Spatiotemporal Changes and Influencing Factors of the Coupled Production–Living–Ecological Functions in the Yellow River Basin, China

Abstract

The imbalance in the “production–living–ecology” function (PLEF) has become a major issue for global cities due to the rapid advancement of urbanization and industrialization worldwide. The realization of PLEF coupling and coordination is crucial for a region’s sustainable development. Existing research has defined the concept of PLEF from the perspective of land function and measured its coupling coordination level using relevant models. However, there is still room for improvement in the indicator system, research methods, and other aspects. This work builds a PLEF coupling coordination evaluation-index system based on the perspective of human habitat using multi-source data in order to examine the spatial differences in PLEF coupling coordination level and the influencing factors in the Yellow River Basin (YRB). Using the modified coupling coordination model, the Moran index, spatial Markov chain model, and geographically weighted random forest model were introduced to analyze its spatial and temporal differentiation and influencing factors. The results found that (a) the level of PLEF coupling coordination in the YRB from 2010 to 2022 has been improving, and the number of severely imbalanced cities has been reduced from 23 to 15, but the level of downstream cities’ coupling coordination is significantly higher than that of upstream cities. The probability of cities maintaining their own level is greater than 50%, and there is basically no cross-level transfer. (b) The Moran index of the PLEF coupling coordination level has risen from 0.137 to 0.229, which shows a significant positive clustering phenomenon and is continually strengthening. The intercity polarization effect is being continually enhanced as seen in the LISA clustering diagram. (c) There is significant heterogeneity between the influencing factors in time and space. In terms of importance level, the series is per capita disposable income (0.416) > nighttime lighting index (0.370) > local general public budget expenditure (0.332) > number of beds per 1000 people (0.191) > NO₂ content in the air (0.110). This study systematically investigates the dynamic evolution of the coupled coordination level of PLEF in the YRB and its influencing mechanism, which is of great practical use.

1. Introduction

The Yellow River Basin (YRB), as a significant food production area and environmental barrier in China, has a complex ecological and economic development context; thus, the conflict between production, living, and ecological roles has worsened as urbanization and industrialization have progressed so quickly. Encouraging the YRB “production–living–ecology” function(PLEF)’s linked and coordinated development is crucial for both sustainable socioeconomic development and the quality of life of the local inhabitants. Its development status is not only related to the eco-friendly progress of the YRB, but also to the process of realizing sustainable development in China and even in Asia [1]. The Chinese government has proposed making joint efforts in the three aspects of economy, community and ecology in order to realize sustainable development [2]. In order to achieve high-quality development of the region, it is crucial to uncover the law of coupled and coordinated growth in the YRB’s PLEF and investigate the dynamic evolution and influence mechanism.

The coupling and coordination level of the PLEF has been extensively researched by scholars from numerous disciplines, such as geography [3], urban planning [4], and ecology. Since the 1980s, land suitability and patterns in Western countries have been evaluated; the connotations, mechanisms, evolution, and role of land functions in social development have been analyzed [5,6]; and land use types have been divided on the basis of economic, production, and ecological functions [7]. The concept of the PLEF has received increasing attention [8]. However, with the development of national spatial planning and mixed land use, scholars have begun to classify the PLEF on the basis of the social attributes and human activities of the land [9] and have increasingly examined PLEF coupling coordination. Studies have been conducted from the perspectives of connotation frameworks [10], classification systems [11], functional identification [12], structural evolution [13], driving mechanisms [14], functional conflicts [15], thereby laying a theoretical foundation for PLEF coupling coordination research. In terms of empirical research, scholars often start from the perspective of urban areas, thereby combining expert questionnaires, spatial conflict indices [9], coupling coordination models [16] and other methods to construct a PLEF evaluation-index system. On the basis of land use data [17], cold and hot spot analysis [18] and triangular coordinate analysis [16] methods have been employed to explore the spatial distribution patterns of the PLEF coupling coordination level, and geographic weighted regression models [19] and obstacle models [20] have been adopted to investigate its influencing mechanism. The results have indicated that the relationship between production and living functions is relatively close and that there are significant differences in the PLEF coupling coordination level between Chinese cities, with some cities experiencing imbalances [19]. Other scholars have conducted county-level research, thereby employing the comprehensive overlap analysis method [21] to evaluate the coupling coordination level of the PLEF and methods such as the geographic detector [22] and multiscale geographically weighted regression [23] models to study the coupling coordination level of the PLEF and its influencing factors on a smaller scale. The results indicate that the influence of economic factors on the coupling coordination level gradually increases. In addition, the coupling coordination level of the PLEF has been investigated comprehensively at the national [24], grid [7], watershed [22,25] and other scales.

The above studies focused mostly on the level of PLEF coupling coordination at the national or provincial level, whereas studies on the YRB are lacking. Moreover, there is still room for improvement in terms of indicator system construction, research scales, and research methods. Specifically, indicator systems have often been constructed from the perspective of functional attributes, thereby neglecting the needs of people and the social attributes of the PLEF. Therefore, from the perspective of the human settlement environment, in this study, a PLEF coupling coordination index-evaluation system was constructed on the basis of five major systems, namely, the natural system, social system, human system, residential system, and supporting system. In terms of research scale, existing studies have focused mainly on national or provincial level research areas, but relatively little city-scale research has been conducted in the YRB. Moreover, there is a particular lack of research on the dynamic evolution of the PLEF coupling coordination level. In terms of research methods, coupling coordination models have been applied in existing studies to evaluate the coupling coordination level of the PLEF. However, owing to the small differences in proportion between the systems, the use of traditional coupling coordination models can lead to a decrease in model validity. On this basis, a modified coupling coordination model was introduced to evaluate the coupling coordination level of the PLEF in the region. In addition, spatial Markov chain models were utilized to reveal the dynamic evolution of the PLEF coupling coordination level, thereby accounting for spatial lag effects. Moreover, a geographically weighted random forest (GWRF) model was adopted to effectively identify the factors influencing the PLEF coupling coordination level in the YRB and to address the high-dimensional nonlinear relationships and spatial heterogeneity between the various influencing factors.

In summary, this study’s goal was to (a) construct an assessment framework for the coupling coordination level of the PLEF in the YRB; to investigate the basic characteristics and key issues of the PLEF in the YRB; and to enrich the theoretical framework of associated studies. (b) Using a spatial Markov chain model, the dynamic evolution mechanism of the PLEF coupling coordination level in the YRB was investigated, and the underlying development patterns were identified. (c) On the basis of considering spatial heterogeneity, the GWRF model was used to reveal the importance of different variables for the coupling coordination level of the PLEF, and their influencing mechanisms were explored, thereby providing empirical suggestions for improving the coupling coordination level and high-quality advancement of the PLEF in the YRB. In Figure 1, the study framework is displayed.

2. Data Processing and Study Area

2.1. Construction of the PLEF Indicator System

In the 1950s, the Greek scholar Sadias first proposed a complete theoretical basis for the human settlement environment, which he believed encompassed five major systems: the ecosystem, the social system, the human system, the residential system, and the supporting system [26]. Among them, the natural system refers to the natural resources and ecological environment closely related to human habitation, whereas the social system reflects the organizational structure and operating laws of human society. The human system mainly includes meeting human material needs and physiological behaviors. The residential system refers to the spatial places that satisfy human habitation needs, and the supporting system primarily includes the necessary infrastructure for human daily life [27,28,29]. The establishment of intensive and efficient regional living environments is an important goal of PLEF coupled and coordinated development, and the significance and connotations of these two aspects are highly compatible [30] (Figure 2). A PLEF coupling coordination level index-evaluation method was developed from the viewpoint of the human settlement environment based on the literature and the actual conditions in the YRB (Figure 2). To ensure the systematic and accurate selection of indicators, Pearson correlation analysis was conducted to eliminate highly correlated variables (Pearson’s > 0.8), resulting in 33 indicators (

Table 1. “production–living–ecology” function(PLEF) coupling coordination evaluation-index system from the perspective of the human settlement environment.

Target Layer	Criterion Layer	Indicator Layer	Indicator Weight	Indicator Attribute
Production function	Agricultural production level	Rate of land reclamation	0.0977	+
		Grain yield	0.0964	+
		Total value of production from fisheries, forestry, agriculture, and animal husbandry	0.105	+
	Nonagricultural production level	Corporate density	0.1048	+
		Financial and insurance industry level	0.1428	+
		Percentage of the tertiary and secondary industries	0.0405	+
	Economic development benefits	Economic density	0.1099	+
		Gross domestic product	0.1175	+
		Total imports and exports	0.1843	+
Living function	Social security	Coverage rate of pension insurance	0.0772	+
		Unemployment rate	0.0985	−
	Public service capability	Education rate	0.064	+
		Number of hospital beds per 10,000 residents	0.0834	+
		Public finance budget expenditure	0.0725	+
	Urban infrastructure level	Per capita road area	0.0714	+
		Per capita electricity consumption	0.1209	+
	Population carrying capacity	Population density	0.0921	+
		Urbanization rate	0.463	+
	Quality of life of the residents	Per capita general consumption expenditure	0.0564	+
		Average salary for employees and staff	0.102	+
		Total consumer goods sales at retail	0.1152	+
Ecological function	Ecological foundation	Forestland region	0.1388	+
		Grassland region	0.1145	+
		Water area	0.1063	+
		Value of ecological services	0.0926	+
	Ecological pressure	Industrial wastewater discharge volume	0.1122	−
		Sulfur dioxide emissions	0.0782	−
		PM2.5	0.0627	−
	Ecological environment condition	Per capita green space in a park	0.0815	+
		Rate of green coverage in developed spaces	0.0234	+
	Ecological environment response	Thorough use of solid waste	0.0988	+
		Harmless treatment rate of garbage	0.0132	+
		Number of environmental professionals	0.0778	+

). The selected indicators exhibited high reliability (Cronbach’s alpha = 0.805 > 0.7), as determined via reliability testing, indicating that they could ensure the reliability of the PLEF coupling coordination measurement results.

The PLEF exhibits multidimensional features. The production function is considered to provide the substance and economic base needed for human survival and development via production and business activities [29]. Given that the YRB is a significant grain-producing region in China, the indicators used should be tailored to the basin’s production characteristics and opportunities for economic change. The production function is characterized by three aspects, namely, agricultural production, nonagricultural production, and economic development benefits, which are reflected in production activities such as agriculture, industry, and finance [31]. The main functions of life are social security, the public service capacity, the urban infrastructure level, the carrying capacity of people, and the quality of life of the residents. Improving the social security system is crucial for enhancing the quality of life of residents and meeting their material and spiritual needs [32,33]. According to the concept of the ecological function, combined with the fragile and sensitive ecological environmental background of the YRB, the ecological function should include four aspects: ecological foundation, environmental stress, condition of the ecological environment, and reaction of the ecological environment [34]. The ecological foundation is the basis of ecological functions, which is important. The ecological pressure reflects the environmental pollution conditions in a given region, and the corresponding ecological environment characterizes the capability of the area to react to ecological and environmental problems (Table 1) and standardize each indicator.

2.2. Selection of Factors Influencing the PLEF Coupling Coordination Level in the YRB

In order to investigate the elements impacting the PLEF coupling coordination level in the YRB, indicators were chosen from the economic level, social development, and terrain environment aspects based on previous research findings [20,35,36] (

Table 2. Factors influencing PLEF coupling coordination.

Criterion Layer	Characterization Layer	Indicator Layer	Symbol	VIF
Economic level	Economic development	Per capita GDP	RGDP	6.532
	Living standard	Per capita disposable income	DPI	7.032
	Expenditure level	Engel’s coefficient	EC	5.331
	Economic vitality	Nighttime light index	NTLI	4.566
	Income	Average salary of staff and workers	WAGE	8.134
Social development	Level of public service facilities	General public budget expenditure	GPBE	2.321
	Urbanization level	Urbanization rate	URBAN	3.465
	Medical resources	Number of hospital beds per thousand residents	HBS	7.923
	Traffic level	Per capita road area	RPA	6.021
	Communication resources	Communication coverage rate	CC	7.971
Terrain and environment	Relief	Slope	AC	3.512
	Vegetation level	Fractional vegetation cover	FCV	2.125
	Hydrological conditions	Hydrological index	HI	5.160
	Air quality	NO2	NO2	2.913
	Land use degree	Land development intensity	LOD	5.459

). The results of a multicollinearity test of the 15 indicators that were chosen showed that there was no multicollinearity because all of the variance inflation factor values were less than 10.

The economic level is a key metric for assessing the health of the local economy and the standard of living of its citizens [37]. Local residents’ economic development and living standards are reflected in the GDP per capita [8], their quality of life is reflected in their disposable income, their consumption level is reflected in their consumption expenditures per capita, the vitality of regional development is reflected in the nighttime light index, and the average salary of on-the-job employees reflects their income level [38]. Social development is the core element of PLEF coupling coordination [39], and the level of community service facilities reflects the completeness of local infrastructure construction, thereby becoming the focus of social attention. The degree of urbanization influences both the quality of the regional natural environment and the growth of regional cities [38]. Medical resources are related to the ability to ensure the health and safety of local residents. The level of transportation reflects the commuting and production efficiency of local residents. Communication resources represent the convenience of external communication for local residents. In terms of the terrain and environment, five indicators were selected, namely, regional average slope, vegetation coverage, hydrological index, NO₂ content in air, and land openness intensity [40], which reflect the regional terrain undulation, vegetation coverage level, hydrological conditions, air quality, and land use degree, respectively. Finally, each indicator was standardized.

2.3. Research Methods

2.3.1. Data Standardization

We used the following formula to normalize the raw data in order to remove the impact of the scale between variables:

$$X_{ij}^{\prime }=(X_{ij}-{\beta }_{ij})/({\alpha }_{ij}-{\beta }_{ij}) \mathrm{if} {\mathrm{X}}_{\mathrm{i}\mathrm{j}} \mathrm{positive\; indicator}$$

(1)

$$X_{ij}^{\prime }=({\alpha }_{ij}-X_{ij})/({\alpha }_{ij}-{\beta }_{ij}) \mathrm{if} {\mathrm{X}}_{\mathrm{i}\mathrm{j}} \mathrm{negative\; indicator}$$

(2)

where X_ij^′ and X_ij represent the standardized and the primitive values of index j in year i, respectively; α_ij and β_ij represent the maximum and minimum values of index j in year i, respectively.

2.3.2. Revised Coupling Coordination Model

In this study, the degrees of mutual promotion and constraint between production, living and ecological functions were explored via coupling models. However, owing to the small differences between the three systems, the use of traditional coupling coordination models will amplify the dependence of coupling coordination on the development level of each system itself, making it difficult to truly reflect the coordination level between the systems [41] and eliminate the value and significance of measuring the coupling coordination level. Therefore, a modified coupling coordination framework was used to overcome this problem. This can be expressed as follows:

$$C=\sqrt{\left\{\left.\left[1-{\displaystyle \frac{\sqrt{\left\{{\left.\left(U_3-U_1\right.\right)}^2\right\}}+\sqrt{\left\{{\left.\left(U_2-U_1\right.\right)}^2\right\}}+\sqrt{\left\{{\left.\left(U_3-U_2\right.\right)}^2\right\}}}{3}}\right.\right]\times \sqrt{{\displaystyle \frac{U_1}{U_3}}\times {\displaystyle \frac{U_2}{U_3}}}\right\}}$$

(3)

$$T={\alpha }_1{U}_1+{\alpha }_2{U}_2+{\alpha }_3{U}_3$$

(4)

$$D=\sqrt{\left\{C\times T\right\}}$$

(5)

where U₁, U₂, and U₃ are the assessment values of the production, living, and ecological levels, U₃ is the assumed maximum value among the three systems, the minimal value is U₁. The coupling degree is denoted by C; T stands for the coordinated development index, and the degree of coupling and coordination between the three systems is indicated by D, the coupling and coordination index of the production, life, and ecological functions. According to existing research, urban development should simultaneously balance production, life, and ecological functions. Therefore, α₁ = α₂ = α₃ = 1/3.

On the basis of the actual situation in cities within the YRB, the coupling coordination index was separated into five categories: severe imbalance, mild imbalance, basic coordination, good coordination, and high coordination (

Table 3. PLEF coupling coordination level types.

Coupling Coordination Index	Coupling Coordination Type	Characteristic
0 < D ≤ 0.3	Serious imbalance	Each function has a very limited degree of growth, and there are substantial mutual constraints and very little coordination between functions.
0.3 < D ≤ 0.4	Mild imbalance	Each function has a comparatively modest degree of development, and there is little coordination between them.
0.4 < D ≤ 0.6	Basic coordination	Each function has a comparatively high degree of growth, and the phenomenon of mutual constraint steadily diminishes while there is some mutual promotion.
0.6 < D ≤ 0.7	Good coordination	Each function has a comparatively high degree of development, which has a positive promoting impact and demonstrates a tendency of coordinated development.
0.7 < D ≤ 1	High coordination	There is a high degree of connection and coordinated development between the functions, and the different functions’ levels of development and mutual reinforcement are noteworthy.

).

2.3.3. Spatial Analysis Method

This research used global Moran’s I to investigate spatial agglomeration phenomena and examine the geographic correlation of the degree of coupling coordination between the production, living, and ecological functions in different cities in the YRB. Spatial correlation patterns were analyzed using local Moran’s I. This can be expressed as follows:

$$\mathrm{Moran}’\mathrm{s} I_{\mathrm{global}}={\displaystyle \frac{{\sum }_{i=1}^n{\sum }_{j=1}^n{W}_{\mathit{ij}}\left.\left(y_i-\overline{y}\right.\right)\left.\left(y_j-\overline{y}\right.\right)}{S^2{\sum }_{i=1}^n{\sum }_{j=1}^n{W}_{\mathit{ij}}}}$$

(6)

$$\mathrm{Moran}’\mathrm{s} I_{\mathrm{local}}={\displaystyle \frac{\left(y_i-\stackrel{\mathrm{-}}{y}\right){\sum }_{j=1}^n{W}_{\mathit{ij}}\left(y_j-\stackrel{\mathrm{-}}{y}\right)}{{\mathrm{S}}^2}}$$

(7)

where n is the quantity of cities; W_ij is the spatial weight matrix in between cities i and j, and the roof proximity weight matrix is used in this study; Y_i and y_j are the PLEF coupling coordination levels in cities i and j, respectively; “y” is the average level of PLEF coupling coordination across all cities; and S² is the sample variance.

2.3.4. Spatial Markov Chain Model

A transition probability matrix is provided using the Markov chain to show each person’s dynamic development process throughout time [42]; the probability distribution and evolution trend of different phenomena in each city can be calculated, and the development characteristics of PLEF coupling coordination in cities in the YRB can be revealed. However, traditional Markov chain models cannot capture the important role of spatial spillover effects in the evolution of the inherent state of a city. Therefore, on the basis of the classic Markov chain model, the notion of the spatial lag was established [43,44] to explore the influence of the coupling and coordination level between the domain and the city on the likelihood of self-transition. In contrast to traditional Markov chain models, conditional transition matrices for k × k were generated, denoted as ${\mathrm{P}}_{ij|\lambda }^{t t+d}$. These matrices indicate the probability of the coupling coordination index of a given city changing from types i to j after d years, under the condition of a type-λ urban spatial lag coupling coordination index in the current year. This can be expressed as follows:

$${Lag}_a=\sum _{b=1}^n{(Y}_b{W}_{ab})$$

(8)

where N represents all of the cities; the spatial weight matrix is called W_ab, and the rook adjacency weight matrix is used herein; and Y_b is the observation value of the PLEF coupling coordination level in city b.

2.3.5. Geographically Weighted Random Forest Model

The geographically weighted regression and random forest models’ benefits are combined in the GWRF model, a spatial machine learning method [45]. The spatial variation coefficient model [46] serves as the foundation for the GWRF model and the research area is divided into multiple smaller regions, after which the random forest model is employed for separate modeling. The data do not need to follow a Gaussian distribution. Hence it is possible to compute the geographic variation link between each city’s explanatory and dependent variables [47] and allowing for spatial instability [48] and better management of spatial heterogeneity phenomena. Moreover, the GWRF surface model can account for nonlinear relationships between high-dimensional variables, and owing to its nonparametric nature, the influence of outliers on the model is reduced. leading to a significant improvement in the model’s prediction performance [49,50]. This can be expressed as follows:

$$Y_i=a\left(u_i,v_i\right)x_i+e$$

(9)

where y_i denotes the dependent variable of the i-th city; a(u_i, v_i) x_i denotes the prediction of the calibrated random forest model at position i; (u_i, v_i) denotes the coordinates of spatial unit i; and e is the spatial error term.

The article uses three indicators, R², mean absolute error (MAE), and root mean squared error (RMSE), as evaluation metrics to assess the goodness of fit and overall simulation performance of the model.

$$R^2=1-{\displaystyle \frac{{\sum }_i{\left(y_i-\widehat{y_i}\right)}^2}{{\sum }_i{\left(y_i-\overline{y}\right)}^2}}$$

(10)

$$MAE={\displaystyle \frac{{\sum }_{\left\{i=1\right\}}^n\left|y_i-\widehat{y_i}\right|}{n}}$$

(11)

$$RMSE=\sqrt{{\displaystyle \frac{1}{n}}\sum _{\left\{i=1\right\}}^n{\left(\widehat{y_i}-y_i\right)}^2}$$

(12)

In this study, the principle of impurity was adopted to improve the original model, and the mean decrease impurity (MDI) was selected to measure the importance of factors.

2.4. Data Sources

In this research, data from 2010 to 2022 were evaluated. To ensure spatial consistency of the data, ArcGIS 10.6 was used to uniformly transform the various data coordinate systems. (a) The China Urban Statistical Yearbook and the China Urban Construction Statistical Yearbook for the 2011–2023 period, together with the national economic and social development bulletins of the different cities, were the primary sources of the statistical data. (b) Land use data were sourced from the Wuhan University China Land Cover Dataset (CLCD) (http://doi.org/10.5281/zenodo.4417809, accessed on 25 June 2024). (c) PM2.5 and NO₂ data were obtained from the China High Air Pollutants website of the Qinghai Tibet Plateau Science Data Center. The Multi-Angle Implementation of Atmospheric Correction (MAIAC) aerosol optical depth (AOD) products of the Moderate Resolution Imaging Spectroradiometer (MODIS) were used to fill in spatially lacking data. (d) Digital elevation model (DEM) data were derived from the 30 m DEM terrain data product of the National Aeronautics and Space Administration (NASA), which are processed and improved on the basis of the Shuttle Radar Topography Mission and provide high accuracy. (e) The National Oceanic and Atmospheric Administration’s (NOAA) National Centers for Environmental Information (NCEI) watershed had meteorological stations from which precipitation data were collected. (f) The National Qinghai Tibet Plateau Data Center provided the vegetation coverage index, and the Chinese region’s 250 m normalized vegetation index dataset was used.

2.5. Study Area

The YRB is an ideal research area for the study of the PLEF coupling coordination level, based on the natural watershed division of the Yellow River Water Resources Committee, considering the completeness of the research area and the availability of the data (Figure 3) [1,51]. First, the YRB serves as a crucial environmental boundary, but its geological structure is fragmented, and the ecological environment is fragile [52]. After a long period of large-scale development, many environmental problems have arisen, including severe land desertification, soil erosion, and even extreme weather events such as sandstorms. Second, the YRB spans three strategic locations in the eastern, main, and western-central regions, with a population that accounts for over 30% of the total population and a total area that accounts for 36.91% of the total area of China. Residents’ living conditions vary greatly, and the higher, medium, and lower reaches differ significantly in terms of development level [53]. Third, the YRB’s ecological environment is delicate and has long been controlled by a crude and energy-intensive development model, which has caused the ecological environment in the area to continuously deteriorate. Nonetheless, the area needs to enhance its comparatively low degree of industry and urbanization growth. Consequently, the inconsistency between the two continues to be a significant concern for the YRB’s future growth [54]. Lastly, the YRB is the primary place where national policies are implemented. The YRB’s land use function in the geographical distribution pattern has been impacted by policies including the Belt and Road Initiative, the Western Development Strategy, and the conversion of agriculture to forests [55]. Furthermore, the government has made it apparent that the YRB is crucial to China’s ecological security, and social and economic advancement, and has designated it as a demonstration region for high-quality development. The aforementioned traits have had a significant influence on the YRB’s evolution. In order to optimize regional ecosystems and achieve high-quality sustainable development, it is crucial to investigate the coordinated evolution and influencing mechanism of the PLEF in the YRB.

3. Result

3.1. Temporal Evolution Characteristics of the PLEF Coupling Coordination Level in the YRB

3.1.1. Temporal Evolution Characteristics of the PLEF in the YRB

Based on the PLEF evaluation-index system constructed in the article, combined with the weights of each index, the PLEF level of each city over the years was calculated, and each functional level was split into five levels using the natural breakpoint approach, gradually decreasing from levels I to V (Figure 4). The findings indicate that the PLEF shows a general geographical differentiation phenomenon, with higher ecological function level differences and low values in the west. The three functions indicate the following: (a) there are many cities with relatively low production levels in the YRB, which are primarily concentrated in the higher elevations of Gansu and Ningxia, with significant differences between East China and West China. Cities in Shandong Province consistently showcase a high level of performance, while the level of urban living functions in Shanxi Province has significantly increased. (b) With the high-level cities mostly found in the provincial capital cities and a sharp decline in low-level cities, particularly in the middle and lower reaches of cities like Baiyin, Ankang, Qingyang, and Linfen, the overall living standard function in the YRB shows a U-shaped change trend. (c) With particularly notable trends in Gansu, Ningxia, and Shanxi Provinces, the YRB’s general ecological function level is on the decline. The southern region has the strongest overall pattern of geographical differentiation, followed by the northern region and other regions. The contiguous regions of Longnan, Hanzhong, Ankang, and Nanyang are home to the majority of high-level cities with ecological functions.

3.1.2. Spatial Differentiation Characteristics of PLEF Coupling Coordination in the YRB

Based on the evaluation of PLEF level, the PLEF coupling coordination level in the YRB from 2010 to 2022 was investigated via the modified coupling coordination model (Figure 5). Both the number of cities above the basic coordination level and the total level of PLEF coupling coordination in the YRB continued to rise, according to the data. In 2010, there were many cities with severe imbalances, which were mainly concentrated in southern Gansu, Ningxia, and Shaanxi Provinces, with only a few cities indicating good coordination levels, mainly Xi’an, Taiyuan, Jinan, Tai’an, Linyi, and other cities. The YRB’s badly unbalanced cities dropped from 23 to 15 in 2022, with Jinan, Xi’an, and Zhengzhou exhibiting excellent coordination. From the standpoint of the pattern of spatial dispersion, the coupling coordination level of the PLEF in the YRB showed a spatial differentiation phenomenon of low values in the upstream region, high values in the downstream region, high values in provincial capital cities, and low values in outlying cities, with a significant siphon effect and a relatively stable distribution pattern. Provincial capitals made up the majority of the cities with strong coupling and coordination levels. The overall coupling and coordination levels in Shandong Province’s cities were higher than those in other YRB cities between 2010 and 2022. However, from 2018 to 2022, the coupling and coordination levels in Liaocheng, Dongying, Heze, and Linyi decreased. Cities with low levels of coupling and coordination were concentrated primarily in upstream areas, with Gansu and Ningxia exhibiting the most significant levels. However, by 2022, cities in Gansu Province developed mild imbalances, whereas cities in Ningxia, except for Yinchuan, still exhibited serious imbalance. The majority of the cities in the YRB’s central regions showed just rudimentary cooperation to look at the PLEF coupling coordination level’s spatial distribution pattern. For 2010, 2014, 2018, and 2022, Moran’s I of the PLEF coupling coordination level in the YRB was computed (

Table 4. Moran’s I of the PLEF coupling coordination level in the Yellow River Basin (YRB). (*, **, *** respectively indicate significant levels at p < 0.1, p < 0.05, and p < 0.01).

	2010	2014	2018	2022
Moran’s I	0.137 *	0.202 ***	0.221 ***	0.229 **
Z	1.7361	2.5747	2.7321	2.7633

). The findings showed that the PLEF coupling coordination level in the YRB had a considerable positive clustering impact, as shown by the positive Moran’s I value. The fact that the Moran’s I value was rising showed that the club impact was considerable and that the clustering effect was continuously growing.

Only clustering kinds may be seen in the global Moran’s I data; specific spatial correlation types cannot be seen. Consequently, areas with p < 0.1 may be defined using local Moran’s I, and local indicators of spatial association (LISA) diagrams were produced (Figure 6). Interestingly, the provinces of Shandong and Henan have the most concentration of high–high clustering regions. On the other hand, Gansu Province’s southern region had a considerable number of low–low clustering zones, suggesting that the PLEF coupling coordination level there was typically low. The number of high–high and low–low clustering cities in the YRB expanded dramatically between 2010 and 2022, and the polarization impact between cities became considerable when looking at the number of associated cities.

3.1.3. Dynamic Transfer Characteristics of the PLEF Coupling Coordination Level in the YRB

The coupling coordination level can be divided into five categories: low level (LL), relatively low level (RLL), medium level (ML), relatively high level (RHL), and high level (HL).

Figure 7 (a) shows that the diagonal values of the traditional Markov chain transition matrix were mostly greater than the nondiagonal values, indicating that the city was influenced by its own level and exhibited intertype path dependence, i.e., a self-locking phenomenon, and that the nondiagonal values differed from 0, indicating a probability of upward or downward transition in the coupling coordination level, without cross-level transition. Moreover cities with low coupling levels could hardly transition upward, with a maximum value of only 20.6%. (b) Both the geographical spillover effect of cities with high coupling coordination levels and the agglomeration of cities with low coupling coordination levels are influenced by the neighborhood state. Notably, reciprocal inhibitory effects result in low coupling coordination levels over the long run when nearby cities with low coupling coordination levels are also low-level cities. The probabilities are 0.8% and 3.1% greater than those when the neighborhood state is not considered. When cities are adjacent to a city with basic or good coordination, the probability remains relatively stable compared with that when neighboring states are not considered. Notably, when high-coordination cities are adjacent to each other, there is a 50% probability that their coupling coordination level decreases due to poor competitive relationships. (c) When a city is severely or mildly imbalanced and its neighboring city is a well or highly coordinated city, the probability of upward transfer is greater than 60%, and there may be a possibility of cross-level transfer. When neighboring cities are also severely or mildly imbalanced cities, the probability of upward transfer is less than 20%, indicating that high-level neighboring cities are more likely to drive improvements in the urban coupling coordination level.

The development trend of the PLEF coupling coordination level in cities and their surrounding cities was examined using ArcGIS (Figure 8). The results revealed that (a) most cities and their neighboring cities remained stable without transfer phenomena. (b) Thirteen cities’ coupling coordination level changed while neighborhood circumstances were steady. Among them, Xi’an and Zhengzhou achieved an upward shift in their coupling coordination level under low-level neighborhood conditions, whereas Zibo and Dezhou achieved an upward shift under high-level neighborhood conditions, mainly due to the influence of Jinan. (c) Eleven cities, including Wuwei, Baiyin, Baoji, and Hanzhong, maintained their level of coupling and coordination while that in their neighboring cities increased. Bayannur and Ulanqab experienced a decrease in their level of coupling and coordination, while the level in their neighboring cities increased. Jinan and Luoyang also exhibited an increase in their level of coupling and coordination, while that in their neighboring cities increased.

3.2. Analysis of the Impact Mechanism of the PLEF Coupling Coordination Level in the YRB

To more accurately identify the importance of the different influencing factors for the coupling coordination level, the GWRF model was used to analyze 15 variables in 62 cities in the YRB from 2010 to 2022. After tenfold validation, various parameters were set as follows: bw = 62, ntree = 500, and mtry = 5. The model’s ultimate goodness of fit R² was 0.82, which suggests a high degree of correlation. Simultaneously having low MAE and MSE (mean MAE = 0.1056, mean RMSE = 0.1335) indicates that the model has small prediction errors, good stability, and strong generalization ability.

3.2.1. Overall Analysis of the Influencing Factors

The relative importance of each influencing factor was determined and was calculated as the percentage of the importance of each influencing factor to the total importance (Figure 9). Overall, social development and economic level were the primary determinants of the PLEF coupling coordination level in the YRB from 2012 to 2022. The significance of economic level declined by 2022, and social development took precedence. A change from depending on economic growth to depending on social development resulted from the urban development model’s metamorphosis and the rise in coupling coordination level. The variables affecting the coupling coordination level of the PLEF in the YRB from 2010 to 2022 showed some variations from the standpoint of criteria. However, the primary parameters affecting the coupling coordination level of the PLEF in the YRB throughout the study period were the number of hospital beds per thousand people, the nighttime light index, the general public budget index, the per capita disposable income, and the NO₂ content.

3.2.2. Heterogeneity Analysis of the Influencing Factors

A spatial distribution map of the significance of the influencing factors was developed based on the GWRF model results in order to more precisely examine the spatial heterogeneity in the coupling coordination level of the PLEF in the YRB under the influence of different factors. The significance was categorized using the natural breakpoint method. The findings showed that there was notable regional variability in the significance of each contributing factor.

(a)

Economic level

Figure 10 shows the spatiotemporal distribution of the significance of the factors influencing the coupling coordination level of the PLEF in the YRB from 2010 to 2022, which is based on the economic level. According to the results, the cities exhibited a significant clustering distribution of high-importance areas under the economic level criterion layer, and the nighttime light index exerted the most significant influence on changes in the coupling coordination level, whereas the importance of the per capita disposable income was the lowest. In particular, there was a trend in mutual transformation between high- and low-value areas, as the high-value areas of the per capita GDP gradually moved from the southern part of Gansu Province downstream of the YRB to the northern part of Shan-xi Province and cities like Jinan, Linyi, and Rizhao in Shandong Province. The spatial pattern and changes in the importance of the per capita general consumption expenditure were similar to those in the importance of the per capita disposable income, but the number of high-value areas gradually decreased. In 2022, only six cities, namely, Ulanqab, Binzhou, Rizhao, Hebi, Puyang, and Datong, were high-value areas. The valuable sections of the nighttime light index mainly occurred in the upper reaches of the YRB, gradually shifting from the northeastern part of Gansu Province to the southern parts of Gansu and Shanxi, with fewer low-value areas, which were spread out only in Baotou, Hohhot, Ulanqab, Dezhou, and Zibo. The spatial pattern of the importance of the average wage of on-the-job workers changed significantly. High-value areas shifted from Zhangye and Wuwei to cities such as Bayannur, Ulanqab, and Hohhot, whereas low-value areas shifted from cities in the middle reaches of the YRB to scattered cities such as Yuncheng, Nanyang, and Pingdingshan.

(b)

Social development

The spatiotemporal distribution of the parameters affecting the PLEF’s coupling and coordination level in the YRB between 2010 and 2022 is shown in Figure 11. The importance of general public budget expenditure was greater than that of the other variables and exhibited a sustained and rapid upward trend. From the viewpoint of influencing factors, the high-value areas of general public budget expenditure continued to spread, gradually expanding from Longnan, Hanzhong, and Qingyang in 2010 to most cities in the lower reaches of the YRB, whereas the low-value areas were concentrated mainly in northwestern Shanxi Province and Shandong Province. The importance level of the urbanization rate indicated a development trend of first reducing and then increasing. In Shandong Province, high-value regions moved to most cities from places like Zhongwei, Guyuan, Shangluo, etc., whereas low-value areas were mostly found in Gansu and Ningxia. The high-value areas of the number of beds per thousand residents were mainly focused in the northern cities of the YRB, with weak transfer phenomena, whereas the low-value areas gradually shifted from the upper reaches of the YRB to cities in the middle and lower reaches, such as Taiyuan, Luoyang, Binzhou, Dongying and other cities. The distribution of per capita road space remained relatively stable, with high-value regions primarily found in urban centers like Anyang, Puyang, Jinan, and Linyi, while low-value regions were mostly located in the upper reaches of the YRB. The significance of the communication coverage rate has stayed fairly consistent. In 2010, there were few high-value areas, and by 2022, the southern part of Shaanxi Province, Yuncheng, and Jincheng had gradually developed into high-value areas, whereas the number of low-value areas continued to decrease, thereby shifting from Gansu and Ningxia Provinces to cities such as Bayannur, Baotou, and Nanyang.

(c)

Terrain environment

Figure 12 illustrates the spatiotemporal distribution of the significance of the factors affecting the coupling coordination level of the PLEF in the YRB from 2010 to 2022, focusing on the terrain environment. Overall, the importance of the terrain environment for the changes in the coupling coordination level was relatively low. The importance of the NO₂ content in air was the highest. Specifically, the importance of the urban slope was relatively low, with a value below 0.1, and the high-value regions gradually shifted from the lower reaches to the upper reaches of the YRB. The high-value areas of the vegetation coverage were concentrated mainly in the northern cities of the YRB, such as Bayannur, Baotou, and Ulanqab, whereas the low-value areas mostly occurred in the Ankang–Shangluo–Nanyang area. The high-value areas of the hydrological index were concentrated mainly in Shandong Province in the lower reaches of the YRB, whereas the low-value areas were distributed mostly in Gansu, Ningxia, Shaanxi, and Inner Mongolia Provinces in the middle and upper reaches of the YRB. The areas with a high NO₂ content in air gradually shifted northward from Tianshui, Dingxi, and Hanzhong to cities such as Ordos, Hohhot, and Baotou, whereas the areas with a low NO₂ content showed the opposite trend, shifting from cities in the northern part of the YRB to cities such as Shangluo, Sanmenxia, Linyi, and Rizhao. In 2010, the high-value areas of the land development intensity were distributed in most cities in Gansu and Ningxia, whereas in 2022, only four cities in Zhangye, Datong, Hohhot, and Ulanqab exhibited high value. The low-value areas were mainly concentrated in cities in the southern part of the YRB.

3.2.3. Influence Mechanism of PLEF Coupling Coordination Level in the YRB

The coupling coordination level of the PLEF in the YRB is jointly influenced by the urban economic level, social development, and terrain environment and exhibits significant heterogeneity in both time and space. On the basis of the GWRF results, the impact mechanism of PLEF coupling coordination in the YRB was explained (Figure 13). The economic level is a key factor influencing coupling coordination development of the PLEF in the YRB. The economic level, as an important material foundation for coordinated development of the PLEF, continues to increase with increasing economic level, regional industrial structure efficiency, and regional production functions. Moreover, it can increase residents’ incomes and promote infrastructure improvement. Social progress is the main factor influencing coupled and coordinated development of the PLEF in the YRB. The improvement in urban infrastructure is the main aspect of social development and exerts a significant promoting effect on enhancing PLEF coupling and coordinated development. Sound infrastructure can improve the quality of life and happiness of residents while also increasing the production efficiency and vitality and stimulating economic development [56,57]. The terrain environment is the fundamental factor influencing PLEF coupling and coordinated development in the YRB. The level of vegetation coverage, air quality, and land use can directly affect the regional ecological environment quality. A favorable ecology can provide residents with a more suitable living environment and further improve their quality of life. The achievement of efficient and sustainable production functions, continuous improvement in living functions, and diverse and stable ecological functions is important for continuously increasing the coupling and coordination level of the PLEF in the YRB and promoting a balanced and virtuous cycle between the systems [58].

4. Discussion and Conclusions

4.1. Discussion

One major obstacle preventing the region’s high-quality growth is the YRB’s low degree of PLEF coupling coordination. In order to uncover the dynamic evolution and impact mechanism of the PLEF coupling coordination level in the basin, an evaluation-index system for the coupling coordination level of the PLEF in the YRB was built, its spatial differentiation pattern was investigated, and spatial Markov chain and GWRF models were used. These methods were based on the research paradigm of the pattern process mechanism and from the perspective of a human settlement environment.

Many cities and regions around the world are undergoing rapid urbanization processes. In the Yangtze River Delta region of China, its urbanization process was earlier than that of YRB, therefore, the coupling coordination level of PLEF is higher than that of YRB, and the gap between cities is also smaller [59]. Willemen [60] found through her research on rural areas in the Netherlands that there is overlap and high coupling between PLEF functions. However, in Antalya, Türkiye, with the progress of urbanization, its original PLEF coupling and coordination status is rapidly changing, facing the problem that the original urban development model is difficult to maintain [61,62]. The YRB, as a demonstration area for high-quality development, exhibits a low level of PLEF coupling and coordination due to its unique development background. It has suffered from a series of problems, such as spatial resource waste, ecological system imbalance, and social development imbalance throughout the years [63]. To increase the coupling and coordination level of the PLEF in the YRB and promote high-quality regional development, on the basis of the above research, the following policy recommendations are proposed.

First, the coupling coordination level of the PLEF in the YRB is generally low, which is clearly related to the slow development process and low efficiency in cities within the YRB. Owing to long-term weak production development and inefficient economic growth, improvements in living standards have been slow. In addition, under the influence of the fragile ecological environment of the YRB and the constraints of various policies on ecological protection, the level of production and living function development in the basin is lower than the level of ecological function development [2]. In response to this phenomenon, cities in the YRB can be divided into two categories: high-ecological-function-level cities and low-production- and living-function-level cities and low-ecological-function-level cities and high-production- and living-function-level cities. For cities with higher ecological levels but lower production and living standards, it is necessary to ensure that the ecological environment is not severely damaged, while effectively improving production and living standards, improving urban construction levels, increasing residents’ income, and enhancing residents’ well-being. For cities with high production and living standards but low ecological levels, efforts should be made to improve the quality of the urban ecological environment, by actively adjusting and optimizing industrial structure, developing green industries, and promoting the optimization of the ecological environment quality in order to realize a resource-saving and environmentally friendly economic development model [64]. Moreover, the government should formulate environmental protection policies that meet local needs, enhance environmental pollution control capabilities, and increase capital investment.

Secondly, through the study of the dynamic evolution mechanism of PLEF coordination level in the Yangtze River Delta, it was found that there are significant self-locking and club effects, which are not conducive to transformation from low-level cities to high-level cities. At the same time, it was found that the high level of coupling coordination is mainly concentrated in provincial capital cities or regional center cities, such as Xi’an, Zhengzhou, Jinan and other cities. To solve this problem, high-level cities should leverage their own advantages, establish and improve new development models with surrounding cities, and also establish a collaborative development model throughout the entire basin [9,14]. Notably, when cities with high coupling and coordination levels are neighboring cities, their probability of downward migration is greater. Therefore, it is necessary for cities with high coupling and coordination levels to avoid homogeneous competition, so that they engage in positive interaction, leverage their own advantages, and drive surrounding low-level cities.

Finally, there were significant differences in the impact mechanism of the PLEF coordination level in the YRB from 2010 to 2022, with the main factors shifting from the economic level to social development and from heavy development to the livelihood of people. However, as an underdeveloped region, the urbanization process of some cities in the YRB is still relatively lagging behind, such as Qingyang, Guyuan, Zhangye and other cities. These cities should increase urban public finance expenditures, actively improve the level of urban public service facilities, and narrow the gap with other regions. Moreover, owing to the heterogeneity among the various influencing factors, each region should focus on improving the coupling and coordination level of the PLEF according to its inherent characteristics [4].

Within the context of the new era, the coupling and coordination level of the PLEF in the YRB is closely related to high-quality regional development. In this study, a modified coupling and coordination model was adopted to explore the coupling and coordination levels of the PLEF in the basin, investigate its spatial evolution characteristics and influencing mechanisms, and provide empirical evidence for coupling and coordination development of the PLEF in the basin. There exists a complex game relationship between production, living and ecological functions, and its specific mechanism should be explored further. Although the use of spatial Markov chain models can effectively explain the spatial dynamic evolution mechanism of PLEF coupling coordination level, the model may not accurately represent the actual situation due to its inability to consider heterogeneity and significant external influences, as well as ignoring the dynamic changes over time. Owing to data availability limitations, there is still room for further optimization of the indicator system established in this study. Further research is needed on the mechanism and development model of PLEF coupling and coordinated development in the YRB to provide more detailed and practical policy recommendations for achieving high-quality development.

4.2. Conclusions

From the perspective of the PLEF level, the ecological function level in the YRB is the highest, and the PLEF gap is reflected mainly along the east–west and north–south directions. The high-value areas of the production and living functions are concentrated mainly in Shandong and Henan Provinces, whereas the high-value areas of the ecological functions largely occur in the southern part of Shaanxi Province and Bayannur and Ulanqab cities in Inner Mongolia.

From the perspective of the spatiotemporal pattern of the coupling coordination level, the coordination level of the PLEF in the YRB continuously increased from 2010 to 2022, with an overall spatial distribution of high values in the eastern part, followed by the middle and western parts, with significant polarization effects. Cities with good or high coordination are concentrated mainly in provincial capitals such as Jinan, Zhengzhou, and Xi’an, whereas severely imbalanced cities are distributed mostly in cities such as Guyuan, Ankang, Qingyang, and Bayannur.

From the perspective of dynamic evolution characteristics, there is a significant dynamic migration phenomenon in the coordination level of the PLEF in the YRB from 2010 to 2022. The probability of coupling coordination level transfer varies greatly among cities in different areas, but most cities and their neighboring cities remain stable without transfer phenomena, showing significant path dependence and self-locking phenomena.

From the perspective of influencing mechanisms, there is significant spatial heterogeneity among the different influencing factors, which is reflected mainly along the east–west and north–south directions. Social development has gradually become the main factor influencing the level of coupling coordination, with the per capita disposable income, the nighttime light index, the general public budget index, and the number of hospital beds per thousand residents as key factors influencing the coupling coordination level of the PLEF in the YRB.