Research on the Coupling and Coordination of Land Ecological Security and High-Quality Agricultural Development in the Han River Basin

Abstract

This study aims to investigate the coupling and harmonization between land ecological security (LES) and high-quality agricultural development (HAD) in the Han River Basin (HRB), China, with the objective of promoting harmonious coexistence between agriculture and ecosystems. Using 17 cities in the HRB as the research objects, an evaluation index system of two systems, LES and HAD, was constructed, analyzed, and evaluated via projective tracer modeling for multiple intelligent genetic algorithms (MIGA-PTM). The degree of coupling coordination (DCC) was used to quantitatively evaluate the coupling coordination development status of the two systems, the obstacle model (OM) was used to identify the main influencing factors, and the gray predictive model first-order univariate model (GM (1, 1)) was used to predict the DCC of the LES and HAD from 2025 to 2040. The results show the following: (1) the LES and HAD levels of the 17 cities in the HRB tended to increase during the study period, and there was a large gap between cities; (2) the spatial distributions of the DCCs of the LES and HAD in the HRB were uneven, with high values in the southern and low values in the central and northern parts, and the overall degree of coupling tended to fluctuate. The overall DCC showed a fluctuating upward trend; (3) the degree of obstacles, per capita water resources, greening coverage, and rate of return on financial expenditure are the main influencing factors; and (4) the prediction results of GM (1, 1) indicate that the LES and HAD of the HRB will be close to reaching the intermediate stage of coupling in 2035. This research offers critical insights into sustainable development practices that facilitate the alignment of agricultural growth with ecological preservation.

1. Introduction

Land is not only a natural complex on Earth’s surface, but also the material basis on which human society depends for its survival and development [1]. Land ecological security (LES) is defined as the maintenance of stability in the structure and function of the land’s natural ecosystem, ensuring the sustainable use of land resources while avoiding or mitigating irreversible damage resulting from human activities [2]. LES is a prerequisite for and a guarantee of food, economic, and social stability, and it is the basis for achieving ecological security [3]. With global population growth and the expansion of economic activities, ecological and environmental problems such as land degradation and soil pollution are becoming increasingly prominent. In particular, for the world’s largest developing country, China, rapid industrialization and urbanization have put unprecedented pressure on land ecosystems, which not only threatens ecological security, but also challenges the sustainable development of agriculture [4]. High-quality agricultural development (HAD) is defined as a development model driven by scientific and technological innovation aimed at achieving efficient, resource-saving, and environmentally sustainable agricultural production [5]. Since the Han River basin (HRB) is an ecological function area and agricultural production base, the coordinated relationship between the LES and the HAD is particularly important. The importance of this issue lies in the fact that the state of land resources in the HRB directly affects the ecological security of the basin and is the cornerstone for maintaining sustainable regional development. The ecological security and agricultural development status of cities along the HRB are of great concern and are typical of wider issues [6]. Consequently, research examining the spatial and temporal evolutionary characteristics and coupling coordination of LES and HAD in the HRB is of significant strategic and practical importance for ensuring regional ecological security, improving agricultural production efficiency, and fostering sustainable economic and social development.

As the HRB is an ecological function area and agricultural production area, the evaluation of LES and HAD in this zone is highly practical for achieving the coordinated development of regional ecological protection and agricultural modernization. Currently, scholars’ evaluation methods for LES and HAD mainly focus on the entropy weight TOPSIS method [7,8], the principal component analysis method [9,10], and the fuzzy evaluation model [11,12]. For example, Li and Liu introduced an entropy-weighted TOPSIS model to assess the HAD in China [13]. Chen and Zhang used the entropy weight comprehensive evaluation method to construct 15 agricultural green development evaluations to evaluate the HAD in Mianyang city [14]. However, traditional evaluation methods have shortcomings when LES and HAD evaluations are performed. In terms of global search capability, traditional methods such as TOPSIS, PCA, and fuzzy analysis tend to focus more on local optimization or specific types of data processing rather than extensive searching in the whole solution space [15]. In terms of flexibility and adaptability, traditional methods of tuning and optimization tend to be limited by specific algorithmic structures and data processing [16]. In terms of parallel processing capability, it may be difficult for traditional methods to utilize parallel computing power effectively, especially when dealing with large-scale data, which may encounter performance bottlenecks [17]. On the basis of some of the shortcomings of the above evaluation methods, this study used the Multi-Intelligent Group Algorithm for Optimized Projection Tracing Model (MIGA-PTM) for the evaluation of the LES and HAD systems in the HRB. MIGA-PTM is an optimization technique that combines multiple intelligent systems (MIS), genetic algorithms (GA), and projective tracer models (PTMs) for complex data analysis and pattern recognition problems. To address the evaluation problem of LES and HAD, the MIGA-PTM used in this study is a multifaceted optimization of the traditional model. In terms of global search capability, the MIGA-PTM can search for optimal solutions in a large range and reduce the risk of falling into local optima through the global search property of genetic algorithms. This is particularly important when dealing with complex land use, ecological security, and agricultural development assessment problems, which usually involve multiple variables and complex constraints. In terms of parallel processing, the MIGA system allows multiple intelligences to work in parallel, which improves the computational efficiency, especially when dealing with large-scale datasets, such as the 35 evaluation indices in this study, or complex systems, compared with traditional algorithms. The MIGA-PTM can obtain faster results. In terms of adaptability, the MIGA-PTM can adapt to various types of data distributions, including nonlinear and non-normally distributed data, which makes it more flexible and effective in LES and HAD nonlinear multidimensional data processing.

Research on coupled and coordinated studies of the HRB LES and other systems is highly practical for achieving the coordinated development of regional ecological conservation and agricultural modernization. Scholars have conducted many studies on LES and its coupled coordination, and many systematic theories and methods have been developed. The literature has focused on exploring the relationships between LES and urbanization [18,19], carbon efficiency [20], economic development [21], food security [22], etc. For example, Guo et al. used a charge-coupled device (CCD) model to investigate the degree of coupling coordination (DCC) between the LES and the economic quality subsystem from 2007 to 2018 [21]. Hong et al. used an ultra-efficient relaxation-based measurement model that considers undesirable outputs and a coupled coherence model to investigate the coordination of LES and carbon emissions in 30 provinces in China [20]. HAD refers to improving the total factor productivity of agriculture and promoting the transformation of agriculture into a quality-benefit and innovation-driven industry based on the premise of ensuring national food security and an effective supply of agricultural products [23,24]. Agriculture is an important economic pillar in the HRB, and the realization of HAD is highly important for enhancing the regional economic development level and increasing farmer incomes [25]. The development of HAD promotes LES; with advancements in HAD, agricultural production methods are progressively transitioning to sustainable, low-carbon, and circular practices. Conversely, the development of LES ensures HAD; favorable LES conditions provide a stable environment and material foundation for HAD. Consequently, within the framework of ecological civilization construction, the relationship between LES and HAD necessitates comprehensive investigation to facilitate the sustainable development of land and agriculture. In the literature, scholars mostly measure LES and HAD separately, and there are few studies on the coupling and coordination between LES and HAD. On the one hand, coupled synergistic research can explore paths and strategies to promote HAD under the premise of safeguarding ecological safety. On the other hand, this study helps to assess the impact of land-use change on agricultural and ecological security and provides a basis for risk management and emergency response. Therefore, this study selected the HRB as the research object, used the improved MLGA-PTM to analyze the spatiotemporal space of the LES and HAD, studied the coupling and coordination relationships between the LES and HAD, used the obstacle model (OM) to analyze the influencing factors, used the gray predictive model first-order univariate model (GM (1, 1)) to predict the DCC, and proposed an optimization strategy. This study is highly important for helping LES and HAD reveal intrinsic correlations and laws, promoting sustainable agricultural development, safeguarding regional ecological security, providing decision-making support, and promoting coordinated regional development.

The rest of this paper is structured as follows: Section 2 provides a comprehensive literature review that synthesizes current research trends in LES and HAD, emphasizes the unique characteristics of the HRB, and assesses the necessity and urgency of addressing the challenges associated with LES and HAD in the region. Section 3 describes the research areas, indicator data, and evaluation index systems used in this study. Section 4 provides an overview of the modeling process for the MIGA-PTM method, natural breakpoint grading method, OM, DCC, and GM (1, 1). Section 5 provides a detailed verification of the MIGA-PTM built using these two systems. The LES and MAD of the HRB were measured via the MIGA-PTM, and the temporal and spatial analyses were combined with the natural breakpoint method and the standard elliptic difference method. The coupling and coordination relationships between the two systems were calculated via DCC, and the factors hindering coupling and coordination were identified and analyzed via OM. Finally, GM (1, 1) was used to predict the LES of the HRB and the DCC of the HAD. Section 6 and Section 7 include a discussion of the results and a concise summary of the findings. The technical roadmap used in this study is illustrated in Figure 1.

2. Literature Review

The HRB, one of China’s most vital economic and ecological regions, faces pressing challenges related to LES and the pursuit of HAD [26]. As urbanization accelerates and agricultural practices intensify, the ecological balance is increasingly threatened. Therefore, it is imperative to address the challenges of LES and sustainable HAD in the HRB.

LES encompasses the capacity of land to sustain its ecological functions while supporting economic activities. In the HRB, rapid industrialization and agricultural expansion have resulted in significant environmental degradation, including soil erosion, water pollution, and loss of biodiversity [27]. The degradation of natural resources not only undermines ecosystem stability but also threatens food security and the livelihoods of local populations. Tscharntke et al. argued that the interplay between ecological integrity and agricultural viability is crucial for sustainable development, highlighting the need for immediate attention to ecological security [28]. HAD emphasizes sustainable practices that enhance productivity while safeguarding ecological health. In the context of the HRB, this development paradigm promotes techniques such as precision agriculture, organic farming, and agroecological practices [29]. These methods aim to optimize resource use and minimize environmental impacts, thereby ensuring long-term HAD. However, current agricultural practices often prioritize short-term economic gains, leading to soil degradation, reduced water quality, and increased greenhouse gas emissions [30]. The urgent need for policy interventions that promote sustainable agricultural practices is therefore critical.

Recent studies have employed various methodologies to assess LES and HAD. Remote sensing and geographic information systems (GISs) are commonly used to monitor land use changes and ecological health [31]. Additionally, socio-economic analyses explore the relationship between land management practices and community well-being, highlighting the socio-political dimensions of ecological security [32]. Another trend involves the exploration of policy frameworks that integrate ecological and agricultural objectives. Research by Avinash et al. proposes a multi-stakeholder approach, emphasizing collaboration among the government, farmers, and NGOs to achieve sustainable outcomes [33].

Despite the progress made, there are still gaps in the literature. First, research that integrates ecological, economic, and social factors into a cohesive land-use planning framework remains insufficient. Most studies tend to isolate these factors. Second, many evaluation studies on LES and HAD rely on traditional models that struggle to account for nonlinear relationships and often overlook spatial heterogeneity. Finally, many studies are still conducted in isolation, independently examining LES and HAD. There is an urgent need to consider the simultaneous coordination of LES and HAD. The necessity and urgency of addressing LES and HAD in the HAD are evident, considering the region’s ecological and socioeconomic significance. In light of the identified necessity and urgency of addressing LES and HAD in the HRB, along with existing research gaps, this paper proposes a novel evaluation method to assess both LES and HAD and investigate the coupling relationship between these two systems. This article offers a case study for evaluating the coupling of LES and HAD in a basin and can inform decision-making and sustainable practices that benefit the basin environment.

3. Study Area and Data

3.1. Study Area

The HRB, an important tributary of the Yangtze River, flows through southern Shaanxi, southern Henan, and Hubei, and it is rich in diverse geographical features [34]. The Han River is approximately 1532 km long, with a watershed area of approximately 135,000 square kilometers [35]. The HRB includes the following major cities: Shiyan, Xiangyang, Jingmen, Suizhou, Xiaogan, Tianmen, Qianjiang, Xiantao, Wuhan, Shennongjia Forestry District, Luoyang, Sanmenxia, Zhumadian, Hanzhong, Ankang, and Shangluo (Figure 2). The HRB spans the Qinling and Daba Mountains and other mountain ranges and has complex and varied topography, including mountains, hills, plains, and other landform types. The upper reaches of the HRB are characterized by a humid subtropical climate with abundant precipitation and high river runoff [36]. The downstream area of the HRB is characterized by high summer rainfall and low winter rainfall, resulting in significant seasonal variations in runoff. The HRB has significant hydrological characteristics and is rich in water resources, which are important for irrigated agriculture, industrial production, and residential life in the basin [37].

3.2. Data Sources

In this study, 35 evaluation indicators for 2010–2022 were selected as the measurement standards for the 17 cities in the HRB. The data were obtained from authoritative sources, such as the China Urban Statistical Yearbook, the China Environmental Statistical Yearbook, the statistical yearbooks of provinces and cities, the National Economic and Social Development Statistical Bulletin of Provinces and Cities, and the Water Resources Bulletin of Provinces and Cities. Some missing data were replaced by fitted values for all historical years.

3.3. Establishment of an Evaluation Indicator System

3.3.1. LES Evaluation Indicator System

On the basis of the principles of goal orientation, system comprehensiveness, and operability, with reference to previous studies [38,39,40], and considering the objective characteristics of the HRB, we constructed an LES evaluation index system based on the pressure–state–response (PSR) model, which consists of three criterion layers and 15 indicator layers (

Table 1. Evaluation index system of LES in the HRB.

Criterion Layer	Index Layer	Calculation	Attribute
Pressure	X1 Population density	Total urban population/Area	–
	X2 Urbanization rate	Urban resident population/Total resident population	–
	X3 Economic density	Economic output/Area size	+
	X4 Natural population growth rate	Statistical data	–
	X5 GDP per capita	Total GDP/Total regional population	–
State	X6 Cultivated land area per capita	Total arable land/Total population	+
	X7 Construction land area per capita	Total built-up land area/Total population	–
	X8 Green coverage	Area covered by vegetation/Total area of the region	+
	X9 Parkland area per capita	Total parkland area/Total population	+
	X10 water resources per capita	Total water resources/Total population	+
Response	X11 Centralized urban sewage treatment rate	Statistical data	+
	X12 Comprehensive industrial solid waste utilization rate	Statistical data	+
	X13 Nonhazardous domestic waste disposal rate	Statistical data	+
	X14 Percentage of tertiary sector	Statistical data	+
	X15 Energy consumption per unit of GDP	Statistical data	–

). The evaluation index system was divided into a target layer, criterion layer, and index layer.

3.3.2. System of Indicators for Evaluating HAD

On the basis of the principles of goal orientation, system comprehensiveness, and operability, with reference to previous studies [41,42,43,44], and considering the objective characteristics of the HRB, we constructed an evaluation index system for HDA, which consists of five guideline layers and 20 indicator layers (

Table 2. Evaluation index system of the HAD in the HRB.

Criterion Layer	Index Layer	Calculation	Attribute
Quality and efficiency level	Y1 Agricultural productivity	Value added of agriculture, forestry, and fisheries/Total output of agriculture, forestry, and fisheries	+
	Y2 Agricultural economic effects	Gross agricultural output/Total sown food area	+
	Y3 Rate of return on fiscal expenditures	Value added of primary sector/Local fiscal expenditure	+
	Y4 Labor productivity	Gross value of agricultural, forestry, livestock, and fisheries production/Rural workers	+
Security of supply	Y5 Level of mechanization per capita	Gross value of agricultural, forestry, livestock, and fisheries production/Rural workers	+
	Y6 Effective irrigation rate	Area of land effectively irrigated/Total cultivated area	+
	Y7 Electrification level	Rural electricity consumption/Rural population	+
	Y8 Funding for agricultural science and technology activities	Internal expenditure on R&D funding in RMB 10,000,000 × (gross output value of agriculture, forestry, animal husbandry, and fisheries/gross domestic product)	+
Co-ordinated development	Y9 Level of urban‒rural income coordination	Per capita disposable income of urban residents/Per capita disposable income of rural residents	-
	Y10 Level of urban‒rural consumption coordination	Per capita disposable income of urban residents/Per capita disposable income of rural residents	-
	Y11 Industrial harmonization index	Secondary and tertiary industry output/Primary industry output	-
	Y12 Level of regional coordination	Agricultural GDP per capita in metropolitan areas/Provincial agricultural GDP per capita	+
Green development	Y13 Fertilizer application intensity	Fertilizer application/Cultivated land area	-
	Y14 Pesticide application intensity	Pesticide application/Cultivated land area	-
	Y15 Intensity of application of agricultural films	Agricultural film use/Area sown to crops	-
	Y16 Comprehensive livestock and poultry manure utilization rate	Statistical data	+
Shared development	Y17 Enrichment level of the rural population	Per capita expenditure on education, culture, and recreation/Per capita consumption expenditure	+
	Y18 Rural Engel coefficient	Food expenditure/Consumption expenditure per rural inhabitant	-
	Y19 Level of rural health care	Statistical data	+
	Y20 Level of farmers’ income	Per capita net income of farmers	+

).

4. Methods

4.1. Projective Tracer Modeling for Multi-Intelligent Genetic Algorithms

4.1.1. Projective Tracer Models

The projection tracer model (PTM) is a multivariate statistical analysis method used to explore low-dimensional structures in high-dimensional datasets. The core aim of the model is to find one or better projection directions such that the projection in these directions maximizes some statistical property of the data [45,46]. The core idea of the projection finding model is to find a low-dimensional projection that best represents the original high-dimensional data structure [47]. The calculation steps are as follows:

Step 1: Normalization.

Let $x\ast (i,j)\{i=1, 2, \dots , n; j=1, 2, \dots , p\}$ be the sample value for each indicator, where $n$ and $p$ are the sample capacity and the number of indicators, respectively.

$$\begin{array}{c}x\left(i,j\right)={\displaystyle \frac{x\ast \left(i,j\right)-x_{\min \left(j\right)}}{x_{\max \left(j\right)}-x_{\min \left(j\right)}}}\end{array}$$

(1)

$$\begin{array}{c}x\left(i,j\right)={\displaystyle \frac{x_{\max \left(j\right)}-x\ast \left(i,j\right)}{{x_{\max \left(j\right)}-x}_{\min \left(j\right)}}}\end{array}$$

(2)

where $x\_max\left(j\right)$ and $x\_min\left(j\right)$ represent the maximum and minimum values of the indicator, respectively, and where $x(i,j)$ represents the normalized eigenvalue sequence of the indicator.

Step 2: Set the projection function $Q\left(a\right)$.

The PP model transforms p-dimensional data $x\ast (i,j)\{j=1, 2, \dots , p\}$ into a one-dimensional projected value $z\left(i\right)$ of $a=\left\{a\right(1),a(2),\dots ,a(p\left)\right\}$.

$$\begin{array}{c}z\left(i\right)={\displaystyle \sum _{j=1}^p}a\left(j\right)x\left(i,j\right),(i=1, 2, \dots , n)\end{array}$$

(3)

$$\begin{array}{c}Q\left(a\right)=S_z{D}_z\end{array}$$

(4)

$$\begin{array}{c}{S}_z=\sqrt{{\displaystyle \frac{{\displaystyle {\sum }_{i=1}^n}{\left(z\left(i\right)-E\left(z\right)\right)}^2}{n=1}}}\end{array}$$

(5)

$$\begin{array}{c}{D}_z={\displaystyle \sum _{i=1}^n}{\displaystyle \sum _{j=1}^n}{\left(R-r\left(i,j\right)\right)}^{\ast }u\left(R-r\left(i,j\right)\right)\end{array}$$

(6)

where $a$ is the unit-length vector, $S_z$ represents the standard deviation of $z\left(i\right)$, $D_z$ represents the local density of the projected value $z\left(i\right)$, and $E\left(z\right)$ represents the mean value of $z\left(i\right)(i=1, 2, \dots , n)$. $R$ is the window radius of the localized density, and $R$ is taken to be 0.1, according to previous studies [48]. $r\left(i,j\right)$ denotes the distance between samples, $r\left(i,j\right)=\left|z\left(i\right)-z\left(j\right)\right|$. $U\left(t\right)$ is a unit step function when $t\ge 0$, $u\left(t\right)$ is 1, and when $t<0$, $u\left(t\right)$ is 0.

Step 3: Set the maximum projection function.

For a given set of samples, the projection indicator function $Q\left(a\right)$ varies only with the projection direction $a$.

The objective function is maximized as follows:

$$\begin{array}{c}Max:Q\left(a\right)=S_z{D}_z\end{array}$$

(7)

Restrictive conditions:

$$\begin{array}{c}s.t.{\displaystyle \sum _{j=1}^p}{a}^2\left(j\right)=1\end{array}$$

(8)

Step 4: Calculate the index.

After the best projection direction $a^{\ast }$ is substituted into Formula (3), the projection value $z^{\ast }\left(i\right)$ of each sample point is obtained. The $z^{\ast }\left(i\right)$ value is the calculated value of the index.

4.1.2. Multi-Intelligent Genetic Algorithms (MIGAs)

Multiple intelligent genetic algorithms (MIGAs) combine the features of a multi-intelligent system (MIS) and a genetic algorithm (GA) to solve optimization problems [49]. The MIGA is encoded with a vector of projection directions, treating each projection direction as an intelligent body stored in an ${ℒ}_x$ grid environment. When the program is initialized, all the intelligences are given an energy equal to the opposite value of the projection indicator. When the program runs, the energy of the intelligence in each neighborhood increases to achieve evolution through competition, Gaussian mutation, and self-learning. The calculation process is as follows: let ${ℒ}_{\mathit{t}}$ be the tth generation intelligent body lattice, where ${ℒ}_{\mathit{t}\mathbf{+}\mathbf{1}\mathbf{/}\mathbf{3}}$ and ${ℒ}_{\mathit{t}\mathbf{+}\mathbf{2}\mathbf{/}\mathbf{3}}$ are the intermediate generation intelligent body lattices between ${ℒ}_{\mathit{t}}$ and ${ℒ}_{\mathit{t}\mathbf{+}\mathbf{1}}$. $Bes{t}^t$ is the optimal intelligence among ${ℒ}_0, {ℒ}_1,\dots ,{ℒ}_t$ $Cbes{t}^t$ is the optimal intelligence in ${ℒ}_{\mathit{t}}$; $Energy({ℒ}_{\mathit{t}})$ is the energy of the intelligence; and ${\mathcal{P}}_c$ and ${\mathcal{P}}_m$ are preset parameters that control the execution of the domain orthogonal crossover operator and Gaussian variational operator, respectively.

(1) Initialize the parameters and smart body grid ${ℒ}_{\mathbf{0}}$, calculate the energy of each smart body, find the smart body with the largest energy $Cbes{t}^0$, and t = 0 to start the iterative loop.

(2) The neighborhood competition operator is executed for each smart body, and a new smart body lattice ${ℒ}_{\mathit{t}\mathbf{+}\mathbf{1}\mathbf{/}\mathbf{3}}$ is obtained.

(3) If U (0, 1) < ${\mathcal{P}}_c$, perform an orthogonal crossover operator on the intelligence in ${ℒ}_{\mathit{t}\mathbf{+}\mathbf{1}\mathbf{/}\mathbf{3}}$ to obtain the intelligence lattice ${ℒ}_{\mathit{t}\mathbf{+}\mathbf{2}\mathbf{/}\mathbf{3}}$.

(4) If U (0, 1) > ${\mathcal{P}}_c$, execute the Gaussian variation operator on the intelligences in ${ℒ}_{\mathit{t}\mathbf{+}\mathbf{2}\mathbf{/}\mathbf{3}}$ to obtain the intelligence lattice ${ℒ}_{\mathit{t}\mathbf{+}\mathbf{1}}$.

(5) Find $Cbes{t}^t$ from ${ℒ}_{\mathit{t}\mathbf{+}\mathbf{1}}$ to perform the self-learning operator.

(6) If $Energy(Cbes{t}^t)$ > $Energy(Bes{t}^t-1)$, then $Bes{t}^t$ = $Cbes{t}^t$; otherwise, $Bes{t}^t$ = $Bes{t}^t-1$.

(7) Determine whether the number of evolutionary generations reaches the maximum number of iterations N; if it reaches N, output $Bes{t}^t$; otherwise, assign t + 1 to t and go to step (2).

(8) According to the $Bes{t}^t$ projection indicator function, the results of the model operation are the output.

Combining the PTM model described in Section 4.1.1, we constructed a MIGA-PTM evaluation model to analyze the levels of LES and HAD in the HRB. Using the evaluation index system for LES and HAD, which includes the standard layer and the indicator layer established in Section 2, along with the collected data, the MIGA-PTM was employed to calculate the output, resulting in the optimal projection eigenvalue, specifically the LES and HAD evaluation values for the HRB. The primary objective of this analysis is to quantify the respective development levels of LES and HAD in the HRB and to explore the coupling relationship between them by integrating the quantified evaluation values. This provides a reference for policymakers to formulate strategies that enhance the coordination between LES and HAD in the HRB.

4.2. Natural Breakpoint Categorization (NBC)

Natural breakpoint classification (NBC) is also known as the Jenks optimization method [50]. The goal of NBC is to minimize the variance within each category while maximizing the variance between the different categories [51].

Step 1: The total sum of squared deviations (SDHN) is computed for an array of a particular class in the classification results, and a set of results is denoted as HI, with a mean value, as follows:

$$\begin{array}{c}SDHN={\displaystyle \sum _{i=1}^n}{\left(X_i-\overline{x}\right)}^2\end{array}$$

(9)

Step 2: The sum of squares of the total deviation of the class (SDFN) is calculated for the combination of each range in the classification results, and the smallest value is obtained and denoted as ${\mathrm{S}\mathrm{D}\mathrm{F}\mathrm{N}}_{min}$. The dataset consisting of $N$ elements is divided into $q$ different classes, which in turn yields $q$ different subsets. The total deviation sum of squares for each subset, ${\mathrm{S}\mathrm{D}\mathrm{F}\mathrm{N}}_i$, ${\mathrm{S}\mathrm{D}\mathrm{F}\mathrm{N}}_j$ ..., ${\mathrm{S}\mathrm{D}\mathrm{F}\mathrm{N}}_n$, and sum ${\mathrm{S}\mathrm{D}\mathrm{F}\mathrm{N}}_1$, is computed as follows:

$$\begin{array}{c}{\mathrm{S}\mathrm{D}\mathrm{F}\mathrm{N}}_1={\mathrm{S}\mathrm{D}\mathrm{F}\mathrm{N}}_i+{\mathrm{S}\mathrm{D}\mathrm{F}\mathrm{N}}_j+\cdots +{\mathrm{S}\mathrm{D}\mathrm{F}\mathrm{N}}_n\end{array}$$

(10)

Similarly, it is possible to divide the classification result into other cases containing different subsets and compute the values of ${\mathrm{S}\mathrm{D}\mathrm{F}\mathrm{N}}_2, \dots , {\mathrm{S}\mathrm{D}\mathrm{F}\mathrm{N}}_c^{k_n}$. The class thus obtained is the best classification, and the WRCC level is graded by the natural suspension method so that the WRCC level can be evaluated in a more intuitive way.

4.3. Coupled Coordination Degree Model (CCDM)

4.3.1. Coupling Degree Model (CDM)

The degree of coupling (DC) is used to measure the extent to which different elements or subsystems of a system are interdependent or interact with each other [52]. In a system, the DC is low if the components are relatively independent and have less mutual influence [53]. The degree of orderly development of each subsystem of LES–HDA in the HRB and the degree of orderly development of each subsystem of LES-HDA in the HRB can be analyzed. The CDM is calculated as follows:

$$C={\left(\frac{n}{\underset{i=1}{\mathsf{\Pi }}{U}_i}\right)}^{\frac{1}{n}}$$

(11)

where n is the number of subsystems, U is the comprehensive index of the subsystems, C is the degree of LES-HAD coupling, and the value of C is [0, 1]. A larger C indicates that the coupling state of the two systems is better, and a smaller C indicates that the coupling state of the two systems is worse and tends to be disordered.

4.3.2. Degree of Coupling Coordination (DCC)

The degree of coupling coordination (DCC) is used to analyze the interaction and coordinated development between two or more systems [54]. The higher the DCC, the more the two systems converge in terms of the development goals, speed, and effect. The DCC is calculated as follows:

$$T=\sum _{i=1}^n{\omega }_i{U}_i$$

(12)

where T is the comprehensive evaluation index of the DCC. $\omega$ is the corresponding coefficient of each subsystem, and the subsystems have the same importance; therefore, the coefficients of the subsystems are considered to be 0.5.

$$D=\sqrt{CT}$$

(13)

where D is the DCC and C is the DC. Referring to related studies [55,56], the DCC is divided into grades, and the criteria for the division are shown in

Table 3. Criteria for classifying the DCC.

Interval of DCC	Level	Status of DCC
[0.0~0.1)	1	Extreme disorder
[0.1~0.2)	2	Severe disorder
[0.2~0.3)	3	Moderate disorder
[0.3~0.4)	4	Mildly disorder
[0.4~0.5)	5	Nearly disorder
[0.5~0.6)	6	Barely coordinated
[0.6~0.7)	7	Elementary coordination
[0.7~0.8)	8	Intermediate coordination
[0.8~0.9)	9	Good coordination
[0.9~1.0]	10	Quality coordination

.

4.4. Obstacle Model (OM)

The obstacle model (OM) assesses the extent to which potential barriers or constraints in a system affect system performance or goal achievement [57]. This analysis helps identify and quantify obstacles that may prevent the system from reaching its optimal state or goal.

Step 1: Determine the factor contribution $F_{ij}$:

$$F_{ij}=z^{\ast }\left(i\right)$$

(14)

Step 2: Calculation of the deviation $I_{ij}$ for each indicator for each year:

$$I_{ij}=1-X_{ij}$$

(15)

In the formula, $X_{ij}$ is the value normalized for each indicator.

Step 3: Calculation of the degree of impediment:

$$Q_{ij}={\displaystyle \frac{F_j{I}_{ij}}{{\displaystyle \sum _j}{F}_j{I}_{ij}}}\ast 100\%$$

(16)

4.5. Gray Predictive Model First-Order Univariate Model (GM (1, 1))

The gray prediction model first-order univariate model gray model (1, 1)) is used for the prediction of sequences with incomplete information, particularly when the data sequence is short and uncertain [58]. The basic idea of the GM (1, 1) model is to describe the trend of the system by constructing a differential equation and then transforming the original data sequence into a more easy-to-analyze sequence via the accumulated generation (AGO) sequence, thus enabling the prediction of system behavior [59]. In this study, GM (1, 1) is integrated with the LES from the period of 2010–2022 for the HRB and the DCC of the HAD to forecast the DCC for the period of 2025–2040. The formula for GM (1, 1) is as follows:

Step 1: After establishing the GM (1, 1) model, the gray color can be obtained via the differential-equation-predicted discrete time response function, which is calculated as follows:

$$\begin{array}{c}{X}^{\left(1\right)}\left(k+1\right)=\left(X^{\left(0\right)}\left(1\right)-{\displaystyle \frac{b}{a}}\right)e^{-ak}+{\displaystyle \frac{b}{a}},k=1,2,\cdots ,n-1\end{array}$$

(17)

where $X^{(1)}(k+1)$ is the cumulative predicted value, $X^{(0)}(1)$ is the original series, a is the developmental gray number, b is the gray role quantity, n is the number of data points, and e is an irrational constant.

Step 2: Check the difference test:

$$S_1=\sqrt{{\displaystyle \frac{{\displaystyle \sum _{n=1}^n}{\left(x^{\left(0\right)}\left(k\right)-{\overline{x}}^{\left(0\right)}\right)}^2}{n-1}}}$$

(18)

$$S_2=\sqrt{{\displaystyle \frac{{\displaystyle \sum _{n=1}^n}{\left({\Delta }^{\left(0\right)}\left(k\right)-{\overline{\Delta }}^{\left(0\right)}\right)}^2}{n-1}}}$$

(19)

$$C={\displaystyle \frac{S_2}{S_1}}$$

(20)

$$\begin{array}{c}p=\left\{\right|{\Delta }^{\left(0\right)}\left(k\right)-{\overline{\Delta }}^{\left(0\right)}|<0.6745S_1\}\end{array}$$

(21)

where S₁ is the standard deviation of the original series $X^{(0)}$, S₂ is the standard deviation of the residual series ${\Delta }^{(0)}$, C is the variance ratio, P is the probability of small residuals, and n is the amount of data. C is used for the model accuracy level test, and the smaller the better, the smaller the better, and the better, the value of C ≤ 0.35; thus, the model accuracy is excellent, and the larger the better, the better, and the better, the value of p > 0.95; thus, the model accuracy is excellent. The classifications are listed in

Table 4. GM (1, 1) precision divisions.

Variance Ratio (C)	Small Residual Probability (p)	Model Accuracy
(0, 0.35)	(0.95, 1.00)	Excellent
(0.35, 0.50)	(0.80, 0.95)	Pass
(0.50, 0.65)	(0.70, 0.80)	Barely Pass
(0.65, 1.00)	(0, 0.70)	Substandard

.

5. Results

5.1. Characteristics of the Spatial and Temporal Evolution of the LES in the HRB

5.1.1. Model Building and Validation

After the standardization of the indicator system, SPSS 22 software was used to calculate the data, and the data for the HRB municipalities were obtained from the credibility of the average Cronbach’s alpha value of 0.821, indicating that the internal consistency of the indicator system is high and that the system is credible.

In this study, the MIGA-PTM LES evaluation model was constructed, and after testing and referring to the existing research results [60], the values of some of the parameters in the model were as follows: the size of the smart body grid Lsize = 30, the competition probability Po = 0.2, the crossover probability Pc = 0.1, and the variation probability Pm = 0.1. After testing, the model cycle iteration number reaches 100, and the sample projection value is no longer optimized; therefore, the maximum number of cycle iterations Gen = 100. To ensure the accuracy of the evaluation of the program, the evaluation needs to be tested before the MIGA-PTM model is established. Considering that there are many indicators in the evaluation of LES and that the validation process is complicated, five indicators (K₁–K₅) with better representativeness and higher sensitivity were selected. In accordance with previous studies [61], three additional dummy indicators, K₁′ ≡ 1, K₂′ = 1−K₂, and K₃′ = K_3, were randomly added to test the model via correlation analysis between variables to ensure that the true global optimal solution was found. The results of the model test runs are listed in

Table 5. Verified values of the metrics.

Year	Inspection Indicators
	K1	K2	K3	K4	K5	K1’	K2’	K3′
2010	0.6253	0.4541	0.521	0.4521	0.785	1	0.5459	0.521
2011	0.652	0.4421	0.5121	0.4451	0.761	1	0.5579	0.5121
2012	0.6211	0.3854	0.5102	0.4325	0.751	1	0.6146	0.5102
2013	0.61235	0.2854	0.501	0.4251	0.749	1	0.7146	0.501
2014	0.6105	0.5842	0.4855	0.431	0.732	1	0.4158	0.4855
2015	0.6108	0.5645	0.474	0.4115	0.712	1	0.4355	0.474
2016	0.5884	0.4875	0.472	0.405	0.681	1	0.5125	0.472
2017	0.5654	0.4658	0.445	0.395	0.632	1	0.5342	0.445
2018	0.5455	0.4458	0.435	0.384	0.748	1	0.5542	0.435
2019	0.5355	0.4658	0.432	0.375	0.751	1	0.5342	0.432
2020	0.5344	0.4785	0.415	0.365	0.755	1	0.5215	0.415
2021	0.5385	0.5021	0.411	0.362	0.702	1	0.4979	0.411
2022	0.5125	0.481	0.398	0.359	0.757	1	0.519	0.398

.

In Table 5 and

Table 6. Projection index values under various scenario presets.

Scenarios	Projection Vectors
	a1	a2	a3	a4	a5	a1′	a2′	a3′
Scenario 1	0.352	0.432	0.489	0.383	−0.251
Scenario 2	0.465	0.485	0.452	0.395	−0.274	−0.0056
Scenario 3	0.463	0.515	0.438	0.418	−0.285	0.0081	−0.521	0.443

, a₁–a₅ and a₁′, a₂,’ and a₃′ are the optimal projection score vectors for the validation metrics K₁–K₅ and K₁,’ K₂′, and K₃,’, respectively. On the basis of the five screened validation indicators, two validation scenarios of increasing the dummy indicator K′1 and increasing the dummy indicators K₁,’ K₂′, and K₃′ were added. The results of the model validation operations are shown in Table 6. As shown in Table 6, when the three dummy indicators K₁′, K₂′, and K₃′ are added, a₁′ ≈ 0, a₂′≈ −a₂, and a₃′≈ a₃, which indicates that the established model is reasonable and effective. When deleting dummy indicators R₂′ and R₃′ and adding only dummy indicator R₁′, a₁′ ≈ 0, and the coefficients of the projection vectors of the indicators with sample values that are all equal are equal to zero. The projected vector coefficients of the indicators after normalization by cost and benefit type are opposite to each other, and the two indicators with identical values have the same projected vector coefficients, which indicates that the programming and parameter settings of the MIGA-PTM model are effective and that the optimization process is able to find the globally optimal solution.

5.1.2. Changes in the Temporal Dimension of LES in the HRB

To study the LES situation of cities and towns in the HRB from the period of 2010–2022, the validated MIGA-PTM model was used to calculate the projected eigenvalues of the LES. Referring to previous studies [62,63], the LES level is divided into five levels: I (safe level), II (good level), III (critical level), IV (sensitive level), and V (dangerous level). The LES level of each of the 17 municipalities in the HRB during the study period was obtained by evaluating the value of the value, and the average value of the sum of the evaluation data of the obtained areas was used to calculate the LES level of the Hubei, Shaanxi, and Henan sections (Figure 3).

According to Figure 3, the LES of the HRB showed an overall upward trend. As far as the Hubei section is concerned, the average value of the ten municipalities had a minimum of 1.405 and a maximum of 1.661, and most of them were at the critical level (III) or above, showing an increasing trend, and the overall level was relatively high. For the Shaanxi section, the average value of the three urban areas had a minimum of 1.326 and a maximum of 2.306, with three municipalities at a good level (II) or above, and the overall level was the best. The growth rate was faster from 2010 to 2014, declined further from 2014 to 2020, and showed an increasing trend from 2020 to 2022, with an overall fluctuating increasing trend. As far as the Henan section is concerned, the average value of the four cities was 1.22 at the lowest and 1.685 at the highest, with large internal variations. The overall level was poor, but the growth status was stable. In terms of the mean values, the lowest value was 1.354, and the highest value was 1.611, and the level went through two stages, from good to excellent, with the overall level around the critical level (III), and maintaining growth status.

5.1.3. Changing Spatial Dimensions of LES

To study the spatial evolution pattern of the LES level in the HRB, we analyzed the spatial distribution of the assessed value of the HDA level in each city in 2010, 2014, 2018, and 2022; used the standard deviation ellipse (SDE) method to externally map the characteristics of spatiotemporal changes in the LES level during the 12-year period; and processed ArcGIS 10.4.1 to draw the spatial distribution map and standard deviation ellipse map (Figure 4 and Figure 5).

In terms of spatial dimensions, the three cities in the Shaanxi section of the HRB and Shiyan, the Shennongjia Forest District, and the cities of Xiantao and Zhumadian in the Hubei section have relatively good LES values, with LES levels in the Class II good grade range. The Shaanxi and Hubei sections of the HRB in the western part of the HRB have high forest coverage, which helps regulate the climate and maintain water sources and biodiversity [64]. The four cities of Hanzhong, Ankang, Shangluo, and Shiyan have forest coverage rates of 45.4%, 45.3%, 45.2%, and 45.8%, respectively, whereas the Shennongjia forest area has a high rate of 86.5%, which is higher than the national average [65]. Sanmenxia, Luoyang, and Nanyang in the Henan section of the HRB and Wuhan, Xiaogan, and Xiangyang in the Hubei section have poor LES. Sanmenxia city, Luoyang city, and Nanyang city have high population densities, with population densities of approximately 200, 250, and 300 people per square kilometer or more, respectively, far exceeding the provincial average, leading to overdevelopment and increased land use intensity. Wuhan, Xiaogan, and Xiangyang cities have accelerated urbanization and industrialization, with the expansion of residential areas taking up a large amount of agricultural land, leading to a lower land ecological function [66]. The land ecological safety of Jingmen city, Tianmen city, Suizhou city, and Qianjiang city in the central part of the Hubei section of the HRB was in the critical level III interval. The reason for this is the acceleration of industrialization and urbanization in these four cities. Moreover, these four cities face soil pollution problems, with soil heavy metal pollution covering areas of 100 square kilometers, 80 square kilometers, 120 square kilometers, and 90 square kilometers, respectively [67].

Figure 5 shows that the azimuth angle fluctuates and increased from a minimum value of 118.84° in 2010 to a maximum value of 121.06° in 2022, and the ellipse rotated clockwise, which means that the level of ecological security of urban land in the southeastern or northwestern direction developed relatively quickly. The longitude of the elliptical center of the standard deviation of LES in the HRB is at the east longitude 111.79°–111.95°, the latitude varies between 32.05° and 32.17° N, and the center of the standard deviation ellipse is roughly located in Xiangyang.

5.2. Characteristics of the Spatial and Temporal Evolution of the HAD in the HRB

5.2.1. Model Building and Validation

After the standardization of the indicator system, SPSS software was used to calculate the data according to the software used to address the HRB cities and municipalities. The data were obtained from the credibility of the average Cronbach’s alpha value of 0.803, indicating that the internal consistency of the indicator system is high and that the system is credible.

In this study, we constructed the MIGA-PPC HDA evaluation model. After testing and referring to the existing research results, some of the parameters in the model were found to have the following values: smart body grid size Lsize = 30, competition probability Po = 0.2, crossover probability Pc = 0.1, and variance probability Pm = 0.1. After testing, the model iteration number reached 90, and the sample projection value no longer continued to optimize the model. Therefore, the maximum number of loop iterations is Gen = 90. To ensure the accuracy of the program evaluation, the MIGA-PPC model should be tested before evaluation. From the evaluation index system, five indicators (T1–T5) with good representativeness and high sensitivity were selected, and three dummy indicators, T₁′ ≡ 1, T₂′ = 1−T₂, and T₃′ = T₃, were randomly added to test the model via a correlation analysis of the variables to ensure that the real global optimal solution was found. The standardized data are presented in

Table 7. Verifying the values of the metrics.

Year	Inspection Indicators
	T1	T2	T3	T4	T5	T1′	T2′	T3′
2010	0.485	0.689	0.369	0.445	0.747	1	0.311	0.369
2011	0.461	0.685	0.354	0.475	0.725	1	0.315	0.354
2012	0.455	0.605	0.352	0.481	0.711	1	0.395	0.352
2013	0.429	0.654	0.344	0.491	0.704	1	0.346	0.344
2014	0.41	0.645	0.341	0.505	0.688	1	0.355	0.341
2015	0.405	0.638	0.332	0.488	0.654	1	0.362	0.332
2016	0.389	0.585	0.328	0.471	0.662	1	0.415	0.328
2017	0.375	0.574	0.324	0.455	0.641	1	0.426	0.324
2018	0.361	0.562	0.318	0.448	0.633	1	0.438	0.318
2019	0.378	0.524	0.385	0.432	0.613	1	0.476	0.385
2020	0.381	0.511	0.374	0.428	0.585	1	0.489	0.374
2021	0.357	0.485	0.398	0.424	0.584	1	0.515	0.398
2022	0.345	0.477	0.381	0.415	0.532	1	0.523	0.381

.

On the basis of the five screened validation indicators, two validation scenarios (Scenarios 2 and 3) were set up to increase the dummy indicator T′1 and to increase the dummy indicators T′1, T′2, and T′3, respectively, and the results of the validation algorithms are shown in

Table 8. Projection index values under scenario presets.

Scenario	Projection Vectors
	b1	b2	b3	b4	b5	b1′	b2′	b3′
Scenario 1	0.475	0.285	0.335	−0.415	0.521
Scenario 2	0.451	0.344	0.371	−0.325	0.514	−0.0044
Scenario 3	0.388	0.351	0.388	−0.338	0.476	0.0061	−0.348	0.329

.

In Table 7 and Table 8, b₁ to b₅ and b₁′, b₂,’ and b₃′ are the optimal projection score vectors for the validation metrics T₁ to T₅ and T₁,’ T₂′, and T₃,’, respectively. On the basis of the five screened validation indicators, two validation scenarios of increasing the dummy indicator K′1 and increasing the dummy indicators T₁,’ T₂, ‘and T₃′ were added. The results of the model validation operations are shown in Table 8. As shown in Table 6, when the three dummy indicators T₁′, T₂′, and T₃′ are added, b₁′ ≈ 0, b₂′ ≈ −b₂, and b₃′ ≈ b₃, which indicates that the established model is reasonable and effective. When deleting dummy indicators T₂′ and T₃′ and adding only dummy indicators T₁′, b₁′ ≈ 0, and the coefficients of the projection vectors of the indicators with all equal sample values are equal to zero. The projected vector coefficients of the indicators after normalization by cost and benefit type are opposite to each other, and the two indicators with identical values have the same projected vector coefficients, which indicates that the programming and parameter settings of the MIGA-PPC model are effective and that the optimization process is able to find the globally optimal solution.

5.2.2. Time Series Analysis of the HAD in the HRB

To study HDA in the HRB from 2010 to 2022, the MLGA-PPC model was used to calculate the projected eigenvalue of HDA, that is, the level of the HDA. Combined with previous studies [68,69], the HDA level was divided into five levels: I (primary), II (intermediate), III (good), IV (advanced), and V (leading).

As shown in Figure 6, the overall level of HAD showed an increasing trend. In the case of the Hubei section, the average of the 10 municipalities ranged from a low of 1.636 to a high of 2.202, passing through three stages of intermediate, good, and advanced, with a high overall level and a steady state of growth over the decade. For the Shaanxi section, the average value of the three cities had a minimum of 1.571 and a maximum of 1.804, with an overall fluctuating upward trend. Having passed through two stages, from intermediate to advanced, with an average overall level and faster growth in the later stages, the average value of the three-city average for the Shaanxi section was basically below the average evaluation level. As far as the Henan section is concerned, the average value of the four cities ranged from a minimum of 1.105 to a maximum of 1.592, passing through two stages, from primary to intermediate, with an overall lower level and slower growth status. The mean ranged from a low value of 1.424 to a high value of 1.866, ranging from intermediate to good, with an overall average level but an increasing trend.

5.2.3. Analysis of the Spatial Distribution of HAD in the HRB

To investigate the spatial evolution of the level of HDA in the HRB, we analyzed the spatial distribution of the evaluation values in each city in 2010, 2014, 2018, and 2022, used the SDE method to explain the characteristics of spatial and temporal variations during the 12 years, and used ArcGIS 10.4.1 to process and plot the spatial distribution map and standard deviation ellipse map (Figure 7 and Figure 8).

In the spatial dimension, the four cities in the Henan section of the HRB were in a poor HDA state. From 2010 to 2022, the per capita water resources of the four cities in the Henan section declined from approximately 1985 cubic meters to approximately 1544 cubic meters, and the ratio of investment in agricultural scientific research to GDP declined from 0.2 percent to 0.15 percent [70,71]. The three cities in the Shaanxi section of the HRB are in good positions for the HAD, with grades in the Class III range or above. Between 2010 and 2022, three cities in Shaanxi invested CNY 5.133 billion in agricultural development, CNY 3.965 billion in Ankang, and CNY 3.522 billion in Shangluo. The agricultural infrastructure of the three cities in Shaanxi has continuously improved, and the effective irrigated area has increased from 1,477,400 mu in 2010 to 2,205,200 mu in 2022 [72,73]. With the combined effects of policy, infrastructure, science and technology, industrial integration, ecological protection, and market expansion, agriculture in Hanzhong, Ankang, and Shangluo in the Shaanxi section of the HRB achieved high-quality development from 2010 to 2022. Shiyan city, Xiangyang city, Jingzhou city, and Suizhou city in the central and western parts of the Hubei section of the HRB are in a relatively good state of HAD, with grades belonging to the Level II interval or above. The government has built and renovated many new irrigation channels and reservoirs to improve irrigation and water use efficiency. For example, Shiyan city added nearly 1 million mu of effective irrigation area, Xiangyang city added nearly 800,000 mu, Jingzhou city added approximately 700,000 mu, and Suizhou city added approximately 600,000 mu [74].

Combined with Figure 8, the azimuth angle fluctuated from a minimum value of 119.99° in 2010 to a maximum value of 121.07° in 2014, and the ellipse rotated clockwise, indicating that the HAD developed faster in the southeast or northwest direction during that time period. The azimuth angle decreased from 121.07° in 2014 to 118.06° in 2018, and the ellipse rotated counterclockwise, indicating that the HAD in that time period developed faster in the northeast and southwest directions. The azimuth angle increased from 118.06° in 2018 to 118.26° in 2022, and the ellipse rotated clockwise, indicating that the HAD in that period developed faster in the southeast and northwest directions. The longitude of the center of the standard deviation ellipse for HDA in the HRB was between 111.81° and 111.91° east, the latitude was between 31.88° and 31.94° north, and the center of the standard deviation ellipse was roughly located in Xiangyang.

5.3. The DCC and Its Spatial and Temporal Evolution Characteristics

Figure 9 shows the heatmap of the coupled coordination of LES and HDA in the HRB, and Figure 10 shows the temporal and spatial evolution trends. The combination of Figure 9 and Figure 10 reflects the level of coupled coordination between the LES and HDA in the HRB from 2010 to 2022. The coupled coordination of LES and HDA in the HRB from 2010 to 2022 showed a fluctuating trend of change, with the cities in the Hubei section showing an initial increase and then a decrease, and the development of two-system coupled coordination was average. Specifically, the level of two-system coupling coordination in Wuhan, Jingmen, and Suizhou city decreased, and the level of coupling coordination in Qianjiang city increased. The DCCs of the two systems in the Shaanxi and Hubei sections showed similar trends and then increased from 2010 to 2014, decreased in 2016, and increased in 2018. However, the two-system DCC of the Shaanxi section in 2022 was greater than that in 2018, with a clear rebound trend, whereas that of the Henan section increased from 2010 to 2014, decreased in 2016, and increased in 2018. However, the two-system DCC decreased again from 2018 to 2022. According to the heatmap in Figure 9, Sanmenxia city has the most significant color change, with a change amplitude of 126.86%; Luoyang, Zhumadian, Hanzhong, Suizhou, Jingmen, and Wuhan have the next most significant changes, with changes of 65.32%, 26.52%, 18.84%, 18.54%, 15.21%, and 14.33%, respectively, followed by Xiantao, Shangluo, Shennongjia Forestry District, Qianjiang city, and Tianmen city, with changes in magnitude of 7.04, 4.7, 4.62, 4.14, and 4.01%, respectively; and Xiangyang, Shiyan, Xiaogan, Ankang, and Nanyang city are at the lowest level, with magnitudes of change of 3.54, 2.41, 0.74, 0.44, and 0.18%, respectively.

Figure 10 shows the spatial trend of the DC between the LES and HDA in the HRB. The spatial distribution of the DC between the LES and HDA in the HRB in 2010 shows that it is high in the south and low in the north, and that of the 17 municipalities, only Jingmen, Suizhou, and Shennongjia Forestry Area were in the well-coordinated stage. Among the 17 cities, only Jingmen, Suizhou, and Shennongjia Forestry Districts were in the good coordination stage, whereas Xiantao, Tianmen, Shangluo, Ankang, and Hanzhong were in the intermediate coordination stage. Wuhan, Xiangyang, Shiyan, Xiaogan, and Qianjiang were in the primary coordination stage, and Luoyang and Sanmenxia were in the severe dysfunctional stage. In 2014, Suizhou was in the high-quality coordination stage; Xiangyang and Shiyan were in the intermediate coupling coordination stage; Wuhan and Shennongjia Forestry Area were in the barely coordinated stage; Hanzhong and Ankang were in the good coupling coordination stage; Sanmenxia was in the severely coupled coordination stage; Wuhan and Shennongjia Forestry Area were in the barely coordinated stage; Hanzhong and Ankang were in the good coupling coordination stage; and Sanmenxia was in the severely dysfunctional stage. Sanmenxia city rose from the severe coupling coordination stage to the barely coupling coordination stage, Zhumadian city dropped to a mild coupling coordination stage, and the coupling coordination ranks of the remaining cities remained unchanged compared with those in 2010. In 2018, Shennongjia Forestry Zone and Ankang City were in a good coupling coordination stage; Zhumadian and Luoyang City rose by two grades to the barely coupling coordination stage; Xiangyang, Shiyan, and Hanzhong City dropped to the primary coupling coordination stage; Sanmenxia and Nanyang City dropped from the barely coordinated stage to the severely dislocated stage and to on the verge of the dislocated stage, respectively; Suizhou City dropped to the primary coordinated stage; Tianmen City dropped to the barely coordinated stage; and the remaining cities’ DCC levels remained unchanged from 2014. In 2022, Shennongjia Forestry District and Xiantao city are in a well-coordinated stage, and the Tianmen and Qianjiang municipalities rose to the intermediate coordinated stage; Jingmen and Ankang dropped to the primary coordinated stage; Luoyang and Zhumadian dropped to the moderate stage; Ankang city dropped to the primary coordination stage; Sanmenxia rose to the mild coordination stage; Nanyang city rose to the barely coordinated stage; and the remaining municipalities’ coupled coordination levels remained unchanged from 2018.

5.4. Analysis of Factors Influencing DCC

According to the OM calculation of the barrier degree of the DCC of the LES and HDA of the HRB in 2022, the barrier factors with significant impacts in each city were screened out, and the barrier factor rankings of the top five barriers in each province and city are listed in

Table 9. Top five obstacle factors.

City	Top Five Indicator Level Barrier Factors
Wuhan	Y9	Y3	X8	X10	X6
Xiangyang	Y9	Y3	X9	X10	Y1
Shiyan	Y9	Y3	X9	X10	Y6
Xiaogan	Y9	Y3	X8	X10	Y5
Jingmen	Y9	Y3	X9	X10	X4
Xiantao	Y9	Y3	X9	X1	X4
Tianmen	Y9	Y3	X9	X10	X4
Qianjiang	Y9	Y3	X9	X10	X4
Suizhou	Y9	Y6	X9	X10	X5
Shennongjia	Y9	Y2	Y3	X4	X1
Hanzhong	Y9	Y6	Y3	X9	X12
Ankang	Y9	Y6	Y3	X9	X10
ShangLuo	Y9	Y6	Y3	X9	X10
Xuoyang	Y5	Y6	Y3	X10	X12
Sanmenxia	Y5	Y6	X9	X10	X12
Zhumadian	Y5	Y6	X9	X10	X5
Nanyang	Y5	Y16	X9	X10	X5

.

A total of 85 samples were screened for the number of barrier factors that ranked in the top five for the 17 municipalities in the HRB in 2022. Of these, 43 were from the LES, 42 were from the AHD, and the number of major barrier factors in the two systems was essentially the same. In the ranking of the number of barrier factors at the indicator level for the 17 municipalities, X10 (water resources per capita) appeared 14 times, with a share of 16.47%; X9 (park green space per capita), Y9 (urban‒rural income coordination level), and Y3 (return on fiscal expenditure) all appeared 13 times; Y6 (effective irrigation rate) occurred 8 times, with a share of 9.41%; Y5 (level of mechanization per capita) and X4 (natural population growth rate) appeared 5 times, accounting for 5.88%; X12 (integrated industrial solid waste utilization rate) and X5 (GDP per capita) each appeared three times, accounting for 3.53%; X12 (population density) and X5 (green coverage) appeared twice, accounting for 2.35% of the total; and X6 (population density), Y1 (agricultural productivity), Y2 (economic effect of agricultural products), and Y16 (comprehensive utilization rate of livestock and poultry manure) each occurred once.

5.5. GM (1, 1) Model Predictive Analysis

The calculation of the DC between the LES and HAD in the HRB via the GM (1, 1) is described in this subsection. First, a rank-ratio test was carried out to judge the applicability of the data series for the model construction. According to the calculation, the level ratio value in the interval [0.982, 1.0098] indicates that the data are suitable for model construction. The level ratio value of the original data was in the interval [0.483–1.74], which did not pass the level ratio test; therefore, the level transformation was carried out, that is, a level transformation value of 1.00 was added to the original value, and the level ratio test values of the level transformed data were in the standard range of the interval [0.857, 1.166], which means that these data are suitable for GM (1, 1) model construction. Next, an a posteriori difference test was performed, and a test C value of 0.219 < 0.35 implies that the model accuracy class was very good. In addition, the small error probability p value of 0.833 < 0.95 implies that the model accuracy was satisfactory. Finally, the average relative error of the HRB was calculated as 0.042. The average relative error was less than 0.05, indicating that the model was accurate. The calculated average residual value was 0.033, which was less than 0.2 and passed the residual test. These four tests show that the prediction of this model is reasonable and can be useful for predicting the DCC of LES and HDA.

Using the GM (1, 1) prediction model, the change in the value of the DCC of the subsystem coordination degree of LES and HDA was predicted from 2025 to 2040 (

Table 10. DCC projections for LES and AHQ in the HRB, 2025–2040.

Year	Real Value	Projected Value	Residual	Relative Error
2010	0.103	0.103	0	0
2011	0.200	0.225	−0.025	−0.125
2012	0.171	0.203	−0.032	−0.187
2013	0.354	0.258	0.096	0.271
2014	0.520	0.455	0.065	0.125
2015	0.644	0.577	0.067	0.104
2016	0.588	0.506	0.082	0.139
2017	0.655	0.532	0.123	0.188
2018	0.558	0.642	−0.084	−0.151
2019	0.569	0.648	−0.079	−0.139
2020	0.685	0.612	0.073	0.107
2021	0.688	0.620	0.068	0.099
2022	0.689	0.610	0.079	0.115
2025		0.682
2028		0.696
2030		0.744
2035		0.792
2040		0.771

). The DC will increase from 0.61 to 0.792 from 2025 to 2035, which is close to the intermediate level of coordination. However, the value will decrease by 2040, which may be related to damage to the ecological balance caused by water resource pollution, eutrophication, soil erosion, etc., and the energy crisis caused by the large consumption of energy in the process of development [22], which requires the HRB to pay more attention to the management and protection of land resources, improve the technical level of agricultural production, and promote the coupling and coordination of the ecological security system of the land and the HDA system.

6. Discussion

Through a comprehensive evaluation, the potential threats and risk points of the LES in the basin can be identified, providing a scientific basis for the development of targeted ecological protection measures. Coupled and coordinated research on these two systems can better balance the relationship between the environment and development, ensuring the promotion of HDA at the same time, without compromising ecological security, to achieve coordinated economic and ecological development. The HRB LES and HAD coupling and coordination studies can be compared with other international watershed studies, which can help to explore global LES and HAD issues and provide Chinese experiences and cases for global LES and HAD coordination studies.

This paper analyzes the connotations, geographical characteristics, and relevant theoretical studies of the LES and HAD methods and proposes an LES and HAD evaluation index system with a universal HRB. Considering that the two evaluation systems are composite systems involving multiple variables and complex constraints, which are inherently open and uncertain, and that traditional evaluation methods have many problems, such as strong subjectivity and weak global search ability, this study adopted and constructed the MIGA-PTM model to evaluate the two systems for empirical research on the current status of LES and HAD in the HRB. Spatial and temporal divergence patterns were analyzed via ArcGIS tools for NBC classification and the SED method. The DCC was used to explore the coupling and coordination relationships between the LES and HAD of the HRB, and the OM was used to explore the main influencing factors. Finally, a corresponding prediction study was conducted via GM (1, 1).

In summary, the spatial distribution of the DCCs of LES and HAD in the HRB shows an imbalance of high in the south and low in the north, with a worse situation in the Henan section and an increasing trend in the overall degree of coupling. The LES levels of the HRB exhibited an increasing trend, and a substantial disparity was observed between cities, which aligns closely with the findings of Su et al. [75]. While their research focused solely on the LES levels of the HRB, our study also emphasized the assessment of HAD, the DC between the two systems, and the relevant influencing factors. The expanded scope of this study not only underscores the significance of the research findings, but also establishes a foundation for future investigations into the HRB. According to the analysis of the degree of obstacles, per capita water resources, greening coverage, and the rate of return on financial expenditures are the main influencing factors. The results of GM (1, 1) indicate that the LES and HAD of the HRB will be close to reaching the middle stage of coupling in 2035 and decline in 2040.

Combined with these findings, a series of policy mechanisms must be explored to enhance the coordination between LES and HAD in order to address their interaction. The first is integrated land use planning: policymakers should implement comprehensive land use planning that prioritizes ecological conservation while accommodating agricultural requirements [76]. The second is the use of agroecological practices: encouraging the adoption of agroecological practices can significantly enhance land ecological security [77]. The third is the use of monitoring and assessment frameworks: establishing robust monitoring and assessment frameworks can facilitate the evaluation of the effectiveness of implemented policies [78]. In summary, addressing the challenges of LES and HAD necessitates a multifaceted approach that includes various policy mechanisms. By adopting integrated and sustainable strategies, it is possible to promote a harmonious relationship between agricultural productivity and ecological health, thereby ensuring long-term benefits for both the environment and local communities.

7. Conclusions

7.1. Research Conclusions

In this study, we constructed an indicator system with two subsystems, LES and HAD, with 17 cities in the HRB as the research objects, and analyzed and evaluated them via the MIGA-PTM. The DCC was used to quantitatively evaluate the coupled and coordinated development of the two systems, the barrier model (OM) was used to identify their main influencing factors, and GM (1, 1) was used to predict the coupled and coordinated degrees of LES and HAD from 2025 to 2040. The results are as follows:

(1) The LES and HAD levels in the HRB significantly vary among cities.

(2) The spatial distribution of the LES and HAD dynamics in the HRB reveals an imbalance characterized by high in the south and low in the north, indicating a fluctuating upward trend.

(3) The key factors influencing the degree of DCC obstacles include per capita water resources, the green coverage rate, the rate of return on fiscal expenditure, and the coordination level of urban and rural residents’ income.

(4) The LES and HAD levels in the HRB are projected to approach moderation by 2035, followed by a decline in 2040.

7.2. Limitations

Research on the coupling and coordination of LES and HAD in the HRB is a complex and important topic that involves a number of disciplinary fields, such as ecology, economics, and geography. However, the current study has several limitations and shortcomings, which are reflected in the following aspects.

(1) The accuracy and coverage of the data are crucial; certain key indicators or regions may be excluded from the analysis because of missing data, leading to limited integrity in the study’s conclusions.

(2) The complexity of environmental change is influenced by intricate feedback mechanisms among the elements of the coupled system, which may vary over time and across different environmental contexts, thereby increasing the overall complexity of the research.

(3) The lack of interdisciplinary integration poses a challenge; LES and HAD encompass various fields, including ecology, economics, and geography, and insufficient integration among these disciplines may hinder the comprehensiveness and depth of the research.

7.3. Future Implications

In response to the limitations noted in Section 7.2, future research should focus on the following aspects:

(1) First, the optimization of data acquisition and processing should prioritize strengthening data collection and updating, improving data accuracy and resolution, and perfecting data integration and sharing mechanisms.

(2) Second, monitoring and prediction related to environmental change should be strengthened, and feedback mechanisms should be explored.

(3) Third, strengthening interdisciplinary cooperation and establishing an integrated research framework should be emphasized.