Prediction and Early Warning of Water Environmental Carrying Capacity Based on Kernel Density Estimation Method and Markov Chain Model

Abstract

Water environmental carrying capacity (WECC) is an important support for social and economic development and is closely related to regional production and consumption patterns. Exploring the level of WECC and its evolution trend is very urgent for the scientific formulation of targeted early warning control strategies. Therefore, this study first constructs the index system of WECC with a DPSIR model, and conducts the quantitative evaluation by combining the Kantiray Weighting method and the TOPSIS method. Then, the Kernel Density Estimation method and the Markov Chain model are applied to explore the spatiotemporal variation characteristics of WECC and predict its evolution trend. Finally, a case study of 17 municipal administrative regions in Hubei Province is carried out. The main findings are as follows: (1) The WECC status in Hubei Province during 2013–2022 was generally satisfactory and showed a trend of fluctuating improvement. (2) The spatial agglomeration effect of WECC in Hubei Province was significant, showing a distribution pattern of “high-high” agglomeration and “low-low” agglomeration. The improvement of the WECC in eastern Hubei was obvious, while that in central Hubei was slower, and the cities with a lower level of WECC had a more significant improvement effect. (3) Overall, the WECC of cities in Hubei Province tends to shift to a higher level. In a short period of time, the grade improvement of urban WECC in Hubei Province is more likely to occur between adjacent grades. With the increase in time span, the probability of this transition rises gradually. This study has proposed a set of methods for the evaluation and prediction of WECC status, which can provide important decision-making guidance for the early warning and regulation of regional differentiated WECC.

1. Introduction

The water environment system plays an irreplaceable role in maintaining regional ecological balance and safeguarding human social and economic activities. With the increase in the intensity and scope of human activities, the pressure on the natural water system continues increasing, resulting in more prominent problems of water resources shortage, water environment pollution and water ecological damage, and water sustainability is facing more severe challenges [1]. According to the United Nations World Water Development Report 2024, roughly half of the world’s population currently experiences severe water scarcity for at least part of the year and one quarter of the world’s population face extremely high levels of water stress [2]. More than three billion people are expected to face water shortages by 2050 [3]. Ecological and environmental problems are fundamentally caused by unsustainable patterns of human production and consumption. Due to the unreasonable industrial development model, significant problems such as excessive consumption of water resources [4], deterioration of water ecology [5] and destruction of the water environment [6] have emerged across many parts of China. The main crisis of water security in China has changed from an insufficient supply to excessive environmental carrying capacity, and some regions or river basins are facing the risk of overloading or serious overloading of the water environmental carrying capacity (WECC) [7,8]. To address these issues, China is actively promoting water environment management [9,10]. Therefore, the research on WECC is in line with both the current situation of water security and the development direction of water ecological civilization.

WECC is the application of the concept of environmental carrying capacity in the field of water systems research. Current WECC research mainly involves concept definition, evaluation, prediction and early warning. At present, the concept of WECC has not reached a consensus, and its specific connotation remains to be further clarified. Many studies have confused WECC with the concepts of water environmental capacity [11,12,13], water resources carrying capacity [14] and water ecological carrying capacity [15]. Overall, there are significant differences in emphasis between the four concepts mentioned above. WECC emphasizes the integrated nature of the water system itself [16], while the other three concepts focus on carrying capacity in terms of water quantity [17], water quality [18] and water ecology [19], respectively. In terms of WECC evaluation research, most of the existing international research focuses on water sustainability [20,21,22] and takes the carrying capacity as the standard to measure the environmental security of water resources [23,24]. Few studies directly carried out the evaluation of WECC. For example, Anamaghi et al. [21] built an evaluation model of WECC based on the concept of uncertain toughness. Dadashi et al. [22] explored the environmental carrying capacity of a watershed based on the multi-criteria assessment method. Comparatively, Chinese scholars have carried out a larger number of evaluation studies on WECC, mainly concerned with its size, state and development and utilization potential [25]. As the specific value of WECC needs to be supported by highly accurate data, and the development and utilization potential evaluation of WECC needs to know both the size and the state of WECC, the actual research mainly focuses on the state evaluation of WECC considering the feasibility of actual research [26,27]. The research area of WECC involves different geographical scales such as river basins [8,28,29], lakes [30,31], provinces [26,27] and cities [32,33], and most of them exit in arid and semi-arid regions [34]. As for the research methods regarding WECC, it mainly includes the index evaluation method and complex system analysis method. Most of the index evaluation methods are based on the pressure and supporting force criteria of water environment, and construct the index system from the perspectives of economic society, resources and environment, etc. The commonly used methods include the Principal Component Analysis [32], the Entropy–TOPSIS [35], the Artificial Neural Network [36], etc. Without considering the internal structure and complex interactions of the system, the index evaluation method can reflect the relative state of WECC and is sufficient to guide the government’s macro-control policies [26,27]. In contrast, the complex system analysis methods, mainly including the System Dynamics (SD) model [37] and the Multi-objective Programming method [38], always require huge amounts of data as support and thus fail to be applied to the rapid evaluation of water system bearing capacity. This study focuses on analyzing the WECC state and its evolution law, that is, objectively evaluating the extent to which pressure from regional human activities on natural water systems exceeds WECC, and exploring its dynamic evolution. Therefore, considering the operability and accuracy of the research results, the index system method should be chosen to carry out the evaluation of WECC.

With the intensification of global climate change and human activities, the establishment of dynamic trend predictions and early warning mechanisms of WECC is of great significance to cope with the possible risk of water environment overloading in advance [39]. At present, there are relatively few early warning studies on WECC, which can be divided into short-term warning and long-term warning according to the early warning period. Short-term early warning research mostly adopts the Regression Prediction model [34,40], prosperity index [39], the Neural Network [41] and other, and the SD model [37,42] is usually used in long-term early warning studies and multi-scenario simulation. However, technical and methodological frameworks for WECC early warning remain underdeveloped, and systematic research is still lacking. In addition, current studies of WECC under deterministic design conditions have yielded statically determined results. Nevertheless, the structure and internal associations of the water system are complex, and the elements involved in each subsystem have stochastic uncertainties [43], which need to be analyzed from the perspective of dynamic uncertainty analysis. In summary, it is urgent and vital to carry out early warning research on the state of integrated WECC from the perspective of uncertainty analysis.

Accordingly, this paper takes 17 municipal administrative regions in Hubei Province of China as the research object and carries out three main studies in turn: (1) On the basis of clear definition of WECC, the DPSIR (Driving Force–Pressure–State–Impact–Response) theory was applied to construct its comprehensive evaluation index system. (2) Then, the state of WECC in Hubei Province was quantitatively assessed by combining the Kantiray Weighting method and the TOPSIS method. The kernel density estimation method was used to analyze the distribution location, distribution pattern, distribution ductility, and distribution polarization trend of WECC, in order to explore its dynamic changes. (3) The Markov Chain model was established to quantitatively study the uncertainty of the stochastic process of WECC, and the corresponding early warning was provided. In general, the research contributions of this paper are as follows: (1) An evaluation index system of WECC was constructed, which comprehensively considered the availability of water resources, the self-purification capacity of the water environment, the stability of the water ecology, and the sustainable development needs of society and economy. This provides a new perspective and method for the scientific evaluation of WECC. (2) The absolute difference in regional WECC was further discussed by applying the Kernel Density Estimation method, which better revealed the spatiotemporal variation characteristics of WECC. (3) The state prediction model of WECC based on the Markov Chain model is proposed, providing an important basis for the formulation of early warning regulation strategies.

The structure of this study is as follows. Section 1 introduces the research background and significance, literature review, research content and innovation points. Section 2 introduces the study area, methods, definition of WECC, index system and data sources. Section 3 presents the main findings of the study, and Section 4 discusses the important findings and summarizes the limitations of the study. Section 5 presents key conclusions and feasible policy recommendations.

2. Materials and Methods

2.1. Research Area

As shown in Figure 1, Hubei Province is situated in the middle of China, spanning a range of latitudes between 29°01′53″ N and 33°6′47″ N, and longitudes between 108°21′42″ E and 116°07′50″ E. Hubei Province administers 17 municipal administrative regions, including 12 prefecture-level cities, namely Wuhan, Huangshi, Shiyan, Yichang, Xiangyang, Ezhou, Jingmen, Xiaogan, Jingzhou, Huanggang, Xianning, and Suizhou; one autonomous prefecture, the Enshi Tujia and Miao Autonomous Prefecture (Enshi); and four county-level administrative units directly under the provincial government, which are Xiantao, Qianjiang, Tianmen, and the Shennongjia Forestry District (Shennongjia). Hubei Province encompasses an area of 185,900 square kilometers, representing 1.94% of the total land area of China. The total area of Hubei Province is constituted by mountains (56%), hills (24%), and plains and lakes (20%). With the longest history of the Yangtze River, Hubei Province is known as “the province of thousands of lakes.” It consists of three major waterways including the Yangtze River, the Han River and the Qing River. The three rivers connect the province’s Jianghan Lake Group and lake reservoir wetlands, providing excellent soil and water conditions for the Jianghan Plain. This supports Hubei Province to become the main grain producing area of China, an important water conservation area and an important ecological barrier in the Yangtze River Basin. In addition, Hubei Province has a good industrial base and is an important manufacturing base in China, with its economic strength leading the central part of China. In 2023, the GDP of Hubei Province was 558.036 billion CNY, ranking 7th in China. The resident population at the end of the year was 58.38 million, and the per capita GDP was 95,538 CNY, ranking 9th in China.

Figure 1. Map of Hubei Province, China.

Hubei Province is the core region of the Yangtze River Basin, and it is relatively rich in water resources, but its spatial and temporal distribution is uneven, and the pressure on the water environment is relatively significant [44]. In light of the accelerated socio-economic development, Hubei Province has witnessed an upward trend in overall water pollution discharge within the basin. Surface pollution has emerged as a more pronounced concern, accompanied by a heightened risk of sudden water environmental hazards and cumulative environmental pollution [45]. These include the eutrophication of water bodies and the occurrence of black-odorous water bodies in urban areas, among other issues. The implementation of the Yangtze River Protection Strategy alleviated the water environment pressure in the Yangtze River Basin to a certain extent [46]. However, it is evident that further strengthening of water environment management is necessary. Therefore, it is of great significance to objectively evaluate the current situation and evolutionary trend of the WECC level in Hubei Province, in order to implement the comprehensive management of the watershed and promote the high-quality development of Hubei Province [47].

2.2. Methods

2.2.1. Kantiray Weighting Method

The Kantiray Weighting method is an objective weighting method that utilizes the product of the correlation coefficient matrix $R$ and the standard deviation matrix $S$ for the calculation of indicator weights. The method combines the information content assignment with the relevance assignment. The greater the information content and the more relevant it is to the evaluation objective, the greater the weight of the indicator. The Kantiray Weighting method is more rigorous than other commonly used objective weighting methods that only consider information content or relevance [48]. The specific steps are as follows:

(1)

Standardized processing

In this paper, the $\max -\min$ method is used to standardize the indicator $y_{ij}$. To eliminate the impact of different quantities on the indicator, the expression of the standardized value $x_{ij}$ is shown below; from top to bottom is the standardized processing formula for positive and negative indicators.

$$x_{ij}=\left\{\begin{array}{c}{\displaystyle \frac{y_{ij}-\min y_{ij}}{\max y_{ij}-\min y_{ij}}}\times 0.99+0.01\\ {\displaystyle \frac{\max y_{ij}-y_{ij}}{\max y_{ij}-\min y_{ij}}}\times 0.99+0.01\end{array}\right.$$

(1)

where $y_{ij}$ is the original data value, $x_{ij}$ is the standardized indicator value, and $i$, $j$ denote the $i-th$ evaluation object and the $j-th$ evaluation indicator, respectively.

(2)

Calculating the matrix of correlation coefficients of the indicators $R$

$$r_{ik}={\displaystyle \frac{{\displaystyle \sum _{j=1}^n\left(x_{ij}-{\overline{x}}_i\right)}\left(x_{kj}-{\overline{x}}_k\right)}{\sqrt{{\displaystyle \sum _{j=1}^n{\left(x_{ij}-{\overline{x}}_i\right)}^2}}\sqrt{{\displaystyle \sum _{j=1}^n{\left(x_{kj}-{\overline{x}}_k\right)}^2}}}}$$

(2)

where $x_{ij}$, $x_{kj}$ respectively indicates that the indicator $i$, $k$ in the $j-th$ evaluation of the value of the object $\left(j=1,2,\cdots ,n\right)$; $m$ for the total number of evaluation indicators $n$ for the total number of evaluation of the object; $x_i$, $x_k$ respectively represents the standardized score of the indicator $i$, $k$ average; $r_{ik}$ denotes the correlation coefficient of the indicator $i$, $k$. The corresponding correlation coefficients can be included in the corresponding position of the matrix to obtain the correlation coefficient matrix $R$.

$$R=\left(\begin{array}{cccc}{r}_{11}& r_{12}& \cdots & r_{1m}\\ r_{21}& r_{22}& \cdots & r_{2m}\\ ⋮& ⋮& \ddots & ⋮\\ r_{m1}& r_{m2}& \cdots & r_{mm}\end{array}\right)$$

(3)

(3)

Calculating the standard deviation diagonal matrix $S$

$${\sigma }_i=\sqrt{{\displaystyle \frac{{\displaystyle \sum _{j=1}^n{\left(x_{ij}-\overline{x}\right)}^2}}{n-1}}}$$

(4)

where ${\sigma }_i$ is the standard deviation of the $i-th$ indicator. The standard deviation matrix $S$ is obtained by including the corresponding standard deviation in the corresponding position on the diagonal of the matrix.

$$S=\left(\begin{array}{cccc}{\sigma }_1& 0& \cdots & 0\\ 0& {\sigma }_2& \cdots & 0\\ ⋮& ⋮& \ddots & ⋮\\ 0& 0& \cdots & {\sigma }_m\end{array}\right)$$

(5)

(4)

Calculating the eigenvector $b$ corresponding to the largest eigenvalue $\lambda$ of the matrix $R\times S$. Set $b=\left(b_1,b_2,\cdots ,b_i,\cdots ,b_m\right)$, normalize the eigenvector values to obtain $b^{*}=\left(b_1^{*},b_2^{*},\cdots ,b_i^{*},\cdots ,b_m^{*}\right)$, and the weight $w_i$ of indicator $i$ is the absolute value of the $i-th$ component of $b^{*}$.

$$w_i=\left|b_i^{*}\right|=\left|{\displaystyle \frac{b_i}{\sqrt{{\displaystyle \sum _{i=1}^m{b}_i^2}}}}\right|$$

(6)

2.2.2. TOPSIS Model

TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) model is a model for decision evaluation of systems with multiple solutions and indicators, where the evaluation objects are ranked by constructing a positive ideal solution and a negative ideal solution, and the distance of each evaluation solution from the positive and negative ideal solutions is calculated. The comprehensive evaluation value is then calculated [49]. This method is logical and clear, easy and flexible to apply, and has been widely used in index synthesis evaluation studies [50,51]. The specific steps are as follows:

(1)

Constructing the weighting matrix

$$Z=\left(\begin{array}{cccc}{z}_{11}& z_{12}& \dots & z_{1n}\\ z_{21}& z_{22}& \dots & z_{2n}\\ ⋮& ⋮& \ddots & ⋮\\ z_{m1}& z_{m2}& \dots & z_{mn}\end{array}\right)=\left(\begin{array}{cccc}{x}_{11}{w}_1& x_{12}{w}_1& \dots & x_{1n}{w}_1\\ x_{21}{w}_2& x_{22}{w}_2& \dots & x_{2n}{w}_2\\ ⋮& ⋮& \ddots & ⋮\\ x_{m1}{w}_m& x_{m2}{w}_m& \dots & x_{mn}{w}_m\end{array}\right)$$

(7)

(2)

Determining the positive and negative ideal solutions.

$Z^{+}$ is the “positive ideal solution” of the evaluation system and $Z^{-}$ is the “negative ideal solution” of the evaluation system.

$$Z^{+}=\left\{\underset{1\le i\le m}{\max }{z}_{ij}|i=1,2,\cdots ,m\right\}=\left\{z_1^{+},z_2^{+},\cdots ,z_m^{+}\right\}$$

(8)

$$Z^{-}=\left\{\underset{1\le i\le m}{\min }{z}_{ij}|i=1,2,\cdots ,m\right\}=\left\{z_1^{-},z_2^{-},\cdots ,z_m^{-}\right\}$$

(9)

(3)

Calculating the distance

The distance between the indicators and the positive and negative ideal solutions is calculated by using the Euclidean Distance calculation method, as shown in Equations (10) and (11). Where $D_j^{+}$ is the distance between the $i-th$ indicator and the “positive ideal solution” $Z_i^{+}$, and $D_j^{-}$ is the distance between the $i-th$ indicator and the “negative ideal solution” $Z_i^{-}$.

$$D_j^{+}=\sqrt{{\displaystyle \sum _{i=1}^m{\left(z_i^{+}-z_{ij}\right)}^2}}$$

(10)

$$D_j^{-}=\sqrt{{\displaystyle \sum _{i=1}^m{\left(z_i^{-}-z_{ij}\right)}^2}}$$

(11)

(4)

Calculating comprehensive evaluation value $T_j$

$$T_j={\displaystyle \frac{D_j^{-}}{D_j^{+}+D_j^{-}}}$$

(12)

$T_j$ is the comprehensive assessment value of the WECC of the $j-th$ evaluation object in the range of $(0,1]$. The closer the integrated assessment value is to 1, the better the level of WECC; the further from 1, the worse.

2.2.3. Kernel Density Estimation Method

Kernel Density Estimation is a nonparametric method used to estimate the density function of a random variable, which has a weaker spatial dependence on the model and more stable results than the parametric methods. Thus, it is usually used to describe the phenomenon of uneven distribution of the study population [52,53]. In this paper, the frequently used Gaussian kernel function is used as the kernel function for estimation with the following expression.

$$\left\{\begin{array}{c}f\left(x\right)={\displaystyle \frac{1}{n}}{\displaystyle \sum _{i=1}^n{K}_h\left(x-x_i\right)}={\displaystyle \frac{1}{nh}}{\displaystyle \sum _{i=1}^{n}K\left(\frac{x-x_i}{h}\right)}\\ K\left(x\right)={\displaystyle \frac{1}{\sqrt{2\pi }}}\exp \left(-{\displaystyle \frac{x^2}{2}}\right)\end{array}\right.$$

(13)

In this context, $n$ refers to the number of observations. Similarly, $x$ represents the value of observation, $x_i$ denotes the mean value of all observations, $K\left(x\right)$ is the kernel density function, and $h$ is the bandwidth. A larger bandwidth results in a smoother density function and a lower level of precision in the estimate. Conversely, a smaller bandwidth produces a less smooth density function and a higher level of precision in the estimate. Therefore, it is preferable to use a smaller bandwidth when estimating the actual density.

This paper analyzes the results of the Kernel Density Estimation method primarily through the following aspects. The distribution location reflects the high and low levels of WECC. The distribution pattern is used to judge the spatial difference and polarization of WECC levels, while the height and width of the curve peaks together reflect the size of the spatial difference, and the number of crests reflects the degree of distribution polarization. The distribution extension is used to depict the size of the spatial difference between the city with the highest level of WECC in Hubei Province and the other cities in the interior of the province. The longer the trailing to the right, the greater the difference between the cities.

2.2.4. Markov Chain Model

The Markov Chain model uses a transition probability matrix to predict how individual regional phenomena change over time. That is to say, dividing continuous discrete values into multiple types to reflect the distribution and evolution of different types of time and state. With the advantages of being easy to implement, it can be used to predict by focusing only on the current state and state transfer probabilities [54]. Assume that the stochastic process $\left\{x_t,t\in T\right\}$, any period $T$ and any states $j$, $i$, and $i_n\left(n=0,1,\cdots ,t-2\right)$, satisfy Equation (14). The probability distribution of $X$ for state $j$ in period $t$ depends only on state $i_{t-1}$ in period $t-1$ and is independent of state $\left(i_0,i_1\cdots ,i_{t-2}\right)$ in period $\left(0,1,\cdots ,t-2\right)$.

$$P\left\{X_{\left(t\right)}=j\left|X_{\left(t-1\right)}=i,X_{\left(t-2\right)}=i_{t-2}\right.,\cdots ,X_{\left(0\right)}=i_0\right\}=P\left\{X_{\left(t\right)}=j\left|X_{\left(t-1\right)}=i\right.\right\}=P_{ij}$$

(14)

$P_{ij}$ denotes the probability that the WECC level of a city during the sample examination period is shifted from state $i$ in period $t-1$ to state $j$ in period $t$; $n_{ij}$ represents the number of times the WECC level is shifted from the state $i$ to the state $j$ occurs; and $n_i$ is the total number of times the state $i$ of the WECC level occurs. $P_{ij}$ can be obtained using the great likelihood estimation method with the following formula:

$$P_{ij}={\displaystyle \frac{n_{ij}}{n_i}}$$

(15)

2.3. Evaluation Index System Construction

WECC refers to the threshold of the intensity of human social activities that a natural water system can carry with minimum water resources consumption and water pollution discharge under a specific level of economic development, social governance and technological innovation [26,55]. It includes not only the quantitative and qualitative requirements of water resources but also involves the health of the water ecosystem and the protection of biodiversity. The evaluation of WECC requires comprehensive consideration of the availability of water resources, the self-purification capacity of the water environment, the stability of the water ecosystem and the needs of sustainable socio-economic development. Sustainable development can be used as the guiding ideology and theoretical basis for WECC research, while WECC research is the concrete embodiment and application of sustainable development theory in the field of water system management. This study refers to the DPSIR model to construct an evaluation index system of the WECC (Figure 2).

Figure 2. Theoretical framework based on DPSIR model.

The DPSIR model was constructed in the 1990s by the European Environmental Agency (EEA). It is developed from the “Driving Force–State–Response” (DSR) model of the Organization of Economic Corporation and Development (OECD) and the United Nations Commission on Sustainable Development (UNCSD). The DPSIR model is easy to communicate, and expresses the relationship between economic, social and environmental factors in an integrated and comprehensive cause-and-effect way [56]. It has been widely used in the study of human and environmental issues, including water resource management, integrated ocean management, biodiversity, climate change, agricultural activities, land use, and so on. “Driving Force” in the model reflects social, demographic, economic and other activities to meet human needs, which are potential contributors to changes in the water environment. “Pressure” refers to human behaviors that are intentionally or unintentionally applied to the water environment as a result of the drive and are direct contributors to environmental change. “State” refers to pressure-induced changes in ecosystems that are often negative and undesired. Some pressures cause change directly, and some pressures require ecosystems to change indirectly through a series of natural processes. “State” can be characterized by physical, chemical, or biological variables. “Impact” captures the benefits and ultimate well-being that humans derive as a result of changes in ecosystem quality and functioning, including provisioning, regulating, and cultural services provided by ecosystem services, as well as impacts that reflect well-being, such as health, education, income, disease, and happiness. “Response” represents the solutions that societies and governments make to stop, compensate for, and mitigate water environment change. These solutions can act on four dimensions: “Driving Force”, “Pressure”, “State” and “Impact.” In DPSIR, the most important thing is how to respond and maintain sustainable development through decision-making. Response behavior, on the other hand, should be based on the problems existing in the development of the system and make timely responses to ensure the efficient, rapid and coordinated operation of the human-land system.

According to the principle of selecting indicators with representativeness and accessibility, synthesizing the existing research results [33,57] and combining the connotation of WECC in the previous section, the evaluation index system of WECC is determined, as shown in Table 1. The weights of the variables are determined using the Kantiray Weighting method.

Table 1. Evaluation index system of WECC.

Target Layer	Standard Layer	Indicator Layer	Unit	Subsystem	Attribute	Weight
WECC	Driving Force (0.2016)	Urbanization rate (D1)	%	Social economy	+	0.0671
		GDP per capita (D2)	CNY/person	Social economy	+	0.0659
		GDP growth rate (D3)	%	Social economy	+	0.0112
		Population density (D4)	person/km2	Social economy	−	0.0574
	Pressure (0.2158)	COD emissions (P1)	Ton	Water environment	−	0.0692
		Ammonia nitrogen emissions (P2)	ton	Water environment	−	0.0699
		Per capital domestic water consumption (P3)	104 m3/person	Water resources	−	0.0177
		Industrial water consumption (P4)	108 m3	Water resources	−	0.0457
		Ecological water consumption (P5)	108 m3	Water ecology	−	0.0133
	State (0.2180)	Water yield modulus (S1)	104 m3/km2	Water resources	+	0.0724
		Water resources per capita (S2)	m3/person	Water resources	+	0.0256
		Annual precipitation (S3)	mm	Water resources	+	0.0721
		Total year-end storage capacity of large and medium-sized reservoirs (S4)	108 m3	Water resources	+	0.0275
		Soil erosion area (S5)	km2	Water ecology	−	0.0204
	Impact (0.1805)	Rate of good water quality at cross-sections (I1)	%	Water environment	+	0.0876
		water consumption per unit of industrial added value (I2)	m3/104 CNY	Water resources	−	0.0613
		Parkland area per capita (I3)	m2	Water ecology	+	0.0291
		Green coverage rate of built-up area (I4)	%	Water ecology	+	0.0025
	Response (0.1841)	Sewage treatment rate (R1)	%	Water environment	+	0.0364
		Percentage of investment in water and environmental utilities (R2)	%	Water environment	+	0.0686
		Share of tertiary sector (R3)	%	Social economy	+	0.0791

2.4. Data Sources

The data covered in this study are obtained from authoritative public databases. D₁, D₄, R₂, I₃, I₄, R₁ are obtained from the Statistical Yearbook of Hubei Province and its municipal administrative regions (2014–2023). Values of indicators such as D₂, D₃, R₃, etc. are taken from the Statistical Bulletin of National Economic and Social Development (2013–2022). P₃, P₄, P₅, S₁, S₂, S₃, S₄, I₂ and other indicators are derived from the Water Resources Bulletin for Hubei Province and its municipal administrative regions (2013–2022). Values for P₁, P₂, and I₁ are obtained from the Bulletin of Environmental Conditions of Hubei Province and its municipal administrative regions (2013–2022). S₅ is derived from the Soil and Water Conservation Bulletin (2013–2022).

3. Results

3.1. Evaluation of the WECC

The indicator weights calculated by the Kantiray Weighting method are shown in Table 1. Among these 21 indicators, the top three factors that have the greatest impact on the WECC of Hubei Province are, respectively, I₁ (0.0876), R₃ (0.0791) and S₁ (0.0724). They belong to the water environment subsystem, the socio-economic subsystem and the water resources subsystem, respectively. In terms of the standard layer, the sum of the weights of the indicators in the state standard layer (0.2180) and the pressure standard layer (0.2158) ranked in the top two. This indicates that the level of regional WECC depends to a certain extent on the current state of the water environment, and that human production and living activities have a greater impact on WECC. The response standard layer (0.1841) has relatively low weights, which to some extent indicates that the overall comprehensive management of the water environment in Hubei Province still needs to be further strengthened. Among the indicators in the driving force standard layer, D₁ (0.0671) has the greatest impact on the WECC of Hubei Province. This may be due to the fact that as the level of urbanization increases, the town planning becomes more scientific and rational, and the water resource management is optimized, thus ensuring the sustainable use of water resources. P₁ (0.0692) and P₂ (0.0699), which are the main pollutants contributing to the damage of the water environment, cause high pressure on the water environment. Among the indicators reflecting the state, S₁ (0.0724) and S₃ (0.0724) account for a large proportion. This indicates that the absolute amount of water resources can significantly affect the level of WECC, and reasonable water resources management can help maintain the stability and diversity of the ecosystem, thus improving WECC. According to the indicators in the impact standard layer, I₁ (0.0876) has the greatest impact on the WECC of Hubei Province. It can directly reflect the impact of water system state change on the river water environment quality. As for the response standard layer, it can be seen that adjusting the industrial structure and increasing R₃ (0.0791) can greatly influence the level of WECC in Hubei Province. Meanwhile, R₂ (0.0686) has also achieved certain results in Hubei Province in recent years.

In order to visualize the status of WECC in each prefecture-level city in Hubei Province, this study adopts the Natural Breakpoint Method to classify the WECC level of Hubei Province into five types (Table 2) [58]. In order to distinguish each type more significantly, red, orange, yellow, blue and green are used to indicate the five types of WECC, respectively.

Table 2. Evaluation standard of the WECC level.

Level	Low	Relatively Low	Middle	Relatively High	High
Evaluation index value	[0, 0.44)	[0.44, 0.51)	[0.51, 0.56)	[0.56, 0.61)	[0.61, 1]
Color	Red	Orange	Yellow	Blue	Green

The WECC calculation results for Hubei Province, from 2013 to 2022, are shown in Table 3. During this 10-year period, there are 170 WECC results in 17 municipal administrative regions. Of these, the low level of WECC occurred the least, at only 6.47% (11/170). The relatively high level of WECC occurred most frequently at 30.59% (52/170). The percentage of relatively low level, medium level, and high level of WECC is 24.7% (42/170), 24.12% (41/170), and 14.12% (24/170), respectively.

Table 3. WECC of cities in Hubei Province from 2013 to 2022.

	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022
Wuhan	0.3981	0.4325	0.4659	0.6059	0.5633	0.5682	0.5796	0.6221	0.5962	0.5884
Huangshi	0.5017	0.5305	0.5333	0.5842	0.5769	0.5507	0.5633	0.6344	0.6105	0.5909
Shiyan	0.4867	0.5039	0.5145	0.5419	0.5815	0.5558	0.5808	0.6168	0.6404	0.5889
Yichang	0.4774	0.4815	0.4993	0.5608	0.5876	0.5662	0.5668	0.6354	0.6014	0.6032
Xiangyang	0.3612	0.3982	0.4276	0.4950	0.5340	0.5157	0.5219	0.5728	0.5798	0.5476
Ezhou	0.4869	0.5096	0.5369	0.6033	0.5330	0.5323	0.5452	0.6277	0.5599	0.5681
Jingmen	0.4780	0.4756	0.4649	0.5247	0.4985	0.5028	0.5058	0.5822	0.5508	0.5381
Xiaogan	0.4241	0.4462	0.4523	0.4865	0.4829	0.4765	0.4903	0.5948	0.5382	0.5741
Jingzhou	0.3420	0.3503	0.3689	0.4523	0.4648	0.4953	0.5287	0.5933	0.5708	0.5525
Huanggang	0.4375	0.4805	0.4983	0.5817	0.5275	0.5219	0.5290	0.6322	0.5812	0.5736
Xianning	0.5319	0.5593	0.5720	0.5933	0.5889	0.5741	0.5923	0.6698	0.6585	0.6151
Suizhou	0.4775	0.4754	0.5048	0.5303	0.5252	0.5195	0.4610	0.5577	0.5705	0.5654
Enshi	0.5314	0.5465	0.5523	0.6082	0.6049	0.6013	0.6005	0.6803	0.6639	0.6224
Xiantao	0.4808	0.4240	0.4638	0.4935	0.4990	0.5506	0.5573	0.6639	0.5639	0.6007
Qianjiang	0.4835	0.4720	0.5139	0.5829	0.5453	0.5616	0.5604	0.6215	0.5885	0.5970
Shennongjia	0.5635	0.6188	0.5770	0.6201	0.6688	0.6140	0.6359	0.6763	0.6955	0.6483
Tianmen	0.5046	0.4960	0.5242	0.5109	0.4520	0.4527	0.4555	0.5726	0.5316	0.5323

Note: The five colors represent the five WECC levels. Red, orange, yellow, blue, and green indicate “low”, “relatively low”, “middle”, “relatively high” and “high”, respectively.

In terms of time, the overall development of WECC in Hubei Province during the period from 2013–2022 is favorable. Substantial increases in the level of WECC occurred in each city in 2016 and 2020. Compared with 2013, the WECC level of each prefecture-level city in 2022 had some improvement, and the cities at the low level and relatively low levels of WECC are gradually reduced in Hubei Province. All of the cities at the low level of WECC in Hubei Province occurred in 2013–2015. Low-level cities peaked in 2013 (5/17), declining annually thereafter. The largest number of cities were at a relatively low level of WECC in 2013 and 2014 (9/17) and disappeared after 2019. More cities in Hubei Province began to reach relatively high levels of WECC after 2015. A high level of WECC was mainly concentrated in 2020 (11/17). In terms of cities, the city with the best WECC status is Shennongjia, while Jingzhou has the worst WECC level. Low level of WECC is mainly concentrated in Wuhan, Xiangyang and Jingzhou. Relatively high level of WECC is mainly concentrated in Wuhan and Yichang. Wuhan and Jingzhou have made great improvement in the WECC level during this decade, and the WECC index has increased by 0.1904 and 0.2105, respectively.

3.2. Dynamic Evolution of WECC

3.2.1. Evolution of Spatial and Temporal Patterns of the WECC

In order to further explore the evolution of spatiotemporal patterns of the WECC level in Hubei Province, this paper utilizes the ArcGIS Pro 3.0.1 to map the spatial distribution. The results are shown in Figure 3.

Figure 3. Spatial distribution map of WECC in Hubei Province in (a) 2013, (b) 2016, (c) 2019 and (d) 2022.

During the period from 2013–2022, the WECC level in Hubei Province continued to evolve, with obvious non-equilibrium characteristics but an improving trend. The spatial agglomeration effect of WECC was significant, showing the spatial distribution characteristics of “high–high” agglomeration and “low–low” agglomeration. Among them, the WECC of the Wuhan metropolitan area in eastern Hubei Province had improved significantly, while the WECC level in central Hubei Province had increased slowly. From the perspective of time, the WECC level of cities in Hubei Province improved most significantly from 2013 to 2016, and almost every city realized the improvement of WECC level. After that, the speed of improvement gradually slowed down, and some cities stagnated or even regressed. In 2013, the WECC of Hubei Province as a whole showed a ladder-like distribution trend of “high in the west and low in the east.” Among them, the Wuhan metropolitan area was generally at the low or relatively low levels of WECC. In 2016–2019, the eastern and western regions of Hubei Province began to converge towards synergistic development, with major cities reaching relatively high and high levels of WECC, while most of the cities in central China remained at medium or relatively low levels of WECC. The level of WECC was gradually developing towards a distribution characterized by “high on both sides and low in the middle.” Xiangyang, Jingmen, Tianmen and Jingzhou in the central region of Hubei Province, maintained the medium level of WECC, showing a distribution characteristic of a vertical ribbon. Both the eastern and western cities of Hubei Province have reached relatively high or high levels of WECC.

3.2.2. Dynamic Evolution of the WECC Distribution

The Kernel Density Estimation can analyze the absolute difference in WECC in Hubei Province through the evolution of wave peaks. It is applied to clarify the distribution location, distribution pattern, distribution ductility, and distribution polarization trend of WECC in Hubei Province. Combining the four features of the curves of the Kernel Density Estimation, we can find that while the WECC level of Hubei Province maintained an overall upward trend from 2013 to 2022, the water environment management level of each city had been developed synergistically, and the differences in the WECC level among cities were gradually reduced. The specific change trend is shown in Figure 4.

Figure 4. Estimation map of core density of WECC in Hubei Province.

(1) In terms of the distribution location, the kernel density curve of WECC shifted somewhat leftward only in 2017 and 2021. As a whole, it showed a fluctuating rightward shifting trend during the period of 2013–2022, and underwent a significant rightward shift in 2022 compared with that of 2013. This meant that the overall WECC level in Hubei Province during the research period showed a fluctuating upward trend, with a yearly increase in the periods 2013–2016, 2018–2020 and 2021–2022. Although the overall WECC level of Hubei Province decreased to a certain extent in 2017 and 2021, it still had a significant increase compared with 2013. This is consistent with the analysis of the previous results.

(2) In terms of the distribution trend, the kernel density curve of WECC showed a distribution trend in which the height of the main peak decreases and then increases, and the width becomes wider and then narrower. This meant that the absolute difference in the level of WECC in Hubei Province had a tendency to narrow after expanding to a certain level. Combined with the results of the previous evaluation, it can be seen that this was due to the overall improvement of the level of WECC in Hubei Province; especially the cities that were at low or relatively low levels of WECC at the beginning have later risen to the top and achieved a large improvement.

(3) From the viewpoint of distribution ductility, before 2020, there was an obvious right trailing phenomenon in the kernel density curve of WECC, and the distribution ductility of the curve showed a gradually converging trend. This indicated that there were areas with a high WECC level during this period and the cities with a low WECC level have improved in Hubei Province. It reflects that the ecological environment development tended to be coordinated. Until after 2020, the right trailing phenomenon of the kernel density curve disappeared, the level of WECC of each prefecture-level city was generally at a relatively high level, and the gap was minimized.

(4) In terms of the distribution polarization trend, the overall polarization phenomenon of the WECC level in Hubei Province was more obvious in 2013 and 2017. The main peak values were all higher than the side peaks, and the polarization phenomenon showed a trend of “weakening-strengthening-gradually weakening.” Most of the kernel density distribution curves showed a single-peak distribution. This indicated that in most years, there was only one obvious peak in the distribution characteristics of WECC in each prefecture-level city in Hubei Province. That is, the WECC of most cities in Hubei Province was concentrated in a particular interval, and there was no obvious multilevel differentiation.

3.3. Prediction and Early Warning of the WECC

Since the WECC level is characterized by abrupt change and stability, it is important to explore the probability of state transfer and the moment of occurrence for prevention and regulation. Based on the WECC level of Hubei Province from 2013 to 2022, the probability matrices of five WECC state transfers of each prefectural city in Hubei Province under different time horizons were calculated by using the Markov Chain model analysis to investigate the state transfer law of WECC. The time span T is set as 1, 2, 3, 4 and 5 years. It is worth noting that the values of the main diagonal of the state transfer probability matrix indicate the probability of a smooth transfer of the WECC level, and the values of the off-diagonal indicate the probability of an upward or downward transfer of the WECC level. The measurement results are shown in Table 4.

Table 4. Markov Chain model transfer probability matrix of WECC in Hubei Province.

Time Span/Year	State Level	Low	Relatively Low	Medium	Relatively High	High
1	Low	45.45%	54.55%	0	0	0
	Relatively Low	2.38%	57.14%	26.19%	14.29%	0
	Medium	0	8.11%	45.94%	37.84%	8.11%
	Relatively High	0	0	23.81%	54.76%	21.43%
	High	0	0	4.76%	38.10%	57.14%
2	Low	18.18%	63.64%	9.09%	9.09%	0
	Relatively Low	0	30.95%	42.86%	26.19%	0
	Medium	0	12.12%	30.30%	42.43%	15.15%
	Relatively High	0	0	20.59%	50.00%	29.41%
	High	0	0	0	50.00%	50.00%
3	Low	0	54.55%	18.18%	27.27%	0
	Relatively Low	0	16.67%	45.24%	35.71%	2.38%
	Medium	0	12.50%	21.88%	50.00%	15.62%
	Relatively High	0	0	6.90%	51.72%	41.38%
	High	0	0	0	0	100%
4	Low	0	27.27%	54.55%	18.18%	0
	Relatively Low	0	23.68%	39.47%	34.21%	2.63%
	Medium	0	3.70%	18.52%	74.07%	3.70%
	Relatively High	0	0	0	27.27%	72.73%
	High	0	0	0	0	100%
5	Low	0	18.18%	45.46%	36.36%	0
	Relatively Low	0	17.65%	35.29%	35.29%	11.77%
	Medium	0	0	15.00%	55.00%	30.00%
	Relatively High	0	0	5.88%	47.06%	47.06%
	High	0	0	0	0	100%

According to the measurement results, the probability values on the diagonal of the time span are not significantly higher than those on the non-diagonal, except for the case of the time span of 1 year. This indicates that the level of WECC in Hubei Province during the period of 2013–2022 was not stable, and the probability of maintaining the initial state was small with high internal mobility. The “lower triangle” area below the diagonal of the state transfer probability matrix for each time span is mostly 0, and the greater the time span, the more frequent the probability of 0. This indicates that the probability of downward transfer of WECC level in Hubei Province is relatively low over time. The probability of WECC level transfer occurring between neighboring types is significantly higher over a short period of time. This implies that the improvement of WECC level is a gradual process, and it is difficult to achieve a “leaping” transfer in a short time.

Specifically, for a time span of 1 year, the probability values on the diagonal are significantly higher than those on the non-diagonal, except for the “low → relatively low” level of WECC transfer type. This reflects that there is a high probability that the WECC level in Hubei Province will remain in its initial state during this time span. The probability of a reverse shift is 42.86% and 23.81% for high level and relatively high levels of WECC cities, respectively. This proves that there is also a risk of a short-term cliff fall for both of these types. In the time span of 2 and 3 years, the improvement of WECC level is mainly concentrated in “low level → relatively low level”, “relatively low level → medium level” and “medium level → higher level.” The average transfer probability of the three is 59.1%, 44.05% and 46.22%, respectively. This indicates that the level of WECC can often be realized within a short period of time up one level, while the probability of leapfrogging is low and it is difficult to reach the high level of WECC quickly. The probability of an upward shift in the WECC level increases further when the time span is 4 and 5 years. The probability of downward transfer is only 3.7% for “medium level → relatively low level” in the 4-year period and 5.88% for “relatively high level → medium level” in the 5-year period. Meanwhile, the probability of leapfrogging increases further, mainly in the categories “low → middle” and “relatively low → high”, with average transfer probabilities of 50% and 34.75%, respectively.

Based on the results of the Markov Chain model, the overall development trend of WECC in Hubei Province can be predicted. If the WECC of a city is at a low level in the initial year, the probability that the city will continue to maintain a low level of WECC after a one-step transfer is 45.45%, and the probability of an upward transfer is 54.55%. The probability of a smooth transfer of the WECC level of the city is 0 when the duration is 5 years, which means that an upward transfer of the WECC level of the city must occur when the duration is 5 years. If the WECC of a city is at a relatively low or medium level in the initial year, the probability of a smooth transfer of the WECC level of the city after a one-step transfer is 57.14% and 45.94%, and the probability of an upward transfer is 40.48% and 45.95%, respectively. When the duration is 5 years, the probability of smooth transfer of the WECC level of the city decreases to 17.65% and 15%, respectively, and the probability of upward transfer increases to 82.35% and 85%. We can find that the level of WECC of each prefecture city in Hubei Province tends to shift to a relatively high level from 2013 to 2022, and the probability of such a shift is gradually increasing with the increase in the time span.

This paper applies the Markov Chain prediction model to forecast WECC level trends in Hubei Provinces’ cities, using 2022 WECC levels as the initial state. According to the state transfer probability matrix P, the state probability distribution vector for the years 2023–2027 is calculated, and the probability that each city is in each state level can be observed intuitively. From the prediction results in Figure 5, it can be seen that according to the existing development trend, in 2026, the probabilities of high level of WECC in Wuhan, Huangshi, Shiyan, Yichang, Ezhou, Xiaogan, Huanggang, Suizhou, Xiantao and Qianjiang, which have a relatively high initial state, are higher than 70%. Therefore, based on the confidence level criterion, we can conclude that the WECC of these cities will be at a high level in 2026. However, in 2027, these cities have equal probabilities of reaching relatively high and high levels of WECC. This suggests that these cities need to take care of the risks of WECC reduction in that year. The cities with medium initial status of WECC, such as Xiangyang, Jingmen, Jingzhou, and Tianmen, have a probability of being at a relatively high level in 2026 that is higher than 70%. Based on the confidence level criterion, the WECC of these cities will be at a relatively high level in 2026. In 2027, there is a probability of 30% that the WECC of this type of city will be at a high level. The cities with initial WECC status at a high level, such as Xianning, Enshi Prefecture, and Shennongjia, have a certain probability of decreasing to a relatively high level in 2023 and 2024, but will almost certainly remain at a high level after 2025.

Figure 5. Probability distribution of different state levels for 17 cities in 2023–2027.

To assess the reliability of the Markov Chain model’s predictions, this study uses the actual data from 2023 (Table 5) to verify the prediction results. The prediction results for 2023 are the probability distributions for different WECC levels for each city. The focus of the verification is whether the actual observed states in 2023 have a high probability in the prediction results. The results show strong statistical consistency. In most cities, the WECC level observed in 2023 is exactly the one that was assigned a high probability by the model. Although in some cities, the WECC is not completely at the highest probability level, it is still within the range considered by the prediction results and has a high probability level. This indicates that the model can provide reliable estimates of the risk or possibility of transitions between different carrying capacity states for cities.

Table 5. WECC of cities in Hubei Province in 2023.

City	WECC	City	WECC	City	WECC
Wuhan	0.5483	Jingmen	0.5617	Enshi	0.6712
Huangshi	0.6225	Xiaogan	0.5934	Xiantao	0.5596
Shiyan	0.6401	Jingzhou	0.5501	Qianjiang	0.6211
Yichang	0.6462	Huanggang	0.5976	Shennongjia	0.7247
Xiangyang	0.5921	Xianning	0.6182	Tianmen	0.5645
Ezhou	0.5962	Suizhou	0.6064

Note: The colors represent the WECC levels. Yellow, blue, and green respectively indicate middle level, relatively high level and high level of WECC.

4. Discussion

4.1. WECC Evaluation and Prediction Analysis

From the comprehensive evaluation results, the overall condition of WECC in Hubei Province was acceptable during the period of 2013–2022, and it has certain potential for development and utilization. However, there will still be some fluctuation in the short term, and it must be strictly controlled and moderately developed. Some similar studies have verified these findings. For example, Song et al. [59] established an Improved Catastrophe Evolution Model based on the Mann–Kendall (M-K) Test and Correlation Analysis to study the water resources carrying capacity in Hubei Province from 2005 to 2016. Lu et al. [60] combined the Principal Component Analysis and the Information Entropy methods to calculate the water resources carrying capacity of Hubei Province from 2008 to 2020. The results of both studies were approximately the same as the contemporaneous results of this study. Furthermore, Lu et al. [60] found that the water resources carrying capacity in Hubei Province fluctuates greatly. This is due to the fact that their studies focused more on the quantity of water resources, and water resources carrying capacity is affected more by precipitation and water consumption. As the province with the longest runoff mileage of the Yangtze River, the reservoir area of the Three Gorges Project, and the core water source area of the South-to-North Water Diversion Central Route Project, Hubei Province has assumed an important responsibility in the ecological protection of rivers and lakes in China [60]. Owing to long-term adherence to comprehensive watershed management and continuous promotion of scientific and technological innovation, Hubei Province has relatively good WECC status [61]. The level of WECC for each prefecture-level city in Hubei Province increased most significantly from 2013 to 2016, and the rate of increase slowed down after 2016. Zhang et al. [62] evaluated the regional carrying capacity of the Yangtze River Economic Belt from 2004 to 2019 and found that the peak appeared in 2016. This is consistent with the results of this paper, and the outcome is due to a number of factors. First, the initial policies and inputs were effective. During the period 2013–2016, Hubei Province introduced a series of policies and measures targeting water environment governance and protection and increased related investment. These policies and financial support have directly promoted the work of cities in water environment management, water resources protection and water ecology restoration, and have significantly improved the WECC status. However, after the WECC has been raised to a certain stage, the difficulty of making further improvements increases significantly. For example, some regions may face difficulties in water source protection, pollution source control and ecological restoration. The solution of these problems will require a longer period of time and continuously large investment. In addition, technological bottlenecks and management systems may also be constraints in the process of water environment management. As governance advances, the technical and managerial requirements become more demanding, but existing technical and managerial institutions may struggle to meet the demands of higher levels of governance. Economic and population growth will put increasing pressure on the water environment, while the improvement of water environment management capacity is constrained. This has led to a gradual slowdown in the rate of improvement of the WECC level.

From the results of spatial and temporal pattern analysis, during the period of 2013–2022, the level of WECC in Hubei Province has gradually changed from the ladder-like distribution of “high in the west and low in the east” to the distribution of “high in the east and west and low in the center.” Wang et al. [63] used an improved ecological footprint model for water resources to calculate the water resources ecological pressure index for cities in Hubei Province in 2005, 2010 and 2015. It found that the water resources pressure in the eastern part of Hubei Province was greater than that in the western part, and the sustainability of water resources utilization in the western canyon area was higher than that in the Jianghan Plain. The conclusions from this study are largely the same as those of this paper. In the eastern region, represented by Wuhan, the pressure on the water environment is excessive compared to the WECC provided by the natural water system due to the rapid growth of population and GDP. This is reflected in the overall relatively low WECC levels in the region at the beginning of the study time period. Ding et al. [64] studied the water ecological carrying capacity of the East Lake in Wuhan from 2002 to 2012 and came up with similar results. In contrast, the WECC level of the Wuhan urban agglomeration increased the fastest from 2013 to 2022, and the water environment condition was generally sustainable in recent years, which is generally consistent with the findings of Song and Hu [65] on the ecological and environmental carrying capacity of the Wuhan urban agglomeration. The region is able to rapidly improve the level of WECC under high water environment pressure, and exploring the reasons can provide some inspiration for similar regions to improve the WECC level. The equilibrium state of urban resource–environment carrying capacity is determined by the resource subsystem and the natural environment subsystem against the pressure from human activities [66]. The state of the WECC is determined by the interaction between the pressures exerted on the water system by human activities and the magnitude of the carrying capacity that the water system can provide. Therefore, the enhancement of WECC status can be analyzed in terms of increasing the size of the WECC and reducing the pressure exerted on the water system by human activities. From the perspective of increasing the WECC, the development of human society, especially the impact of technological progress, will continue to break through the limiting factors constraining sustainable development, and new upper limits of carrying capacity will emerge on top of the existing carrying capacity [67]. Wuhan is a good example. During the period 2013–2022, it adhered to the strategic positioning of “Ecological Priority and Green Development” and achieved fruitful results in ecological and environmental management. For example, water pollutant emissions have been significantly reduced, the level of urban greening has been markedly improved, and the efficiency of water resource utilization has been significantly enhanced. Simultaneously, the continuous progress of technologies and improvement of the social governance level also push Wuhan to a large extent to improve its own WECC level.

In this paper, we find that the level of WECC tends to shift to a relatively high level in various cities in Hubei Province. Similar studies corroborate our result. Zhu et al. [68] calculated the water resources carrying capacity of Hubei Province from 2001 to 2017 using the Cloud Modeling method and found that the water resources carrying capacity of Hubei Province is bound to develop in a better direction. Wang et al. [63] concluded that the water resources ecological carrying capacity of Hubei Province is on the rise. The conclusion can equally be analyzed in terms of increasing the size of WECC and reducing the pressure of human activities on the water system. On the one hand, technological progress can break through the upper limit of WECC. As a major education province in China, Hubei Province has many scientific research institutions and universities, and it has fully utilized its advantages to develop the level of technological innovation, which has led to rapid technological progress. The application of advanced technologies in water environment management, such as big data and the Internet of Things, information technology, wetland restoration and landscape restoration, and other ecological restoration technologies, has effectively enhanced the level of WECC in Hubei Province. On the other hand, the improvement of water environment management has reduced the pressure of human activities on the water system. Accordingly, China began to implement the strategy of Yangtze River Conservation in 2016 and began to fully implement the River and Lake Chief System in 2017. As an important province in the Yangtze River Economic Belt and the first province to coordinate the River Chief System and Lake Chief System, Hubei Province has rapidly and vigorously carried out water environment management throughout the region, reducing the pressure on the water environment caused by economic and social development. Therefore, according to the above evolution law, there exists an overall trend of steady growth in the WECC of Hubei Province, and the state of WECC in each subregion will have a high probability of being dominated by a relatively high level.

In this paper, a comparison of the probability of WECC state transfer across time spans reveals that the probability of WECC level transfer occurring between neighboring types is higher for short periods of time. As the time span increases, the probability that this transfer will occur for the WECC level increases gradually. The results of this study verified that the level of WECC itself has a certain degree of stability. It usually will fluctuate within a range, but the overall state remains constant. The improvement of WECC status requires a certain amount of time or energy accumulation, just like the electron leap in physics. This characteristic of WECC is reflected in a number of WECC evaluation studies [33,69]. However, the probability of a leapfrog growth or decline of WECC level within a short period is not zero. This suggests that the migration of the WECC state also has a certain degree of suddenness, as in the case of a volcano’s energy buildup to a sudden eruption. Meanwhile, by predicting the probability of the distribution of WECC status for each city from 2023 to 2027, this study finds that most of the cities in Hubei Province will experience an increase in WECC level in 2026, while they will be at risk of a decrease in 2027. This suggests that the time required for WECC improvement in Lakeland to develop in line with the existing good scenarios is roughly four years, and that special attention needs to be paid to guarding against the risk of decline in the fifth year. A review of the literature reveals that similar results have been obtained in studies on WECC in other regions. For example, Jin et al. [70] used the Unconfirmed Measurement Theory (UMT) model to evaluate the WECC in the Gansu section of the Yellow River Basin from 2015 to 2020. The results showed that the overall trend of WECC in the region was also in a gradual increase. In the fourth year (2018), the WECC of the study area increased from grade III to grade IV, realizing an increase of one grade, while in the 5th year (2019) the WECC decreased to grade III (critical load) compared to 2018. Zhang et al. [71] calculated the water resources–water environmental carrying capacity of Yutang District, Xiangtan City, Hunan Province, from 2005 to 2015 using Principal Component Analysis. The results showed that in the case of the water resources–water environmental carrying capacity of Yutang District increasing year by year, it increased from grade III (poor) to grade II (good) in the 4th year (2009) after 2005 and improved from grade III (poor) to grade II (good) after 2009. This provides a reference for cities to formulate water environment management plans.

4.2. Policy Suggestions

(1) Water conservation and water pollution control should be further emphasized to reduce the pressure of human activities on the water supply environment. Since water pollution accompanies the utilization of water resources, the policymakers should optimize the allocation of water resources, improve water use efficiency and reduce the waste of water resources by implementing rigid constraints on water resources. Regional government should increase investment in the construction and upgrading of wastewater treatment facilities, actively support enterprises to independently research and develop wastewater treatment technologies, improve wastewater treatment rates and reuse rates, and reduce the impact of pollutants on the water environment. The relevant government departments should actively implement ecological restoration projects, such as wetland restoration and revegetation of riverbanks, to enhance the self-purification capacity of water bodies and improve the quality of the water environment.

(2) The regional industrial development model should be adapted to WECC. On the basis of clarifying the status and development trend of the WECC, the related government departments should reasonably set the optimal industrial development scale of different regions and resolutely promote regional differentiated management under the constraint of the WECC. For regions with lower levels of WECC, it is necessary to further strengthen the control of the scale of socio-economic development and the supervision of industrial transfers, and to formulate reasonable industrial access policies to ensure that the structure of the socio-economic system is compatible with the carrying capacity of the regional water environment. For some regions with higher level of WECC, it is necessary to accelerate the improvement of economic strength and the level of social governance on the premise of adhering to WECC constraint requirements and the advanced development of the industrial structure, in order to further expand the support for the scale of socio-economic development.

(3) A monitoring and early warning system for WECC status should be established. The relevant government departments at all levels should actively carry out dynamic monitoring and timely assessment of WECC, and scientifically study and delineate early warning control lines and response lines of WECC according to the assessment results. Policymakers should establish reward-punishment assessment standards, enhance WECC early warning responses, and strengthen law enforcement supervision.

4.3. Limitations and Future Research

Although this study constructed a new set of index systems to comprehensively evaluate the WECC status of cities in Hubei Province and accurately predicted its future development trend, there are still the following shortcomings that deserve further study. (1) This study does not identify the specific factors driving WECC state transitions, precluding accurate early warning. In the future, the causal relationships within the complex system of WECC should be further studied, and attempts should be made to combine remote sensing and image recognition technology to provide clear early warning signals, not just forecasting the changes in WECC status. (2) The spatial variability of resource and environmental elements and conditions is large and needs to be refined to small scales. This approach enhances the practical utility of evaluation and early warning results. Due to the availability of data, this study stays at the city level, and the research scale can be further narrowed down if richer data becomes available. (3) The self-purification capacity of water bodies is a crucial internal mechanism for regulating the WECC and water quality. Future research can further integrate parameters that directly reflect the self-purification capacity, thereby providing a deeper understanding of the intrinsic driving mechanism of WECC.

5. Conclusions

In this paper, on the basis of clarifying the definition of WECC, a new integrated research framework is constructed for dynamically assessing and predicting WECC in 17 municipal administrative regions in Hubei Province. Specifically, this paper first utilizes the DPSIR method to construct a comprehensive evaluation index system of WECC. Then the Kantiray Weighting method and the TOPSIS method are combined to quantitatively evaluate the status of WECC. Finally, the spatiotemporal characteristics and evolution trend of WECC are explored by combining the Kernel Density Estimation method and the Markov Chain model. The main conclusions are as follows:

(1) During the period 2013–2022, the overall status of WECC in Hubei Province is acceptable, with relatively high levels occurring most frequently. Shennongjia has the best WECC status and Jingzhou has the worst. The differences in WECC levels among the cities in Hubei Province are not significant and have a tendency to narrow. The improvement of WECC is closely related to the share of tertiary sector, such as environmental engineering services, water pollution treatment technologies, and modern ecological monitoring industries. Additionally, the main pollutants COD and ammonia nitrogen have a greater impact on the level of WECC.

(2) Since 2016, the spatial distribution pattern of WECC in Hubei Province has evolved from a general “high in the west and low in the east” to a new pattern of “high in the east and west and low in the middle.” The central region is relatively slower than the east and west. In terms of time, the level of WECC as a whole shows a fluctuating upward trend of “rising-falling-rising.”

(3) The future WECC of Hubei Province has an overall trend of steady growth. While the state of WECC has a certain stability, there is also a certain probability of sudden transfer. Hubei Province needs to pay special attention to prevent the risk of decline in 2027 and intervene in advance. The improvement of WECC has regularity, and the “leaping” transfer of WECC in a short period of time is not common. The improvement of WECC level requires continuous and stable investment.