly in coal, gas, oil, and salt resources.

The region is a typical ecological transition zone, with the northern part characterized by wind–sand grasslands, covering 42% of the area, and the southern part by hilly terrain on the Loess Plateau, covering 58%. The overall area faces severe soil erosion. Additionally, the region has an average annual precipitation of 393.90 mm, and the per capita water resources amount to 740 m³, only 33% of the Chinese average and below the internationally recognized water scarcity threshold of 1000 m³ per capita [20]. Moreover, the fragmented terrain exacerbates soil erosion, making water resources difficult to develop and utilize effectively. Water scarcity has become a bottleneck for regional economic and social development and the construction of ecological civilization in Yulin City, hindering its transition from high-speed to high-quality economic development.

2.2. Data Collection

This study collected original data on various factors from 2011 to 2020 for 12 counties and districts (including county-level cities) in Yulin City. The data sources for economic and social development factors primarily include the “Statistical Yearbook of Yulin City” and the “Statistical Bulletin of National Economy and Social Development of Yulin City” (http://tjj.yl.gov.cn/list/tjgb, accessed on 20 April 2024). Data related to water resources and ecological environment factors, including water usage, were sourced from the “Water Resources Bulletin of Yulin City” (http://slj.yl.gov.cn/d-show-528.html, accessed on 20 April 2024) and other officially published datasets. Some factors were derived from original data calculations; for instance, the ecological environment water usage rate was calculated as the ecological environment water consumption divided by the total water consumption.

2.3. WRCC Evaluation System and Indicator Determination

The relationship between WRCC and urban development, as well as the natural endowment of water resources, forms the underlying framework of regional WRCC, which is intricately intertwined with various complex systems. Quantitative assessment of WRCC enables an objective evaluation of the prospects for sustainable development and utilization of water resources. Based on the evaluation results, adjustments and responses can be made to the layout and regulation of the regional ecological environment and socio-economic development to achieve spatial balance and sustainable development. Establishing a comprehensive and systematic indicator system [16] is essential to ensure objective evaluation of regional WRCC. Therefore, this study begins with WRCC and focuses on three subsystems: the population and social system, the economic development system, and the ecological environment system. These choices are made based on regional characteristics and data availability [21,22]. Within these three systems, 15 potential evaluation factors for regional WRCC were initially selected. Using the Friedman test, KMO test, and Bartlett’s spherical test, correlation analysis and information contribution rate analysis were conducted to validate the effectiveness of the indicators and eliminate redundant information. Finally, 10 indicators were determined to constitute the WRCC evaluation indicator system for this study.

2.4. Entropy Weight Method

Indicator weights can reflect the relative importance of different indicators, and determining these weights scientifically and reasonably is crucial for methodological soundness. Information entropy is a measure of the degree of uncertainty or confusion inherent in the data, and through the precise calculation of the information entropy value (i.e., entropy) of each evaluation indicator, it is possible to quantify the degree of variation and uncertainty of these indicators in the evaluation system and achieve the determination of the weights of each indicator. The method applied in this process is the entropy weighting method. The entropy weight method can objectively determine the weights of indicators based on the judgment matrix composed of evaluation indicator values, using entropy values to reflect the importance of each indicator to the overall evaluation system [8].

For the evaluation indicator system of WRCC, there are both positive and negative indicators. Positive indicators indicate that higher values lead to better evaluation results, while negative indicators work inversely. Additionally, the physical dimensional differences between indicators can influence the calculation results. Therefore, this study employs standardized linear interpolation to normalize both positive and negative indicators. We assume there are $m$ indicators: $a$₁, $a$₂,……, $a$_m; and each indicator has $n$ values, where $n$ represents the number of administrative regions.

For positive indicators:

$$u_{ij}={\displaystyle \frac{a_{ij}-{min}_j\left\{a_{ij}\right\}}{{max}_j\left\{a_{ij}\right\}-{min}_j\left\{a_{ij}\right\}}}$$

(1)

For negative indicators:

$$u_{ij}={\displaystyle \frac{{max}_j\left\{a_{ij}\right\}-a_{ij}}{{max}_j\left\{a_{ij}\right\}-{min}_j\left\{a_{ij}\right\}}}$$

(2)

where $a_{ij}$ represents the $j$-th indicator value of the $i$-th sample, $u_{ij}$ denotes the standardized and normalized value after dimensionless processing, and $i$-th and $j$-th are indices that distinguish between different indicators and samples, respectively. Furthermore, positive indicators signify that an increase in their values is advantageous to the evaluation outcome, whereas negative indicators exhibit the opposite effect.

Entropy can be used to measure its variability, computed as:

$$h_i=-k\sum _{j=1}^n({\displaystyle \frac{1+u_{ij}}{\sum _{j=1}^n\left(1+u_{ij}\right)}})\ln ({\displaystyle \frac{1+u_{ij}}{\sum _{j=1}^n\left(1+u_{ij}\right)}})$$

(3)

where $h_i$ is the $i$-th entropy value and $k=1/\mathrm{l}\mathrm{n}\left(n\right)$, $n$ represents the number of administrative regions, $k$ > 0.

Entropy weight, determined by the entropy weighting method, is calculated as:

$${\omega }_i={\displaystyle \frac{1-h_i}{m-\sum _{j=1}^m{h}_i}}$$

(4)

where ${\omega }_i$ is the weight of the $i$-th indicator, with a total of 10 indicators used in this study.

2.5. Improved Fuzzy Comprehensive Evaluation Model

The fuzzy comprehensive evaluation model is a method based on the synthesis principle of fuzzy relationships, which quantifies difficult-to-quantify factors to achieve research on complex systems. It demonstrates outstanding performance in dealing with uncertainties, providing evaluation results that can qualitatively distinguish and quantitatively rank. This method is highly applicable to the evaluation of water resources carrying capacity, which itself is built upon multiple systems and indicators. This approach calculates the affiliation matrices of the different factors for the different levels of the target, establishing a fuzzy comprehensive evaluation result matrix. After normalization through weighted fuzzy vector operations, a comprehensive evaluation value can be obtained [23].

In this study, the factor set $R$ = {$a$₁, $a$₂,⋯⋯, $a$_m}, m = 10, and evaluation set V = [V₁, V₂,⋯⋯, V_n], $n$ = 3 are established. Using a linear change-based method to determine the degree of affiliation, we address the issue of leapfrogging that may occur in a single linear change, which can lead to subjective bias in results. In this paper, the membership degree of values falling at the midpoint of interval V₂ is set to 1, while the membership degrees on both sides of the interval are set to 0.5. For the two marginal intervals V₁ and V₃, the affiliation degree increases for values farther away from the interval, with critical points set at 0.50. The critical values are defined as follows: $k_3$ for the boundary between V₃ and V₂, $k_1$ for the boundary between V₂ and V₁, and $k_2$ as the midpoint of interval V₂. The determination of $k$ values adopts a dynamic adjustment method to eliminate the subjective partitioning impact of the evaluation set. For positive indicators, $k_1$ takes the minimum factor value, $k_2$ takes the average factor value, and $k_3$ takes the maximum factor value. For negative indicators, the opposite applies.

Taking positive indicators as an example, the affiliation function for each level is calculated as follows (inverse operations apply for negative indicators):

$$\begin{gathered} r_{i1}=\left\{\begin{array}{c}0.5\left(1+{\displaystyle \frac{a_{ij}-k_1}{a_{ij}-k_2}}\right) \left(a_{ij}\ge k_1\right)\\ 0.5\left(1+{\displaystyle \frac{k_1-a_{ij}}{k_1-k_2}}\right) \left(k_2\le a_{ij}\le k_1\right)\\ 0 \left(u_i<k_2\right)\end{array}\right. \\ {r}_{i2}=\left\{\begin{array}{c}0.5\left(1-{\displaystyle \frac{a_{ij}-k_1}{a_{ij}-k_2}}\right) \left(a_{ij}\ge k_1\right)\\ 0.5\left(1+{\displaystyle \frac{k_1-u_{ij}}{k_1-k_2}}\right) \left(k_2\le a_{ij}\le k_1\right)\\ 0.5\left(1+{\displaystyle \frac{a_{ij}-k_3}{k_2-k_3}}\right) \left(k_3\le a_{ij}\le k_2\right)\\ 0.5\left(1-{\displaystyle \frac{k_3-a_{ij}}{k_2-a_{ij}}}\right) \left(a_{ij}<k_3\right)\end{array}\right. \\ {r}_{i3}=\left\{\begin{array}{c}0 \left(a_{ij}\ge k_2\right)\\ 0.5\left(1-{\displaystyle \frac{a_{ij}-k_3}{k_2-k_3}}\right) \left(k_3\le a_{ij}<k_2\right)\\ 0.5\left(1+{\displaystyle \frac{k_3-a_{ij}}{k_2-a_{ij}}}\right) \left(a_{ij}\le k_3\right)\end{array}\right. \end{gathered}$$

(5)

where $a_{ij}$ represents the original data matrix and $r_{ij}$ denotes the affiliation degree of each indicator for different levels.

According to the principle of fuzzy matrix operations, after normalization, the regional water resources carrying capacity score $\alpha$ is obtained using the formula:

$$B=W·R=\left({\omega }_1, {\omega }_2,\dots ,{\omega }_i,\dots {\omega }_m\right)\left[\begin{array}{ccc}{r}_{11}& \cdots & r_{1\mathrm{n}}\\ ⋮& r_{\mathrm{i}\mathrm{j}}& ⋮\\ r_{m1}& \cdots & r_{mn}\end{array}\right]=(b_1,b_2,\dots ,b_n)$$

(6)

where $W$is the matrix of indicator weights; $r_{ij}$ is the relative affiliation degree of the $i$-th factor in the $j$-th level of the evaluation set; $R$ is the corresponding affiliation degree matrix; and $B$ is the result of fuzzy operations.

The normalization formula is:

$$\alpha ={\displaystyle \frac{\sum _{i=1}^3{b}_j^k{a}_i}{\sum _{j=1}^3{b}_j^k}}$$

(7)

where $\alpha$ is the water resources carrying capacity score based on the fuzzy computation matrix. A higher $\alpha$ value indicates poorer water resources carrying capacity. For arid areas, $k$ is typically set to 1. $a$ is used to highlight the advantage level to enhance the distinguishability of the evaluation level scores. In this study, $a_i=[0.05,0.5,0.95]$ is chosen accordingly. $b_j$ represents the result of fuzzy relation operations.

2.6. The TOPSIS Model

Determining WRCC through a single method alone is challenging regarding validating its reasonableness. Therefore, this study opts to correct the evaluation results using the TOPSIS model, which offers advantages such as strong objectivity, clear mechanistic principles, scientific calculation, and computational simplicity. The basic principle of TOPSIS is based on the normalized original data matrix, where the optimal and worst solutions among limited options are identified. By computing the distances between the evaluation solutions and these optimal and worst solutions, the relative proximity of the evaluation object to the optimal solution is determined as the basis for evaluation. The model operates on the premise that the distance to the positive ideal solution is minimized while the distance to the negative ideal solution is maximized. This is achieved by calculating the Euclidean distance between each evaluation indicator and the critical threshold vectors, resulting in the ranking of regional WRCC based on proximity calculations [24].

The normalized processed data $u_{ij}$, multiplied by the entropy weight ${\omega }_j$ for each evaluation indicator, yields the weighted evaluation data $x_{ij}$. Here, the maximum value represents the positive ideal solution $x_j^{+}$, and the minimum value represents the negative ideal solution $x_j^{-}$. The formula for calculating the Euclidean distance between the positive and negative ideal solutions and the weighted evaluation data is as follows:

$$d_i^{+}=\sqrt{\sum _{j=1}^n{(x_{ij}-x_j^{+})}^2}$$

(8)

$$d_i^{-}=\sqrt{\sum _{j=1}^n{(x_{ij}-x_j^{-})}^2}$$

(9)

The formula for calculating proximity is as follows:

$${\mathsf{\beta }}_i={\displaystyle \frac{d_i^{-}}{d_i^{+}{+d}_i^{-}}}$$

(10)

where ${\mathsf{\beta }}_i$ is the evaluation result of the TOPSIS model, with values ranging from 0 to 1, with closer to 1 representing better water carrying capacity.

2.7. Geodetector

Regional WRCC is influenced by multiple systems. For Yulin City, significant spatial differences in resources and social development exist among different districts. Additionally, rapid economic development over time also affects the driving forces impacting WRCC differently at different times. Geodetector has received widespread attention for exploring the driving forces behind WRCC and spatial balance, showing promising results in applications [25]. Geodetector was proposed to detect spatial differentiation and underlying driving forces. Its core idea is that if an independent variable significantly affects a dependent variable, its spatial distributions should exhibit similarity. Geodetector calculates and compares q values of individual factors and their interaction values to determine whether there is interaction, and if so, the strength, direction, linearity, or non-linearity of the interaction [26]. It includes modules for factor detection, interaction detection, risk monitoring, and ecological detection. This study focuses on the factor detection and interaction detection modules, using Geodetector to analyze the driving effects of different factors on the spatio-temporal pattern evolution of regional WRCC. The initial data of the dependent variables are classified into five categories using K-means clustering analysis as inputs for Geodetector.

The measurement q is commonly used to express detection results, and its expression is as follows:

$$q=1-{\displaystyle \frac{1}{N{\sigma }^2}}\sum _{h=1}^m{N}_h{\sigma }_h^2$$

(11)

where $m$ represents the layer of variable $Y$ or factor $X$ in this study, $X$ represents the set of indicators influencing regional WRCC, and $Y$ signifies the quantitative assessment of the actual WRCC within that region. $N_h$ and $N_{ }$ are the number of units in layer $h$ and in the entire region, respectively; ${\sigma }^2$ and ${\sigma }_h^2$ are the variances of variable $Y$ in layer $h$ and in the entire region, respectively; and $q$ is the explanatory power of independent variable $X$ on dependent variable $Y$, ranging from 0 to 1, where a higher value of $q$ indicates stronger explanatory power of the corresponding factor on the dependent variable.

3. Results

3.1. WRCC Evaluation Index Selection

3.1.1. Factor Selection Based on Correlation Analysis

To ensure a true reflection of the current situation of the WRCC system, the selection of indicators should be comprehensive, non-redundant, and avoid overlapping information, thereby reducing correlations among indicators. This study initially selected 15 indicators from the systems of demographic society, economic development, and the ecological environment, constructing the initial set of influencing factors (

Table 1. Initial index system of WRCC.

System	Factors	Units
Demographic Society	Urbanization Rate	(%)
	Population Density	(person/km2)
	Urban Population	(10,000 people)
	Per Capita Water Resources	(m3/person)
	Per Capita Domestic Water Consumption in Urban Areas	(m3/person/day)
Economic Development	GDP Per Capita GDP	(RMB)
	Water Consumption per Ten Thousand Yuan of Industrial Output	(m3/10,000 RMB)
	Irrigation Water Use per Unit Area	(m3/mu)
	Industrial Water Consumption	(10,000 m3)
Ecological and Environment	Total Water Resources	(100 million m3)
	Water Supply	(10,000 m3)
	Water Use	(10,000 m3)
	Precipitation	(mm)
	Ecological Water Use Rate	(%)

).

The effectiveness of the indicators selected for evaluating WRCC in this study was verified using Friedman, KMO, and Bartlett tests [27]. The Friedman test tested the null hypothesis (H0) that all indicators have the same overall distribution. The test yielded a significance level of 0.00 < 0.01, rejecting H0, indicating significant differences in the distribution of selected indicators. The KMO test resulted in 0.48 < 0.50, suggesting weak or no significant correlations among factors, which meets the redundancy reduction requirement for model construction. Bartlett’s test resulted in Sig < 0.05, indicating that the indicators are suitable for factor analysis. Ultimately, indicators with correlation coefficients greater than 0.90 and those contributing to the bottom 20% of cumulative information were excluded from the selected WRCC evaluation indicators for this study (

Table 2. Multi-dimensional index system of water resources carrying capacity.

Criterion Layer	Indicator Layer	Units	Number	Attribute
Demographic Society	Urbanization Rate	(%)	a1	+
	Population Density	(person/km2)	a2	-
	Per Capita Water Resources	(m3/person)	a3	+
Economic Development	GDP Per Capita GDP	(RMB)	a4	+
	Water Use per Ten Thousand Yuan of Output	(m3/10,000 RMB)	a5	-
	Irrigation Water Use per Unit Area	(m3/mu)	a6	-
Ecological and Environment	Total Water Resources	(100 million m3)	a7	+
	Water Supply	(10,000 m3)	a8	+
	Precipitation	(mm)	a9	+
	Ecological Water Use Efficiency	(%)	a10	+

).

Determination of Indicator Weights Based on Entropy Weight Method

Based on the established WRCC evaluation indicator system, the weights of each indicator were determined using the entropy weight method, as shown in

Table 3. WRCC indicator weight table.

Criteria Layer	Indicator Layer	Weight
		(2011)	(2012)	(2013)	(2014)	(2015)	(2016)	(2017)	(2018)	(2019)	(2020)
Demographic Society	Urbanization Rate	0.13	0.13	0.13	0.13	0.13	0.14	0.13	0.13	0.13	0.13
	Population Density	0.02	0.01	0.01	0.01	0.01	0.02	0.01	0.01	0.01	0.02
	Per Capita Water Resources	0.13	0.13	0.13	0.14	0.13	0.13	0.13	0.14	0.13	0.14
Economic Development	Per Capita GDP	0.13	0.13	0.13	0.13	0.14	0.15	0.14	0.14	0.14	0.13
	Water Use per Ten Thousand Yuan of Output	0.01	0.01	0.01	0.02	0.02	0.01	0.01	0.01	0.01	0.01
	Irrigation Water Use per Unit Area	0.01	0.01	0.01	0.01	0.01	0.01	0.02	0.02	0.02	0.01
Ecological and Environment	Total Water Resources	0.14	0.14	0.14	0.14	0.14	0.14	0.14	0.14	0.14	0.14
	Water Supply	0.13	0.13	0.13	0.14	0.14	0.14	0.13	0.14	0.14	0.14
	Precipitation	0.17	0.17	0.17	0.15	0.14	0.13	0.16	0.15	0.16	0.15
	Ecological Water Use Efficiency	0.13	0.13	0.14	0.14	0.14	0.14	0.13	0.13	0.13	0.13

. It can be seen that the differences between indicators are relatively small, indicating balanced development across various dimensions. Moreover, the stability of regional development is reflected in the unchanged weights of indicators from 2011 to 2020.

The weights of each indicator reflect their importance to the target layer. Specifically, precipitation (0.17 to 0.13) holds a higher weight in the evaluation of Yulin City’s WRCC, highlighting the region’s dependency on precipitation in arid areas. In contrast, indicators such as population density (0.02 to 0.01), water use per ten thousand yuan of output (0.01), and irrigation water use per unit area (0.02 to 0.01) have lower weights. These lower-weight indicators are primarily concentrated in the demographic and social and economic development systems, indicating that natural endowment conditions of water resources are relatively important in the assessment of regional WRCC.

3.2. Analysis of Spatio-Temporal Evolution of WRCC

To achieve the coupling of two evaluation methods without losing the accuracy of their evaluation results and to preserve their evaluation values as objectively and accurately as possible, this study couples the TOPSIS score and fuzzy comprehensive score through average weighting: ${\gamma }_i=(1-{\alpha }_i+{\mathsf{\beta }}_i)/2$. Here, ${\gamma }_i$ represents the comprehensive evaluation result of regional WRCC, with higher values indicating better regional water carrying capacity conditions.

3.2.1. Analysis of WRCC Development Trends

The evaluation of regional WRCC over multiple years can reflect the development status of the region. This study evaluated the WRCC from 2011 to 2020 and obtained the comprehensive evaluation results and grades of regional WRCC. The overall carrying capacity scores in the region range from [0.13 to 0.82]. This range is evenly divided into five grades, each corresponding to different degrees of development direction (

Table 4. Criteria for comprehensive evaluation levels of regional WRCC.

Level	Carrying Capacity	Index	Development Direction
I	Extremely poor	0.10~0.25	Restoration
II	Poor	0.25~0.40	Not suitable for development
III	Moderate	0.40~0.55	Maintain current development status
IV	Good	0.55~0.70	Suitable for moderate development
V	Excellent	0.70~0.85	Suitable for development

).

To analyze the WRCC in Yulin City from 2011 to 2020, we applied the fuzzy comprehensive evaluation model and TOPSIS model by integrating regional indicators with entropy weighting. This approach yielded respective scores for each year, culminating in an overall comprehensive evaluation score γ_i as shown in Figure 2. From 2011 to 2020, Yulin City’s WRCC improved from Level II to Level III, transitioning from an unsustainable development status to a sustainable maintenance status.

The region’s WRCC fluctuated around 0.35 (Level II) from 2011 to 2013, with a notable increase to 0.41 (Level III) in 2014. Subsequently, the overall WRCC remained stable around at 0.40 until 2020. Notably, from 2011 to 2014, the average annual water resource total in Yulin City was 182 million m³, increasing to 200 million m³ from 2014 to 2020. Concurrently, the regional ecological environment water usage rate rose from 0.47 to 1.40, reflecting improvements in water resource conditions, increased water conservation, and greater governmental investments in ecological environments. A primary factor contributing to these improvements could be attributed to Yulin City’s strict implementation of water resource management regulations in 2014, enhancing water resource management practices. This underscores the effectiveness of urban water resource management policies in improving WRCC, corroborated by significant outcomes observed post-policy implementation. The peak carrying capacity in 2017 (0.44) may be attributed to higher precipitation that year, demonstrating the critical role of rainfall in enhancing WRCC in arid regions [7].

3.2.2. Analysis of Temporal Evolution of WRCC

By inputting index values of various districts and counties into the evaluation model with the entropy weight, the WRCC of each district and county from 2011 to 2020 was derived (Figure 3). It can be observed that the differences between the results of the fuzzy comprehensive evaluation model and TOPSIS evaluation are minor, validating the applicability of both methods and enabling mutual calibration between the models.

During the period from 2011 to 2020, different development trends were observed among various districts and counties. Yuyang, Shenmu, Fugu, Suide, Jiaxian, Qingjian, and Zizhou remained relatively stable. Jingbian, Dingbian, and Hengshan showed significant increases in carrying capacity scores, with Dingbian exhibiting the highest growth rate (10.20%). This growth was mainly attributed to increased precipitation after 2014, leading to an increase in total water resources, coupled with improvements in agricultural water use efficiency due to the development of water-saving agriculture. Hengshan District transitioned from a state of undevelopability to one capable of sustaining current development (0.26 to 0.41), influenced by regional urbanization progress and the optimization of industrial water use patterns. Mizhi and Jiaxian experienced an increase in irrigation water use per unit area, leading to a decline in agricultural water use efficiency and a deterioration in regional WRCC.

Based on Table 2, the classification of WRCC for each district and county (Figure 4) shows that in 2011, only three counties including Yuyang had a carrying capacity level reaching Grade 3 or above. By 2020, however, more than half of the counties had achieved Grade 3 or above. This indicates a significant trend in the development of regional WRCC. However, future development will still be constrained by the current total water resources available.

3.2.3. Spatial Pattern Analysis of Water Resources Carrying Capacity

To explore the regional distribution characteristics of water resources carrying capacity (WRCC), Yulin City was divided into two parts based on industrial homogeneity: the northern six counties (including Yuyang, Shenmu, Hengshan, Dingbian, Jingbian, and Fugu) and the southern six counties (comprising Suide, Mizhi, Jiaxian, Wubu, Qingjian, and Zizhou [28]). The WRCC evaluation results (Figure 5d) exhibit significant spatial disparities. Yuyang (0.67), Shenmu (0.75), and Fugu (0.55) perform relatively well, closely linked to their economic development and industrial layout. Conversely, Qingjian (0.27) and Suide (0.27) show lower levels of carrying capacity, similarly influenced by economic development and urbanization speed. Zizhou (0.24) demonstrates the lowest capacity, facing significant inadequacies in ecological water use and a much lower urbanization rate (40.74%) compared to the 2020 average for Chinese counties (63.89%). These factors may negatively impact the evaluation of regional water resources carrying capacity, contributing to their poorer performance.

The spatial pattern of WRCC has undergone significant changes over time. In 2011 (Figure 5a), the WRCC in the northern six counties exhibited a northeast high and southwest low distribution pattern, spanning levels 1 to 5. The southern six counties showed a semi-circular distribution with high values in the southwest and low values in the center, spanning levels 1 to 2, indicating relatively minor spatial differences compared to the northern cities. By 2015 (Figure 5b), the spatial distribution pattern in the northern six counties remained stable, but with reduced spatial disparities in levels, indicating a trend towards equilibrium. Concurrently, WRCC distribution in the southern six counties evolved into a pattern with higher levels at the north and south edges and lower levels in the central areas, with unchanged level spans. In 2020 (Figure 5d), the northern six counties exhibited reduced spatial disparities in WRCC levels, with spans ranging from moderate to high (levels 3 to 4). The southern six counties showed a new pattern with lower levels in the north and higher levels in the south, with increased spatial disparities in level spans ranging from level 1 to 3, with some areas showing improved WRCC levels.

From 2011 to 2020, Yulin City’s overall WRCC spatially demonstrated higher levels in the northern six counties compared to the southern six counties. In terms of the average WRCC values (Figure 5d), the highest point in this area was in Shenmu in the northeast, achieving a level IV evaluation (0.75), while the lowest point was in Zizhou in the south, at only level I (0.24). WRCC deeply reflects the comprehensive performance of various indicators, where the northern six counties excel in per capita GDP, the urbanization rate, water consumption per unit output value, and water supply volume compared to the southern six counties, primarily due to their more developed economic foundation and industrial structure. However, the southern six counties exhibit relatively better water resource conditions, yet the effective conversion of resources into developmental impetus is hindered by limitations in economic development and industrial structure. Moreover, Yulin City, as a typical arid area energy city, heavily relies on fossil resources such as oil for regional economic development. The current industrial structure layout closely correlates with the distribution of energy regions, contributing to the uneven spatial development of WRCC. Particularly in the northern regions, the highly homogeneous economic structure and intense competition pose potential future risks to regions that originally exhibited good carrying capacity performance.

3.3. Regional Drivers of Water Resources Carrying Capacity

3.3.1. Significance Analysis of Driving Factors

Using Geodetector, the explanatory power of various factors on regional water resources carrying capacity (WRCC) was calculated (Figure 6). The average explanatory power of WRCC is as follows: the demographic social system (0.70~0.72~0.73) shows a steady increase over time; the economic development system (0.35~0.50~0.41) initially rises and then declines; the ecological and environmental system (0.72~0.70~0.55) declines annually. Over time, the influence of the demographic and social system on WRCC was gradually strengthened, while the influence of the ecological environment system showed a weakening trend. Overall, the explanatory power of all factors on Yulin City’s WRCC mainly decreased. Moreover, each system exhibits factors with strong explanatory power for WRCC, and these factors show significant differences across different systems.

Specifically, in 2011, the strongest explanatory factors were the ecological water consumption rate (a10), total water resources (a7), and per capita water resources (a3), focusing on water resource endowment conditions. This reflects a period when Yulin City’s overall economic development was just beginning and the energy industries were in a rising phase. Water resources were sufficient to support social development, leading to relatively balanced regional economic and social spaces with minimal urban development disparities. Thus, water resources carrying capacity was mainly influenced by water resource quantity and ecological water use. By 2015, the main driving factors for regional water resources carrying capacity shifted to water supply volume (a8), per capita GDP (a4), and the urbanization rate (a1). This stage indicates that sustainable development and the utilization of water resources were primarily influenced by regional economic development. Additionally, other non-dominant factors within the economic development system significantly enhanced their explanatory power for regional WRCC. In 2020, the strongest explanatory factors were water supply volume (a8), total water resources (a7), and per capita GDP (a4). This indicates that with accelerated economic development and urbanization processes, water resource management constraints and widening economic disparities have significantly impacted WRCC. Particularly noteworthy is the increasing influence of China’s strictest water resource management system and related water extraction management regulations on regional WRCC. It is worth noting that the contribution of the economic system to the carrying capacity of water resources is similarly emphasized in the study on the carrying capacity of water resources in Guangdong Province [22]. The findings all point to the centrality of economic factors in the evaluation of water resources carrying capacity. This commonality further confirms the far-reaching impacts of economic development on water use and the possible universal significance of economic factors in different geographical and social contexts.

From a spatial perspective, the water resources carrying capacity (WRCC) in the northern and southern regions is influenced by different driving factors. As shown in

Table 5. Explanatory power of drivers for WRCC in northern and southern six counties.

Units	Years	a1	a2	a3	a4	a5	a6	a7	a8	a9	a10
North 6 counties	2011	0.92	0	0.54	0.98	0.52	0.95	0.79	0.83	0.81	0.98
	2015	0.55	0.47	0.73	0.94	0.66	0.10	0.90	0.90	0.08	0.52
	2020	0.66	0.02	0.97	0.45	0.42	0.81	0.95	0.32	0.01	0.96
South 6 counties	2011	0.46	0.32	0.43	0.15	0.97	0.41	0.08	0	0.64	0
	2015	0.83	0.14	0.44	0.66	0.80	0.50	0.20	0.17	0.43	0.78
	2020	0.45	0.38	0.02	0.27	0.92	0.06	0.27	0.02	0.92	0.31

, the demographic social system’s role in driving WRCC in the northern six counties gradually weakened, while the ecological and environmental system showed a trend of gradual enhancement, and the economic development system experienced a process of decline from strong to weak. This pattern reveals that as the northern six counties undergo rapid economic development, the government has gradually recognized the importance of ecological environment protection and increased relevant investments, making the ecological environment system’s impact on WRCC increasingly significant.

In contrast, the economic development system in the southern six counties maintained a high driving role in WRCC from 2011 to 2020, while the driving force of the ecological and environmental system showed a trend of decline. Although the demographic and social system overall showed a slight decline, it is noteworthy that the urbanization rate factor (0.83) exhibited a strong driving effect in 2020. This indicates that the water resources carrying capacity in the southern six counties continues to face challenges under economic development pressure, and issues of development differentiation still exist. Under the unified government-led water resources management policies, the spatial balance has improved to some extent, but the direct impact of this change on WRCC remains relatively limited. The different internal driving factors between the northern and southern parts reflect the different driving effects on WRCC under different development paths. Moreover, the explanatory power of the dominant factors in both spatial and temporal dimensions has relatively decreased, indicating that the relationships between factors across systems have become more complex.

3.3.2. Analysis of Interaction Effects of Driving Factors

Using Geodetector to analyze the interaction of water resources carrying capacity (WRCC) indicators in Yulin City in 2011, 2015, and 2020, we explored the interaction effects of various driving factors, as shown in Figure 7. In most cases, the combined explanatory power (q-value) of two factors exceeds that of a single factor. Taking the ecological environment water use rate in 2020 as an example, the explanatory power of a single factor is only 0.08, while the average explanatory power of the interaction of two factors reaches as high as 0.84, with the highest value reaching 0.99 and the lowest value at 0.25, which is significantly higher than that of a single factor.

From a temporal perspective, in 2011, the interaction explanatory power of irrigated water consumption per unit area (a6) and precipitation (a9) in Yulin City was 1.00, indicating that factors with weak explanatory power as single factors may exert stronger explanatory power through synergy. Similarly, in 2020, the interaction explanatory power of water consumption per unit GDP (a5) with population density (a2) and water consumption per unit GDP (a5) with precipitation (a9) were both 1.00. Despite the single-factor explanatory power of water consumption per unit GDP being only 0.21, its interaction with other factors significantly enhanced its explanatory power. Additionally, we observed that the comprehensive average explanatory power of interaction factors fluctuated across the three years, rising from 0.91 to 0.95 and then dropping back to 0.91. This change reflects how arid regions dependent on energy benefit from resource-driven economic development but are also constrained by water resources during their development process.

From a spatial perspective, both northern and southern regions exhibit consistent enhancement effects after factor interaction. However, the combinations of enhanced interactive factors also show significant regional differences, with notably greater enhancement in the northern six counties compared to the southern six counties. As seen in Figure 8, the effects of various factors within the population and social system of the northern six counties are all strengthened after the interaction, and the ecological environment water use rate (a10) also displays a similar trend. This phenomenon can be explained in conjunction with previous studies, indicating that the northern six counties are currently transitioning from a phase of rapid economic development to a phase emphasizing high-quality development focusing on ecological environmental quality. On the other hand, as shown in Figure 9, the interaction effects of water consumption per unit GDP in the southern six counties from 2011 to 2020 consistently exhibit very high levels. Additionally, interactions between population density and precipitation, as well as precipitation and irrigated water consumption per unit area, also demonstrate a significant strengthening trend. This indicates that the southern six counties are maintaining continuous regional economic growth by increasing resource consumption, currently facing challenges of sustainable development due to uneven resource distribution.

4. Discussion

4.1. Regional WRCC Development Differences

A comparative analysis of the WRCC evaluations of the best-performing Shenmu County-level City and the worst-performing Zizhou County (Figure 10) highlights the disparities in their WRCC development. From 2011 to 2020, Shenmu’s average WRCC score was 0.75, achieving a V-level rating, indicating its potential to support sustainable water resource development and utilization in the future. The linear fit result for Shenmu shows a positive slope, suggesting an overall improving trend in its WRCC. The 95% confidence and prediction intervals stabilize between 0.55 and 0.90, demonstrating that its WRCC remains at a good level or higher.

Geographically, Shenmu is located in the northern six counties, serving as the core area of the national northern Shaanxi energy and chemical industry base. It is not only a leading economic area in Shaanxi Province but also holds a significant position in the western country economies of China. In the analysis of driving factors, we found that the ecological environment water use rate and precipitation are the strongest single-element drivers. In the interaction of factors (Figure 11), combinations such as the urbanization rate and total water resources, per capita GDP and precipitation, and per capita GDP and the ecological environment water use rate exhibit the strongest driving forces. In contrast, Zizhou County’s WRCC had an average score of 0.24 from 2011 to 2020, with a rating of only I-level, indicating that its WRCC urgently needs restoration and cannot continue to support development activities. The linear fit results for Zizhou County show a negative slope, revealing a declining trend in its WRCC. The 95% confidence and prediction intervals range from 0.10 to 0.40, indicating that its WRCC is far below the medium standard.

Located in the southern six counties, Zizhou County is an ecological function county and an agricultural development benchmark. The driving factors for its WRCC differ significantly from those of Shenmu City. In the single-element analysis, precipitation and urbanization rates show the strongest driving effects. In the interaction of factors, combinations of precipitation and per capita GDP, water consumption per unit area of irrigation, total water resources, and water supply form the strongest driving forces.

A common factor between the two regions is the strong driving force of precipitation, primarily due to the sensitivity of these arid areas to rainfall, aligning with the overall detection results of Yulin City. The arid zone is highly sensitive to precipitation, as it directly affects water supply and is critical for all water-demanding sectors such as agriculture and industry. In Zizhou County, as an agricultural area, precipitation is directly related to crop growth and agricultural harvests. Shenmu City, on the other hand, is highly urbanized but its economy is dependent on the fossil energy industry. Precipitation directly affects the cost of industrial water, and the high water consumption characteristics of the industry make the impact significant. At the same time, as the ecological environment has a greater impact on the carrying capacity of water resources, precipitation is indispensable in maintaining the ecological balance of the arid zone and is an important driver for the economy and ecology of Shenmu City. However, their differences are also evident. Shenmu City’s primary driving factors focus on the economic development system, particularly GDP-related indicators, where socio-economic development has boosted the WRCC. With economic prosperity, there is also an increased emphasis on ecological civilization construction, with government and enterprises committed to industrial transformation and clean production, positively supporting the improvement of WRCC.

On the other hand, Zizhou County is more influenced by ecological environment system factors. The agricultural development needs underscore the importance of ensuring a healthy ecological environment. However, due to its relatively low productivity, it struggles to support large-scale economic development, and its current WRCC is insufficient to meet the needs of social development. Therefore, accelerating industrial construction and achieving agricultural modernization may be effective ways to enhance Zizhou County’s WRCC.

4.2. Impact of Management Policies on Regional WRCC

This study thoroughly investigates the WRCC (WRCC) of Yulin City from 2011 to 2020. By constructing a mathematical model, the research systematically evaluates regional WRCC across three dimensions: population and society, economic development, and the ecological environment. A detailed analysis of the driving effects of various factors on the spatiotemporal distribution of WRCC has been conducted. The results of this evaluation aim to provide robust data support and strategic recommendations for relevant decision-makers and managers.

Based on the findings of this research, it is recommended that water resource management departments focus on regional water resource conditions. Building water diversion projects and other hydraulic infrastructure can achieve spatial equilibrium in water resource distribution. For areas with severely inadequate or weak WRCC, rapid adjustments in industrial structure are necessary to promote healthy and stable regional economic development. This should be performed while ensuring that the ecological environment remains unharmed and by fostering the development of industries with local characteristics. Conversely, regions currently exhibiting excellent WRCC should focus on improving water resource utilization efficiency, strictly controlling total pollutant discharge and water usage, and continuously strengthening ecological environment protection. Additionally, accelerating innovation in industrial structure is crucial to ensure that economic sustainable development is promoted while steadily enhancing WRCC, thereby fostering a virtuous cycle of regional sustainable development.

The Yulin city government should be sensitive to the dynamics of precipitation, which is a strong driving indicator, and plan and implement water allocation strategies based on these dynamics. Given the natural challenges of precipitation scarcity in arid regions, the government should actively promote and support water-saving technologies at the policy level. In agriculture, the focus should be on promoting drought-tolerant crop varieties and adjusting irrigation patterns to be more precise and efficient, reducing the high dependence of agriculture on precipitation fluctuations. Meanwhile, for industrial areas, the government should introduce incentives to promote water recycling across industries, agriculture, and urban sectors, which will not only help alleviate the pressure on water resources but is also crucial for achieving sustainable development. Furthermore, optimizing the city’s water supply network and upgrading drainage facilities will ensure that when precipitation fluctuates, the stability of the water supply system and the efficiency of the drainage system are both improved, thus maximizing water recycling efficiency. This series of initiatives aims to address and fulfill the strict management of regional water resources, including total volume control, water use efficiency improvement, and water functional area pollution limitation.

Yulin City has successfully transformed from an agricultural region to an industrial powerhouse and from a poverty-stricken area to a prosperous one, largely through the development and utilization of fossil resources. Various greening and restoration measures have also achieved significant success [29]. However, Yulin City faces future sustainable development challenges, such as regional development imbalances, ecological environment degradation, and the urgency of industrial transformation. In planning for the future, the Yulin City government must increase investment in ecological environment protection, improve water use efficiency, and enhance regional WRCC through water diversion projects and water-saving measures to support sustainable social development.

4.3. Strengths and Limitations

This study focuses on Yulin City in Shaanxi Province, China, an area characterized by significant north–south differences. As a representative of typical arid regions and energy cities, Yulin’s need for socio-economic development and ecological environmental protection is particularly urgent. Currently, research on WRCC (WRCC) from a multi-system perspective at a small scale to guide future societal development is insufficient. Therefore, the results of this study are expected to provide valuable contributions to WRCC research in similar regions.

In terms of the research methodology, existing studies on WRCC often rely on a single method for spatial difference analysis and seldom address spatial correlation. To address this, our study integrates the fuzzy comprehensive evaluation method and the TOPSIS method for a comprehensive assessment. This approach effectively avoids the potential errors associated with a single method and enhances the rationality and scientific validity of the evaluation results. Additionally, this study employs Geodetector to deeply explore the main driving factors of WRCC. By comparing the results of the obstacle degree model’s driving factor detection analysis (using the year 2020 as an example, as shown in

Table 6. Obstacle degree driving factor analysis results.

Areas	a1	a2	a3	a4	a5	a6	a7	a8	a9	a10
Yuyang	0.081	0.082	0.082	0.082	0.083	0.084	0.080	0.079	0.085	0.084
Shenmu	0.081	0.081	0.079	0.078	0.082	0.084	0.080	0.082	0.085	0.083
Fugu	0.082	0.084	0.084	0.083	0.082	0.084	0.084	0.084	0.083	0.080
Hengshan	0.083	0.081	0.083	0.083	0.084	0.084	0.084	0.083	0.084	0.084
Jingbian	0.082	0.082	0.083	0.083	0.083	0.082	0.083	0.083	0.083	0.085
Dingbian	0.084	0.081	0.082	0.084	0.087	0.081	0.082	0.081	0.085	0.084
Suide	0.084	0.086	0.085	0.085	0.083	0.086	0.085	0.085	0.082	0.084
Mizhi	0.084	0.086	0.085	0.085	0.084	0.084	0.085	0.085	0.084	0.083
Jiaxian	0.086	0.084	0.085	0.085	0.083	0.084	0.085	0.085	0.084	0.084
Wubu	0.083	0.086	0.084	0.084	0.082	0.082	0.085	0.085	0.082	0.079
Qingjian	0.085	0.083	0.084	0.084	0.083	0.082	0.085	0.085	0.080	0.084
Zizhou	0.085	0.084	0.084	0.085	0.084	0.082	0.084	0.085	0.082	0.085

), the Geodetector results are more distinctive and can significantly reflect the differentiation of regional WRCC driving factors. This achieves a systematic evaluation analysis from county to north–south to the entire city, forming a “point-line-surface” pattern. Furthermore, this study improves the affiliation function in the fuzzy comprehensive evaluation to address issues related to linear changes leading to transgressive phenomena and the subjectivity arising from interval setting.

Despite conducting a reasonable and objective evaluation analysis of Yulin’s WRCC and its driving factors from 2011 to 2020, some limitations remain. In terms of indicator selection, although the study fully considered regional conditions and referred to previous research, it was constrained by the available data. Future studies could increase the number of indicators to provide a more comprehensive evaluation of regional WRCC. This study primarily focuses on the current and past WRCC evaluation and lacks an in-depth exploration of the future evolution of WRCC. Therefore, future research is recommended to use machine learning and data-driven predictive models to study the evolution of future WRCC and to set up scenario simulations to explore the impact of different factors on future WRCC changes. Moreover, this study focuses on a relatively small research scale. Future studies could consider expanding the research scope to the Loess Plateau region or the middle and upper reaches of the Yellow River to comprehensively explore WRCC issues.

5. Conclusions

The main conclusions of this study are as follows:

• The weights of various evaluation indicators for WRCC show that the natural endowment of water resources is more critical in the region. Precipitation has a higher weight, while socio-economic indicators such as population density, water use per unit of output value, and irrigation water use per unit area have lower weights. This suggests relatively balanced development in water resources carrying capacity in the region, albeit with a strong dependence on precipitation.
• From 2011 to 2020, Yulin City transitioned significantly from an undeveloped state to a sustainable development state in terms of WRCC, increasing from 0.35 to 0.40. The improvement in WRCC since 2014 was significantly influenced by relevant water resource management policies issued by the Yulin City government. Spatial differentiation is more pronounced than temporal differences, which is particularly evident in the north–south divide within the region.
• Over time, the main driving factors of regional WRCC have been continuously changing, reflecting the multifaceted influences on the WRCC system, which can lead to fluctuations. Currently, the primary driving factors in Yulin City include water supply, total water resources, and per capita GDP, indicating that development in arid areas is primarily constrained by water resources. The phenomenon of weakening driving factors also suggests progress toward spatial equilibrium in the region. Moreover, significant regional disparities exist within Yulin City, providing essential bases for subsequent policy formulation.