Ecological well-being performance in a Chinese urban agglomeration: Spatiotemporal analysis and policy insights from an Orange-based machine learning framework

Ecological well-being performance in a Chinese urban agglomeration: Spatiotemporal analysis and policy insights from an Orange-based machine learning framework | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Ecological well-being performance in a Chinese urban agglomeration: Spatiotemporal analysis and policy insights from an Orange-based machine learning framework Rui Zhang, Zuo Zhang, Yuanyuan Zhu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7839392/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 14 Feb, 2026 Read the published version in Applied Spatial Analysis and Policy → Version 1 posted You are reading this latest preprint version Abstract From the global perspective of human–nature sustainable development, ecological well-being performance (EWP) captures the efficiency dimension of sustainability by linking ecological consumption with human well-being. For a rapidly developing country, how to balance development and ecology in the spatial dimension is not only related to spatial coordination and sound policy-making, but also determines the sustainability of growth. However, existing studies often overlook spatial disparities, typological differentiation, and nonlinear determinants. To address these gaps, this study developed a machine learning-based multi-method framework on the Orange visual programming platform, integrating spatiotemporal analysis and interpretable machine learning, with a focus on a typical region of China as the study object. This framework was applied to examine EWP in the middle reaches of the Yangtze River urban agglomeration (MRYRUA) during 2005–2022. The results show an overall improvement but with persistent spatial disparities, notable typological differences among cities, and key drivers dominated by industrial structure and government expenditure. By revealing the disparities behind aggregate progress, this study contributes to more precise estimation results and a clearer understanding of driving mechanisms, while accurately restoring spatial patterns, thereby laying a solid foundation for scientifically formulating multi-scale regional development policies. Ecological well-being performance (EWP) Spatiotemporal analysis Interpretable machine learning Urban agglomeration (Yangtze River China) Sustainability policy Orange Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 1 Introduction The stability of the Earth system is deeply intertwined with human well-being (Rockström et al., 2021 ). Since the Industrial Revolution, rapid industrialization and high-consumption lifestyles have accelerated ecological degradation and resource depletion, intensifying risks such as climate change, biodiversity loss, and pollution (Folke et al., 2021 ). At the same time, persistent inequalities in poverty, health, and education continue to challenge global equity (UN Environment Programme, 2019 ). To address these dual crises, the Sustainable Development Goals (SDGs) have called for improving human well-being within planetary boundaries (UN Global Compact, 2015 ). However, recent assessments (UN Sustainable Development Solutions Network, 2024 ) suggest that about 83% of SDG targets may not be achieved by 2030, highlighting the challenge of enhancing human well-being within ecological ceiling. This global difficulty underscores the importance for rapidly developing countries to design scientific and multi-scale regional policies. In China, rapid urbanization and industrialization over the past four decades have brought unprecedented economic growth, but also severe environmental pressures (Ma et al., 2025 ). In response, national strategies such as Ecological Civilization Construction (2012), New-type Urbanization (2014), and Rural Revitalization Strategy (2017) aim to align economic growth with ecological protection and social sustainability, striving to better balance rapid development with ecological conservation, immediate development needs with long-term sustainability, and regional development disparities. Urban agglomerations, as integrated spaces of multiple cities and urban–rural systems, are not only key arenas for economic growth, population concentration, and ecological risk management, but also exhibit greater complexity in spatial governance and policy-making (Fang et al., 2018 ; Wang et al., 2022 ). Among them, the middle reaches of the Yangtze River urban agglomeration (MRYRUA), located at the central junction linking east–west and north–south China and stretching along the country’s largest inland river, is a rapidly expanding yet ecologically fragile region. It exemplifies the trade-offs between development and environment in emerging megaregions (Zhang et al., 2023 ). In this context, improving resource efficiency, safeguarding ecosystems, and enhancing well-being have become pressing issues. Ecological well-being performance (EWP), as an indicator of how efficiently ecological consumption is transformed into human well-being (Zhu et al., 2022 ), provides an effective framework for analyzing these issues. The concept of EWP originated from steady-state economics (Daly, 1974 ), but quantitative research gained traction after the introduction of ecological footprint theory (Rees, 1992 ). Subsequently, several similar concepts combining human well-being and environmental impact have emerged, such as the happy planet index (Marks et al., 2006 ), national economic performance (Common, 2007 ), environmental efficiency of well-being (Knight & Rosa, 2011 ). Early studies typically defined EWP as a single-ratio relationship between human well-being and ecological consumption (Knight, 2014 ; Zhu et al., 2015 ). In recent years, model-based methods, especially Data Envelopment Analysis (DEA), have been widely adopted to accommodate multiple input and output indicators (Zhang et al., 2021 ), enabling more detailed assessment of ecological inputs and well-being outputs. As measurement methods matured, numerous studies have examined the temporal and spatial characteristics of EWP across multiple geographic scales, including national (Zhang et al., 2018 ), provincial (Li et al., 2018 ), and urban levels(Bian et al., 2020 ; Hu et al., 2021 ). Temporally, some studies have explored EWP trends and dynamic changes (Behjat & Tarazkar, 2021 ; Zhang et al., 2022 ). Spatially, researchers have investigated the spatial characteristics and regional disparities of EWP using spatial correlation networks (Zhao et al., 2022 ), spatial autocorrelation analysis (Xia & Li, 2022 ) and spatial convergence models (Deng et al., 2021 ). Another major research area focuses on identifying the determinants of EWP. Previous studies have comprehensively examined various factors affecting EWP (Li et al., 2018 ; Zhu et al., 2022 ). Several studies have specifically examined the effects of green finance policies (Wang & Gao, 2024 ), digital economy (Yang et al., 2023 ), environmental regulations (Shao et al., 2024 ), low-carbon city pilot policies (Han et al., 2025 ) and new-type urbanization (Zhang et al., 2024 ). Traditional econometric approaches such as multiple linear regression, panel Tobit regression (Long et al., 2017 ), and spatial econometric models (Fang & Xiao, 2019 ) have been employed. Overall, existing studies provide valuable insights into the measurement, spatiotemporal features, and drivers of EWP. However, significant gaps remain. First, current assessments rarely differentiate EWP types. Given that EWP is a relative indicator, similar scores can obscure substantial disparities in input–output configurations or overlook unsustainable situations, such as excessive ecological consumption or inadequate human well-being, which may breach safe and just boundaries (Rockström et al., 2021 ). This points to the necessity of a more nuanced classification of ecology–well-being transformation types. Second, existing studies rarely investigate the nonlinear and heterogeneous effects of factors influencing EWP. Traditional linear regression models, in particular, struggle to capture such complexities within the human–environment system underlying the ecology–well-being transformation. To address these gaps, this study proposes a comprehensive analytical framework that integrates spatiotemporal analysis, K-means clustering, and interpretable machine learning to examine spatiotemporal patterns, types, and drivers of EWP across the MRYRUA. The framework is implemented on Orange ( https://orangedatamining.com/ ), an open-source visual programming platform that facilitates efficient data processing and modeling (Zhang et al., 2023 ). While Orange has been widely applied in biomedicine (Godec et al., 2019 ), education (Yağcı, 2022 ), and urban research (Zhang et al., 2024 ), its potential in sustainability analytics remains largely unexplored. In summary, building on the above four sections, this study focuses on rapidly developing, spatially integrated and ecologically fragile urban agglomerations, analyzing the relationship between human well-being and ecological protection from a spatial perspective to support sound regional policy and ultimately achieve balanced and sustainable development. Against this background, the objectives of this study are threefold. First, to construct and apply an Orange-based analytical framework for investigating the spatiotemporal divergence and driving mechanisms of EWP in the MRYRUA. Second, to classify cities into EWP types by jointly considering inputs, outputs, and efficiency, guided by the safe and just framework, so as to uncover differences in efficiency and balance. Third, to employ interpretable machine learning techniques to capture nonlinear relationships and regional heterogeneity that cannot be adequately addressed by traditional regression models. Finally, in terms of evaluation and analytical methods, this study also seeks to contribute to the application of a multi-model analysis platform characterized by strong visualization, high integrative capacity, and interactive features. The remainder of this paper is organized as follows: Section 2 introduces the framework, methodology, study area, and data; Section 3 presents the spatiotemporal evolution of EWP and its drivers; Section 4 discusses the findings; and Section 5 concludes and provides the policy insights. 2 Framework and methods 2.1 Framework design Natural ecosystem and socio-economic system are deeply interdependent. Natural ecosystem provides essential services that support human survival, development, and well-being (Costanza et al., 2016 ; Chen et al., 2024 ), while human activities consume natural resources, generate waste, and reshape the environment (Daly, 1974 ). From a human–environment systems perspective, this process can be understood as the transformation of ecological consumption into human well-being, with strong regional variation shaped by factors such as urbanization, technological innovation, industrial transformation, globalization, and government regulation (Gao et al., 2024 ). Ecological well-being performance (EWP), as an efficiency-oriented metric, captures the conversion relationship between ecological consumption and human well-being. Within this coupled system, ecological consumption and human well-being are constrained by an ecological ceiling, which represents the limits of environmental carrying capacity, and a social foundation, which reflects the minimum thresholds necessary for human development and well-being (Fanning & Raworth, 2025 ). Sustainable improvement in EWP therefore needs to occur within these ecological and social boundaries. Given the complexity and diversity of this transformation across regions and over time, a multidimensional analytical perspective is essential. Specifically, spatiotemporal analysis reveals temporal evolution and spatial heterogeneity, typological differentiation identifies structural imbalances that similar EWP scores may obscure, and interpretable machine learning captures nonlinear and heterogeneous drivers. Building on the coupled human–environment systems conceptual framework, this study develops a structured and scalable Orange-based analytical framework that integrates these three dimensions to uncover uneven and multi-type divergences in EWP (Fig. 1 ). This conclusion highlights the necessity of combining spatiotemporal, typological, and machine learning perspectives as a key consideration for the framework proposed in the following part. The analytical framework is built on Orange, a no-code visual programming platform developed by the University of Ljubljana (Štajdohar & Demšar, 2013 ), and integrates the GIS platform and the Super-SBM model. It consists of four key modules, including temporal analysis, spatial analysis, clustering, and influencing factors. The Super-SBM model is used to measure EWP values. The GIS tools provide technical support for spatial data organization and visualization. All analytical modules are powered by eight categories of Orange widgets ( https://orangedatamining.com/widget-catalog/ ), each offering specific functions and allowing interconnection to accommodate various analytical tasks (Zhang, 2022 ; Zhang et al., 2024 ). The four core modules are as follows: (1) Temporal analysis module examines dynamic trends of EWP over time, capturing long-term variation and stage characteristics. It incorporates widgets such as line plots, scatter plots, sieve diagrams, and box plots. (2) Spatial analysis module visualizes the spatial distribution using the Geo Map widget. Regional disparities are assessed using a customized Theil index widget, developed for this study using Orange’s open-source architecture. The index is calculated following the approach outlined in Wang et al. ( 2021 ). (3) Clustering module applies K-means clustering to jointly analyze ecological inputs, well-being outputs, and overall efficiency, thereby identifying distinct EWP types and revealing heterogeneous transformation pathways across cities. (4) Influencing factors module employs interpretable machine learning to identify drivers of EWP, quantify their contributions, and capture nonlinear and heterogeneous effects, offering a deeper understanding of the mechanisms behind EWP variation. 2.2 Materials and methods 2.2.1 Study area and data The middle reaches of the Yangtze River urban agglomeration (MRYRUA) is a large-scale urban cluster that consists of the Wuhan Metropolitan Area (WMA), the Changsha-Zhuzhou-Xiangtan City Group (CZXCG), and the Poyang Lake City Group (PLCG) (Fig. 2 ). It lies in the middle reaches of the Yangtze River, China’s largest river, where river basin governance and ecological governance are deeply intertwined. Covering approximately 317,000 km² and comprising 31 prefecture-level cities, it represents China’s largest urban agglomeration by land area. Spanning three provinces and encompassing cities of different administrative levels as well as both urban and rural areas, the region faces greater challenges of cross-regional spatial governance and policy coordination. By the end of 2022, the population had reached 127 million, accounting for about 9% of the national total. MRYRUA plays a critical role in national strategies such as the Rise of Central China, New-type Urbanization, and Beautiful China initiatives (Zhang et al., 2023 ). However, it still faces significant pressures from rapid population growth, resource consumption, and environmental challenges. Given these conditions, this study selected the MRYRUA as a representative case for city-level spatiotemporal analysis of EWP. The findings are expected to support sustainable urbanization in China and provide insights for other rapidly urbanizing regions in developing countries. Prefecture-level cities within the MRYRUA were chosen as the basic spatial units for analysis. From the perspective of spatial units and data sources, this design ensures consistency and comparability across cities, while also reflecting the practical basis for EWP measurement at the urban scale. Socioeconomic data covering the period from 2005 to 2022 were primarily sourced from the China City Statistical Yearbook, China Urban Construction Statistical Yearbook, the Hunan, Hubei, and Jiangxi Statistical Yearbooks, as well as municipal yearbooks, socioeconomic development statistical bulletins. Missing values were supplemented using linear interpolation or trend extrapolation methods. The vector data of city boundaries were sourced from the Resource and Environmental Science Data Platform ( www.resdc.cn ). 2.2.2 EWP indicators Building on the theoretical foundation of EWP and previous studies (Hu et al., 2021 ; Xia & Li, 2022 ; Zhu et al., 2022 ; Han et al., 2025 ), this study constructs an evaluation indicator system for EWP (Table 1 ). From a multiple-input and multiple-output perspective, it considers resource consumption and environmental degradation as cost-type input, while economic, social, and environmental dimensions of human well-being are treated as benefit-type output. Table 1 Evaluation indicators of ecological well-being performance Dimensions Primary indicators Secondary indicators Indicator description Input-ecological consumption Resource consumption Per capita built-up area (m²/person) Reflects land use intensity. Per capita energy consumption (tce/person) Reflects energy use intensity. Per capita water consumption (m³/person) Reflects water use intensity. Environmental damage Per capita industrial wastewater discharge (t/person) Reflects the damage to the water environment caused by wastewater discharge. Per capita SO₂ emissions (t/person) Reflects the damage to the atmospheric environment caused by exhaust emissions. Per capita municipal solid waste disposal (kg/person) Reflects the damage to the soil environment caused by solid waste disposal. Output - human well-being Thriving economy Per capita disposable income of urban residents (CNY/person) Indicates the income-related well-being of urban residents Per capita disposable income of rural residents (CNY/person) Indicates the income-related well-being of rural residents Progressive society Ratio of per capita disposable income: rural to urban residents Indicates the level of social equity and urban–rural common prosperity Average years of schooling (years) Reflects the well-being benefits associated with access to education. Hospital Beds per 10,000 Population Reflects the well-being benefits related to healthcare access and coverage Eco-friendly environment Per capita green space area (m²/person) Reflects individual-level environmental well-being through access to green public spaces. Green coverage rate in built-up areas (%) Reflects the ecological livability of urban areas through overall green coverage. The EWP indicator system is refined by integrating insights from previous research and introducing targeted improvements, ensuring that it is more closely aligned with the realities of the study region while fully considering alignment with national SDG goals. To better capture economic prosperity, it replaces the commonly used GDP with urban and rural per capita income (Guo & Qian, 2021 ). For progressive society, it incorporates a social equity indicator measured by the urban–rural income ratio. This indicator is aligns with SDG 10 on reducing inequality (UN Sustainable Development Solutions Network, 2024 ), yet has been largely overlooked in prior EWP studies. Notably, manufactured capital, such as labor and physical capital—commonly treated as intermediate inputs in ecological economic efficiency studies—is not excluded from the indicator system, as it ultimately originates from natural capital (Daly, 2005 ). 2.2.3 Super-SBM model The Slack-Based Measure (SBM) model evaluates the relative efficiency of decision-making units through linear programming. It accommodates multiple inputs and outputs without requiring a predefined production function, making it widely applicable in efficiency analysis (Han et al., 2025 ). However, the traditional SBM model cannot distinguish among fully efficient units. The Super-SBM model addresses this limitation by enabling the ranking of efficient units (Tone, 2002 ). Therefore, this study adopts the Super-SBM model to evaluate the EWP of the MRYRUA. The model is formulated as follows: In the Super-SBM model, ρ * represents the EWP score; M and N denote the number of input and output variables, respectively, and I is the total number of decision-making units (DMUs). x m and y n represent the input and output vectors, while x mk and y nk denote the specific values of the m -th input and n -th output for the k -th DMU. S m − and S n + are the input and output slack variables, respectively. λ i is the weight variable associated with the i -th DMU. If ρ *≥1, the DMU is considered fully efficient, meaning that both input and output slacks are equal to zero. If ρ *<1, the DMU exhibits inefficiency due to the presence of input excess or output shortfall. 2.2.4 K-means clustering and interpretable machine learning K-means is a widely used unsupervised learning algorithm that partitions a dataset into a specified number of clusters, aiming to maximize similarity within clusters and differences between clusters (Rousseeuw, 1987 ). By avoiding predefined classification thresholds, the unsupervised clustering method helps circumvent the subjectivity often inherent in manual or rule-based groupings. It is useful for identifying patterns in cities based on their input–output–efficiency characteristics, thereby providing insights into variations in resource use and social well-being. This study uses the K-means clustering algorithm to examine ecological consumption, human well-being, and EWP across the MRYRUA. Machine learning models are widely applied in regression, clustering, prediction, and dimensionality reduction tasks, and have seen increasing use across diverse fields such as geography, economics, public administration, and climate science(Lundberg et al., 2020 ; Trok et al., 2024 ). Using multiple models helps avoid issues like overfitting or underfitting associated with single models and improves prediction accuracy, especially when dealing with complex nonlinear relationships and spatiotemporal heterogeneity in geographic data (Malone et al., 2014 ). Therefore, this study employs four machine learning models to explore the key factors influencing EWP in the MRYRUA. These models include: (1) Multiple Linear Regression (MLR), a traditional method for modeling linear relationships, serves as the baseline model in this study. (2) Adaptive Boosting (AdaBoost), an ensemble method, improves performance by iteratively focusing on difficult-to-predict samples and combining weak learners into a strong model (Friedman, 2001 ). (3) Random Forest (RF) constructs multiple randomized decision trees and aggregates their results, offering robustness and strong resistance to overfitting (Varian, 2014 ). (4) Extreme Gradient Boosting (XGBoost) enhances gradient boosting by incorporating second-order optimization, regularization, and parallel processing to improve efficiency and reduce overfitting(Li, 2022 ). To ensure reliable and generalizable results, all models were trained using 5-fold cross-validation, with hyperparameters optimized via random search. Model interpretability is critical in understanding the underlying drivers of EWP. To this end, Shapley Additive Explanations (SHAP) is employed to quantify the contribution of individual features to model outputs (Lundberg & Lee, 2017 ). Based on cooperative game theory, SHAP provides a robust framework for feature attribution, enabling interpretation of complex “black-box” models at both global and local levels. It ensures consistent and reliable explanations by assigning higher SHAP values to more influential features. To further capture nonlinear relationships, this study follows the approach of Fu et al. ( 2025 ) by combining SHAP with Generalized Additive Models (GAM).Specifically, the SHAP values are regressed on their actual variable values using GAM to generate smooth response curves, which reveal the ranges of positive and negative effects as well as the turning points where the influence of a factor fundamentally changes. 3 Results 3.1 Spatiotemporal characteristics of EWP 3.1.1 Temporal evolution characteristics Temporal analysis is of particular significance in this study as it reveals dynamic changes in the relationship between ecological consumption and human well-being, especially for rapidly developing countries and regions. Based on Orange’s temporal analysis module, the time-series evolution of EWP in the MRYRUA from 2005 to 2022 is analyzed. During this period, the average annual EWP value increased by 37.45% (from 0.558 to 0.767), indicating a marked overall improvement. As shown in Fig. 3 (a), EWP followed a trajectory of early fluctuations followed by steady growth, with all three sub-agglomerations improving—particularly the CZCXG region. Figure 3 (b) shows a clear upward shift in EWP distributions over time, consistent with the line plot. The spread of data points also indicates notable spatial and temporal heterogeneity. Figure 3 (c), a sieve diagram, confirms this trend: in 2005–2013, low EWP values (< 0.38) dominated; from 2014–2017, moderate values (0.38–0.52) increased; and by 2018–2022, high EWP values (≥ 0.74) became significantly more prevalent. These results suggest a two-stage trajectory of EWP development in the MRYRUA—initial fluctuations followed by sustained improvement, as discussed below. (1) Fluctuation phase (2005–2013): During this period, EWP followed a volatile trajectory with an overall downward trend. From 2005 to 2008, EWP declined year by year, then rebounded after 2008, but reached a low point again in 2011. This phase coincided with rapid urbanization and industrialization in the study area. During this time, cities placed disproportionate emphasis on economic growth and adhered to an extensive development pattern. This pattern, characterized by high input, intensive energy consumption, and severe pollution, contributed to the decline in EWP. The 2008 global financial crisis had a significant impact on energy-intensive and pollution-heavy industries. As a result, resource input and environmental pollution decreased, which in turn contributed to a temporary improvement in EWP in 2009. (2) Sustained improvement phase (2014–2022): Beginning in 2014, EWP increased markedly, maintaining an overall upward trajectory despite intermittent fluctuations and ultimately reaching its peak in 2022. Following the elevation of Ecological Civilization Construction to a national-level strategy, green development and ecological protection accelerated. Guided by policies such as “prioritizing ecological protection over large-scale development,” the MRYRUA increased efforts in environmental protection and made visible progress in green and low-carbon development. The launch of the New-type Urbanization strategy in 2014 also introduced a series of policies focused on improving public well-being. Driven by both national and local initiatives, EWP steadily improved. In 2020, the COVID-19 pandemic and the associated global economic slowdown led to reductions in resource consumption and pollution, which in turn contributed to a temporary spike in EWP. 3.1.2 Spatial patterns and regional disparities Spatial pattern analysis is crucial for understanding regional ecological protection, well-being enhancement and governance coordination. To analyze the spatial patterns and regional disparities of EWP, the study period was divided into four approximately balanced intervals based on data availability and policy phases: 2005–2008, 2009–2012, 2013–2016, and 2017–2022. This approach helps to minimize the influence of anomalies or irregularities in individual years. The average EWP for each period represents general performance level during that timeframe. Figure 4 illustrates the spatiotemporal evolution of EWP across MRYRUA from 2005 to 2022. From a spatial perspective, EWP in the MRYRUA characterized by higher values in central areas and lower values in the periphery, indicating a marked spatial polarization. In the early period, high EWP values were relatively sparse but exhibited spatial clustering in and around core cities like Wuhan and Nanchang. During 2009–2012, a sharp rise in EWP occurred in parts of PLCG. From 2013–2016, core cities like Wuhan and Changsha began to show strong improvements, while border cities such as Xianning, Huangshi, Jiujiang, and Yueyang lagged behind, forming a “central collapse” pattern. By 2017–2022, a more polycentric structure emerged, with multiple high-EWP zones expanding outward. Nevertheless, a persistent “high core–low periphery” pattern remains, underscoring uneven development within the agglomeration and the need for coordinated regional strategies. The Theil index and its decomposition were used to assess intra-regional disparities in EWP across the MRYRUA. As shown in Fig. 5 , the overall Theil index declined by 75.13% (from 0.197 to 0.049) over the study period—a 75.24% reduction—reflecting a substantial decrease in regional inequality. The decomposition results indicate that over 80% of the total inequality originated within sub-agglomerations rather than between them. Among the three, the WMA exhibited the highest Theil index, followed by the PLCG and CZXCG. The WMA also contributed most to overall inequality, consistently above 45% and reaching around 60% in recent years. The PLCG ranked second, though its share declined over time. The CZXCG contributed the least, with a slight downward trend. These findings suggest that intra-regional disparities within the WMA are both the most pronounced and the primary source of EWP inequality across the region. 3.2 Typological features of EWP A clustering analysis was conducted to examine city-level patterns of ecological consumption, human well-being, and EWP across the MRYRUA. Average data from 2018 to 2022 were used to reflect recent development trends and to reduce short-term fluctuations. Using the Silhouette Score method, four clusters were identified as optimal, achieving the highest score (0.212), indicating the best balance between cohesion and separation. Accordingly, the cities were grouped into four distinct clusters (Fig. 6 ). C1-type cities demonstrate moderate EWP despite below-average ecological consumption and human well-being. Representing a “low input, low output, and moderate efficiency” pattern, this cluster includes 15 cities. As shown in Fig. 6 , these cities are characterized by relatively low levels of land, energy, and waste inputs, as well as limited income and green space provision. Despite such minimal inputs, they achieve moderate EWP scores, indicating a certain degree of ecological efficiency. However, the persistently low well-being indicators suggest that ecological resources are not fully leveraged to support broader social development. C2-type cities show low EWP, marked by high ecological consumption but limited well-being gains—a “high input, low output, and low efficiency” pattern. As shown in Fig. 6 , these cities demonstrate excessive resource consumption—particularly in water use, land use, and waste emissions—yet fail to achieve corresponding improvements in well-being. Although they perform well in certain aspects such as urban–rural income equality and green space provision, most social development indicators remain weak, especially in education and healthcare. These four small cities likely suffer from inefficient resource use or dependence on resource-intensive industries, underscoring the need to optimize ecological inputs to enhance well-being outcomes. C3-type cities—Changsha, Wuhan, and Nanchang—achieve the highest EWP, following a “moderate input, high output, and high efficiency” pattern. Despite variability in ecological consumption (e.g., high land consumption but low exhaust water and gas emissions), they lead in income, education, healthcare, and green space. These cities achieve the highest level of well-being with only moderate ecological inputs, making them the most efficient in terms of EWP. This underscores the comparative advantages of large cities in promoting industrial upgrading, enhancing resource efficiency, and ensuring effective public service delivery. C4-type cities exhibit low EWP with a “high input, moderate output, and low efficiency” pattern. This group of nine cities shows above-average levels in most ecological and well-being indicators, though water use and income equality lag behind. High energy use and emissions suggest dependence on resource- and pollution-intensive industries. While better than C2-type cities in income, education, and healthcare, their EWP remains low due to inefficient resource use. These cities are more stable than C2-type cities but require more sustainable development strategies to improve efficiency and reduce environmental pressure. 3.3 Influencing factors of EWP 3.3.1 Influencing factors and model performance The transformation of ecological consumption into human well-being determines EWP, yet this process is complex and influenced by multiple external factors. Building on prior studies (Li et al., 2018 ; Zhu et al., 2022 ; Dong et al., 2024 ), this study identifies a set of key variables to explore the drivers of EWP. Seven key variables are selected to represent key influencing factors: technological innovation, environmental regulation, industrial structure, population density, urbanization, foreign direct investment, and local government expenditure. An interpretable machine learning approach is employed to examine the driving effects of these factors on EWP. Detailed variable information is provided in Table 2 . Table 2 Definition and descriptive statistics of variables Variable type Variable Explanation Obs. Mean SD Min Max Dependent variable EWP Ecological well-being performance 558 0.58 0.26 0.16 1.19 Independent variable Technological innovation (TI) Number of invention patent applications (Patents per 10,000 people) 558 7.92 9. 63 0.00 72.17 Environmental regulation (ER) Proportion of environmental keywords in local government work reports (%)(Bao & Liu, 2022 ) 558 0.80 0.048 0.01 1.02 Industrial structure (IS) Share of secondary industry in GDP (%) 558 48.38 7.55 28.56 78.84 Population density (PD) Population density of the built-up urban area (10,000 people per km²) 558 1.54 1.00 0.39 7.16 Urbanization (UR) Urbanization rate of permanent resident population (%) 558 53.01 11.75 22.62 84.70 Foreign direct investment (FDI) Foreign direct investment as a percentage of GDP (%) 558 2.11 1.58 0.00 8.03 Government expenditure (GE) Government expenditure as a percentage of GDP (%) 558 15.29 5.20 5.75 31.79 Note: Obs. = Number of observations; SD = Standard deviation. Table 3 presents the regression performance of the four machine learning models. XGBoost demonstrates the best performance, exhibiting the lowest error metrics and the highest R², which indicates superior model fit. RF ranks second, performing comparably to XGBoost in terms of MSE and RMSE, but with slightly higher MAE and MAPE. AdaBoost also shows competitive performance, though slightly behind the top two. MLR shows the weakest performance, suggesting that linear models may not adequately capture the underlying data patterns. Table 3 Comparative performance of four machine learning models Model MSE RMSE MAE MAPE R 2 XGBoost 0.021 0.145 0.094 0.176 0.701 RF 0.021 0.146 0.108 0.209 0.697 AdaBoost 0.022 0.15 0.108 0.205 0.679 MLR 0.034 0.185 0.145 0.286 0.511 3.3.2 Results of interpretable machine learning As shown in Fig. 7 , IS and GE consistently rank as the most influential drivers of EWP across all models, whereas UR, ER, PD, and TI fall into the mid-range of importance, and FDI shows the lowest contribution. This suggests that across different models, the overall ranking of feature importance remains relatively stable, yet the mechanisms through which each factor shapes EWP differ considerably. Compared to the linear model, the three ensemble models reveal more diverse and intricate SHAP value patterns. To further investigate these roles, Fig. 8 presents the SHAP–GAM single-feature dependence plots, which reveal the nonlinear relationships and critical thresholds of each driver. The relationships between the seven indicators and EWP are all characterized by significant nonlinearity and threshold effects. With the exception of FDI, all variables show significant fits ( p < 0.001), with R ² values ranging from 0.31 to 0.85, indicating good explanatory power. Technological innovation (TI) has a positive effect on EWP, particularly at higher levels, suggesting the presence of nonlinearities and threshold effects. Figure 7 indicates that the effect is not uniformly strong across all cities. Figure 8 (f) reveals a nonlinear pattern: at lower levels, TI has little discernible effect, but once a certain scale of innovation activity is reached, its contribution increases rapidly. However, when technological innovation becomes excessive, the marginal benefits tend to diminish, indicating that the positive impact of innovation on EWP gradually weakens at higher levels. This suggests that beyond a certain point, further technological input may lead to efficiency saturation, higher costs, or rebound effects, thereby reducing the incremental gains in EWP. Environmental regulation (ER) shows a positive effect on EWP. As indicated in Fig. 7 , its overall contribution is moderate and varies considerably across cities. Figure 8 (d) highlights that when regulation is weak, its impact is often negative or negligible, as insufficient enforcement fails to offset the associated economic costs. In contrast, stronger regulation generates clearer positive effects, suggesting that only beyond a certain intensity does ER promote ecological improvements and welfare gains, in line with the Porter hypothesis. Industrial structure (IS) has a significant negative impact on EWP, indicating that upgrading the industrial mix contributes to better performance. Across all models, higher IS values are predominantly linked to negative SHAP values, whereas lower IS values are associated with positive ones. This pattern indicates that a larger share of secondary industry within the economic structure tends to exert a stronger negative effect on EWP. A heavy reliance on secondary industry is generally linked to elevated pollution levels and intensive energy use, leading to greater environmental degradation. In the past decade, the MRYRUA has undergone industrial upgrading, with a shift away from traditional manufacturing toward cleaner, higher-value sectors. This has reduced ecological pressure and enhanced well-being, contributing to improved EWP. Population density (PD) generally has a positive effect on EWP. High PD values are broadly associated with positive SHAP values, reflecting a stronger contribution to EWP. This implies that higher urban population density can lead to economies of scale and agglomeration effects, which in turn enhance resource efficiency. However, the curve flattens at higher PD levels, suggesting that excessive agglomeration may no longer yield proportional benefits, possibly due to congestion, environmental stress, or diminishing marginal returns. Urbanization (UR) has a negative impact on EWP. In nearly all models, UR shows a negative relationship with EWP, with only a few cities displaying a positive contribution. This suggests that the quality of urbanization in the study area has remained relatively low over the past decade. Rapid urban expansion has often occurred alongside intensive industrialization, frequently driven by low-value-added, resource-intensive, and polluting industries. Moreover, Moreover, improvements in public services such as education and healthcare have lagged behind urban growth, resulting in limited gains in per capita well-being. Government expenditure (GE) has a significant positive impact on EWP. The SHAP values of GE show a symmetric distribution, with higher feature values corresponding to higher SHAP values, underscoring its stable positive contribution. This reflects the important role of public fiscal intervention in advancing environmental quality and well-being. In particular, government spending on environmental protection, social well-being, and green industry support has substantially improved both ecological conditions and residents’ quality of life in the study area. Foreign direct investment (FDI) has a limited impact on EWP. In all models, FDI exhibits a much lower average absolute SHAP value compared to other variables, suggesting limited explanatory power. However, this does not imply that FDI has no impact whatsoever. Existing studies suggest that the influence of FDI may be moderated by competing mechanisms, including the “pollution haven” and “pollution halo” hypotheses, as well as income effects (Xiong et al., 2023 ). In the AdaBoost and RF models, some cities with low FDI levels show positive SHAP values. This suggests that, in certain cases, lower levels of foreign investment may actually be more conducive to improving EWP. 4 Discussion The relationship between the environment and human well-being is a critical global issue (Folke et al., 2021 ; Rockström et al., 2021 ). Enhancing EWP is essential for the sustainability of urban agglomerations. However, previous studies have largely overlooked both the typological classification of EWP and the nonlinear effects of its drivers. There is limited understanding of how cities within the same urban agglomeration differ in their EWP transformation trajectories, and what context-specific drivers shape these divergent pathways. The present study addresses these gaps by developing an integrated framework using Orange, a visual programming platform, to examine the evolution, types, and driving mechanisms of EWP in the MRYRUA. This study area is characterized by rapid development, high spatial density and diversity, and pronounced ecological fragility, which together heighten governance complexity and the challenge of cross-regional policy coordination. Our framework and indicator system explicitly aligns with the SDGs while being tailored to China’s regional development realities, thereby enhancing both international comparability and local applicability. By applying spatiotemporal analysis, K-means clustering, and interpretable machine learning, we show that EWP patterns vary significantly across cities, driven by nonlinear and heterogeneous effects of multiple factors. The key findings and insights are as follows: ໿The evolution of EWP reflects broader shifts in socioeconomic development models and the transformation of human–environment relationships. In the MRYRUA, the trajectory of EWP indicates that the region experienced significant population, resource, and environmental pressures during its early stages of development. This is consistent with previous research findings (Chen et al., 2022 ). However, with the implementation of national strategies such as Ecological Civilization Construction, the relationship between nature and human well-being has gradually shifted from imbalance to coordination. Such a transformation underscores China’s progress in aligning ecological policy and broader development initiatives with the overarching agenda of the SDGs (Cao et al., 2023 ). This transition also parallels the developmental trajectories of mature urban agglomerations and metropolitan regions worldwide, which typically follow a gradual upward trend toward sustainability (Fang et al., 2018 ). Nevertheless, substantial disparities in EWP persist among cities within the MRYRUA. These findings are consistent with previous studies (Zhang et al., 2024 ), underscoring the urgent need to enhance inter-city coordination and foster the diffusion of development benefits across the region. Additionally, beyond this consistency, our study differs from prior work by integrating a multi-model, interpretable machine-learning workflow, enabling more precise estimates and clearer identification of nonlinear thresholds and context-specific mechanisms in spatially integrated, ecologically fragile, high-density, and diverse metropolitan regions. This study adopts a “safe and just space” perspective to classify the ecology–well-being transformation process, revealing significant variation among cities in balancing ecological consumption and human well-being. At present, only a few cities have achieved an optimal state. Most cities exhibit low ecological consumption, yet their well-being levels remain below the average and are potentially nearing the social foundation of justice (Rockström et al., 2021 ). For these cities, priority should be given to improving public services and social protection systems to ensure that basic well-being needs are met. Conversely, cities with high per capita ecological consumption should prioritize green industrial transformation, reduce resource use, and improve efficiency. In doing so, this study also emphasizes the bottom-line thinking of ecological safety during rapid development, as well as the necessity of regional balance and context-specific approaches, thereby laying a foundation for more rational regional policy design. Interpretable machine learning results reveal that drivers of EWP exert nonlinear and heterogeneous effects, depending on city context. Although technological innovation is generally contributes to improved EWP (Yang et al., 2023 ), the analysis reveals that only high levels of technological capability lead to substantial positive outcomes. Given the high costs associated with foundational innovation, it is recommended that technologically advanced cities support less developed counterparts, including leveraging digital means to substitute or complement deficiencies in other proximity dimensions(Cheng et al., 2025 ). Moreover, greater emphasis should be placed on the well-being-enhancing effects of technological innovation, ensuring that modern technologies are effectively implemented in practice to improve residents’ quality of life. The impact of environmental regulation also appears to be both nonlinear and context-specific, highlighting the importance of balancing regulatory costs against the ecological benefits they yield. In the long term, the overall quality of urbanization in the study area remains relatively low. Rapid urban growth has not been matched by corresponding improvements in income, healthcare, or education (Sun et al., 2020 ), and it has been accompanied by rising social inequality (Pandey et al., 2022 ). Therefore, further efforts are needed to promote the social integration of rural migrants and to enhance human well-being across the region. Compared with existing studies, this research offers a novel framework for understanding the multidimensional characteristics of regional EWP. Leveraging the capabilities of the Orange data mining platform, the framework integrates data processing, spatiotemporal visualization, clustering analysis, and multi-model machine learning. By integrating multiple analytical perspectives, the framework captures the spatial, temporal, and structural complexity of how ecological resources are transformed into human well-being. It complements traditional efficiency-based approaches by introducing a typological perspective on EWP, adding a classification dimension to conventional evaluations and going beyond pure efficiency scores to identify multidimensional patterns in input–output relations. It also deepens the understanding of how key drivers interact in nonlinear and context-specific ways. This approach advances conceptual debates on sustainability transitions and urban-regional inequality, complementing recent efforts to apply machine learning to sustainable development contexts (Sha et al., 2024 ). In addition, it offers practical guidance for sustainability policy design in rapidly urbanizing agglomerations, providing a valuable reference for other developing countries to better balance development and protection, coordinate regional collaboration, and pursue sustainable growth. Due to limitations in data availability, this study includes only objective well-being indicators in the measurement of human well-being. As such, the analysis may not fully reflect the experiential or perceptual dimensions of residents’ well-being. Furthermore, the classification of EWP types within the “safe and just space” perspective remains preliminary. Excessive absolute levels of ecological consumption can lead to severe environmental degradation, while insufficient absolute levels of well-being may undermine the equitable right to human development. Therefore, future research should explore potential pathways for enhancing EWP within an absolute spatial boundary that simultaneously ensures ecological safety and social justice. 5 Conclusions and policy insights This study developed a comprehensive analytical framework for ecological well-being performance (EWP) based on the Orange data mining platform. Applying this framework to cities within the middle reaches of the Yangtze River urban agglomeration (MRYRUA) during 2005–2022, we identified spatiotemporal evolution and typological patterns of EWP and uncovered its complex driving mechanisms through interpretable machine learning models. The main conclusions obtained by the study are as follows: (1) EWP in the MRYRUA has generally improved, with overall levels increasing by more than 37% (from 0.558 in 2005 to 0.767 in 2022), shifting from early fluctuations to steady growth under the influence of national strategies. However, pronounced spatial disparities persist, characterized by a “high core–low periphery” pattern and a “central collapse” at sub-agglomeration junctions, particularly within the Wuhan Metropolitan Area. (2) The typological analysis reveals clear differentiation in the ecology–well-being transformation pathways among cities in the MRYRUA. Only a few core cities (e.g., Wuhan, Changsha, Nanchang) achieved a coordinated status between ecological consumption and human well-being, maintaining both high efficiency and acceptable absolute levels, while most cities deviate from this balance—either through excessive ecological inputs or insufficient well-being gains—implying potential sustainability risks. (3) The influencing factors of EWP show significant nonlinear and heterogeneous effects. Industrial structure and fiscal expenditure exert the strongest influences, while technological innovation, environmental regulation, population density, and urbanization show moderate but context-specific roles. FDI has only a limited and variable impact. The findings in this study provide valuable policy insights for sustainable development in the MRYRUA: (1) Implement differentiated and collaborative development policies by city type. Low-efficiency cities should pursue industrial upgrading and resource optimization, while high-efficiency but low-well-being cities should improve public services. Core cities can foster regional collaboration through green technology sharing and management experience, advancing coordinated development across the agglomeration. (2) Optimize industrial and fiscal structures to enhance efficiency and reduce disparities. Shifting from secondary to service- and innovation-based industries, and directing fiscal spending toward environmental protection and social programs, can jointly improve ecological efficiency and human well-being. (3) Focus on the quality rather than the speed of urbanization. Rapid urban expansion has not always brought more equitable well-being. Policies should emphasize inclusive and compact growth, strengthen education and healthcare, and promote social integration for sustainable well-being gains. (4) Advance innovation and environmental regulation through differentiated approaches. Innovation policies should expand technological capacity, support cross-city diffusion, and apply digital tools to improve ecological efficiency and well-being. Governments should calibrate regulation strength to local capacity, ensuring that compliance costs do not outweigh ecological benefits. Declarations Author Contribution Author ContributionsR.Z. drafted the initial manuscript, conducted data analysis, designed the overall framework, prepared the figures, and carried out the survey analysis. Z.Z. provided comprehensive guidance, contributed to framework design, revised the manuscript and figures, and secured project support. Y.Z. contributed to the conceptual design in the early stage, offered suggestions on the initial draft, and provided data. All authors reviewed and approved the final manuscript. Acknowledgement This study was financially supported by the National Natural Science Foundation of China (Nos. 72474082, 72174071, 42471238), the Teaching Reform Research Project of Hubei Province (No. 2024096) and the Fundamental Research Funds for the Central Universities in China (No. CCNU25ZZ275). References Bao, R., & Liu, T. (2022). How does government attention matter in air pollution control? Evidence from government annual reports. Resources, Conservation and Recycling , 185 , 106435. https://doi.org/10.1016/j.resconrec.2022.106435. Behjat, A., & Tarazkar, M.H. (2021). Investigating the factors affecting the ecological well-being performance in Iran from 1994 to 2014. Environment, Development and Sustainability , 23 (9), 13871–13889. https://doi.org/10.1007/s10668-021-01243-0. Bian, J., Zhang, Y., Shuai, C., Shen, L., Ren, H., & Wang, Y. (2020). Have cities effectively improved ecological well-being performance? Empirical analysis of 278 Chinese cities. Journal of Cleaner Production , 245 , 118913. https://doi.org/10.1016/j.jclepro.2019.118913. Cao, M., Chen, M., Zhang, J., Pradhan, P., Guo, H., Fu, B., Li, Y., Bai, Y., Chang, L., Chen, Y., Sun, Z., Xu, Z., Zhu, R., Meadows, M.E., & Lü, G. (2023). Spatio-temporal changes in the causal interactions among Sustainable Development Goals in China. Humanities and Social Sciences Communications , 10 , 450. https://doi.org/10.1057/s41599-023-01952-z. Chen, H., Shi, Y., & He, H. (2024). A sustainable urban framework based on natural capital and ecosystem services accounting. Applied Spatial Analysis and Policy , 18 , 23. https://doi.org/10.1007/s12061-024-09631-5. Chen, W., Gu, T., & Zeng, J. (2022). Urbanisation and ecosystem health in the Middle Reaches of the Yangtze River urban agglomerations, China: A U-curve relationship. Journal of Environmental Management , 318 , 115565. https://doi.org/10.1016/j.jenvman.2022.115565 Cheng, D., Hu, S., & Zeng, G. (2025). Spatio-temporal evolution of green technology innovation networks and its proximity mechanism in the context of digital transformation: The case of Yangtze river delta in China. Applied Spatial Analysis and Policy , 18 , 106. https://doi.org/10.1007/s12061-025-09685-z. Common, M. (2007). Measuring national economic performance without using prices. Ecological Economics , 64 (1), 92–102. https://doi.org/10.1016/j.ecolecon.2007.07.017. Costanza, R., Daly, L., Fioramonti, L., Giovannini, E., Kubiszewski, I., Mortensen, L.F., Pickett, K.E., Ragnarsdottir, K.V., De Vogli, R., & Wilkinson, R. (2016). Modelling and measuring sustainable wellbeing in connection with the UN Sustainable Development Goals. Ecological Economics , 130 , 350–355. https://doi.org/10.1016/j.ecolecon.2016.07.009. Daly, H.E. (1974). The world dynamics of economic growth: The economics of the steady state. American Economic Review , 64 (2), 15–23. Daly, H.E. (2005). Economics in a full world. IEEE Engineering Management Review , 33 (4), 21. https://doi.org/10.1109/EMR.2005.27010. Deng, Y.J., Yang, X., Ma, Q.W., & Wang, K.D. (2021). Regional disparity and convergence of China’s ecological welfare performance level. China Population, Resources and Environment , 31 (4), 132–143. Dong, Y., Sun, Y., & Ding, J. (2024). Decomposition of ecological welfare performance drivers in China. Acta Geographica Sinica , 79 (5), 1337–1354. https://doi.org/10.11821/dlxb202405015. Fang, C., Wang, Z., & Ma, H. (2018). The theoretical cognition of the development law of China’s urban agglomeration and academic contribution. Acta Geographica Sinica , 73 (4), 651–665. https://doi.org/10.11821/dlxb201804005. Fang, S.J., & Xiao, Q. (2019). Research on regional ecological well-being performance and spatial effect in China. China Population, Resources and Environment , 29 (3), 1–10. Fanning, A.L., & Raworth, K. (2025). Doughnut of social and planetary boundaries monitors a world out of balance. Nature , 646 (8083), 47–56. https://doi.org/10.1038/s41586-025-09385-1. Folke, C., Polasky, S., Rockstrom, J., Galaz, V., Westley, F., Lamont, M., Scheffer, M., Osterblom, H., Carpenter, S.R., Chapin, F.S., 3rd, Seto, K.C., Weber, E.U., Crona, B.I., Daily, G.C., Dasgupta, P., Gaffney, O., Gordon, L.J., Hoff, H., Levin, S.A., Lubchenco, J., Steffen, W., & Walker, B.H. (2021). Our future in the Anthropocene biosphere. Ambio , 50 (4), 834–869. https://doi.org/10.1007/s13280-021-01544-8. Friedman, J.H. (2001). Greedy function approximation: A gradient boosting machine. The Annals of Statistics , 29 (5), 1189–1232. https://doi.org/10.1214/aos/1013203451. Fu, Z., Yang, X., Ma, Y., Sun, Y., & Wang, T. (2025). Integrating explainable AI and causal inference to unveil regional air quality drivers in China. Journal of Environmental Management , 390 , 126270. https://doi.org/10.1016/j.jenvman.2025.126270. Gao, Y., Zhang, N., Ma, Q., & Li, J. (2024). How is human well-being related to ecosystem services at town and village scales? A case study from the Yangtze River Delta, China. Landscape Ecology , 39 (7), 126. https://doi.org/10.1007/s10980-024-01925-w. Godec, P., Pancur, M., Ilenic, N., Copar, A., Strazar, M., Erjavec, A., Pretnar, A., Demsar, J., Staric, A., Toplak, M., Zagar, L., Hartman, J., Wang, H., Bellazzi, R., Petrovic, U., Garagna, S., Zuccotti, M., Park, D., Shaulsky, G., & Zupan, B. (2019). Democratized image analytics by visual programming through integration of deep models and small-scale machine learning. Nature Communications , 10 , 4551. https://doi.org/10.1038/s41467-019-12397-x. Guo, Q., & Qian, H. (2021). Negative human capital externalities in well-being: evidence from Chinese cities. Regional Studies , 55 (6), 1046–1058. https://doi.org/10.1080/00343404.2020.1861237. Han, H., Gu, R., & Yang, Y. (2025). Impacts of low-carbon city pilot policy on ecological well-being performance across Chinese cities: A spatial difference-in-difference analysis. Sustainable Cities and Society , 118 , 105864. https://doi.org/10.1016/j.scs.2024.105864. Hu, M., Pang, Y., Jin, T., & Li, Z. (2021). Spatio-temporal evolution of EIWB and influencing factors: An empirical study from the Yangtze river delta. Applied Spatial Analysis and Policy , 14 , 999–1024. https://doi.org/10.1007/s12061-021-09389-0. Knight, K.W. (2014). Temporal variation in the relationship between environmental demands and well-being: A panel analysis of developed and less-developed countries. Population and Environment , 36 (1), 32–47. https://doi.org/10.1007/s11111-013-0200-1. Knight, K.W., & Rosa, E.A. (2011). The environmental efficiency of well-being: A cross-national analysis. Social Science Research , 40 (3), 931–949. https://doi.org/10.1016/j.ssresearch.2010.11.002. Li, C.Y., Zhang, S.Q., Zhang, W., & Liao, X.C. (2018). Measurement and influencing factors of inter-provincial ecological well-being performance in China. Scientia Geographica Sinica , 39 (12), 1875–1883. https://doi.org/10.13249/j.cnki.sgs.2019.12.005. Li, Z. (2022). Extracting spatial effects from machine learning model using local interpretation method: An example of SHAP and XGBoost. Computers, Environment and Urban Systems , 96 , 101845. https://doi.org/10.1016/j.compenvurbsys.2022.101845. Long, L.J., Wang, X., & Guo, B. (2017). Evaluation of urban ecological well-being performance based on revised DEA model: A case study of 35 major cities in China. Journal of Natural Resources , 32 (4), 595–605. https://doi.org/10.11849/zrzyxb.20160515. Lundberg, S.M., Erion, G., Chen, H., DeGrave, A., Prutkin, J.M., Nair, B., Katz, R., Himmelfarb, J., Bansal, N., & Lee, S.I. (2020). From local explanations to global understanding with explainable ai for trees. Nature Machine Intelligence , 2 , 56–67. https://doi.org/10.1038/s42256-019-0138-9. Lundberg, S.M., & Lee, S.-I. (2017). A unified approach to interpreting model predictions. In Proceedings of the 31st International Conference on Neural Information Processing Systems , edited by S. M. Lundberg and S.-I. Lee. Long Beach, California, USA: Curran Associates Inc. Ma, H., Wang, K., & Fang, C. (2025). How does new-type urbanization promote green development efficiency in China? The mediating role of governments, enterprises, and residents. Geography and Sustainability , 6 (3), 100241. https://doi.org/10.1016/j.geosus.2024.10.001. Malone, B.P., Minasny, B., Odgers, N.P., & McBratney, A.B. (2014). Using model averaging to combine soil property rasters from legacy soil maps and from point data. Geoderma , 232 , 34–44. https://doi.org/10.1016/j.geoderma.2014.04.033. Marks, N., Abdallah, S., Simms, A., & Thompson, S. (2006). The (un)Happy Planet Index: An index of human well-being and environmental impact. Pandey, B., Brelsford, C., & Seto, K.C. (2022). Infrastructure inequality is a characteristic of urbanization. Proceedings of the National Academy of Sciences of the United States of America , 119 (15), e2119890119. https://doi.org/10.1073/pnas.2119890119. Rees, W.E. (1992). Ecological footprints and appropriated carrying capacity: What urban economics leaves out. Environment and Urbanization , 4 (2), 121–130. https://doi.org/10.1177/095624789200400212. Rockström, J., Gupta, J., Lenton, T.M., Qin, D., Lade, S.J., Abrams, J.F., Jacobson, L., Rocha, J.C., Zimm, C., Bai, X., Bala, G., Bringezu, S., Broadgate, W., Bunn, S.E., DeClerck, F., Ebi, K.L., Gong, P., Gordon, C., Kanie, N., Liverman, D.M., Nakicenovic, N., Obura, D., Ramanathan, V., Verburg, P.H., van Vuuren, D.P., & Winkelmann, R. (2021). Identifying a safe and just corridor for people and the planet. Earth’s Future , 9 (4), e2020EF001866. https://doi.org/10.1029/2020EF001866. Rousseeuw, P.J. (1987). Silhouettes: A graphical aid to the interpretation and validation of cluster analysis. Journal of Computational and Applied Mathematics , 20 , 53–65. https://doi.org/10.1016/0377-0427(87)90125-7. Sha, A., Madhan, S., Karthikeya, M., Megha, R., Dhanush, K.R., Swain, D., Gopakumar, G., & Geetha, M. (2024). Machine learning for sustainable development: Ranking villages for rural development initiatives. Applied Spatial Analysis and Policy , 18 , 6. https://doi.org/10.1007/s12061-024-09606-6. Shao, S., Ge, L., & Zhu, J. (2024). How to achieve the harmony between humanity and nature: environmental regulation and environmental welfare performance from the perspective of geographical factors. Journal of Management World (8), 119–114. https://doi.org/10.19744/j.cnki.11-1235/f.2024.0088. Štajdohar, M., & Demšar, J. (2013). Interactive network exploration with Orange. Journal of Statistical Software , 53 (6), 1–24. https://doi.org/10.18637/jss.v053.i06. Sun, L., Chen, J., Li, Q., & Huang, D. (2020). Dramatic uneven urbanization of large cities throughout the world in recent decades. Nature Communications , 11 (1), 5366. https://doi.org/10.1038/s41467-020-19158-1. Tone, K. (2002). A slacks-based measure of super-efficiency in data envelopment analysis. European Journal of Operational Research , 143 (1), 32–41. https://doi.org/10.1016/S0377-2217(01)00324-1. Trok, J.T., Barnes, E.A., Davenport, F.V., & Diffenbaugh, N.S. (2024). Machine learning-based extreme event attribution. Science Advances , 10 (34), eadl3242. https://doi.org/10.1126/sciadv.adl3242. UN Environment Programme. (2019). Global environment outlook–GEO-6: Healthy planet, healthy people . Cambridge: Cambridge University Press. UN Global Compact. (2015). Sustainable development goals: 17 goals to transform our world. https://www.un.org/en/exhibits/page/sdgs-17-goals-transform-world. UN Sustainable Development Solutions Network. (2024). Sustainable Development Report 2024. https://dashboards.sdgindex.org/. Varian, H.R. (2014). Big data: New tricks for econometrics. Journal of Economic Perspectives , 28 (2), 3–28. https://doi.org/10.1257/jep.28.2.3. Wang, R., Wang, Q., & Yao, S. (2021). Evaluation and difference analysis of regional energy efficiency in China under the carbon neutrality targets: Insights from DEA and Theil models. Journal of Environmental Management , 293 , 112958. https://doi.org/10.1016/j.jenvman.2021.112958. Wang, S., Huang, L., Xu, X., & Li, J. (2022). Spatio-temporal variations in ecological spaces and their ecological carrying status in China’s mega-urban agglomerations. Journal of Geographical Sciences , 32 (9), 1683–1704. https://doi.org/10.1007/s11442-022-2018-1. Wang, X., & Gao, C. (2024). Does green finance policy help to improve carbon reduction welfare performance? Evidence from China. Energy Economics , 132 , 107452. https://doi.org/10.1016/j.eneco.2024.107452. Xia, M., & Li, J. (2022). Assessment of ecological well-being performance and its spatial correlation analysis in the Beijing-Tianjin-Hebei urban agglomeration. Journal of Cleaner Production , 362 , 132621. https://doi.org/10.1016/j.jclepro.2022.132621. Xiong, H., Liu, Y., Kuang, M., & Li, Y. (2023). Nonlinear effects of socio-economic factors on urban haze in China: Evidence from spatial econometric smooth transition autoregressive regression approach. Journal of Environmental Management , 347 , 118991. https://doi.org/10.1016/j.jenvman.2023.118991. Yağcı, M. (2022). Educational data mining: prediction of students’ academic performance using machine learning algorithms. Smart Learning Environments , 9 (1), 11. https://doi.org/10.1186/s40561-022-00192-z. Yang, L., Ma, Z., & Xu, Y. (2023). How does the digital economy affect ecological well-being performance? Evidence from three major urban agglomerations in China. Ecological Indicators , 157 , 111261. https://doi.org/10.1016/j.ecolind.2023.111261. Zhang, D., Guo, Z., Niu, X., Gao, Y., Wang, C., & Xu, H. (2022). Comprehensive evaluation on sustainable development based on planetary pressures and ecological well-being performance: A case study on the belt and road regions. Journal of Cleaner Production , 376 , 134211. https://doi.org/10.1016/j.jclepro.2022.134211. Zhang, Q., Kong, Q., Zhang, M., & Huang, H. (2024). New-type urbanization and ecological well-being performance: A coupling coordination analysis in the middle reaches of the Yangtze River urban agglomerations, China. Ecological Indicators , 159 , 111678. https://doi.org/10.1016/j.ecolind.2024.111678. Zhang, S., Zhu, D., Shi, Q., & Cheng, M. (2018). Which countries are more ecologically efficient in improving human well-being? An application of the Index of Ecological Well-being Performance. Resources, Conservation and Recycling , 129 , 112–119. https://doi.org/10.1016/j.resconrec.2017.10.015. Zhang, X., Fan, H., Liu, F., Lv, T., Sun, L., Li, Z., Shang, W., & Xu, G. (2023). Coupling coordination between the ecological environment and urbanization in the middle reaches of the Yangtze River urban agglomeration. Urban Climate , 52 , 101698. https://doi.org/10.1016/j.uclim.2023.101698. Zhang, Y., Mao, Y., Jiao, L., Shuai, C., & Zhang, H. (2021). Eco-efficiency, eco-technology innovation and eco-well-being performance to improve global sustainable development. Environmental Impact Assessment Review , 89 , 106580. https://doi.org/10.1016/j.eiar.2021.106580. Zhang, Z. (2022). Orange tutorial: Data mining and analysis with “building blocks” and “linking game” . Wu Han: China University of Geosciences Press. Zhang, Z., Dou, Y., Liu, X., & Gong, Z. (2023). Multi-hierarchical spatial clustering for characteristic towns in China: An Orange-based framework to integrate GIS and Geodetector. Journal of Geographical Sciences , 33 (3), 618–638. https://doi.org/10.1007/s11442-023-2098-6. Zhang, Z., Zhang, M., Song, X., & Li, Z. (2024). Image-based machine learning and cluster analysis for urban road network: employing Orange for codeless visual programming. Geo-spatial Information Science , 1–18. https://doi.org/10.1080/10095020.2024.2377212. Zhao, L., Cao, N., Han, Z., & Gao, X. (2022). Evolution characteristics and formation mechanism of spatial correlation network of ecological well-being performance in China. Journal of Natural Resources , 37 (12), 3183–3200. https://doi.org/10.31497/zrzyxb.20221211. Zhu, D., Zhang, S., & Sutton, D.B. (2015). Linking Daly’s Proposition to policymaking for sustainable development: Indicators and pathways. Journal of Cleaner Production , 102 , 333–341. https://doi.org/10.1016/j.jclepro.2015.04.070. Zhu, Y., Zhang, R., Gu, J., & Gao, Z. (2022). Spatiotemporal evolution and driving mechanism of ecological well-being performance in the urban agglomeration of the middle reaches of the Yangtze River under the carbon peaking and carbon neutrality goals. Progress in Geography , 41 (12), 2231–2243. https://doi.org/10.18306/dlkxjz.2022.12.004. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 14 Feb, 2026 Read the published version in Applied Spatial Analysis and Policy → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7839392","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":535110634,"identity":"69ec939d-d03e-48b3-97e5-4c8786f0af3c","order_by":0,"name":"Rui Zhang","email":"","orcid":"","institution":"Central China Normal University","correspondingAuthor":false,"prefix":"","firstName":"Rui","middleName":"","lastName":"Zhang","suffix":""},{"id":535110635,"identity":"15f65ea6-6d08-49b9-92a2-5869bd5cc579","order_by":1,"name":"Zuo Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2UlEQVRIiWNgGAWjYDACCQST8UFCRQ1pWpgNHpw5RpoWNsmHLcyEdcjPbn728Mufw3L8M9KvVSQ2sDHwt3cn4NXCOOeYubEMz2FjiRs5ZTcSd8gwSJw5uwGvFmaJBDNpCYnDiRskctJuJJ5hYzCQyMWvhU0i/Zu0hAFES0FiGzNhLTwSOWaSHxJAWtKPMRClRUIip0ya4UC6scSZN0BHnjnGQ9Av8jPSt0n++GMtx9+e/vDjj4oaIKMXvxYQYOaBuNEATBJUDgKMP8AU+wOiVI+CUTAKRsHIAwA2qUZ4/OKyvgAAAABJRU5ErkJggg==","orcid":"","institution":"Central China Normal University","correspondingAuthor":true,"prefix":"","firstName":"Zuo","middleName":"","lastName":"Zhang","suffix":""},{"id":535110637,"identity":"02830319-dc84-4f44-90f0-b2eef0c778e5","order_by":2,"name":"Yuanyuan Zhu","email":"","orcid":"","institution":"Central China Normal University","correspondingAuthor":false,"prefix":"","firstName":"Yuanyuan","middleName":"","lastName":"Zhu","suffix":""}],"badges":[],"createdAt":"2025-10-12 08:38:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7839392/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7839392/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s12061-026-09800-8","type":"published","date":"2026-02-14T15:57:15+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":94823636,"identity":"770c4ad1-f567-4231-8af8-fbcc0250a847","added_by":"auto","created_at":"2025-10-31 06:47:42","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":14111248,"visible":true,"origin":"","legend":"","description":"","filename":"ASAPManuscriptZR25.10.12.docx","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/d66aa9cb6f791a2b5b2c0715.docx"},{"id":94771289,"identity":"0f2017e2-fcc5-4c05-a683-d9bb27544c01","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5492,"visible":true,"origin":"","legend":"","description":"","filename":"396c47faec564ca89680c6003e2803de.json","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/978ca24d015ab0f944780ca1.json"},{"id":94771303,"identity":"cef255e2-dffb-47b0-a4d6-62d7a42d3ca1","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":185824,"visible":true,"origin":"","legend":"","description":"","filename":"396c47faec564ca89680c6003e2803de1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/299685336fd9363c07fe272b.xml"},{"id":94771290,"identity":"158297a4-a818-4649-b76f-0ff2a5cd41fc","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"jpeg","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2430602,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/cc546b65967e6b8b64e0e0f6.jpeg"},{"id":94825263,"identity":"3363e06b-b6a0-425a-9ff7-c22c8ae8e88d","added_by":"auto","created_at":"2025-10-31 06:50:01","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1640593,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/7bed6860a6d247d067874bc9.jpeg"},{"id":94825293,"identity":"89d78d0a-e91b-4d8b-9ff4-d1b1ba726f81","added_by":"auto","created_at":"2025-10-31 06:50:05","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":642957,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/86434911877a9123b2ea81e9.jpeg"},{"id":94771301,"identity":"7872cb07-8118-45fe-a858-f7110f90c62a","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3723566,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/810a74027bd9c3e28d5e9d2c.jpeg"},{"id":94771295,"identity":"74e83bd4-f7f6-4902-8e54-642aa8820d3c","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1676853,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/1b53c536dc4e19014e143dd5.jpeg"},{"id":94825041,"identity":"d9cb02bd-3f55-4e51-ab28-e70c38fa7329","added_by":"auto","created_at":"2025-10-31 06:49:46","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":691478,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/501a9867c2e8fb13fad189b9.jpeg"},{"id":94823903,"identity":"7780cfaf-7e75-4652-a744-189f96404a85","added_by":"auto","created_at":"2025-10-31 06:48:16","extension":"jpeg","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1876017,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/80a172915dad811f07c954e6.jpeg"},{"id":94824658,"identity":"9e459754-8e7b-4e12-83a2-fd8687fcfa57","added_by":"auto","created_at":"2025-10-31 06:49:11","extension":"jpeg","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1197315,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/14a4c6e6d76ce36d2fcc068a.jpeg"},{"id":94824974,"identity":"3b7fb4d2-49e3-41e7-8e0e-45335e406caa","added_by":"auto","created_at":"2025-10-31 06:49:38","extension":"wmf","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3482,"visible":true,"origin":"","legend":"","description":"","filename":"image3.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/c19cc57a25ddb8a9bc01b0c9.wmf"},{"id":94771316,"identity":"28eaf601-a318-4ba7-b158-147feaa43872","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"wmf","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5160,"visible":true,"origin":"","legend":"","description":"","filename":"image4.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/a5e62cc86ed265c9197cdc15.wmf"},{"id":94771307,"identity":"bca9579a-5e8d-4176-bc8a-fa6d9f2816a3","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":423478,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/cf29d8ca5b6d94641975d828.png"},{"id":94771304,"identity":"a1a40f95-75ed-4555-ab41-3544dace8033","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":281617,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/5f74294c497f585d13a92cb1.png"},{"id":94771299,"identity":"bb62297a-5b54-4d74-b30b-e532ea14e499","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":113787,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/f42e117ba5bf93964a6fd0c1.png"},{"id":94771309,"identity":"5e1d644d-3148-4393-b3fb-19b029f4a514","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":686077,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/fc227e5561043fcb14aca198.png"},{"id":94771315,"identity":"9117c5a5-61ed-4443-a76b-423f1b777936","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":226914,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/46d6ed33adb9ad1f0ca4b57c.png"},{"id":94771312,"identity":"ebb56140-9939-4bfb-b7c5-997961cce6af","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":145561,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/3457f5a285b01157d5fe8b68.png"},{"id":94824583,"identity":"972986fa-f845-4f1a-8789-1caa9e66e944","added_by":"auto","created_at":"2025-10-31 06:49:09","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":323497,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/4449c8c2e0ab698e09ecb2fc.png"},{"id":94824566,"identity":"e8beacf4-d6be-4de7-88ef-566ff83a9257","added_by":"auto","created_at":"2025-10-31 06:49:07","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":217084,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/621b32c86927560c40310b70.png"},{"id":94824946,"identity":"488d636c-d2c9-445f-8edc-bc3a7a63fa51","added_by":"auto","created_at":"2025-10-31 06:49:37","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1981,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/96424fbd36495b81788fa164.png"},{"id":94771308,"identity":"16e00706-1788-4d14-ad4e-cf1f49f95bfb","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4323,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/dbe39e550f448bcb6e939ce2.png"},{"id":94771317,"identity":"1b72f9ef-ef72-4a02-9518-24faeb5623e9","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"xml","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":185053,"visible":true,"origin":"","legend":"","description":"","filename":"396c47faec564ca89680c6003e2803de1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/a52b82e2f45169fcc5a6458e.xml"},{"id":94825061,"identity":"d22ab8b7-b097-4411-89c0-52878a4263c6","added_by":"auto","created_at":"2025-10-31 06:49:46","extension":"html","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":195887,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/a6845c5622a6c9517db29aa1.html"},{"id":94771286,"identity":"1912a19d-44ff-4320-a109-aa6b8ef564d9","added_by":"auto","created_at":"2025-10-30 14:02:58","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":431191,"visible":true,"origin":"","legend":"\u003cp\u003eIntegrated conceptual and analytical framework for ecological well-being performance\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/8ede7b844b0ed7596791f1c8.png"},{"id":94771288,"identity":"6303b61e-1636-4e20-a081-3db0228817b1","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":277532,"visible":true,"origin":"","legend":"\u003cp\u003eLocation of study area\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/028022c21713a556ce42d150.png"},{"id":94771287,"identity":"c8c29858-2b4d-4636-9a9a-eca008d246f4","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":300354,"visible":true,"origin":"","legend":"\u003cp\u003eTemporal changes of EWP in MRYRUA. \u003cem\u003eNote\u003c/em\u003e: In (c) sieve diagram, rectangle size denotes the expected frequency, the number of small squares indicates observed values. Blue and red represent positive and negative deviations from expectations, respectively, with darker shades indicating larger discrepancies.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/046175e01d35183f18357f6c.png"},{"id":94824573,"identity":"e346f176-166c-4112-a198-f8ddb5bbf7a8","added_by":"auto","created_at":"2025-10-31 06:49:08","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":501203,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial pattern of EWP in MRYRUA (2005–2022)\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/fb7fdc17ad61fd97743dc0cf.png"},{"id":94823820,"identity":"b6d390b6-30a2-47db-9cf9-fadc4f76067b","added_by":"auto","created_at":"2025-10-31 06:48:06","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":231338,"visible":true,"origin":"","legend":"\u003cp\u003eThe Theil index of EWP in MRYRUA (2005–2022)\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/e717ac5cc30a491930d557ad.png"},{"id":94771292,"identity":"2047a307-bcc4-4f10-9691-c09b1b44ed36","added_by":"auto","created_at":"2025-10-30 14:02:59","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":295485,"visible":true,"origin":"","legend":"\u003cp\u003eCity clustering based on EWP, ecological consumption and human well-being. \u003cem\u003eNote\u003c/em\u003e: (\u003cstrong\u003eb\u003c/strong\u003e) shows the results after Z-score normalization\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/c5c3e72f6a6373fe88a71722.png"},{"id":94825024,"identity":"e939a6e8-46d4-4dbf-a2bd-2fd5c465a6bf","added_by":"auto","created_at":"2025-10-31 06:49:42","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":273878,"visible":true,"origin":"","legend":"\u003cp\u003eSHAP summary plots and feature importance rankings across four models\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/5e5491cd41644110908ca790.png"},{"id":94823887,"identity":"4ee65baa-cf75-42fa-b944-80efd1e2f921","added_by":"auto","created_at":"2025-10-31 06:48:14","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":216420,"visible":true,"origin":"","legend":"\u003cp\u003eSHAP–GAM single-feature dependence and nonlinear effects of key drivers on EWP\u003c/p\u003e\n\u003cp\u003eNote：The blue solid line represents the GAM smooth trend, and the dark band represents the 95% confidence interval; The green/pink shadows represent the positive/negative marginal contributions to EWP, respectively.\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/e1fe76db42ffba0a5721ae22.png"},{"id":102785397,"identity":"e1b09729-afd8-4cec-8aeb-10e64de907c6","added_by":"auto","created_at":"2026-02-16 16:06:17","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3186705,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7839392/v1/4b7b79eb-37cc-4106-8824-b1088a7650f5.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Ecological well-being performance in a Chinese urban agglomeration: Spatiotemporal analysis and policy insights from an Orange-based machine learning framework","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eThe stability of the Earth system is deeply intertwined with human well-being (Rockstr\u0026ouml;m et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Since the Industrial Revolution, rapid industrialization and high-consumption lifestyles have accelerated ecological degradation and resource depletion, intensifying risks such as climate change, biodiversity loss, and pollution (Folke et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). At the same time, persistent inequalities in poverty, health, and education continue to challenge global equity (UN Environment Programme, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). To address these dual crises, the Sustainable Development Goals (SDGs) have called for improving human well-being within planetary boundaries (UN Global Compact, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). However, recent assessments (UN Sustainable Development Solutions Network, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) suggest that about 83% of SDG targets may not be achieved by 2030, highlighting the challenge of enhancing human well-being within ecological ceiling. This global difficulty underscores the importance for rapidly developing countries to design scientific and multi-scale regional policies.\u003c/p\u003e\u003cp\u003eIn China, rapid urbanization and industrialization over the past four decades have brought unprecedented economic growth, but also severe environmental pressures (Ma et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). In response, national strategies such as Ecological Civilization Construction (2012), New-type Urbanization (2014), and Rural Revitalization Strategy (2017) aim to align economic growth with ecological protection and social sustainability, striving to better balance rapid development with ecological conservation, immediate development needs with long-term sustainability, and regional development disparities. Urban agglomerations, as integrated spaces of multiple cities and urban\u0026ndash;rural systems, are not only key arenas for economic growth, population concentration, and ecological risk management, but also exhibit greater complexity in spatial governance and policy-making (Fang et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Wang et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Among them, the middle reaches of the Yangtze River urban agglomeration (MRYRUA), located at the central junction linking east\u0026ndash;west and north\u0026ndash;south China and stretching along the country\u0026rsquo;s largest inland river, is a rapidly expanding yet ecologically fragile region. It exemplifies the trade-offs between development and environment in emerging megaregions (Zhang et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). In this context, improving resource efficiency, safeguarding ecosystems, and enhancing well-being have become pressing issues. Ecological well-being performance (EWP), as an indicator of how efficiently ecological consumption is transformed into human well-being (Zhu et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), provides an effective framework for analyzing these issues.\u003c/p\u003e\u003cp\u003eThe concept of EWP originated from steady-state economics (Daly, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e1974\u003c/span\u003e), but quantitative research gained traction after the introduction of ecological footprint theory (Rees, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e1992\u003c/span\u003e). Subsequently, several similar concepts combining human well-being and environmental impact have emerged, such as the happy planet index (Marks et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), national economic performance (Common, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), environmental efficiency of well-being (Knight \u0026amp; Rosa, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Early studies typically defined EWP as a single-ratio relationship between human well-being and ecological consumption (Knight, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Zhu et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). In recent years, model-based methods, especially Data Envelopment Analysis (DEA), have been widely adopted to accommodate multiple input and output indicators (Zhang et al., \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), enabling more detailed assessment of ecological inputs and well-being outputs. As measurement methods matured, numerous studies have examined the temporal and spatial characteristics of EWP across multiple geographic scales, including national (Zhang et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), provincial (Li et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), and urban levels(Bian et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Hu et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Temporally, some studies have explored EWP trends and dynamic changes (Behjat \u0026amp; Tarazkar, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Spatially, researchers have investigated the spatial characteristics and regional disparities of EWP using spatial correlation networks (Zhao et al., \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), spatial autocorrelation analysis (Xia \u0026amp; Li, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) and spatial convergence models (Deng et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Another major research area focuses on identifying the determinants of EWP. Previous studies have comprehensively examined various factors affecting EWP (Li et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Zhu et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Several studies have specifically examined the effects of green finance policies (Wang \u0026amp; Gao, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), digital economy (Yang et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), environmental regulations (Shao et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), low-carbon city pilot policies (Han et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) and new-type urbanization (Zhang et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Traditional econometric approaches such as multiple linear regression, panel Tobit regression (Long et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), and spatial econometric models (Fang \u0026amp; Xiao, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) have been employed.\u003c/p\u003e\u003cp\u003eOverall, existing studies provide valuable insights into the measurement, spatiotemporal features, and drivers of EWP. However, significant gaps remain. First, current assessments rarely differentiate EWP types. Given that EWP is a relative indicator, similar scores can obscure substantial disparities in input\u0026ndash;output configurations or overlook unsustainable situations, such as excessive ecological consumption or inadequate human well-being, which may breach safe and just boundaries (Rockstr\u0026ouml;m et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). This points to the necessity of a more nuanced classification of ecology\u0026ndash;well-being transformation types. Second, existing studies rarely investigate the nonlinear and heterogeneous effects of factors influencing EWP. Traditional linear regression models, in particular, struggle to capture such complexities within the human\u0026ndash;environment system underlying the ecology\u0026ndash;well-being transformation. To address these gaps, this study proposes a comprehensive analytical framework that integrates spatiotemporal analysis, K-means clustering, and interpretable machine learning to examine spatiotemporal patterns, types, and drivers of EWP across the MRYRUA. The framework is implemented on Orange (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://orangedatamining.com/\u003c/span\u003e\u003cspan address=\"https://orangedatamining.com/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), an open-source visual programming platform that facilitates efficient data processing and modeling (Zhang et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). While Orange has been widely applied in biomedicine (Godec et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), education (Yağcı, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), and urban research (Zhang et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), its potential in sustainability analytics remains largely unexplored.\u003c/p\u003e\u003cp\u003eIn summary, building on the above four sections, this study focuses on rapidly developing, spatially integrated and ecologically fragile urban agglomerations, analyzing the relationship between human well-being and ecological protection from a spatial perspective to support sound regional policy and ultimately achieve balanced and sustainable development. Against this background, the objectives of this study are threefold. First, to construct and apply an Orange-based analytical framework for investigating the spatiotemporal divergence and driving mechanisms of EWP in the MRYRUA. Second, to classify cities into EWP types by jointly considering inputs, outputs, and efficiency, guided by the safe and just framework, so as to uncover differences in efficiency and balance. Third, to employ interpretable machine learning techniques to capture nonlinear relationships and regional heterogeneity that cannot be adequately addressed by traditional regression models. Finally, in terms of evaluation and analytical methods, this study also seeks to contribute to the application of a multi-model analysis platform characterized by strong visualization, high integrative capacity, and interactive features. The remainder of this paper is organized as follows: Section \u003cspan refid=\"Sec2\" class=\"InternalRef\"\u003e2\u003c/span\u003e introduces the framework, methodology, study area, and data; Section \u003cspan refid=\"Sec9\" class=\"InternalRef\"\u003e3\u003c/span\u003e presents the spatiotemporal evolution of EWP and its drivers; Section \u003cspan refid=\"Sec17\" class=\"InternalRef\"\u003e4\u003c/span\u003e discusses the findings; and Section \u003cspan refid=\"Sec18\" class=\"InternalRef\"\u003e5\u003c/span\u003e concludes and provides the policy insights.\u003c/p\u003e"},{"header":"2 Framework and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003e\u003cstrong\u003e2.1 Framework design\u003c/strong\u003e\u003c/h2\u003e\n \u003cp\u003eNatural ecosystem and socio-economic system are deeply interdependent. Natural ecosystem provides essential services that support human survival, development, and well-being (Costanza et al., \u003cspan class=\"CitationRef\"\u003e2016\u003c/span\u003e; Chen et al., \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e), while human activities consume natural resources, generate waste, and reshape the environment (Daly, \u003cspan class=\"CitationRef\"\u003e1974\u003c/span\u003e). From a human\u0026ndash;environment systems perspective, this process can be understood as the transformation of ecological consumption into human well-being, with strong regional variation shaped by factors such as urbanization, technological innovation, industrial transformation, globalization, and government regulation (Gao et al., \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). Ecological well-being performance (EWP), as an efficiency-oriented metric, captures the conversion relationship between ecological consumption and human well-being. Within this coupled system, ecological consumption and human well-being are constrained by an ecological ceiling, which represents the limits of environmental carrying capacity, and a social foundation, which reflects the minimum thresholds necessary for human development and well-being (Fanning \u0026amp; Raworth, \u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e). Sustainable improvement in EWP therefore needs to occur within these ecological and social boundaries. Given the complexity and diversity of this transformation across regions and over time, a multidimensional analytical perspective is essential. Specifically, spatiotemporal analysis reveals temporal evolution and spatial heterogeneity, typological differentiation identifies structural imbalances that similar EWP scores may obscure, and interpretable machine learning captures nonlinear and heterogeneous drivers. Building on the coupled human\u0026ndash;environment systems conceptual framework, this study develops a structured and scalable Orange-based analytical framework that integrates these three dimensions to uncover uneven and multi-type divergences in EWP (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). This conclusion highlights the necessity of combining spatiotemporal, typological, and machine learning perspectives as a key consideration for the framework proposed in the following part.\u003c/p\u003e\n \u003cp\u003eThe analytical framework is built on Orange, a no-code visual programming platform developed by the University of Ljubljana (\u0026Scaron;tajdohar \u0026amp; Dem\u0026scaron;ar, \u003cspan class=\"CitationRef\"\u003e2013\u003c/span\u003e), and integrates the GIS platform and the Super-SBM model. It consists of four key modules, including temporal analysis, spatial analysis, clustering, and influencing factors. The Super-SBM model is used to measure EWP values. The GIS tools provide technical support for spatial data organization and visualization. All analytical modules are powered by eight categories of Orange widgets (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://orangedatamining.com/widget-catalog/\u003c/span\u003e\u003c/span\u003e), each offering specific functions and allowing interconnection to accommodate various analytical tasks (Zhang, \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e; Zhang et al., \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). The four core modules are as follows:\u003c/p\u003e\n \u003cp\u003e(1) \u003cstrong\u003eTemporal analysis module\u003c/strong\u003e examines dynamic trends of EWP over time, capturing long-term variation and stage characteristics. It incorporates widgets such as line plots, scatter plots, sieve diagrams, and box plots.\u003c/p\u003e\n \u003cp\u003e(2) \u003cstrong\u003eSpatial analysis module\u003c/strong\u003e visualizes the spatial distribution using the Geo Map widget. Regional disparities are assessed using a customized Theil index widget, developed for this study using Orange\u0026rsquo;s open-source architecture. The index is calculated following the approach outlined in Wang et al. (\u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003e(3) \u003cstrong\u003eClustering module\u003c/strong\u003e applies K-means clustering to jointly analyze ecological inputs, well-being outputs, and overall efficiency, thereby identifying distinct EWP types and revealing heterogeneous transformation pathways across cities.\u003c/p\u003e\n \u003cp\u003e(4) \u003cstrong\u003eInfluencing factors module\u003c/strong\u003e employs interpretable machine learning to identify drivers of EWP, quantify their contributions, and capture nonlinear and heterogeneous effects, offering a deeper understanding of the mechanisms behind EWP variation.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n \u003ch2\u003e2.2 Materials and methods\u003c/h2\u003e\n \u003cdiv id=\"Sec5\" class=\"Section3\"\u003e\n \u003ch2\u003e\u003cstrong\u003e2.2.1\u003c/strong\u003e Study area and data\u003c/h2\u003e\n \u003cp\u003eThe middle reaches of the Yangtze River urban agglomeration (MRYRUA) is a large-scale urban cluster that consists of the Wuhan Metropolitan Area (WMA), the Changsha-Zhuzhou-Xiangtan City Group (CZXCG), and the Poyang Lake City Group (PLCG) (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). It lies in the middle reaches of the Yangtze River, China\u0026rsquo;s largest river, where river basin governance and ecological governance are deeply intertwined. Covering approximately 317,000 km\u0026sup2; and comprising 31 prefecture-level cities, it represents China\u0026rsquo;s largest urban agglomeration by land area. Spanning three provinces and encompassing cities of different administrative levels as well as both urban and rural areas, the region faces greater challenges of cross-regional spatial governance and policy coordination. By the end of 2022, the population had reached 127 million, accounting for about 9% of the national total. MRYRUA plays a critical role in national strategies such as the Rise of Central China, New-type Urbanization, and Beautiful China initiatives (Zhang et al., \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). However, it still faces significant pressures from rapid population growth, resource consumption, and environmental challenges. Given these conditions, this study selected the MRYRUA as a representative case for city-level spatiotemporal analysis of EWP. The findings are expected to support sustainable urbanization in China and provide insights for other rapidly urbanizing regions in developing countries.\u003c/p\u003e\n \u003cp\u003ePrefecture-level cities within the MRYRUA were chosen as the basic spatial units for analysis. From the perspective of spatial units and data sources, this design ensures consistency and comparability across cities, while also reflecting the practical basis for EWP measurement at the urban scale. Socioeconomic data covering the period from 2005 to 2022 were primarily sourced from the China City Statistical Yearbook, China Urban Construction Statistical Yearbook, the Hunan, Hubei, and Jiangxi Statistical Yearbooks, as well as municipal yearbooks, socioeconomic development statistical bulletins. Missing values were supplemented using linear interpolation or trend extrapolation methods. The vector data of city boundaries were sourced from the Resource and Environmental Science Data Platform (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ewww.resdc.cn\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e\n \u003ch2\u003e\u003cstrong\u003e2.2.2\u003c/strong\u003e EWP indicators\u003c/h2\u003e\n \u003cp\u003eBuilding on the theoretical foundation of EWP and previous studies (Hu et al., \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e; Xia \u0026amp; Li, \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e; Zhu et al., \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e; Han et al., \u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e), this study constructs an evaluation indicator system for EWP (Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). From a multiple-input and multiple-output perspective, it considers resource consumption and environmental degradation as cost-type input, while economic, social, and environmental dimensions of human well-being are treated as benefit-type output.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab1\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eEvaluation indicators of ecological well-being performance\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eDimensions\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ePrimary indicators\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSecondary indicators\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eIndicator description\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"6\"\u003e\n \u003cp\u003eInput-ecological consumption\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" rowspan=\"3\"\u003e\n \u003cp\u003eResource consumption\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePer capita built-up area (m\u0026sup2;/person)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eReflects land use intensity.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePer capita energy consumption (tce/person)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eReflects energy use intensity.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePer capita water consumption (m\u0026sup3;/person)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eReflects water use intensity.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"3\"\u003e\n \u003cp\u003eEnvironmental damage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePer capita industrial wastewater discharge (t/person)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eReflects the damage to the water environment caused by wastewater discharge.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePer capita SO₂ emissions (t/person)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eReflects the damage to the atmospheric environment caused by exhaust emissions.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePer capita municipal solid waste disposal (kg/person)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eReflects the damage to the soil environment caused by solid waste disposal.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"7\"\u003e\n \u003cp\u003eOutput - human well-being\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\" rowspan=\"2\"\u003e\n \u003cp\u003eThriving economy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePer capita disposable income of urban residents (CNY/person)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndicates the income-related well-being of urban residents\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePer capita disposable income of rural residents (CNY/person)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndicates the income-related well-being of rural residents\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"3\"\u003e\n \u003cp\u003eProgressive society\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRatio of per capita disposable income: rural to urban residents\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndicates the level of social equity and urban\u0026ndash;rural common prosperity\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAverage years of schooling (years)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eReflects the well-being benefits associated with access to education.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eHospital Beds per 10,000 Population\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eReflects the well-being benefits related to healthcare access and coverage\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\" rowspan=\"2\"\u003e\n \u003cp\u003eEco-friendly environment\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePer capita green space area (m\u0026sup2;/person)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eReflects individual-level environmental well-being through access to green public spaces.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGreen coverage rate in built-up areas (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eReflects the ecological livability of urban areas through overall green coverage.\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003eThe EWP indicator system is refined by integrating insights from previous research and introducing targeted improvements, ensuring that it is more closely aligned with the realities of the study region while fully considering alignment with national SDG goals. To better capture economic prosperity, it replaces the commonly used GDP with urban and rural per capita income (Guo \u0026amp; Qian, \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e). For progressive society, it incorporates a social equity indicator measured by the urban\u0026ndash;rural income ratio. This indicator is aligns with SDG 10 on reducing inequality (UN Sustainable Development Solutions Network, \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e), yet has been largely overlooked in prior EWP studies. Notably, manufactured capital, such as labor and physical capital\u0026mdash;commonly treated as intermediate inputs in ecological economic efficiency studies\u0026mdash;is not excluded from the indicator system, as it ultimately originates from natural capital (Daly, \u003cspan class=\"CitationRef\"\u003e2005\u003c/span\u003e).\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec7\" class=\"Section3\"\u003e\n \u003ch2\u003e2.2.3 Super-SBM model\u003c/h2\u003e\n \u003cp\u003eThe Slack-Based Measure (SBM) model evaluates the relative efficiency of decision-making units through linear programming. It accommodates multiple inputs and outputs without requiring a predefined production function, making it widely applicable in efficiency analysis (Han et al., \u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e). However, the traditional SBM model cannot distinguish among fully efficient units. The Super-SBM model addresses this limitation by enabling the ranking of efficient units (Tone, \u003cspan class=\"CitationRef\"\u003e2002\u003c/span\u003e). Therefore, this study adopts the Super-SBM model to evaluate the EWP of the MRYRUA. The model is formulated as follows:\u003c/p\u003e\n \u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\n \u003cdiv class=\"EquationNumber\"\u003e\u003cimg src=\"data:image/png;base64,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\"\u003e\u003c/div\u003e\n \u003c/div\u003e\n \u003cp\u003eIn the Super-SBM model, \u003cem\u003e\u0026rho;\u003c/em\u003e* represents the EWP score; \u003cem\u003eM\u003c/em\u003e and \u003cem\u003eN\u003c/em\u003e denote the number of input and output variables, respectively, and \u003cem\u003eI\u003c/em\u003e is the total number of decision-making units (DMUs). \u003cem\u003ex\u003c/em\u003e\u003csub\u003e\u003cem\u003em\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003ey\u003c/em\u003e\u003csub\u003e\u003cem\u003en\u003c/em\u003e\u003c/sub\u003e represent the input and output vectors, while \u003cem\u003ex\u003c/em\u003e\u003csub\u003e\u003cem\u003emk\u003c/em\u003e\u003c/sub\u003e and \u003cem\u003ey\u003c/em\u003e\u003csub\u003e\u003cem\u003enk\u003c/em\u003e\u003c/sub\u003e denote the specific values of the \u003cem\u003em\u003c/em\u003e-th input and \u003cem\u003en\u003c/em\u003e-th output for the \u003cem\u003ek\u003c/em\u003e-th DMU. \u003cem\u003eS\u003c/em\u003e\u003csub\u003e\u003cem\u003em\u003c/em\u003e\u003c/sub\u003e\u003csup\u003e\u0026minus;\u003c/sup\u003e and \u003cem\u003eS\u003c/em\u003e\u003csub\u003e\u003cem\u003en\u003c/em\u003e\u003c/sub\u003e\u003csup\u003e+\u003c/sup\u003e are the input and output slack variables, respectively. \u0026lambda;\u003csub\u003e\u003cem\u003ei\u003c/em\u003e\u003c/sub\u003e is the weight variable associated with the \u003cem\u003ei\u003c/em\u003e-th DMU. If \u003cem\u003e\u0026rho;\u003c/em\u003e*\u0026ge;1, the DMU is considered fully efficient, meaning that both input and output slacks are equal to zero. If \u003cem\u003e\u0026rho;\u003c/em\u003e*\u0026lt;1, the DMU exhibits inefficiency due to the presence of input excess or output shortfall.\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e\n \u003ch2\u003e2.2.4 K-means clustering and interpretable machine learning\u003c/h2\u003e\n \u003cp\u003eK-means is a widely used unsupervised learning algorithm that partitions a dataset into a specified number of clusters, aiming to maximize similarity within clusters and differences between clusters (Rousseeuw, \u003cspan class=\"CitationRef\"\u003e1987\u003c/span\u003e). By avoiding predefined classification thresholds, the unsupervised clustering method helps circumvent the subjectivity often inherent in manual or rule-based groupings. It is useful for identifying patterns in cities based on their input\u0026ndash;output\u0026ndash;efficiency characteristics, thereby providing insights into variations in resource use and social well-being. This study uses the K-means clustering algorithm to examine ecological consumption, human well-being, and EWP across the MRYRUA.\u003c/p\u003e\n \u003cp\u003eMachine learning models are widely applied in regression, clustering, prediction, and dimensionality reduction tasks, and have seen increasing use across diverse fields such as geography, economics, public administration, and climate science(Lundberg et al., \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e; Trok et al., \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). Using multiple models helps avoid issues like overfitting or underfitting associated with single models and improves prediction accuracy, especially when dealing with complex nonlinear relationships and spatiotemporal heterogeneity in geographic data (Malone et al., \u003cspan class=\"CitationRef\"\u003e2014\u003c/span\u003e). Therefore, this study employs four machine learning models to explore the key factors influencing EWP in the MRYRUA. These models include:\u003c/p\u003e\n \u003cp\u003e(1) Multiple Linear Regression (MLR), a traditional method for modeling linear relationships, serves as the baseline model in this study. (2) Adaptive Boosting (AdaBoost), an ensemble method, improves performance by iteratively focusing on difficult-to-predict samples and combining weak learners into a strong model (Friedman, \u003cspan class=\"CitationRef\"\u003e2001\u003c/span\u003e). (3) Random Forest (RF) constructs multiple randomized decision trees and aggregates their results, offering robustness and strong resistance to overfitting (Varian, \u003cspan class=\"CitationRef\"\u003e2014\u003c/span\u003e). (4) Extreme Gradient Boosting (XGBoost) enhances gradient boosting by incorporating second-order optimization, regularization, and parallel processing to improve efficiency and reduce overfitting(Li, \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e). To ensure reliable and generalizable results, all models were trained using 5-fold cross-validation, with hyperparameters optimized via random search.\u003c/p\u003e\n \u003cp\u003eModel interpretability is critical in understanding the underlying drivers of EWP. To this end, Shapley Additive Explanations (SHAP) is employed to quantify the contribution of individual features to model outputs (Lundberg \u0026amp; Lee, \u003cspan class=\"CitationRef\"\u003e2017\u003c/span\u003e). Based on cooperative game theory, SHAP provides a robust framework for feature attribution, enabling interpretation of complex \u0026ldquo;black-box\u0026rdquo; models at both global and local levels. It ensures consistent and reliable explanations by assigning higher SHAP values to more influential features. To further capture nonlinear relationships, this study follows the approach of Fu et al. (\u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e) by combining SHAP with Generalized Additive Models (GAM).Specifically, the SHAP values are regressed on their actual variable values using GAM to generate smooth response curves, which reveal the ranges of positive and negative effects as well as the turning points where the influence of a factor fundamentally changes.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e"},{"header":"3 Results","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003e3.1 Spatiotemporal characteristics of EWP\u003c/h2\u003e\n \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e\n \u003ch2\u003e3.1.1 Temporal evolution characteristics\u003c/h2\u003e\n \u003cp\u003eTemporal analysis is of particular significance in this study as it reveals dynamic changes in the relationship between ecological consumption and human well-being, especially for rapidly developing countries and regions. Based on Orange\u0026rsquo;s temporal analysis module, the time-series evolution of EWP in the MRYRUA from 2005 to 2022 is analyzed. During this period, the average annual EWP value increased by 37.45% (from 0.558 to 0.767), indicating a marked overall improvement. As shown in Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e(a), EWP followed a trajectory of early fluctuations followed by steady growth, with all three sub-agglomerations improving\u0026mdash;particularly the CZCXG region. Figure \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e(b) shows a clear upward shift in EWP distributions over time, consistent with the line plot. The spread of data points also indicates notable spatial and temporal heterogeneity. Figure \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e(c), a sieve diagram, confirms this trend: in 2005\u0026ndash;2013, low EWP values (\u0026lt;\u0026thinsp;0.38) dominated; from 2014\u0026ndash;2017, moderate values (0.38\u0026ndash;0.52) increased; and by 2018\u0026ndash;2022, high EWP values (\u0026ge;\u0026thinsp;0.74) became significantly more prevalent. These results suggest a two-stage trajectory of EWP development in the MRYRUA\u0026mdash;initial fluctuations followed by sustained improvement, as discussed below.\u003c/p\u003e\n \u003cp\u003e(1) Fluctuation phase (2005\u0026ndash;2013): During this period, EWP followed a volatile trajectory with an overall downward trend. From 2005 to 2008, EWP declined year by year, then rebounded after 2008, but reached a low point again in 2011. This phase coincided with rapid urbanization and industrialization in the study area. During this time, cities placed disproportionate emphasis on economic growth and adhered to an extensive development pattern. This pattern, characterized by high input, intensive energy consumption, and severe pollution, contributed to the decline in EWP. The 2008 global financial crisis had a significant impact on energy-intensive and pollution-heavy industries. As a result, resource input and environmental pollution decreased, which in turn contributed to a temporary improvement in EWP in 2009.\u003c/p\u003e\n \u003cp\u003e(2) Sustained improvement phase (2014\u0026ndash;2022): Beginning in 2014, EWP increased markedly, maintaining an overall upward trajectory despite intermittent fluctuations and ultimately reaching its peak in 2022. Following the elevation of Ecological Civilization Construction to a national-level strategy, green development and ecological protection accelerated. Guided by policies such as \u0026ldquo;prioritizing ecological protection over large-scale development,\u0026rdquo; the MRYRUA increased efforts in environmental protection and made visible progress in green and low-carbon development. The launch of the New-type Urbanization strategy in 2014 also introduced a series of policies focused on improving public well-being. Driven by both national and local initiatives, EWP steadily improved. In 2020, the COVID-19 pandemic and the associated global economic slowdown led to reductions in resource consumption and pollution, which in turn contributed to a temporary spike in EWP.\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e\n \u003ch2\u003e3.1.2 Spatial patterns and regional disparities\u003c/h2\u003e\n \u003cp\u003eSpatial pattern analysis is crucial for understanding regional ecological protection, well-being enhancement and governance coordination. To analyze the spatial patterns and regional disparities of EWP, the study period was divided into four approximately balanced intervals based on data availability and policy phases: 2005\u0026ndash;2008, 2009\u0026ndash;2012, 2013\u0026ndash;2016, and 2017\u0026ndash;2022. This approach helps to minimize the influence of anomalies or irregularities in individual years. The average EWP for each period represents general performance level during that timeframe. Figure \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e illustrates the spatiotemporal evolution of EWP across MRYRUA from 2005 to 2022.\u003c/p\u003e\n \u003cp\u003eFrom a spatial perspective, EWP in the MRYRUA characterized by higher values in central areas and lower values in the periphery, indicating a marked spatial polarization. In the early period, high EWP values were relatively sparse but exhibited spatial clustering in and around core cities like Wuhan and Nanchang. During 2009\u0026ndash;2012, a sharp rise in EWP occurred in parts of PLCG. From 2013\u0026ndash;2016, core cities like Wuhan and Changsha began to show strong improvements, while border cities such as Xianning, Huangshi, Jiujiang, and Yueyang lagged behind, forming a \u0026ldquo;central collapse\u0026rdquo; pattern. By 2017\u0026ndash;2022, a more polycentric structure emerged, with multiple high-EWP zones expanding outward. Nevertheless, a persistent \u0026ldquo;high core\u0026ndash;low periphery\u0026rdquo; pattern remains, underscoring uneven development within the agglomeration and the need for coordinated regional strategies.\u003c/p\u003e\n \u003cp\u003eThe Theil index and its decomposition were used to assess intra-regional disparities in EWP across the MRYRUA. As shown in Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e, the overall Theil index declined by 75.13% (from 0.197 to 0.049) over the study period\u0026mdash;a 75.24% reduction\u0026mdash;reflecting a substantial decrease in regional inequality. The decomposition results indicate that over 80% of the total inequality originated within sub-agglomerations rather than between them. Among the three, the WMA exhibited the highest Theil index, followed by the PLCG and CZXCG. The WMA also contributed most to overall inequality, consistently above 45% and reaching around 60% in recent years. The PLCG ranked second, though its share declined over time. The CZXCG contributed the least, with a slight downward trend. These findings suggest that intra-regional disparities within the WMA are both the most pronounced and the primary source of EWP inequality across the region.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003e3.2 Typological features of EWP\u003c/h2\u003e\n \u003cp\u003eA clustering analysis was conducted to examine city-level patterns of ecological consumption, human well-being, and EWP across the MRYRUA. Average data from 2018 to 2022 were used to reflect recent development trends and to reduce short-term fluctuations. Using the Silhouette Score method, four clusters were identified as optimal, achieving the highest score (0.212), indicating the best balance between cohesion and separation. Accordingly, the cities were grouped into four distinct clusters (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eC1-type cities demonstrate moderate EWP despite below-average ecological consumption and human well-being. Representing a \u0026ldquo;low input, low output, and moderate efficiency\u0026rdquo; pattern, this cluster includes 15 cities. As shown in Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e, these cities are characterized by relatively low levels of land, energy, and waste inputs, as well as limited income and green space provision. Despite such minimal inputs, they achieve moderate EWP scores, indicating a certain degree of ecological efficiency. However, the persistently low well-being indicators suggest that ecological resources are not fully leveraged to support broader social development.\u003c/p\u003e\n \u003cp\u003eC2-type cities show low EWP, marked by high ecological consumption but limited well-being gains\u0026mdash;a \u0026ldquo;high input, low output, and low efficiency\u0026rdquo; pattern. As shown in Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e, these cities demonstrate excessive resource consumption\u0026mdash;particularly in water use, land use, and waste emissions\u0026mdash;yet fail to achieve corresponding improvements in well-being. Although they perform well in certain aspects such as urban\u0026ndash;rural income equality and green space provision, most social development indicators remain weak, especially in education and healthcare. These four small cities likely suffer from inefficient resource use or dependence on resource-intensive industries, underscoring the need to optimize ecological inputs to enhance well-being outcomes.\u003c/p\u003e\n \u003cp\u003eC3-type cities\u0026mdash;Changsha, Wuhan, and Nanchang\u0026mdash;achieve the highest EWP, following a \u0026ldquo;moderate input, high output, and high efficiency\u0026rdquo; pattern. Despite variability in ecological consumption (e.g., high land consumption but low exhaust water and gas emissions), they lead in income, education, healthcare, and green space. These cities achieve the highest level of well-being with only moderate ecological inputs, making them the most efficient in terms of EWP. This underscores the comparative advantages of large cities in promoting industrial upgrading, enhancing resource efficiency, and ensuring effective public service delivery.\u003c/p\u003e\n \u003cp\u003eC4-type cities exhibit low EWP with a \u0026ldquo;high input, moderate output, and low efficiency\u0026rdquo; pattern. This group of nine cities shows above-average levels in most ecological and well-being indicators, though water use and income equality lag behind. High energy use and emissions suggest dependence on resource- and pollution-intensive industries. While better than C2-type cities in income, education, and healthcare, their EWP remains low due to inefficient resource use. These cities are more stable than C2-type cities but require more sustainable development strategies to improve efficiency and reduce environmental pressure.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003ch2\u003e3.3 Influencing factors of EWP\u003c/h2\u003e\n \u003cdiv id=\"Sec15\" class=\"Section3\"\u003e\n \u003ch2\u003e3.3.1 Influencing factors and model performance\u003c/h2\u003e\n \u003cp\u003eThe transformation of ecological consumption into human well-being determines EWP, yet this process is complex and influenced by multiple external factors. Building on prior studies (Li et al., \u003cspan class=\"CitationRef\"\u003e2018\u003c/span\u003e; Zhu et al., \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e; Dong et al., \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e), this study identifies a set of key variables to explore the drivers of EWP. Seven key variables are selected to represent key influencing factors: technological innovation, environmental regulation, industrial structure, population density, urbanization, foreign direct investment, and local government expenditure. An interpretable machine learning approach is employed to examine the driving effects of these factors on EWP. Detailed variable information is provided in Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eDefinition and descriptive statistics of variables\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVariable type\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVariable\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eExplanation\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eObs.\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMean\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSD\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMin\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMax\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDependent variable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEWP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEcological well-being performance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e558\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.19\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndependent variable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTechnological innovation (TI)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNumber of invention patent applications (Patents per 10,000 people)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e558\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e7.92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e9. 63\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e72.17\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEnvironmental regulation (ER)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eProportion of environmental keywords in local government work reports (%)(Bao \u0026amp; Liu, \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e558\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.048\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.02\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndustrial structure (IS)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eShare of secondary industry in GDP (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e558\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e48.38\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e7.55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e28.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e78.84\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePopulation density (PD)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePopulation density of the built-up urban area (10,000 people per km\u0026sup2;)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e558\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.54\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e7.16\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eUrbanization (UR)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eUrbanization rate of permanent resident population (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e558\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e53.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e11.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e22.62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e84.70\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eForeign direct investment (FDI)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eForeign direct investment as a percentage of GDP (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e558\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2.11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.58\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.00\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e8.03\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGovernment expenditure (GE)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGovernment expenditure as a percentage of GDP (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e558\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e15.29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.75\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e31.79\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003ctfoot\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"8\"\u003eNote: Obs. = Number of observations; SD\u0026thinsp;=\u0026thinsp;Standard deviation.\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tfoot\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003eTable\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e presents the regression performance of the four machine learning models. XGBoost demonstrates the best performance, exhibiting the lowest error metrics and the highest R\u0026sup2;, which indicates superior model fit. RF ranks second, performing comparably to XGBoost in terms of MSE and RMSE, but with slightly higher MAE and MAPE. AdaBoost also shows competitive performance, though slightly behind the top two. MLR shows the weakest performance, suggesting that linear models may not adequately capture the underlying data patterns.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab3\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eComparative performance of four machine learning models\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eModel\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMSE\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eRMSE\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMAE\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMAPE\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eR\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eXGBoost\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.145\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.094\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.176\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.701\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.021\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.146\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.108\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.209\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.697\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAdaBoost\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.108\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.205\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.679\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMLR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.034\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.185\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.145\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.286\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.511\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec16\" class=\"Section3\"\u003e\n \u003ch2\u003e3.3.2 Results of interpretable machine learning\u003c/h2\u003e\n \u003cp\u003eAs shown in Fig. \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e, IS and GE consistently rank as the most influential drivers of EWP across all models, whereas UR, ER, PD, and TI fall into the mid-range of importance, and FDI shows the lowest contribution. This suggests that across different models, the overall ranking of feature importance remains relatively stable, yet the mechanisms through which each factor shapes EWP differ considerably. Compared to the linear model, the three ensemble models reveal more diverse and intricate SHAP value patterns. To further investigate these roles, Fig. \u003cspan class=\"InternalRef\"\u003e8\u003c/span\u003e presents the SHAP\u0026ndash;GAM single-feature dependence plots, which reveal the nonlinear relationships and critical thresholds of each driver. The relationships between the seven indicators and EWP are all characterized by significant nonlinearity and threshold effects. With the exception of FDI, all variables show significant fits (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), with \u003cem\u003eR\u003c/em\u003e\u0026sup2; values ranging from 0.31 to 0.85, indicating good explanatory power.\u003c/p\u003e\n \u003cp\u003eTechnological innovation (TI) has a positive effect on EWP, particularly at higher levels, suggesting the presence of nonlinearities and threshold effects. Figure \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e indicates that the effect is not uniformly strong across all cities. Figure \u003cspan class=\"InternalRef\"\u003e8\u003c/span\u003e(f) reveals a nonlinear pattern: at lower levels, TI has little discernible effect, but once a certain scale of innovation activity is reached, its contribution increases rapidly. However, when technological innovation becomes excessive, the marginal benefits tend to diminish, indicating that the positive impact of innovation on EWP gradually weakens at higher levels. This suggests that beyond a certain point, further technological input may lead to efficiency saturation, higher costs, or rebound effects, thereby reducing the incremental gains in EWP.\u003c/p\u003e\n \u003cp\u003eEnvironmental regulation (ER) shows a positive effect on EWP. As indicated in Fig. \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003e, its overall contribution is moderate and varies considerably across cities. Figure \u003cspan class=\"InternalRef\"\u003e8\u003c/span\u003e(d) highlights that when regulation is weak, its impact is often negative or negligible, as insufficient enforcement fails to offset the associated economic costs. In contrast, stronger regulation generates clearer positive effects, suggesting that only beyond a certain intensity does ER promote ecological improvements and welfare gains, in line with the Porter hypothesis.\u003c/p\u003e\n \u003cp\u003eIndustrial structure (IS) has a significant negative impact on EWP, indicating that upgrading the industrial mix contributes to better performance. Across all models, higher IS values are predominantly linked to negative SHAP values, whereas lower IS values are associated with positive ones. This pattern indicates that a larger share of secondary industry within the economic structure tends to exert a stronger negative effect on EWP. A heavy reliance on secondary industry is generally linked to elevated pollution levels and intensive energy use, leading to greater environmental degradation. In the past decade, the MRYRUA has undergone industrial upgrading, with a shift away from traditional manufacturing toward cleaner, higher-value sectors. This has reduced ecological pressure and enhanced well-being, contributing to improved EWP.\u003c/p\u003e\n \u003cp\u003ePopulation density (PD) generally has a positive effect on EWP. High PD values are broadly associated with positive SHAP values, reflecting a stronger contribution to EWP. This implies that higher urban population density can lead to economies of scale and agglomeration effects, which in turn enhance resource efficiency. However, the curve flattens at higher PD levels, suggesting that excessive agglomeration may no longer yield proportional benefits, possibly due to congestion, environmental stress, or diminishing marginal returns.\u003c/p\u003e\n \u003cp\u003eUrbanization (UR) has a negative impact on EWP. In nearly all models, UR shows a negative relationship with EWP, with only a few cities displaying a positive contribution. This suggests that the quality of urbanization in the study area has remained relatively low over the past decade. Rapid urban expansion has often occurred alongside intensive industrialization, frequently driven by low-value-added, resource-intensive, and polluting industries. Moreover, Moreover, improvements in public services such as education and healthcare have lagged behind urban growth, resulting in limited gains in per capita well-being.\u003c/p\u003e\n \u003cp\u003eGovernment expenditure (GE) has a significant positive impact on EWP. The SHAP values of GE show a symmetric distribution, with higher feature values corresponding to higher SHAP values, underscoring its stable positive contribution. This reflects the important role of public fiscal intervention in advancing environmental quality and well-being. In particular, government spending on environmental protection, social well-being, and green industry support has substantially improved both ecological conditions and residents\u0026rsquo; quality of life in the study area.\u003c/p\u003e\n \u003cp\u003eForeign direct investment (FDI) has a limited impact on EWP. In all models, FDI exhibits a much lower average absolute SHAP value compared to other variables, suggesting limited explanatory power. However, this does not imply that FDI has no impact whatsoever. Existing studies suggest that the influence of FDI may be moderated by competing mechanisms, including the \u0026ldquo;pollution haven\u0026rdquo; and \u0026ldquo;pollution halo\u0026rdquo; hypotheses, as well as income effects (Xiong et al., \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). In the AdaBoost and RF models, some cities with low FDI levels show positive SHAP values. This suggests that, in certain cases, lower levels of foreign investment may actually be more conducive to improving EWP.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e"},{"header":"4 Discussion","content":"\u003cp\u003eThe relationship between the environment and human well-being is a critical global issue (Folke et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Rockstr\u0026ouml;m et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Enhancing EWP is essential for the sustainability of urban agglomerations. However, previous studies have largely overlooked both the typological classification of EWP and the nonlinear effects of its drivers. There is limited understanding of how cities within the same urban agglomeration differ in their EWP transformation trajectories, and what context-specific drivers shape these divergent pathways. The present study addresses these gaps by developing an integrated framework using Orange, a visual programming platform, to examine the evolution, types, and driving mechanisms of EWP in the MRYRUA. This study area is characterized by rapid development, high spatial density and diversity, and pronounced ecological fragility, which together heighten governance complexity and the challenge of cross-regional policy coordination. Our framework and indicator system explicitly aligns with the SDGs while being tailored to China\u0026rsquo;s regional development realities, thereby enhancing both international comparability and local applicability. By applying spatiotemporal analysis, K-means clustering, and interpretable machine learning, we show that EWP patterns vary significantly across cities, driven by nonlinear and heterogeneous effects of multiple factors. The key findings and insights are as follows:\u003c/p\u003e\u003cp\u003e໿The evolution of EWP reflects broader shifts in socioeconomic development models and the transformation of human\u0026ndash;environment relationships. In the MRYRUA, the trajectory of EWP indicates that the region experienced significant population, resource, and environmental pressures during its early stages of development. This is consistent with previous research findings (Chen et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, with the implementation of national strategies such as Ecological Civilization Construction, the relationship between nature and human well-being has gradually shifted from imbalance to coordination. Such a transformation underscores China\u0026rsquo;s progress in aligning ecological policy and broader development initiatives with the overarching agenda of the SDGs (Cao et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). This transition also parallels the developmental trajectories of mature urban agglomerations and metropolitan regions worldwide, which typically follow a gradual upward trend toward sustainability (Fang et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Nevertheless, substantial disparities in EWP persist among cities within the MRYRUA. These findings are consistent with previous studies (Zhang et al., \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), underscoring the urgent need to enhance inter-city coordination and foster the diffusion of development benefits across the region. Additionally, beyond this consistency, our study differs from prior work by integrating a multi-model, interpretable machine-learning workflow, enabling more precise estimates and clearer identification of nonlinear thresholds and context-specific mechanisms in spatially integrated, ecologically fragile, high-density, and diverse metropolitan regions.\u003c/p\u003e\u003cp\u003eThis study adopts a \u0026ldquo;safe and just space\u0026rdquo; perspective to classify the ecology\u0026ndash;well-being transformation process, revealing significant variation among cities in balancing ecological consumption and human well-being. At present, only a few cities have achieved an optimal state. Most cities exhibit low ecological consumption, yet their well-being levels remain below the average and are potentially nearing the social foundation of justice (Rockstr\u0026ouml;m et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). For these cities, priority should be given to improving public services and social protection systems to ensure that basic well-being needs are met. Conversely, cities with high per capita ecological consumption should prioritize green industrial transformation, reduce resource use, and improve efficiency. In doing so, this study also emphasizes the bottom-line thinking of ecological safety during rapid development, as well as the necessity of regional balance and context-specific approaches, thereby laying a foundation for more rational regional policy design.\u003c/p\u003e\u003cp\u003eInterpretable machine learning results reveal that drivers of EWP exert nonlinear and heterogeneous effects, depending on city context. Although technological innovation is generally contributes to improved EWP (Yang et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), the analysis reveals that only high levels of technological capability lead to substantial positive outcomes. Given the high costs associated with foundational innovation, it is recommended that technologically advanced cities support less developed counterparts, including leveraging digital means to substitute or complement deficiencies in other proximity dimensions(Cheng et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Moreover, greater emphasis should be placed on the well-being-enhancing effects of technological innovation, ensuring that modern technologies are effectively implemented in practice to improve residents\u0026rsquo; quality of life. The impact of environmental regulation also appears to be both nonlinear and context-specific, highlighting the importance of balancing regulatory costs against the ecological benefits they yield. In the long term, the overall quality of urbanization in the study area remains relatively low. Rapid urban growth has not been matched by corresponding improvements in income, healthcare, or education (Sun et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), and it has been accompanied by rising social inequality (Pandey et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Therefore, further efforts are needed to promote the social integration of rural migrants and to enhance human well-being across the region.\u003c/p\u003e\u003cp\u003eCompared with existing studies, this research offers a novel framework for understanding the multidimensional characteristics of regional EWP. Leveraging the capabilities of the Orange data mining platform, the framework integrates data processing, spatiotemporal visualization, clustering analysis, and multi-model machine learning. By integrating multiple analytical perspectives, the framework captures the spatial, temporal, and structural complexity of how ecological resources are transformed into human well-being. It complements traditional efficiency-based approaches by introducing a typological perspective on EWP, adding a classification dimension to conventional evaluations and going beyond pure efficiency scores to identify multidimensional patterns in input\u0026ndash;output relations. It also deepens the understanding of how key drivers interact in nonlinear and context-specific ways. This approach advances conceptual debates on sustainability transitions and urban-regional inequality, complementing recent efforts to apply machine learning to sustainable development contexts (Sha et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In addition, it offers practical guidance for sustainability policy design in rapidly urbanizing agglomerations, providing a valuable reference for other developing countries to better balance development and protection, coordinate regional collaboration, and pursue sustainable growth.\u003c/p\u003e\u003cp\u003eDue to limitations in data availability, this study includes only objective well-being indicators in the measurement of human well-being. As such, the analysis may not fully reflect the experiential or perceptual dimensions of residents\u0026rsquo; well-being. Furthermore, the classification of EWP types within the \u0026ldquo;safe and just space\u0026rdquo; perspective remains preliminary. Excessive absolute levels of ecological consumption can lead to severe environmental degradation, while insufficient absolute levels of well-being may undermine the equitable right to human development. Therefore, future research should explore potential pathways for enhancing EWP within an absolute spatial boundary that simultaneously ensures ecological safety and social justice.\u003c/p\u003e"},{"header":"5 Conclusions and policy insights","content":"\u003cp\u003eThis study developed a comprehensive analytical framework for ecological well-being performance (EWP) based on the Orange data mining platform. Applying this framework to cities within the middle reaches of the Yangtze River urban agglomeration (MRYRUA) during 2005\u0026ndash;2022, we identified spatiotemporal evolution and typological patterns of EWP and uncovered its complex driving mechanisms through interpretable machine learning models. The main conclusions obtained by the study are as follows:\u003c/p\u003e\u003cp\u003e(1) EWP in the MRYRUA has generally improved, with overall levels increasing by more than 37% (from 0.558 in 2005 to 0.767 in 2022), shifting from early fluctuations to steady growth under the influence of national strategies. However, pronounced spatial disparities persist, characterized by a \u0026ldquo;high core\u0026ndash;low periphery\u0026rdquo; pattern and a \u0026ldquo;central collapse\u0026rdquo; at sub-agglomeration junctions, particularly within the Wuhan Metropolitan Area.\u003c/p\u003e\u003cp\u003e(2) The typological analysis reveals clear differentiation in the ecology\u0026ndash;well-being transformation pathways among cities in the MRYRUA. Only a few core cities (e.g., Wuhan, Changsha, Nanchang) achieved a coordinated status between ecological consumption and human well-being, maintaining both high efficiency and acceptable absolute levels, while most cities deviate from this balance\u0026mdash;either through excessive ecological inputs or insufficient well-being gains\u0026mdash;implying potential sustainability risks.\u003c/p\u003e\u003cp\u003e(3) The influencing factors of EWP show significant nonlinear and heterogeneous effects. Industrial structure and fiscal expenditure exert the strongest influences, while technological innovation, environmental regulation, population density, and urbanization show moderate but context-specific roles. FDI has only a limited and variable impact.\u003c/p\u003e\u003cp\u003eThe findings in this study provide valuable policy insights for sustainable development in the MRYRUA:\u003c/p\u003e\u003cp\u003e(1) Implement differentiated and collaborative development policies by city type. Low-efficiency cities should pursue industrial upgrading and resource optimization, while high-efficiency but low-well-being cities should improve public services. Core cities can foster regional collaboration through green technology sharing and management experience, advancing coordinated development across the agglomeration.\u003c/p\u003e\u003cp\u003e(2) Optimize industrial and fiscal structures to enhance efficiency and reduce disparities. Shifting from secondary to service- and innovation-based industries, and directing fiscal spending toward environmental protection and social programs, can jointly improve ecological efficiency and human well-being.\u003c/p\u003e\u003cp\u003e(3) Focus on the quality rather than the speed of urbanization. Rapid urban expansion has not always brought more equitable well-being. Policies should emphasize inclusive and compact growth, strengthen education and healthcare, and promote social integration for sustainable well-being gains.\u003c/p\u003e\u003cp\u003e(4) Advance innovation and environmental regulation through differentiated approaches. Innovation policies should expand technological capacity, support cross-city diffusion, and apply digital tools to improve ecological efficiency and well-being. Governments should calibrate regulation strength to local capacity, ensuring that compliance costs do not outweigh ecological benefits.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eAuthor ContributionsR.Z. drafted the initial manuscript, conducted data analysis, designed the overall framework, prepared the figures, and carried out the survey analysis. Z.Z. provided comprehensive guidance, contributed to framework design, revised the manuscript and figures, and secured project support. Y.Z. contributed to the conceptual design in the early stage, offered suggestions on the initial draft, and provided data. All authors reviewed and approved the final manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThis study was financially supported by the National Natural Science Foundation of China (Nos. 72474082, 72174071, 42471238), the Teaching Reform Research Project of Hubei Province (No. 2024096) and the Fundamental Research Funds for the Central Universities in China (No. CCNU25ZZ275).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eBao, R., \u0026amp; Liu, T. (2022). How does government attention matter in air pollution control? Evidence from government annual reports. \u003cem\u003eResources, Conservation and Recycling\u003c/em\u003e, \u003cem\u003e185\u003c/em\u003e, 106435. https://doi.org/10.1016/j.resconrec.2022.106435.\u003c/li\u003e\n \u003cli\u003eBehjat, A., \u0026amp; Tarazkar, M.H. (2021). Investigating the factors affecting the ecological well-being performance in Iran from 1994 to 2014. \u003cem\u003eEnvironment, Development and Sustainability\u003c/em\u003e, \u003cem\u003e23\u003c/em\u003e (9), 13871\u0026ndash;13889. https://doi.org/10.1007/s10668-021-01243-0.\u003c/li\u003e\n \u003cli\u003eBian, J., Zhang, Y., Shuai, C., Shen, L., Ren, H., \u0026amp; Wang, Y. (2020). Have cities effectively improved ecological well-being performance? Empirical analysis of 278 Chinese cities. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e, \u003cem\u003e245\u003c/em\u003e, 118913. https://doi.org/10.1016/j.jclepro.2019.118913.\u003c/li\u003e\n \u003cli\u003eCao, M., Chen, M., Zhang, J., Pradhan, P., Guo, H., Fu, B., Li, Y., Bai, Y., Chang, L., Chen, Y., Sun, Z., Xu, Z., Zhu, R., Meadows, M.E., \u0026amp; L\u0026uuml;, G. (2023). Spatio-temporal changes in the causal interactions among Sustainable Development Goals in China. \u003cem\u003eHumanities and Social Sciences Communications\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e, 450. https://doi.org/10.1057/s41599-023-01952-z.\u003c/li\u003e\n \u003cli\u003eChen, H., Shi, Y., \u0026amp; He, H. (2024). A sustainable urban framework based on natural capital and ecosystem services accounting. \u003cem\u003eApplied Spatial Analysis and Policy\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e, 23. https://doi.org/10.1007/s12061-024-09631-5.\u003c/li\u003e\n \u003cli\u003eChen, W., Gu, T., \u0026amp; Zeng, J. (2022). Urbanisation and ecosystem health in the Middle Reaches of the Yangtze River urban agglomerations, China: A U-curve relationship. \u003cem\u003eJournal of Environmental Management\u003c/em\u003e, \u003cem\u003e318\u003c/em\u003e, 115565. https://doi.org/10.1016/j.jenvman.2022.115565\u003c/li\u003e\n \u003cli\u003eCheng, D., Hu, S., \u0026amp; Zeng, G. (2025). Spatio-temporal evolution of green technology innovation networks and its proximity mechanism in the context of digital transformation: The case of Yangtze river delta in China. \u003cem\u003eApplied Spatial Analysis and Policy\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e, 106. https://doi.org/10.1007/s12061-025-09685-z.\u003c/li\u003e\n \u003cli\u003eCommon, M. (2007). Measuring national economic performance without using prices. \u003cem\u003eEcological Economics\u003c/em\u003e, \u003cem\u003e64\u003c/em\u003e (1), 92\u0026ndash;102. https://doi.org/10.1016/j.ecolecon.2007.07.017.\u003c/li\u003e\n \u003cli\u003eCostanza, R., Daly, L., Fioramonti, L., Giovannini, E., Kubiszewski, I., Mortensen, L.F., Pickett, K.E., Ragnarsdottir, K.V., De Vogli, R., \u0026amp; Wilkinson, R. (2016). Modelling and measuring sustainable wellbeing in connection with the UN Sustainable Development Goals. \u003cem\u003eEcological Economics\u003c/em\u003e, \u003cem\u003e130\u003c/em\u003e, 350\u0026ndash;355. https://doi.org/10.1016/j.ecolecon.2016.07.009.\u003c/li\u003e\n \u003cli\u003eDaly, H.E. (1974). The world dynamics of economic growth: The economics of the steady state. \u003cem\u003eAmerican Economic Review\u003c/em\u003e, \u003cem\u003e64\u003c/em\u003e (2), 15\u0026ndash;23.\u003c/li\u003e\n \u003cli\u003eDaly, H.E. (2005). Economics in a full world. \u003cem\u003eIEEE Engineering Management Review\u003c/em\u003e, \u003cem\u003e33\u003c/em\u003e (4), 21. https://doi.org/10.1109/EMR.2005.27010.\u003c/li\u003e\n \u003cli\u003eDeng, Y.J., Yang, X., Ma, Q.W., \u0026amp; Wang, K.D. (2021). Regional disparity and convergence of China\u0026rsquo;s ecological welfare performance level. \u003cem\u003eChina Population, Resources and Environment\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e (4), 132\u0026ndash;143.\u003c/li\u003e\n \u003cli\u003eDong, Y., Sun, Y., \u0026amp; Ding, J. (2024). Decomposition of ecological welfare performance drivers in China. \u003cem\u003eActa Geographica Sinica\u003c/em\u003e, \u003cem\u003e79\u003c/em\u003e (5), 1337\u0026ndash;1354. https://doi.org/10.11821/dlxb202405015.\u003c/li\u003e\n \u003cli\u003eFang, C., Wang, Z., \u0026amp; Ma, H. (2018). The theoretical cognition of the development law of China\u0026rsquo;s urban agglomeration and academic contribution. \u003cem\u003eActa Geographica Sinica\u003c/em\u003e, \u003cem\u003e73\u003c/em\u003e (4), 651\u0026ndash;665. https://doi.org/10.11821/dlxb201804005.\u003c/li\u003e\n \u003cli\u003eFang, S.J., \u0026amp; Xiao, Q. (2019). Research on regional ecological well-being performance and spatial effect in China. \u003cem\u003eChina Population, Resources and Environment\u003c/em\u003e, \u003cem\u003e29\u003c/em\u003e (3), 1\u0026ndash;10.\u003c/li\u003e\n \u003cli\u003eFanning, A.L., \u0026amp; Raworth, K. (2025). Doughnut of social and planetary boundaries monitors a world out of balance. \u003cem\u003eNature\u003c/em\u003e, \u003cem\u003e646\u003c/em\u003e (8083), 47\u0026ndash;56. https://doi.org/10.1038/s41586-025-09385-1.\u003c/li\u003e\n \u003cli\u003eFolke, C., Polasky, S., Rockstrom, J., Galaz, V., Westley, F., Lamont, M., Scheffer, M., Osterblom, H., Carpenter, S.R., Chapin, F.S., 3rd, Seto, K.C., Weber, E.U., Crona, B.I., Daily, G.C., Dasgupta, P., Gaffney, O., Gordon, L.J., Hoff, H., Levin, S.A., Lubchenco, J., Steffen, W., \u0026amp; Walker, B.H. (2021). Our future in the Anthropocene biosphere. \u003cem\u003eAmbio\u003c/em\u003e, \u003cem\u003e50\u003c/em\u003e (4), 834\u0026ndash;869. https://doi.org/10.1007/s13280-021-01544-8.\u003c/li\u003e\n \u003cli\u003eFriedman, J.H. (2001). Greedy function approximation: A gradient boosting machine. \u003cem\u003eThe Annals of Statistics\u003c/em\u003e, \u003cem\u003e29\u003c/em\u003e (5), 1189\u0026ndash;1232. https://doi.org/10.1214/aos/1013203451.\u003c/li\u003e\n \u003cli\u003eFu, Z., Yang, X., Ma, Y., Sun, Y., \u0026amp; Wang, T. (2025). Integrating explainable AI and causal inference to unveil regional air quality drivers in China. \u003cem\u003eJournal of Environmental Management\u003c/em\u003e, \u003cem\u003e390\u003c/em\u003e, 126270. https://doi.org/10.1016/j.jenvman.2025.126270.\u003c/li\u003e\n \u003cli\u003eGao, Y., Zhang, N., Ma, Q., \u0026amp; Li, J. (2024). How is human well-being related to ecosystem services at town and village scales? A case study from the Yangtze River Delta, China. \u003cem\u003eLandscape Ecology\u003c/em\u003e, \u003cem\u003e39\u003c/em\u003e (7), 126. https://doi.org/10.1007/s10980-024-01925-w.\u003c/li\u003e\n \u003cli\u003eGodec, P., Pancur, M., Ilenic, N., Copar, A., Strazar, M., Erjavec, A., Pretnar, A., Demsar, J., Staric, A., Toplak, M., Zagar, L., Hartman, J., Wang, H., Bellazzi, R., Petrovic, U., Garagna, S., Zuccotti, M., Park, D., Shaulsky, G., \u0026amp; Zupan, B. (2019). Democratized image analytics by visual programming through integration of deep models and small-scale machine learning. \u003cem\u003eNature Communications\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e, 4551. https://doi.org/10.1038/s41467-019-12397-x.\u003c/li\u003e\n \u003cli\u003eGuo, Q., \u0026amp; Qian, H. (2021). Negative human capital externalities in well-being: evidence from Chinese cities. \u003cem\u003eRegional Studies\u003c/em\u003e, \u003cem\u003e55\u003c/em\u003e (6), 1046\u0026ndash;1058. https://doi.org/10.1080/00343404.2020.1861237.\u003c/li\u003e\n \u003cli\u003eHan, H., Gu, R., \u0026amp; Yang, Y. (2025). Impacts of low-carbon city pilot policy on ecological well-being performance across Chinese cities: A spatial difference-in-difference analysis. \u003cem\u003eSustainable Cities and Society\u003c/em\u003e, \u003cem\u003e118\u003c/em\u003e, 105864. https://doi.org/10.1016/j.scs.2024.105864.\u003c/li\u003e\n \u003cli\u003eHu, M., Pang, Y., Jin, T., \u0026amp; Li, Z. (2021). Spatio-temporal evolution of EIWB and influencing factors: An empirical study from the Yangtze river delta. \u003cem\u003eApplied Spatial Analysis and Policy\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e, 999\u0026ndash;1024. https://doi.org/10.1007/s12061-021-09389-0.\u003c/li\u003e\n \u003cli\u003eKnight, K.W. (2014). Temporal variation in the relationship between environmental demands and well-being: A panel analysis of developed and less-developed countries. \u003cem\u003ePopulation and Environment\u003c/em\u003e, \u003cem\u003e36\u003c/em\u003e (1), 32\u0026ndash;47. https://doi.org/10.1007/s11111-013-0200-1.\u003c/li\u003e\n \u003cli\u003eKnight, K.W., \u0026amp; Rosa, E.A. (2011). The environmental efficiency of well-being: A cross-national analysis. \u003cem\u003eSocial Science Research\u003c/em\u003e, \u003cem\u003e40\u003c/em\u003e (3), 931\u0026ndash;949. https://doi.org/10.1016/j.ssresearch.2010.11.002.\u003c/li\u003e\n \u003cli\u003eLi, C.Y., Zhang, S.Q., Zhang, W., \u0026amp; Liao, X.C. (2018). Measurement and influencing factors of inter-provincial ecological well-being performance in China. \u003cem\u003eScientia Geographica Sinica\u003c/em\u003e, \u003cem\u003e39\u003c/em\u003e (12), 1875\u0026ndash;1883. https://doi.org/10.13249/j.cnki.sgs.2019.12.005.\u003c/li\u003e\n \u003cli\u003eLi, Z. (2022). Extracting spatial effects from machine learning model using local interpretation method: An example of SHAP and XGBoost. \u003cem\u003eComputers, Environment and Urban Systems\u003c/em\u003e, \u003cem\u003e96\u003c/em\u003e, 101845. https://doi.org/10.1016/j.compenvurbsys.2022.101845.\u003c/li\u003e\n \u003cli\u003eLong, L.J., Wang, X., \u0026amp; Guo, B. (2017). Evaluation of urban ecological well-being performance based on revised DEA model: A case study of 35 major cities in China. \u003cem\u003eJournal of Natural Resources\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e (4), 595\u0026ndash;605. https://doi.org/10.11849/zrzyxb.20160515.\u003c/li\u003e\n \u003cli\u003eLundberg, S.M., Erion, G., Chen, H., DeGrave, A., Prutkin, J.M., Nair, B., Katz, R., Himmelfarb, J., Bansal, N., \u0026amp; Lee, S.I. (2020). From local explanations to global understanding with explainable ai for trees. \u003cem\u003eNature Machine Intelligence\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e, 56\u0026ndash;67. https://doi.org/10.1038/s42256-019-0138-9.\u003c/li\u003e\n \u003cli\u003eLundberg, S.M., \u0026amp; Lee, S.-I. (2017). A unified approach to interpreting model predictions. In \u003cem\u003eProceedings of the 31st International Conference on Neural Information Processing Systems\u003c/em\u003e, edited by S. M. Lundberg and S.-I. Lee. Long Beach, California, USA: Curran Associates Inc.\u003c/li\u003e\n \u003cli\u003eMa, H., Wang, K., \u0026amp; Fang, C. (2025). How does new-type urbanization promote green development efficiency in China? The mediating role of governments, enterprises, and residents. \u003cem\u003eGeography and Sustainability\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e (3), 100241. https://doi.org/10.1016/j.geosus.2024.10.001.\u003c/li\u003e\n \u003cli\u003eMalone, B.P., Minasny, B., Odgers, N.P., \u0026amp; McBratney, A.B. (2014). Using model averaging to combine soil property rasters from legacy soil maps and from point data. \u003cem\u003eGeoderma\u003c/em\u003e, \u003cem\u003e232\u003c/em\u003e, 34\u0026ndash;44. https://doi.org/10.1016/j.geoderma.2014.04.033.\u003c/li\u003e\n \u003cli\u003eMarks, N., Abdallah, S., Simms, A., \u0026amp; Thompson, S. (2006). The (un)Happy Planet Index: An index of human well-being and environmental impact.\u003c/li\u003e\n \u003cli\u003ePandey, B., Brelsford, C., \u0026amp; Seto, K.C. (2022). Infrastructure inequality is a characteristic of urbanization. \u003cem\u003eProceedings of the National Academy of Sciences of the United States of America\u003c/em\u003e, \u003cem\u003e119\u003c/em\u003e (15), e2119890119. https://doi.org/10.1073/pnas.2119890119.\u003c/li\u003e\n \u003cli\u003eRees, W.E. (1992). Ecological footprints and appropriated carrying capacity: What urban economics leaves out. \u003cem\u003eEnvironment and Urbanization\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e (2), 121\u0026ndash;130. https://doi.org/10.1177/095624789200400212.\u003c/li\u003e\n \u003cli\u003eRockstr\u0026ouml;m, J., Gupta, J., Lenton, T.M., Qin, D., Lade, S.J., Abrams, J.F., Jacobson, L., Rocha, J.C., Zimm, C., Bai, X., Bala, G., Bringezu, S., Broadgate, W., Bunn, S.E., DeClerck, F., Ebi, K.L., Gong, P., Gordon, C., Kanie, N., Liverman, D.M., Nakicenovic, N., Obura, D., Ramanathan, V., Verburg, P.H., van Vuuren, D.P., \u0026amp; Winkelmann, R. (2021). Identifying a safe and just corridor for people and the planet. \u003cem\u003eEarth\u0026rsquo;s Future\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e (4), e2020EF001866. https://doi.org/10.1029/2020EF001866.\u003c/li\u003e\n \u003cli\u003eRousseeuw, P.J. (1987). Silhouettes: A graphical aid to the interpretation and validation of cluster analysis. \u003cem\u003eJournal of Computational and Applied Mathematics\u003c/em\u003e, \u003cem\u003e20\u003c/em\u003e, 53\u0026ndash;65. https://doi.org/10.1016/0377-0427(87)90125-7.\u003c/li\u003e\n \u003cli\u003eSha, A., Madhan, S., Karthikeya, M., Megha, R., Dhanush, K.R., Swain, D., Gopakumar, G., \u0026amp; Geetha, M. (2024). Machine learning for sustainable development: Ranking villages for rural development initiatives. \u003cem\u003eApplied Spatial Analysis and Policy\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e, 6. https://doi.org/10.1007/s12061-024-09606-6.\u003c/li\u003e\n \u003cli\u003eShao, S., Ge, L., \u0026amp; Zhu, J. (2024). How to achieve the harmony between humanity and nature: environmental regulation and environmental welfare performance from the perspective of geographical factors. \u003cem\u003eJournal of Management World\u003c/em\u003e (8), 119\u0026ndash;114. https://doi.org/10.19744/j.cnki.11-1235/f.2024.0088.\u003c/li\u003e\n \u003cli\u003e\u0026Scaron;tajdohar, M., \u0026amp; Dem\u0026scaron;ar, J. (2013). Interactive network exploration with Orange. \u003cem\u003eJournal of Statistical Software\u003c/em\u003e, \u003cem\u003e53\u003c/em\u003e (6), 1\u0026ndash;24. https://doi.org/10.18637/jss.v053.i06.\u003c/li\u003e\n \u003cli\u003eSun, L., Chen, J., Li, Q., \u0026amp; Huang, D. (2020). Dramatic uneven urbanization of large cities throughout the world in recent decades. \u003cem\u003eNature Communications\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e (1), 5366. https://doi.org/10.1038/s41467-020-19158-1.\u003c/li\u003e\n \u003cli\u003eTone, K. (2002). A slacks-based measure of super-efficiency in data envelopment analysis. \u003cem\u003eEuropean Journal of Operational Research\u003c/em\u003e, \u003cem\u003e143\u003c/em\u003e (1), 32\u0026ndash;41. https://doi.org/10.1016/S0377-2217(01)00324-1.\u003c/li\u003e\n \u003cli\u003eTrok, J.T., Barnes, E.A., Davenport, F.V., \u0026amp; Diffenbaugh, N.S. (2024). Machine learning-based extreme event attribution. \u003cem\u003eScience Advances\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e (34), eadl3242. https://doi.org/10.1126/sciadv.adl3242.\u003c/li\u003e\n \u003cli\u003eUN Environment Programme. (2019). \u003cem\u003eGlobal environment outlook\u0026ndash;GEO-6: Healthy planet, healthy people\u003c/em\u003e. Cambridge: Cambridge University Press.\u003c/li\u003e\n \u003cli\u003eUN Global Compact. (2015). Sustainable development goals: 17 goals to transform our world. https://www.un.org/en/exhibits/page/sdgs-17-goals-transform-world.\u003c/li\u003e\n \u003cli\u003eUN Sustainable Development Solutions Network. (2024). Sustainable Development Report 2024. https://dashboards.sdgindex.org/.\u003c/li\u003e\n \u003cli\u003eVarian, H.R. (2014). Big data: New tricks for econometrics. \u003cem\u003eJournal of Economic Perspectives\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e (2), 3\u0026ndash;28. https://doi.org/10.1257/jep.28.2.3.\u003c/li\u003e\n \u003cli\u003eWang, R., Wang, Q., \u0026amp; Yao, S. (2021). Evaluation and difference analysis of regional energy efficiency in China under the carbon neutrality targets: Insights from DEA and Theil models. \u003cem\u003eJournal of Environmental Management\u003c/em\u003e, \u003cem\u003e293\u003c/em\u003e, 112958. https://doi.org/10.1016/j.jenvman.2021.112958.\u003c/li\u003e\n \u003cli\u003eWang, S., Huang, L., Xu, X., \u0026amp; Li, J. (2022). Spatio-temporal variations in ecological spaces and their ecological carrying status in China\u0026rsquo;s mega-urban agglomerations. \u003cem\u003eJournal of Geographical Sciences\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e (9), 1683\u0026ndash;1704. https://doi.org/10.1007/s11442-022-2018-1.\u003c/li\u003e\n \u003cli\u003eWang, X., \u0026amp; Gao, C. (2024). Does green finance policy help to improve carbon reduction welfare performance? Evidence from China. \u003cem\u003eEnergy Economics\u003c/em\u003e, \u003cem\u003e132\u003c/em\u003e, 107452. https://doi.org/10.1016/j.eneco.2024.107452.\u003c/li\u003e\n \u003cli\u003eXia, M., \u0026amp; Li, J. (2022). Assessment of ecological well-being performance and its spatial correlation analysis in the Beijing-Tianjin-Hebei urban agglomeration. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e, \u003cem\u003e362\u003c/em\u003e, 132621. https://doi.org/10.1016/j.jclepro.2022.132621.\u003c/li\u003e\n \u003cli\u003eXiong, H., Liu, Y., Kuang, M., \u0026amp; Li, Y. (2023). Nonlinear effects of socio-economic factors on urban haze in China: Evidence from spatial econometric smooth transition autoregressive regression approach. \u003cem\u003eJournal of Environmental Management\u003c/em\u003e, \u003cem\u003e347\u003c/em\u003e, 118991. https://doi.org/10.1016/j.jenvman.2023.118991.\u003c/li\u003e\n \u003cli\u003eYağcı, M. (2022). Educational data mining: prediction of students\u0026rsquo; academic performance using machine learning algorithms. \u003cem\u003eSmart Learning Environments\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e (1), 11. https://doi.org/10.1186/s40561-022-00192-z.\u003c/li\u003e\n \u003cli\u003eYang, L., Ma, Z., \u0026amp; Xu, Y. (2023). How does the digital economy affect ecological well-being performance? Evidence from three major urban agglomerations in China. \u003cem\u003eEcological Indicators\u003c/em\u003e, \u003cem\u003e157\u003c/em\u003e, 111261. https://doi.org/10.1016/j.ecolind.2023.111261.\u003c/li\u003e\n \u003cli\u003eZhang, D., Guo, Z., Niu, X., Gao, Y., Wang, C., \u0026amp; Xu, H. (2022). Comprehensive evaluation on sustainable development based on planetary pressures and ecological well-being performance: A case study on the belt and road regions. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e, \u003cem\u003e376\u003c/em\u003e, 134211. https://doi.org/10.1016/j.jclepro.2022.134211.\u003c/li\u003e\n \u003cli\u003eZhang, Q., Kong, Q., Zhang, M., \u0026amp; Huang, H. (2024). New-type urbanization and ecological well-being performance: A coupling coordination analysis in the middle reaches of the Yangtze River urban agglomerations, China. \u003cem\u003eEcological Indicators\u003c/em\u003e, \u003cem\u003e159\u003c/em\u003e, 111678. https://doi.org/10.1016/j.ecolind.2024.111678.\u003c/li\u003e\n \u003cli\u003eZhang, S., Zhu, D., Shi, Q., \u0026amp; Cheng, M. (2018). Which countries are more ecologically efficient in improving human well-being? An application of the Index of Ecological Well-being Performance. \u003cem\u003eResources, Conservation and Recycling\u003c/em\u003e, \u003cem\u003e129\u003c/em\u003e, 112\u0026ndash;119. https://doi.org/10.1016/j.resconrec.2017.10.015.\u003c/li\u003e\n \u003cli\u003eZhang, X., Fan, H., Liu, F., Lv, T., Sun, L., Li, Z., Shang, W., \u0026amp; Xu, G. (2023). Coupling coordination between the ecological environment and urbanization in the middle reaches of the Yangtze River urban agglomeration. \u003cem\u003eUrban Climate\u003c/em\u003e, \u003cem\u003e52\u003c/em\u003e, 101698. https://doi.org/10.1016/j.uclim.2023.101698.\u003c/li\u003e\n \u003cli\u003eZhang, Y., Mao, Y., Jiao, L., Shuai, C., \u0026amp; Zhang, H. (2021). Eco-efficiency, eco-technology innovation and eco-well-being performance to improve global sustainable development. \u003cem\u003eEnvironmental Impact Assessment Review\u003c/em\u003e, \u003cem\u003e89\u003c/em\u003e, 106580. https://doi.org/10.1016/j.eiar.2021.106580.\u003c/li\u003e\n \u003cli\u003eZhang, Z. (2022). \u003cem\u003eOrange tutorial: Data mining and analysis with \u0026ldquo;building blocks\u0026rdquo; and \u0026ldquo;linking game\u0026rdquo;\u003c/em\u003e. Wu Han: China University of Geosciences Press.\u003c/li\u003e\n \u003cli\u003eZhang, Z., Dou, Y., Liu, X., \u0026amp; Gong, Z. (2023). Multi-hierarchical spatial clustering for characteristic towns in China: An Orange-based framework to integrate GIS and Geodetector. \u003cem\u003eJournal of Geographical Sciences\u003c/em\u003e, \u003cem\u003e33\u003c/em\u003e (3), 618\u0026ndash;638. https://doi.org/10.1007/s11442-023-2098-6.\u003c/li\u003e\n \u003cli\u003eZhang, Z., Zhang, M., Song, X., \u0026amp; Li, Z. (2024). Image-based machine learning and cluster analysis for urban road network: employing Orange for codeless visual programming. \u003cem\u003eGeo-spatial Information Science\u003c/em\u003e, 1\u0026ndash;18. https://doi.org/10.1080/10095020.2024.2377212.\u003c/li\u003e\n \u003cli\u003eZhao, L., Cao, N., Han, Z., \u0026amp; Gao, X. (2022). Evolution characteristics and formation mechanism of spatial correlation network of ecological well-being performance in China. \u003cem\u003eJournal of Natural Resources\u003c/em\u003e, \u003cem\u003e37\u003c/em\u003e (12), 3183\u0026ndash;3200. https://doi.org/10.31497/zrzyxb.20221211.\u003c/li\u003e\n \u003cli\u003eZhu, D., Zhang, S., \u0026amp; Sutton, D.B. (2015). Linking Daly\u0026rsquo;s Proposition to policymaking for sustainable development: Indicators and pathways. \u003cem\u003eJournal of Cleaner Production\u003c/em\u003e, \u003cem\u003e102\u003c/em\u003e, 333\u0026ndash;341. https://doi.org/10.1016/j.jclepro.2015.04.070.\u003c/li\u003e\n \u003cli\u003eZhu, Y., Zhang, R., Gu, J., \u0026amp; Gao, Z. (2022). Spatiotemporal evolution and driving mechanism of ecological well-being performance in the urban agglomeration of the middle reaches of the Yangtze River under the carbon peaking and carbon neutrality goals. \u003cem\u003eProgress in Geography\u003c/em\u003e, \u003cem\u003e41\u003c/em\u003e (12), 2231\u0026ndash;2243. https://doi.org/10.18306/dlkxjz.2022.12.004.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Ecological well-being performance (EWP), Spatiotemporal analysis, Interpretable machine learning, Urban agglomeration (Yangtze River, China), Sustainability policy, Orange","lastPublishedDoi":"10.21203/rs.3.rs-7839392/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7839392/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eFrom the global perspective of human\u0026ndash;nature sustainable development, ecological well-being performance (EWP) captures the efficiency dimension of sustainability by linking ecological consumption with human well-being. For a rapidly developing country, how to balance development and ecology in the spatial dimension is not only related to spatial coordination and sound policy-making, but also determines the sustainability of growth. However, existing studies often overlook spatial disparities, typological differentiation, and nonlinear determinants. To address these gaps, this study developed a machine learning-based multi-method framework on the Orange visual programming platform, integrating spatiotemporal analysis and interpretable machine learning, with a focus on a typical region of China as the study object. This framework was applied to examine EWP in the middle reaches of the Yangtze River urban agglomeration (MRYRUA) during 2005\u0026ndash;2022. The results show an overall improvement but with persistent spatial disparities, notable typological differences among cities, and key drivers dominated by industrial structure and government expenditure. By revealing the disparities behind aggregate progress, this study contributes to more precise estimation results and a clearer understanding of driving mechanisms, while accurately restoring spatial patterns, thereby laying a solid foundation for scientifically formulating multi-scale regional development policies.\u003c/p\u003e","manuscriptTitle":"Ecological well-being performance in a Chinese urban agglomeration: Spatiotemporal analysis and policy insights from an Orange-based machine learning framework","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-30 14:02:54","doi":"10.21203/rs.3.rs-7839392/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"bd3f1964-7f38-4a76-a6a7-323245537a19","owner":[],"postedDate":"October 30th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-16T16:03:09+00:00","versionOfRecord":{"articleIdentity":"rs-7839392","link":"https://doi.org/10.1007/s12061-026-09800-8","journal":{"identity":"applied-spatial-analysis-and-policy","isVorOnly":false,"title":"Applied Spatial Analysis and Policy"},"publishedOn":"2026-02-14 15:57:15","publishedOnDateReadable":"February 14th, 2026"},"versionCreatedAt":"2025-10-30 14:02:54","video":"","vorDoi":"10.1007/s12061-026-09800-8","vorDoiUrl":"https://doi.org/10.1007/s12061-026-09800-8","workflowStages":[]},"version":"v1","identity":"rs-7839392","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7839392","identity":"rs-7839392","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}