Comprehensive evaluation and stability analysis of ecosystem quality of Yellow River Basin during 1980–2019

Quality and stability are the basic characteristics of ecosystems, reflecting their structure, process, functional integrity, ability to resist disturbance, self‐regulation, and dynamic balance. The quantitative description of ecosystem quality and stability is an indispensable and important task to promote the ecological protection and high‐quality development of the Yellow River Basin as a major national strategy. In this study, an index evaluation system was constructed from the perspective of system governance, considering the special characteristics of the Yellow River Basin, and an entropy weight model was used to establish a multi‐indicator long time series ecosystem quality evaluation index (EQI) to quantitatively describe the ecological quality changes in the Yellow River Basin over 40 years. The dissipative structure theory and Brussels apparatus model were used to establish quantitative indicators and methods, and provide ideas for the study of ecosystem homeostasis transformation in the Yellow River basin. This study found that: (1) the average EQI value of the Yellow River Basin ecosystem for 40 years was 63.96, the maximum and minimum values were 69.65 and 59.45, respectively; the overall quality showed an oscillating trend of improvement, with an annual increase of 0.03, and the spatial distribution was better in the downstream than in the midstream, and in the midstream than in the upstream; (2) The key factors influencing the system changed gradually from the total water consumption and total annual precipitation in the basin in the early stage to the average income of residents in the last 10 years. (3) The system steady‐state transition force values from 1980 to 2019 were all <0, the steady‐state conversion force of the Yellow River Basin ecosystem was low, but the overall trend was oscillating upward, and the overall trend was getting closer to the critical condition of steady‐state conversion.

Therefore, quantitatively evaluating the quality and stability of ecosystems and exploring the driving role of changes in ecosystem quality are important research directions for ecosystem protection and restoration, which will provide important scientific tools and theoretical support for promoting ecological civilization construction and high-quality ecological development .
Ecosystem quality is mainly manifested in production capacity, changes after disturbance, and its impact on human survival and sustainable socioeconomic development, whereas ecosystem stability is the response to external disturbance conditions (Liang et al., 2017). Based on the entropy weight method, Zhao and Zhang (2021) constructed a comprehensive evaluation model for Dongting Lake ecosystem quality using the ecosystem productivity, stability, and carrying capacity indices, and Liang et al. constructed an ecological quality evaluation index for the Shiyang River Basin (Folke et al., 2010). Although these studies considered stability as an aspect of ecosystem quality evaluation, more attention had been paid to a single indicator within a certain time and range, namely, the fluctuation of production capacity (Angeler & Allen, 2016). The composition of the stability index was relatively simple and stability was not regarded as a system. The overall characteristics of multiple internal indicators were considered, and the time series of the data was short.
The stability of an ecosystem refers to its response to disturbances from external conditions, which is reflected in the internal adjustment of the ecosystem (Biggs et al., 2009) Ecological dynamics consider the system to have multiple (local) steady states (Arrow et al., 1996) that cross critical thresholds that characterize the key variables of the system, which can be transformed between different steady states (Filatova et al., 2015;Gunderson et al., 2017). Ecosystem homeostasis transitions involve large-scale, sudden, and long-lasting changes (Wang et al., 2020). Loss of system homeostasis may lead to discontinuous ecosystem functions and serious irreversible consequences, such as soil erosion, desertification, groundwater depletion, and biodiversity loss (Reyers et al., 2018). However, some decision-making systems cannot predict or observe such signals, or may misinterpret such signals, and often ignore the dynamic changes in ecosystems (factors such as system state transition thresholds, loss of buffer capacity, resilience [Wu, 1991]). At present, research on the steady-state transition of a system mainly focuses on the identification of steady-state transitions and the analysis of the driving mechanism. Steady-state transition identification methods include statistical and model analyses (Wu, 1996). However, the statistical analysis method requires long-term data collection, (Zhang et al., 2010) and model analysis mainly selects relevant indicators through system dynamics models, equilibrium models, and other methods, establishes an internal feedback mechanism of the system, and identifies steady-state transitions (Ma et al., 2013). The driving mechanism was analyzed based on the identification of state transitions (Chen, Yu, et al., 2015). Wu (Jingping, 2020; believed that the use of dissipative structure theory to analyze and discuss the stability of ecosystems has the characteristics of rationality and accuracy. The Brussels model is a mathematical model for the quantitative analysis of dissipative structures and has been widely used in business management, finance, and other majors . The influence and evolution analysis changes the internal and external factors of a complex water resource system in the overall state system. Correspondingly, for the quantitative evaluation of ecosystem stability, it can also be avoided based on the Brussels model to establish a method to quantitatively describe the steady-state transformation of ecosystems and provide scientific support for the study of ecosystem and multistability and high-quality development of ecosystems. The Yellow River Basin is a giant complex system that contains rich ecological environmental elements, such as rivers, lakes, forests, grasslands, wetlands, deserts, and Gobi (Wang & Zhao, 2018). These elements and the external environment constantly undergo material circulation and energy flow, which maintain the healthy life of the Yellow River. It plays an important role in and constitutes an important ecological barrier (Wang & He, 2022). The Yellow River stretches thousands of miles, connecting the Qinghai-Tibet Plateau, Loess Plateau, and North China Plain. It is an ecological corridor that spans three terraces (Yin et al., 2022). Intensified human activities and unreasonable development and resource utilization have led to continuous changes in the environment of the Yellow River Basin, with prominent problems, such as ecosystem degradation, reduced water conservation capacity, serious soil erosion, tributary pollution, and shrinking wetlands. In recent years, major projects, such as flood control and disaster reduction, water and sediment control, and soil and water conservation have been continuously developed in the Yellow River Basin. Based on the concept of a virtuous cycle, (Zhu et al., 2014) these measures drive the ecosystem to undergo a steady-state transition. Therefore, paying attention to the steady-state transition of the watershed ecosystem, explore its evolution law, and take restoration measures to avoid the loss of the steady state or promote the steady-state transition and ensure the sustainability of the ecosystem is of great significance.
This study used the entropy weight model to construct an index evaluation system under the premise of fully considering the particularity of the Yellow River Basin ecosystem, established a multi-index long-term series of ecosystem quality evaluation indices, and applied it to a comprehensive quality evaluation of the Yellow River Basin ecosystem. To identify the steady-state transition of the ecosystem in the Yellow River Basin under the new situation, based on the characteristics of the Yellow River, this study used the dissipative structure theory and the Brussels device model to conduct in-depth research and establish an evaluation index and method to quantify the steady-state transition of the ecosystem. The positive and negative feedback mechanisms of the external environment provide ideas for research on the steady-state transformation of the ecosystem in the Yellow River Basin and provide a scientific basis for research on high-quality ecosystem development.

| Study area
The Yellow River is the second largest river in China (Figure 1), with a total length of 5464 km and drainage area of 795,000 km 2 . The Yellow River Basin spans the three major regions of the east, middle, and west of the country and constitutes an important ecological barrier in China (Zhao, Yu, et al., 2017). It is an ecological corridor connecting the Qinghai-Tibet Plateau, Loess Plateau, and North China Plain, and plays an important role in ecological security (Du et al., 2021). The Yellow River flows through nine provinces (regions), and its upper reaches include Qinghai, Sichuan, Gansu, Ningxia, and Inner Mongolia. The middle reaches included Shaanxi and Shanxi. The downstream areas included Henan and Shandong. At the end of 2019, the total population of the nine provinces in the basin was 450 million, accounting for about 32% of the country's total population; the regional GDP was 23.9 trillion yuan, accounting for about 24% of the country's total. It is an important economic zone and energy base in China (Zhang, Lian, et al., 2021).

| Data sources
The construction of an ecosystem quality evaluation index system has been studied by many researchers. Zhang, Chen, et al. (2021) started from the four aspects of structure, function, stability, and stress of comprehensive ecosystem structure, productivity, carbon sequestration, water conservation, soil maintenance, functional stability, and human coercion, and constructed an ecosystem quality index system. Although the impact of human disturbance is considered, stability is still characterized by the stability index of the representative index NPP, and Li et al. (2021) believed that the normalization vegetation index, leaf area index, surface water content, and land surface temperature are key indicators for characterizing ecosystems. They only consider ecosystem quality in a narrow sense, ignoring the impact of human society.
From the perspective of system governance, this study comprehensively considered the ecological environment and human social factors to construct an index evaluation system and fully considered the particularity of the Yellow River Basin itself. When calculating the water demand to maintain the ecological health of the lower Yellow River, the transportation of sediment was also considered as the primary factor. Constructing a system with only general indicators of the ecological system will cause a lack and limitation of the typical characteristics of the Yellow River Basin,  and therefore, the total water volume and sediment volume were selected as the representative indicators. There is a great contradiction between water use, and the utilization rate of water resources also leads to changes in the ecosystem. Therefore, relevant indicators of human economy and society are included in the evaluation system (Xi J P. attends deliberations of Qinghai delegation-News Report-Communist Party of China News, 2021).
This study selected 22 indicators based on the principles of systematicity, comprehensiveness, and availability to construct a comprehensive evaluation system and method for the ecosystem quality of the Yellow River Basin (Yin et al., 2022). Each level selected representative indicators to reflect changes in hydrological conditions, water quality, habitat quality, human activities, economic development, F I G U R E 1 Distribution of the main basins of the Yellow River. CAO ET AL. and so forth. Specific indicator systems are presented in Supporting Information: Table S1. Hydrological condition includes annual runoff, discharge and sediment, load guaranteed rate of ecological base flow at important sections. Water environment factors include water quality compliance rate of important water function areas, the proportion of tributaries whose water quality is better than that of Class III rivers. Ecological factors include habitat quality index, vegetation cover index, water network density index, land stress index, area of soil and water loss control on the Loess Plateau, typical regional wetland area change rate. Climate factor include total annual precipitation, average temperature. Social development factors include permanent residents, urbanization rate, per capita disposable income of urban residents, GDP growth rate, GDP per capita, night light index. Water resources include irrigated area, water consumption rate, total water use in the basin.
The data types of indicators mainly include basic data and industry (professional) data. The basic data are mainly remote sensing, mapping and geographic information data, which is provided by NASA (https://modis.gsfc.nasa.gov/), has a temporal resolution of 8 days and a spatial resolution of 1 km. The industry (professional) data are relatively mature spatial and temporal series data that are widely used from the daily business and research of the Yellow River Basin. The time span of the collected data is 1980-2019, totaling 40 years. Due to certain historical reasons, a small amount of data are missing, and according to different situations, after reasonable analysis, common linear interpolation, spline interpolation, Lagrangian interpolation and gray prediction method are used to fill in the data. For sequences with a small amount of missing data and intermediate data, segmented linear interpolation is used; for sequences with large missing data, spline interpolation or Lagrangian interpolation is used; for missing data at the endpoints, gray prediction method is used to make up the data. Due to the multisource heterogeneity and quality of the data, the currently collected data cannot be interpreted spatially and can only be analyzed basin-wide in the temporal dimension.

| Entropy weight model
The entropy weight model is a commonly used mathematical evaluation model for the comprehensive index method (Du et al., 2021). In this study, the entropy weight model was used to comprehensively evaluate the ecosystem quality. For the specific method employed in this study (Zhang, Lian, et al., 2021): 1. Possibility function: (1) 2. Index entropy: where f ij is the probability function, p ij is the proportion of each standard possibility function value to all values, and n n is the number of standard intervals of the indicator. 3. Entropy weight: where N is the number of corresponding indicators, S i and w i represent the entropy value and weight of corresponding indicators, respectively. 4. Quality evaluation index: The larger the total entropy value S, the lower the quality; thus, we defined the comprehensive ecological quality evaluation index (EQI) as: The ecosystem of the Yellow River Basin is a giant open and complex system that is far from equilibrium and is composed of rivers, the ecological environment, the social economy, and other elements. Based on the environmental development index (EDI) proposed in a previous study to evaluate the ecological environment status of the river basin, this study comprehensively considered the interaction of rivers, ecological environment, social economy, and other elements in the river basin, and constructed a comprehensive evaluation of the ecological quality of the river basin (Zhang, Chen, et al., 2021). EQI was analyzed through entropy weight coupling. The larger the value, the higher the development quality of the watershed, and the moirtuous the circle; the smaller the value, the lower the development quality of the watershed, and there are certain problems in the state.
(4) Contribution degrees: Where W i is Entropy weight, N is the total number of indicators.

| Brussels model
Stability is an important aspect of ecosystem quality. From different levels of an ecosystem, stability can be population, community, system, or ecological function stability; from different temporal and spatial scales, ecosystems can be divided into long-term or short-term stability and so on. Steady-state transition studies mainly involve ecologically medium time scales (decades), which are close to the time scales that humans can easily perceive or manipulate; therefore, it is necessary to pay attention to steady-state transitions to help humans manage ecosystems and ensure their sustainability (Xie, 2013). Holling (Holling, 1973) suggested that an original ecosystem experience many different stable equilibrium states. Under the influence of human activities (resource utilization, pollution, etc.), an ecosystem may change from one stable equilibrium state to another, leading to serious ecological consequences, such as species endangerment. The multistability of a system can be expressed using the stability landscape method that intuitively describes the dynamic characteristics of the system (Figure 2).
Although a steady-state landscape map can use the meaning of the ecological landscape to describe the abstract concept of steady-state transition, it cannot quantitatively describe the steady-state transition within the system. Ecosystems are complex, and their internal components exhibit nonlinear correlation characteristics, fluctuation characteristics of system states, and spatiotemporal heterogeneity within the system (Wang, 2008). Therefore, an ecosystem is considered to be a kind of dissipative structure and a complex system, (Wu, 1991) which satisfies the characteristics of dissipative structure openness far from equilibrium, nonlinearity, fluctuation, and sudden change. Therefore, dissipative structure theory can be used to study the dynamic mechanism of an ecosystem and quantify the internal steadystate transformation driving force of the system. The Brussels device model is a mathematical model proposed by Prigogine that can quantitatively analyze the dissipative structure. Internal positive entropy is the root cause of disorder; internal disorder develops, positive entropy increases, and as order develops, positive entropy decreases; external environmental entropy (i.e., negative entropy) is the source of order, when external factors are unfavorable to ecology, ecological negative entropy decreases, and when external factors are favorable to the ecology, ecological negative entropy increases. The system evolves in an orderly manner under the combined action of the internal positive and external negative entropies. The Brussels machine model can be described as follows: where the transformed model parameter A represents the positive entropy flow of the ecosystem (internal factors of the ecosystem), B represents the negative entropy flow of the system (disturbance caused by human society to the ecosystem), D represents the low dynamic state of the steady-state transition of the ecosystem through the interaction of positive and negative entropy flows, E represents the high dynamic state of the steady-state transition of the ecosystem through the interaction of positive and negative entropy flows, X represents the quantifiable factor of the positive entropy flow indicator of the ecosystem, and Y represents the negative entropy flow of the ecosystem. Therefore, the entropy flow relationship of the coupled ecology-human society system can be described by the Brussels apparatus. According to the equations and corollaries of the Brussels apparatus, at that time, a system can only become a dissipative structure; thus, for an ecology-human society coupled system, the following equation can be used to determine the magnitude of the driving force of the system steady state.

| 243
When the value of   − + B A (1 ) 2 is 0, the critical state of the steady-state transition is ready to transform into a new steady state; when the value is <0, the system is dominated by positive entropy evolution; when the value is >0, the system breaks through the critical state with the introduction of negative entropy and enters another trough position in the stable landscape map, and evolves into a new steady state.
The magnitude of the driving force for the ecological-socioeconomic steady-state transition can be quantified using the Brussels machine model. We achieve Brussels machine model in Python. The specific steps are as follows.
1. The evaluation index was divided, as shown in Supporting Information: Table S1. According to the target layer in Supporting Information: Table S1, the indicators were divided into two parts: internal system (ecosystem) and external system (human social system) indicators (n and m, respectively). 2. The entropy weight model calculated the positive entropy value A of the internal system and the negative entropy value B of the external system: where S i and w i represent the entropy values and weights of the corresponding indicators, respectively. n and m represent the number of internal and external indicators, respectively. 3. Calculation of system steady-state transition forces: 3 | RESULTS

| EQI value change and analysis
The calculated results show that the EQI of the Yellow River Basin ecosystem during 1980-2019 showed an overall oscillatory development for the better, with a mean value of 63.96, maximum value of 69.65 (2018), and minimum value of 59.45 (1986) (Figure 3), showing an overall trend of quality improvement with an increase rate of 0.03/year. After the rapid rise of EQI in 1995, there was a relatively stable and small decline phase, followed by a rapid growth trend in 2014. It can be divided into two phases of "peak-valley-peak" type, that is, 1980-1986-1995 and 1996-2010-2019 phases, both of which display decreasing and then increasing trends. Based on the giant system of the Yellow River basin, the study basin was divided into the upper, middle, and lower reaches for comparative analysis, and average EQI values of 54.61, 66.89, and 67.44 were obtained for the upper, middle, and lower reaches of the basin during 1980-2019, respectively ( Figure 4). The upstream ecosystem quality was the lowest and the downstream ecosystem quality was the highest. In the last four decades, the downstream ecosystem quality increased the most significantly (0.034/year), especially after 1992, when the ecosystem quality increased abruptly, leveled off, and then showed an upward trend after 2012. The midstream ecosystem quality changed at a rate of 0.034/year, with a clear increasing trend until 2000, after which the ecosystem quality developed steadily. The rate of change of the upstream ecosystem quality was the slowest (0.027/year), with a slight decreasing trend between 1995 and 2009, while its ecosystem quality increased yearly after 2010, especially with a clear increasing trend after 2014.

| Degree of contribution of key (highly weighted) indicators to ecosystem quality
The greater the degree of contribution of the composite indicator of the criterion layer to the EQI of an ecosystem, the more sensitive the indicator, and measures need to be taken to manage or strengthen regulation. In general, environmental development quality, hydrological condition, and climatic factors have always had a strong influence (Contribution degree is 35%, 26.7%, 18.3%, respectively) on EQI, while social development factors, Ecological factor and Water resource factor were most important during 1980-1990, and their contribution decreased yearly after 2010, due to the gradual stabilization of social development. Water environment factors (water quality) had the smallest contribution in 1980 and then increased annually, the contribution degree is 13.5% in 2019, making it the largest contributor (Supporting Information: Figure S1).
A statistical analysis of the number of high weights (top three weights) of every indicator in the ecosystem for each year from 1980 to 2019 was carried out, and eight indicators with relatively high frequencies were selected: GDP growth rate (18 times), water quality compliance rate of important water function areas (15 times), wetland area change rate of typical regions (11 times), incoming sand (10 times), average income of residents (8 times), water consumption in the basin (7 times), total annual precipitation (5 times), and water network density (3 times) to further analyze the degree of contribution of important indicators to the coupled system.
The contribution of important indicators to the ecosystem in general shows that during 1980-1990, the total water use in the basin, total annual precipitation, and rate of water quality compliance in important water function areas had larger contribution values; during 1995-2005, the GDP growth rate, total water use in the basin, and rate of change of wetland area in typical regions had a larger influence on the system; and during 2010-2019, the amount of incoming sand and average income of residents became the key factors affecting the system (as shown in Supporting Information: Figure S2).

| Trends in watershed ecosystem stability
The system steady-state transition force values from 1980 to 2019 were all <0 (as shown in Supporting Information: Table S2 and Figure 5), indicating that the steady-state transition force of the Yellow River Basin ecosystem was low, but the overall trend oscillated upward and was closer to reaching the critical conditions for steady-state transition. The positive feedback (positive entropy flow value) of the system showed an upward trend, and the negative feedback (negative entropy flow value) showed a downward trend. The positive feedback changes were relatively small compared to the negative feedback, which indicates that the system did not have a huge rise and fall in internal development over the 40 years, and the development state was relatively stable. The negative feedback showed a large decrease and an upward trend in the absolute value.

| DISCUSSION
The entropy weight model was used to propose a calculation method that can quantitatively describe the watershed ecosystem and provide a scientific basis for ecological protection and high-quality development of the Yellow River Basin. Previous reports have also use entropy weight model to demonstrated that the level of water ecological safety in the Yellow River basin has been increasing year by year (Zuo, 2019). The group of Chinese Academy of Water Sciences applied theories such as happiness concept and demand hierarchy to deeply analyze the connotation of happy river, and constructed the river and lake happiness index and its index system and measurement method (China Institute of Water Resources and Hydropower Research, 2020). Liu et al. applied theories such as happiness concept and hierarchy of needs to construct an evaluation index system for ecological protection and high-quality development of eight prefecturelevel cities in the Yellow River basin based on the connotation and actual needs of ecological protection and high-quality development in the Henan section of the Yellow River basin from three dimensions: economic, social, and environmental (Liu et al., 2023). Wang et al. constructed a set of evaluation index system to evaluate the ecological environment of the Yellow River basin by using hierarchical analysis and spatial principal component method, and conducted a hierarchical mapping and dynamic study (Wang et al., 2004). Li et al. studied the spatial and temporal evolution of the key elements of the Haihe River basin and concluded that human activities have replaced natural conditions as the key factors affecting the Haihe River basin. diminished the majority of ecosystem services (Li, 2013). Previous studies have focused more on indicators of economic and social dimensions, but less on indicators of ecological protection dimensions, especially the relationship between water and other elements in the ecosystem is not deeply studied. Most scholars have introduced the concept of green development in their studies and established ecological status indicators in terms of ecological governance capacity, greening construction situation, resource consumption, and so forth. However, there is a lack of research on the relationship between green development and ecological protection, and the selected indicators are somewhat one-sided and subjective. The research objects and scope of the development quality of watershed systems are being expanded.
Evaluating the development quality of watershed ecosystems at the regional scale is closer to the concept of composite representation of the interaction between resources, environment and society and culture. Therefore, the trend of basin system assessment research should not only include the deepening of research at the system scale, but also the refinement and application of multiple concepts of river health, ecological environment, and social economy, and the advantage of regional integrated research from river-geographyecology perspective, to effectively practice ecological protection and high-quality development of the Yellow River basin. Meanwhile, it is also important to note that the current perceptions of high-quality development in the academic community are not uniform. Different perceptions result in different evaluation systems, which leads to inconsistent evaluation results of the quality development of river basins. Therefore, the evaluation results, especially the ranking, need to be compared and analyzed with other conceptual frameworks and indicator systems to achieve an objective perception of the quality development of the Yellow River Basin.
For the basin in general, 1980-1986-1995 was the stage of accelerating socialist modernization and economic development after the reform and opening up in 1978; the economic development was rough and focused on pursuing GDP growth rate, which caused a series of ecological and environmental problems; the overall trend of EQI was decreasing, but the implementation of the Three North Protection Forest Project in 1979 and Yellow River Protection Forest Project in 1984 promoted the gradual improvement of the ecosystem (Zhao, Yu, et al., 2017); therefore, the EQI was on an upward trend after 1986, and the overall development was more oscillating and disorderly (Du et al., 2021). 1995-2010-2019 economic development experienced a change from rough to intensive during 1995-2010 due to various policies and foreign trade support, rapid economic growth, whereas excessive consumption of resources and environmental pollution was also increasingly serious (Jingping, 2020;Li et al., 2021;. The overall trend was decreasing, and after 2010, human society entered a healthy development stage, focusing on ecological protection and management while pursuing economic growth. The construction of ecological civilization, such as water and soil conservation management on the Loess Plateau and pollution prevention and control battle, all contribute to the growth trend of system EQI (Wang, 2008) (see Figure 3, where the length of the legend represents the length of time for policy and engineering implementation).
In this study, the entropy weight of each indicator of the complex system was again calculated using the entropy weight model as a dynamic change over time. The key factors affecting the system gradually changed from the total water consumption and total annual precipitation in the basin during the early stage to the amount of sand coming from the Yellow River and average income of residents in the last 10 years. Therefore, the overall ecosystem quality in the downstream region was higher than that in the upstream region (Xie, 2013). The upstream region was sparsely populated, economically underdeveloped, and unaffected by sedimentation; therefore, its overall ecosystem quality was low. Water quality became the biggest contributor in 2019, which was caused by the increasingly serious pollution of the dry and tributaries of the Yellow River with rapid economic development (Zhang, Chen, et al., 2021). Although the water quality of the Yellow River Basin, especially the water quality of the mainstream section, has greatly improved through the implementation of policies such as major scientific and technological projects for the control and management of water pollution and the pollution prevention and control battle, the water environment factor is still a sensitive factor at this stage due to the lagging effect, and it is necessary to strengthen supervision to maintain and improve the water quality of the basin (Ma et al., 2013). The contribution of the water resources factor also decreased annually, which was due to the gradual improvement of water resource conservation and efficient utilization systems, strengthening water conservation in agriculture, building a water-saving society, establishing a monitoring and early warning mechanism for water resources carrying capacity in the Yellow River Basin, strengthening rigid constraints on water resources, moderate open source, and accelerating regional water transfer to relieve water supply pressure have been implemented, resulting in reasonable planning and efficient utilization of water resources in the basin (Recommendations of the Central Committee of the Communist Party of China on the formulation of the 14th Five-Year Plan for National Economic and Social Development and the Visionary Goals for 2035), Liang et al., 2017;Wang et al., 2020) (Supporting Information: Figure S1).
The steady-state transition force values of the Yellow River Basin ecosystem during 1980-2019 were <0, but the overall trend oscillated upward and the overall trend was closer to reaching the critical conditions for steady-state transition. The positive feedback (positive entropy flow value) of the system showed an increasing trend, and the negative feedback (negative entropy flow value) showed a decreasing trend, indicating that, under the influence of external changes, policies, development concepts, and restoration projects of the system, the exchange of material, energy, and information favorable to development was carried out within and outside the system, which made the system less disorderly (Zhao & Zhang, 2021). At the same time, according to the calculation results of the Brussels instrument model, the steady-state transition force of the system was increasing, which also indicates that the Yellow River Basin ecosystem will enter a new stage of development, and the stability of the ecosystem will be transformed. The critical point of the transition has been approached to require new measures to drive this transformation to move into a higher-quality steady-state development stage and achieve sustainable development of the ecosystem and the synergistic development of human society (Zhang, Chen, et al., 2021). The Entropy weight model based on information entropy and the dissipative structure-based Brussels instrument model are the two main types of methods for evaluating the development quality and stability of systems. In conclusion, the quantitative method of system development quality identified in this study inherits the form and advantages of the assessment model and realizes the assessment of the spatial and temporal dynamics of ecosystem development quality, however the results have some uncertainties due to the temporal and spatial accuracy of data sources. Future simulation studies need to acquire long time series of field observations and remote sensing data for large scale validation. Meanwhile, the spatial resolution of the initial data is yet to be improved to provide more accurate model-driven data. Entropy weight model and Brussels instrument model analysis can be applied to other basin quality and stability assessment work. The next work focuses on subdividing the subsystem quality assessment of the five key ecological reserves in the Yellow River basin, which can also provide a basis for basin development quality improvement.

| CONCLUSION
This study used the Yellow River Basin as an example, considered the systemic and typical nature of the basin, constructed a comprehensive index system to evaluate the development status of the Yellow River Basin, proposed a calculation method that can quantitatively describe the basin ecosystem using the entropy weight model and dissipation structure, and constructed a physical index that can quantify the steady-state transition of the basin development-system conversion force, which provided a scientific basis for ecological protection and high-quality development of the Yellow River Basin. The following conclusions were drawn: 1. From 1980 to 2019, the average EQI value of the Yellow River basin was 63.96, with a maximum value of 69.65 (2018) and minimum value of 59.45 (1986), with an overall oscillating trend of increasing EQI. The entropy weights of the indicators of the complex system changed dynamically with time. The key factors influencing the system changed gradually from the total water consumption and total annual precipitation in the basin in the early stage to the average income of residents in the last 10 years. 2. The steady-state transition force values of the Yellow River Basin system from 1980 to 2019 were all <0, indicating that the steady-state transition force of the Yellow River Basin ecosystem was low, but the overall trend oscillated upward and approached the critical point of steady-state transition. Further protection of the ecological environment of the Yellow River Basin and improving the synergy between ecological environment improvement and economic development level can lead the Yellow River Basin ecosystem to a high-quality development state and realize harmonious coexistence between humans and nature. Extending the entropy weighting model and Brussels instrument model to other typical basin for development quality and stability assessment is our next focus.

ACKNOWLEDGMENTS
This work was financially supported by The National Key Research and Development Program of China (2022YFC 3204305).

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.
ETHICS STATEMENT None declared.