The Differential Promoting Effect of Urban–Rural Integration Development on Common Prosperity: A Case Study from Guangdong, China

Abstract

Under the background that urban–rural integrated development continuously deepens and the common prosperity goal continuously advances, systematically identifying the actual results of urban–rural integrated development and its influence mechanism on common prosperity holds important significance for understanding regional development differences and optimizing policy implementation paths. Based on land use data, NTL data, and POI facility data from 2013 to 2025, this study comprehensively employs spatial analysis and deep learning methods to conduct an empirical analysis on the spatiotemporal evolution characteristics and coupling relationship of urban–rural integrated development and common prosperity levels from dimensions including urban–rural spatial form evolution, economic activity intensity, and public service facility diversity. The research results indicate that urban–rural integration significantly promotes urban spatial expansion and the improvement in overall economic activity levels during the study period, but the difference in development magnitude among different regions remains obvious. The common prosperity level generally presents a rising trend, but it highly concentrates in the Pearl River Delta and city–county center areas in space, and the promotion effect of urban–rural integration on common prosperity exhibits obvious characteristics of regional heterogeneity, stages, time lags, and diminishing marginal effects. This study considers that urban–rural integration does not inevitably and synchronously transform into an elevation in common prosperity levels. Combining regional development basis and structural conditions to optimize urban–rural integration development paths by region and by stage and to improve the realization quality of common prosperity possesses important practical reference value.

1. Introduction

1.1. Study Background

Regional development imbalance and the urban–rural gap increasingly become key factors restricting high-quality urban development in China [1]. Urban and rural areas in China have shown significant differences in income level, public service supply, development opportunities, and space utilization efficiency for a long time [2]. Therefore, promoting urban–rural integrated development acts as an important path to break the urban–rural dual structure, promote the free flow of factors, and realize regional coordinated development [3]. Meanwhile, in the high-quality development stage, common prosperity gradually becomes a core goal in the modernization process of China. Its connotation expands from income growth to more comprehensive levels such as public service equalization, life quality improvement, and fairness in development opportunities [4,5]. In this context, urban–rural integration functions as a development carrier to promote the realization of common prosperity.

Urban–rural integrated development provides an important spatial carrier and institutional foundation for narrowing the urban–rural gap and enhancing overall welfare levels by promoting the rational flow and efficient allocation of population, capital, and public resources between urban and rural areas [6]. In this sense, urban–rural integration is both an important implementation mechanism for promoting common prosperity and a concrete manifestation of the goal of common prosperity in the spatial dimension [7]. However, existing research, to some extent, still treats the two separately, lacking systematic analysis of the intrinsic mechanisms through which urban–rural spatial structure and development models affect common prosperity [8]. This separation in research perspective constrains a holistic understanding of the relationship between urban–rural integration and common prosperity and leaves the questions of through which pathways and in which dimensions urban–rural integration influences common prosperity to be further explored in depth [9].

1.2. Literature Review

Existing research on urban–rural integrated development has produced a relatively rich body of theoretical and empirical results. The relevant literature typically explores the connotation, measurement methods, and economic and social effects of urban–rural integration from different dimensions [10]. Regarding research content, academia generally understands urban–rural integration as a multidimensional and systematic process, with research focus primarily on economic integration, social integration, and spatial integration [11,12,13]. Among these, economic integration emphasizes the examination of urban–rural industrial synergy, factor mobility, and income linkages [14]. Social integration focuses on narrowing the gaps in the provision of public services, social security, and living conditions [15]. Spatial integration highlights the coordinated evolution of urban–rural functional division, land use structure, and spatial form [16]. In terms of research methods, the existing literature mostly relies on panel data at the provincial or prefecture-level city level. Some studies further use county-level or micro-survey data to construct an urban–rural integration development index through methods such as comprehensive evaluation indicator systems, entropy method, and principal component analysis. On this basis, they apply fixed-effect models, spatial econometric models, or multi-level linear models to analyze its influencing factors and economic and social effects [17,18,19]. Related studies generally agree that urban–rural integration helps promote the rational allocation of factors, enhance regional development efficiency, and, to some extent, alleviate urban–rural income disparities and public service imbalances [20]. Meanwhile, the intrinsic connections between different dimensions of urban–rural integration in existing research and their comprehensive effects need to be systematically characterized within a unified analytical framework. Additionally, most current studies remain at the level of measuring integration levels and analyzing influencing factors, lacking sufficient analysis of the deeper mechanisms of urban–rural integration.

Research on common prosperity has gradually become an important topic in the fields of economics and regional development in recent years. The relevant literature mainly focuses on the definition, measurement methods, and pathways for achieving common prosperity [21,22]. Regarding research content, academia generally understands common prosperity as a comprehensive development goal that goes beyond mere income growth [23]. Research focus covers multiple aspects, including residents’ income distribution patterns, equalization of public services, social security levels, and quality of life [24,25]. Based on this understanding, existing studies typically construct multidimensional indicator systems to characterize the level of common prosperity. Indicator selection includes economic dimensions such as income level, income inequality, and consumption structure [26,27,28], and gradually extends to non-economic dimensions that reflect development quality and opportunity equity, such as education, healthcare, housing, social security, and ecological environment [29,30]. In terms of research methods, the relevant literature mostly relies on provincial or prefecture-level city panel data. It employs methods such as entropy method, principal component analysis, or factor analysis to comprehensively measure multidimensional indicators [31,32]. Based on this, panel regression models, spatial econometric models, or grouped regression methods are used to analyze regional differences, evolutionary characteristics, and influencing factors of common prosperity [33,34]. Overall, existing research generally agrees that China’s level of common prosperity shows a steady upward trend, but regional disparities and urban–rural gaps remain significant. Income distribution structure, fiscal expenditure structure, and public service supply capacity are considered important factors influencing the achievement of common prosperity [35]. The existing research mostly measures the level of common prosperity by constructing multidimensional indicator systems, but it has two major limitations. First, it emphasizes indicator construction while neglecting process analysis, meaning it focuses on the static level of common prosperity but overlooks its dynamic evolution and the phased coupling relationship with urban–rural integration. Second, it emphasizes economic outcomes while neglecting spatial attributes, meaning it lacks a deep explanation of the spatial distribution characteristics of common prosperity and fails to connect with the impact of urban–rural spatial restructuring.

Compared to the relatively abundant research achievements on urban–rural integrated development and common prosperity separately, the existing literature that incorporates both into the same analytical framework to systematically explore the impact of urban–rural integration on common prosperity remains limited [36]. An important reason is that there is a certain degree of overlap between the indicator construction for urban–rural integrated development and common prosperity [37]. For instance, both involve income disparity, public service provision, and improvement in living conditions. This, to some extent, blurs the conceptual boundaries between the two, making it prone to confusion in the connotation of indicators in related research [38,39]. Affected by this, some studies include both urban–rural integration and common prosperity as comprehensive indices in their empirical models but fail to clearly distinguish the logical relationship between urban–rural integration as a spatial development process and common prosperity as a development outcome. Nonetheless, related studies still reveal potential connections between the two from different aspects. The existing literature finds that the flow of factors between urban and rural areas, coordinated industrial development, and the extension of public services to rural areas help improve the urban–rural income structure and residents’ welfare levels [40,41]. This, to some extent, aligns with the goal requirements of common prosperity. However, such research mostly approaches the topic from economic or social dimensions, lacking a systematic characterization of the spatial manifestations of urban–rural integration [42]. It also fails to adequately discuss the potential differential impacts of urban–rural integration on common prosperity at different stages and under varying spatial conditions. Overall, existing studies have not yet constructed an analytical framework that clearly distinguishes yet interconnects urban–rural integration and common prosperity from the perspective of spatial structure and functional evolution. Whether and under what conditions urban–rural integration promotes common prosperity still requires further in-depth exploration [43].

The existing literature systematically explores urban–rural integrated development from multiple dimensions such as factor mobility between urban and rural areas, equalization of public services, and coordinated regional development. It also analyzes the pathways to achieve common prosperity from perspectives such as income distribution, opportunity equity, and welfare security [44,45]. Existing studies have deficiencies at three levels. First, existing studies fail to clarify the internal connection between the spatial restructuring framework of urban–rural integration and the social effect framework of common prosperity. They neither explain how spatial form evolution affects public service equalization nor elucidate how factor mobility acts on the improvement in resident welfare. Second, the analysis paradigm presents the feature of emphasizing description over mechanism. Most studies verify the positive impact of urban–rural integration on common prosperity through linear regression. However, they do not reveal the internal mechanism of the nonlinear interaction between the two. They especially ignore key characteristics such as regional heterogeneity and temporal lag. Finally, in terms of measurement methods, existing research relies heavily on macro-statistical data. This makes it difficult to accurately capture the spatial heterogeneity characteristics of both urban–rural integration and common prosperity. Consequently, the characterization of their coupling relationship is relatively coarse, lacking a refined and dynamic analytical perspective. Therefore, this study conducts a comparative analysis of urban–rural integration outcomes and the structure of common prosperity. It delves into the differential effects and mechanisms through which urban–rural integrated development influences the achievement of common prosperity across different regions and at different developmental stages. This provides a theoretical basis and policy insights for optimizing the pathways of urban–rural integrated development and promoting the tiered and categorized realization of the common prosperity goal.

1.3. Study Questions

Based on the aforementioned research gaps, this study proposes the following three core research questions: First, what differences exist in the spatiotemporal evolution characteristics of urban–rural integration development and the level of common prosperity in Guangdong Province? Second, does the impact of urban–rural integration development on common prosperity exhibit characteristics of regional heterogeneity, phasing, and lagging? Third, through what pathways does urban–rural integration affect the improvement in the common prosperity level, and what are the key constraining factors in its mechanism of action?

This study innovatively integrates land use data, NTL data, and POI facility data. It constructs a space–economy–society tridimensional measurement framework for urban–rural integration and common prosperity, covering the dimensions of spatial structure evolution, economic activity disparity, and public service equalization. This approach achieves a more comprehensive and multidimensional characterization of the study subjects.

1.4. Theoretical Analysis Framework

Urban–rural integration establishes a complete causal chain affecting common prosperity through three progressive dimensions: spatial restructuring, factor mobility, and inclusive resource allocation. Its core logic is breaking the urban–rural dual structure, followed by optimizing the efficiency of resource allocation, and improving residents’ welfare levels. The fundamental carrier of urban–rural functional integration is that urban–rural integration promotes the spatial boundary between urban and rural from fragmentation to penetration, and shifts the land use structure from single-function to multi-function. Urban industrial space extends into the rural area, and rural ecological space interfaces with urban consumption space, forming a functional gradient pattern of urban–township–rural. This spatial restructuring provides the geographical foundation for factor mobility and industrial synergy, and serves as the prerequisite condition for urban–rural integration to influence common prosperity. The core driving force of urban–rural economic synergy originates from the breaking of institutional obstacles by urban–rural integration. By breaking institutional barriers such as household registration and land, urban–rural integration promotes the free factor mobility of labor, capital, and technology between urban and rural areas. Specifically, on the one hand, rural labor enters urban area to obtain higher income, and the accumulation of human capital in turn feeds back into rural development. The efficient allocation of factors narrows the urban–rural income gap, which is the key pathway through which urban–rural integration influences common prosperity.

2. Materials and Methods

2.1. Study Area

This paper selects Guangdong Province as the study area (Figure 1). As the region with the highest level of economic development and the most pronounced regional disparities in China [46], Guangdong exhibits significant internal heterogeneity. On the one hand, the Pearl River Delta region, relying on a highly concentrated industrial system and well-developed infrastructure, has long led the nation in terms of economic development, factor allocation efficiency, and public service provision capacity. On the other hand, eastern, western, and northern Guangdong remain primarily dependent on agriculture and resource-based industries, with relatively weak foundations for urban–rural development and more prominent issues of urban–rural disparities and regional imbalances. This marked intra-regional heterogeneity makes Guangdong Province an ideal sample for examining the progress of urban–rural integration and its impact on common prosperity. Therefore, choosing Guangdong as the study area not only offers strong representativeness and practical relevance but also helps reveal the spatial differences and mechanisms through which urban–rural integration affects common prosperity at the provincial scale.

2.2. Study Data

This study first uses land use classification data to delineate urban and rural space. Then, it employs night-time light (NTL) data to reveal the urban–rural income gap. Finally, it utilizes the diversity of POI data to measure the distribution differences in urban and rural public services and social resources, reflecting the level of common prosperity.

The data sequence in this study spans from 2013 to 2025. It covers 13 continuous years. To ensure time series consistency, all datasets undergo a three-step standardization processing of year calibration, projection unification, and spatial matching. First, we uniformly adopt the WGS-84 geographic coordinate system, which eliminates projection deviations from different data sources. Second, we complete data clipping based on the administrative boundary vector map of Guangdong Province, which ensures that the spatial scope of all data is completely consistent. Third, we conduct lag verification on the latest data of 2025. We then cross-check it with the data of the same period in 2024. This reduces the interference of data outliers on the analysis results. The potential uncertainty in the data of this study mainly originates from three aspects. First, the resolution of remote sensing images limits the classification accuracy of land use classification data. Second, extreme weather and holiday lighting easily disturb NTL data. Third, POI data has platform collection preferences. The coverage rate of commercial POI is higher than that of public welfare POI. The subsequent analysis controls the above uncertainties through methods such as data filtering and outlier removal.

2.2.1. Land Use Data

Land use classification data describes the characteristics of the urban–rural spatial structure and serves as an important foundation for identifying urban–rural spatial differences and the degree of integration. Based on the concept of urban–rural integration development, the transformation of urban–rural space from single-function to multifunction is regarded as a key manifestation of deepening urban–rural integration. Therefore, using the mixed land use index to characterize the urban–rural spatial structure helps measure the level of urban–rural integration from a spatial dimension. This study adopts the China Land Cover Dataset (CLCD) released by Wuhan University as the source of land use classification data [37]. This dataset, based on multi-source remote sensing imagery, offers advantages such as long temporal coverage, high spatial resolution, and a relatively comprehensive classification system. It has been widely applied in research on urban–rural spatial structure, land use change, and regional development. During data processing, this study first clips and reclassifies the CLCD land use data to align with the scope and scale of the Guangdong Province study area. This study obtains land use classification data from 2013 to 2025, as shown in Figure 2.

2.2.2. Night-Time Light (NTL) Data

NTL data is widely used to describe the intensity of human economic activities and the level of regional development, serving as an important proxy variable for reflecting spatial economic imbalances. In the study of urban–rural integration development, the urban–rural income gap is not only reflected in statistical income level differences but also in the distribution patterns of spatial economic activity density and intensity. Based on this understanding, using differences in night-time light brightness to describe the relative gaps in economic activities and income levels within urban–rural space helps identify the actual progress of urban–rural integration from a spatial perspective. More balanced distribution of night-time lights in urban–rural space usually indicates a higher degree of diffusion of economic activities between urban and rural areas, a relatively smaller urban–rural income gap, and a relatively higher level of urban–rural integration. This study uses Visible Infrared Imaging Radiometer Suite (VIIRS) NTL data as the research data source. This data, released by the National Oceanic and Atmospheric Administration (NOAA), offers advantages such as strong temporal continuity, relatively high spatial resolution, and minimal saturation effects. It has been widely applied in economic activity measurement, regional development disparity analysis, and urban–rural spatial research [47]. In terms of data processing, this study first conducts noise removal on the NTL data to eliminate interference from abnormal bright spots and background noise, and performs uniform calibration on data from different years to improve the consistency of the time series. NTL data from 2013 to 2025 is obtained, as shown in Figure 3.

2.2.3. POI Data

POI (Points of Interest) data reveals the infrastructure level, public service provision, and social resource distribution in different regions by reflecting the distribution of various points of interest at specific geographic locations. Since POI data covers multiple fields such as commerce, education, healthcare, culture, and transportation, its diversity and comprehensiveness can comprehensively reflect the coverage of urban–rural public services, thereby revealing the differences in opportunities for urban and rural residents to access public services [48]. Therefore, the diversity of POI data can serve as an important indicator for measuring the level of common prosperity in the process of urban–rural integration. This study uses POI data provided by Amap and Baidu Maps as the research data source. This data is collected through various methods such as satellite remote sensing, geographic information system (GIS) technology, and user big data analysis, featuring high spatial resolution, wide coverage, and strong real-time performance, making them suitable for analyzing the spatial differences in public services and social resources in urban and rural areas. The POI data provided by Amap and Baidu Maps covers different categories such as shopping malls, schools, hospitals, government service agencies, and cultural and entertainment venues, which can reflect the differences in resource allocation between urban and rural areas in various aspects such as economy, education, healthcare, and culture. During data processing, POI data is first cleaned to remove duplicates and invalid entries, and categorized to ensure that each category accurately represents a specific public service or social resource. Next, according to the division of urban and rural areas, POI data is spatially clustered based on geographic coordinates, and the density and diversity indices of different POI categories are calculated. This study obtains POI data from 2013 to 2025, as shown in Figure 4.

2.3. Methods

2.3.1. Theoretical Basis for Indicator Construction

This study centers on the core connotations of urban–rural integration and common prosperity. It selects three categories of core indicators: land use mix degree, urban–rural NTL disparity, and POI facility diversity. Their theoretical basis and representational logic are as follows:

The essence of urban–rural integration is the transformation of urban–rural space from dual segmentation to functional composite [16]. Land use structure is the direct carrier of spatial functions. Traditional urban–rural space possesses a single functional attribute. Cities mainly comprise construction land, and rural areas mainly comprise agricultural land. Urban–rural integration development promotes the interwoven layout of living, industrial, and ecological functions in urban–rural space. It manifests as the diversification and equilibrium of land use types. The core of urban–rural economic integration is narrowing the urban–rural income gap and the difference in economic activity intensity [14]. The land use mixture degree calculated by the Shannon diversity index reflects both the richness and distribution balance of land types. A higher index value indicates more composite spatial functions, fuzzier urban–rural boundaries, and a higher level of spatial integration. The selection of this indicator aligns with the theoretical consensus that spatial reconstruction is the basic prerequisite of urban–rural integration. It accurately portrays the core characteristics of urban–rural spatial integration.

NTL brightness possesses a significant positive correlation with regional economic activity intensity and resident income levels [47]. Studies widely confirm it serves as an effective proxy variable for macroeconomic data. The urban–rural NTL gap index reflects the balance degree of urban–rural economic development by quantifying the light brightness difference between urban and rural areas. A smaller index indicates that the intensity of economic activities between urban and rural areas is closer, the flow of factors is smoother, and the level of economic integration is higher. The selection of this indicator breaks through the limitation of low spatial resolution in traditional income statistical data. It aligns with the theoretical connotation that free economic factor mobility is the key driving force of urban–rural integration.

The core connotation of common prosperity encompasses three major dimensions: the equalization of public services, the improvement in living convenience, and the fairness of development opportunities [23]. The type and distribution of POI facilities directly determine the accessibility of public services and the fairness of opportunities available to residents [48]. The POI facility diversity index covers various types of facilities including education, healthcare, commerce, and transportation. A higher index indicates a more comprehensive supply of regional public services, more balanced living convenience and development opportunities for residents, and a higher level of common prosperity.

2.3.2. Shannon Diversity Index

The Shannon Diversity Index originates from information entropy theory and can comprehensively reflect the quantity and distribution equilibrium of different types of elements in a system. This index measures the diversification level of the overall structure by describing the proportion and uniformity of each type of element, with a higher value indicating a more complex and balanced internal structure of the system [49]. In urban and regional studies, the Shannon Diversity Index is widely used to characterize the mixed degree of land use structure and the diversity of functional elements, possessing a strong theoretical foundation and practical applicability [50]. This study uniformly adopts the Shannon Diversity Index as the diversity measurement method, applying it to calculate both land use mix degree and POI functional diversity, thereby achieving methodological consistency and interpretative coherence of indicators.

Compared to other methods for measuring diversity or mix degree, the Shannon Diversity Index simultaneously considers both the number of types and structural equilibrium. This index not only reflects the quantity of element types but also describes the distribution differences between different types, avoiding information loss that may result from relying solely on type count. Moreover, its mathematical form is concise and highly comparable. The calculation method for the Shannon Diversity Index is clear, and its results are continuous, allowing for horizontal comparisons across different spatial units and different element systems. Additionally, the Shannon Diversity Index has a wide range of applicable objects. This method can be flexibly applied to various element systems, such as land use structure and functional facility distribution, making it suitable for the comprehensive analysis of multi-source data.

The calculation formula for the Shannon Diversity Index is as follows:

$$H=-\sum _{i=1}^S{p}_i\mathrm{l}\mathrm{n} p_i$$

where:

$H$ is the Shannon Diversity Index, $S$ is the number of element types, $p_i$ is the proportion of the $i$-th type of element in the study unit, and $\mathrm{l}\mathrm{n}$ is the natural logarithm.

When only a single type of element exists within the study unit, $H$ takes the value 0; as the number of element types increases and their distribution tends toward equilibrium, the value of $H$ gradually increases.

To eliminate the influence of different type counts on the results, this study normalizes the Shannon Diversity Index:

$$H^{\prime }={\displaystyle \frac{H}{\ln S}}$$

The normalized index ranges from 0 to 1, with a higher value indicating a greater level of diversity.

At the land use level, the Shannon Diversity Index measures the mixed degree of different land use types within a spatial unit. In the specific calculation, $p_i$ represents the proportion of the $i$-th land use type’s area to the total area of the unit, and $S$ represents the number of land use types. The Land Use Shannon Diversity Index reflects the complexity and equilibrium of the land use structure within a spatial unit. Areas with higher index values typically exhibit an interwoven distribution of multiple land use types and possess strong functional composite characteristics, commonly found in urban or urban–rural transition zones. Areas with lower index values are dominated by single or a few land use types, often corresponding to typical urban core areas or traditional rural regions. Therefore, the Land Use Shannon Diversity Index can serve as an important indicator for identifying differences in urban–rural spatial structure and the characteristics of urban–rural transition.

At the functional element level, the Shannon Diversity Index measures the diversity level of POI types within a study unit. Here, $p_i$ represents the proportion of the $i$-th POI type’s count to the total POI count in the unit, and $S$ represents the number of POI functional types. The POI Shannon Diversity Index can reflect the comprehensive supply level of public services, commerce, employment, and living functions within a region. A higher index value indicates a richer functional structure and more complete service types in the area, which helps improve residents’ convenience and opportunity equity. Under the common prosperity research framework, POI functional diversity is regarded as an important indicator for measuring the accessibility of regional public services and the balance of living opportunities. Areas with high diversity typically possess a more complete functional supply system and a stronger foundation for achieving a higher level of common prosperity.

2.3.3. Deep Neural Network (DNN)

This study employs a deep neural network (DNN) to construct a dynamic urban–rural boundary identification model. The DNN, through its multi-layer nonlinear mapping structure, can effectively learn the complex relationships between land use mix characteristics and urban–rural spatial attributes, making it suitable for describing the continuous transition features of urban–rural space from urban to transition zone and to rural. Using the Land Use Shannon Diversity Index as the core input variable, the DNN model, via supervised learning, outputs the urban–rural attribute results for spatial units, achieving adaptive adjustment of urban–rural boundaries [51].

Among various deep learning models, this study selects the DNN. First, the input data for this study is numerical indicators, such as the land use diversity index and its neighborhood characteristics, and does not involve remote sensing imagery or sequence prediction, making the DNN more suitable for processing structured feature data. Second, there is no unified threshold for land use mix degree in urban–rural space. DNN can avoid manually setting classification criteria and automatically learns discrimination rules through a data-driven approach. Finally, compared to models like CNN and Transformer, DNN has a simpler structure, clearer parameter meanings, and is easier to interpret and reproduce, meeting the basic requirement for interpretability in urban–rural spatial research.

The DNN model consists of an input layer, several hidden layers, and an output layer. Its basic computational form is as follows:

$${\mathrm{h}}^{\left(l\right)}=\sigma \left({\mathrm{W}}^{\left(l\right)}{\mathrm{h}}^{\left(l−1\right)}+{\mathrm{b}}^{\left(l\right)}\right)$$

where:

${\mathrm{h}}^{\left(l\right)}$ is the output vector of the $l$-th layer, ${\mathrm{W}}^{\left(l\right)}$ and ${\mathrm{b}}^{\left(l\right)}$ are the weight matrix and bias term of the $l$-th layer, respectively, $\sigma (\cdot )$ is the nonlinear activation function (such as ReLU), and the input layer ${\mathrm{h}}^{\left(0\right)}$ is the Land Use Shannon Diversity Index and its extended features. The output layer uses the Sigmoid or Softmax function to map the model output into urban–rural attribute probability values.

The model input is the Land Use Shannon Diversity Index ($H_{LU}$), which reflects the land use mix degree within a spatial unit and is the core input variable of the model. The model output is the urban–rural attribute probability value:

$$y\in [0, 1]$$

where:

$y$ approaching 1 indicates strong urban attributes, and $y$ approaching 0 indicates strong rural attributes.

To ensure model identification accuracy, this study uses the 5-fold cross-validation method to test the performance of the DNN model. It selects Accuracy, Precision, Recall, and F1-score as the core evaluation metrics. The training results show that the model achieves an accuracy of 89.2% and an F1-score of 0.87 on the validation set, indicating the model has strong discriminative ability for urban–rural spatial attributes. Simultaneously, analysis through the confusion matrix reveals that the model’s identification accuracy for urban–rural transition zones (F1 = 0.85) is significantly higher than traditional methods. This effectively addresses the ambiguity problem of the urban/rural binary classification in transition zones.

Among various deep learning models and traditional classification methods, the core rationale for selecting DNN in this study and its differences from other methods are as follows:

Comparison with the traditional fixed threshold method: The traditional method classifies urban–rural boundaries based on empirically set land use mix degree thresholds, which has the drawbacks of strong subjectivity and poor adaptability. Significant differences in the development foundations of urban and rural areas exist across different regions. A uniform threshold cannot adapt to the spatial characteristics of all areas, easily leading to substantial deviations between boundary demarcation and actual conditions. In contrast, the DNN model autonomously learns discrimination rules based on sample data. It can dynamically adjust boundary determination criteria according to the spatial characteristics of different regions, thereby enhancing the objectivity and accuracy of identification results. Furthermore, the traditional method can only output a binary classification result of urban/rural, failing to depict the continuous characteristics of urban–rural transition zones. The urban–rural attribute probability value (y ∈ [0, 1]) output by the DNN model can accurately reflect the urban–rural transition characteristics of spatial units. This better aligns with the spatial evolution patterns of urban–rural integration.

Comparison with other machine learning models: Compared to traditional machine learning models such as Support Vector Machine (SVM) and Random Forest (RF), DNN possesses stronger nonlinear fitting capability. The relationship between urban–rural spatial attributes and land use mix degree is not a simple linear association; it is also subject to the complex influence of neighboring characteristics such as terrain and transportation. SVM and RF have limited fitting ability for high-dimensional nonlinear data, making them prone to underfitting issues. In contrast, DNN, through the nonlinear mapping of multiple hidden layers, can capture the complex interactive relationships between variables, thereby enhancing the model’s prediction accuracy.

Comparison with other deep learning models: Compared to models such as Convolutional Neural Networks (CNN) and Transformer, DNN is more suitable for the structured data characteristics of this study. CNN excels at processing image-like data and relies on the spatial texture features of remote sensing imagery. Transformer is suited for handling sequential data, but it has a large number of parameters and high computational costs. The input data for this study is numerical structured indicators such as land use mix degree and terrain slope. The DNN structure is simpler, its parameter meanings are clearer, and it facilitates interpretation and method reproducibility. This aligns with the basic requirement for interpretability in urban–rural spatial research.

2.3.4. Urban–Rural Income Gap Index

Under the condition of lacking high spatial resolution income statistics, NTL data is widely used in regional economic development and income level research because it can objectively reflect the intensity of human economic activities and differences in living standards. The spatial variation in night-time light brightness reflects, to a certain extent, the gap in residents’ income levels and economic activity between different regions. This study constructs an urban–rural income gap measurement index based on NTL data to describe the degree of difference in economic development and income levels within urban–rural space, and further reflects the level of urban–rural integration development. A smaller urban–rural income gap indicates closer economic linkages between urban and rural areas and a higher degree of integration development.

Based on the spatial units or urban–rural area division results, this study extracts night-time light brightness information from urban areas and rural areas, respectively, and constructs an urban–rural night-time light difference index to characterize the level of the urban–rural income gap. The fundamental rationale is that a higher night-time light brightness indicates more active regional economic activities and relatively higher resident income levels. A smaller difference in night-time light brightness between urban and rural areas suggests a smaller urban–rural income gap and a higher level of urban–rural integration development.

Within the study area:

$L_u$ is the average night-time light brightness of urban areas, and $L_r$ is the average night-time light brightness of rural areas. The urban–rural income gap index is defined as follows:

$$G_{LU}={\displaystyle \frac{\mid L_u-L_r\mid }{L_u}}$$

where:

$G_{LU}$: the urban–rural income gap index based on night-time lights. A larger value indicates a more distinct difference in urban–rural night-time lights and a greater urban–rural income gap. A smaller value indicates a smaller difference in urban–rural night-time lights and a higher level of urban–rural integration development. To facilitate comparison between different regions, the index results can be further normalized to the range from 0 to 1.

3. Results

3.1. Analysis of Urban–Rural Integration Development Based on Land Use Data and NTL Data

Based on the analysis of land use data for urban–rural spatial changes in 2013, 2017, 2021, and 2025, it can be observed that the urban–rural spatial pattern underwent significant changes during the study period. Overall, the scale of urban built-up land continued to expand, showing distinct outward expansion characteristics, with urban boundaries continuously extending into surrounding areas. In contrast, changes in rural built-up land and agricultural land were relatively slow, with some areas exhibiting a trend of being encroached upon by urban space or undergoing functional transformation. These results indicate that the urban–rural spatial structure underwent substantial adjustment during the study period, with the dominant role of urban space in the overall spatial pattern continuously strengthening. Further analysis of urban–rural spatial differences based on NTL data reveals that, as shown in Figure 5, the intensity of urban–rural economic activities showed a continuous upward trend from 2013 to 2025, reflecting the overall improvement in economic levels and residents’ income levels in the study area. However, this upward trend exhibited significant spatial imbalance among different spatial units. In urban areas, especially the core cities of the Pearl River Delta, the increase in NTL intensity was substantial, demonstrating highly concentrated economic activities and rapid income growth. In contrast, non-core cities and vast rural areas also saw an increase in NTL intensity, but the rate of growth was relatively limited. The disparities between urban and rural areas, as well as among different regions, did not show a corresponding synchronous reduction.

The spatial and economic disparities in the process of urban–rural integration originate from the dual differentiation of regional development foundations and policy orientation. First, inherent differences in development foundations determine the capacity for factor agglomeration. The core cities of the Pearl River Delta leverage their port advantages, industrial foundations, and early policy benefits to form manufacturing services linked industrial clusters. Their capacity to attract factors such as capital, technology, and high-end talent is significantly higher than that of non-Pearl River Delta regions. This factor agglomeration effect further drives the expansion of urban construction land, creating a positive feedback cycle of factor agglomeration to spatial expansion and economic growth. The non-Pearl River Delta regions rely primarily on agriculture and resource-based industries. Their industrial chains are short with low added value, making it difficult to create a siphoning effect for factor agglomeration. The labor force in rural areas even continues to outflow to the Pearl River Delta, resulting in insufficient momentum for rural development. The second issue is that differentiated policy guidance intensifies spatial polarization. During the study period, Guangdong Province’s development policies have long favored the core areas of the Pearl River Delta, with major industrial projects and infrastructure investments concentrated in cities like Guangzhou and Shenzhen, driving a rapid increase in their night-time light intensity. In contrast, policy support for non-Pearl River Delta regions primarily focuses on addressing shortcomings and emphasizing infrastructure improvement, and has a relatively limited effect on promoting industrial upgrading and endogenous economic growth. Furthermore, differences in urban and rural land policies constrain rural development. The efficiency of market-based allocation for urban construction land is high, while the transfer mechanisms for rural land are still not well-developed. This prevents the full release of the economic value of rural land, further widening the gap in economic activity intensity between urban and rural areas.

Overall, urban–rural integration development promoted the improvement in the overall economic level and the enhancement in spatial connections during the study period. However, its actual outcomes are more manifested as a further strengthening of urban advantages, with significant differences persisting between core cities and rural areas in terms of economic activity intensity and income growth.

3.2. Analysis of Common Prosperity Based on POI Data

By measuring the level of common prosperity based on the POI-related facility diversity index, it can be observed that its spatial distribution exhibits significant imbalance characteristics, as shown in Figure 6. Overall, areas with high diversity index values are mainly concentrated in the Pearl River Delta region, and within non-Pearl River Delta areas, they are primarily distributed in municipal and county-level central urban areas. In contrast, vast rural areas far from central urban areas generally have low diversity index values. This spatial pattern indicates that significant gaps still exist between different regions and different spatial levels in terms of public service provision, living convenience, and development opportunities. The degree of achieving common prosperity varies markedly across space. From the perspective of temporal evolution, the POI facility diversity index shows differentiated change characteristics across different periods. During the period from 2017 to 2021, the overall increase in the diversity index was the largest, reflecting the most pronounced rise in the level of common prosperity during this stage. In comparison, changes in the diversity index were relatively moderate during the remaining periods, with limited overall improvement, indicating a relatively weaker degree of improvement in the level of common prosperity during these two phases.

The spatiotemporal pattern differences in POI facility diversity are essentially the result of combined effects from resource allocation logic and the degree of supply–demand matching. The layout of public service facilities follows the principle of efficiency first. Core cities of the Pearl River Delta and central urban areas of cities and counties have high population densities and strong economic strength, giving their facility construction significant scale effects that can attract high-quality resources like education, healthcare, and commerce to agglomerate. In contrast, rural areas have dispersed populations and weak fiscal support capabilities, making facility construction costly with low utilization rates, which leads to insufficient supply of public welfare facilities. This allocation logic forms a public service gradient of central urban areas to townships to rural villages, creating a certain tension with the equalization goal of common prosperity. The period from 2017 to 2021 was a key phase for Guangdong Province in advancing its Action Plan for Equalization of Basic Public Services. Policy dividends drove the extension of facilities such as education and healthcare to counties and key towns, directly leading to a rapid increase in the POI diversity index. In contrast, from 2013 to 2017, policies were in a pilot and exploratory stage, resulting in a slower pace of facility expansion. From 2021 to 2025, a bottleneck of hardware was easy to supplement, but software was difficult to improve emerged. Although facilities were added in rural areas, they lacked professional personnel and operational funding, making it difficult for service quality to align with urban standards, which led to a convergence in the growth rate of the POI diversity index. Additionally, POI facilities in rural areas primarily focus on basic services, with insufficient supply of high-quality services such as elderly care, child care, and culture. This creates a gap with residents’ demand for a better life and constrains the further enhancement in common prosperity.

Overall, although the level of common prosperity showed a general improvement trend during the study period, it exhibited significant imbalances in both spatial distribution and temporal evolution.

3.3. Analysis of the Differential Effects of Urban–Rural Integration Development on Common Prosperity

Based on the local spatial autocorrelation (LISA) method, a cluster analysis of the coupling relationship between urban–rural integration development and the level of common prosperity reveals that the spatial association patterns of urban–rural integration and common prosperity show significant differences across different stages, as illustrated in Figure 7. From 2013 to 2017, HH clusters were mainly concentrated in the Pearl River Delta region, indicating that this area simultaneously exhibited high levels of both urban–rural integration and common prosperity. HL clusters were primarily distributed in municipal and county centers of non-Pearl River Delta areas, showing a certain degree of asynchrony between the advancement of urban–rural integration and the improvement in common prosperity. From 2017 to 2021, the clustering pattern underwent significant changes. First, the HH cluster area in the Pearl River Delta region expanded notably, indicating a further strengthening of the synergistic improvement trend between urban–rural integration and common prosperity. Second, LH clusters in the surrounding areas of the Pearl River Delta increased significantly, reflecting that some regions experienced a relatively rapid improvement in common prosperity even with a relatively lower level of urban–rural integration. Finally, the number of HL clusters in non-Pearl River Delta areas increased, indicating a rise in spatial units where urban–rural integration advanced more quickly but the improvement in common prosperity lagged relatively behind.

From the perspective of stage transition conditions, the promotion effect of urban–rural integration on common prosperity depends on whether the transmission chain of factor mobility, industrial upgrading to universal accessibility of public services is smooth.

From 2013 to 2017, it was the early stage of integration dominated by factor mobility, and the lag of common prosperity became prominent. During this stage, the core of urban–rural integration was the flow of population and capital to cities, driving urban spatial expansion and enhanced economic activities, but the supply of public services does not follow up synchronously. Although non-Pearl River Delta cities and counties advanced urban–rural integration through land development, industrial upgrading lagged, failing to create sufficient high-quality employment opportunities, resulting in slow growth in residents’ income. Coupled with insufficient public service facilities, this formed an HL cluster characterized by high integration and low prosperity. In contrast, core cities in the Pearl River Delta leveraged their industrial advantages to achieve simultaneous progress in factor mobility and public service supply, thus presenting a stable HH cluster pattern.

From 2017 to 2021 is a period characterized by a smooth transmission chain driven by policy, leading to accelerated improvement in common prosperity. With the implementation of Guangdong Province’s public service equalization policy, the focus of urban–rural integration shifts from factor agglomeration to universal access to resources. The industrial spillover effects of Pearl River Delta core cities begin to appear. Surrounding areas, while undertaking industrial transfers, receive more investment in public service resources, driving an increase in LH clusters. Meanwhile, Pearl River Delta core cities further consolidate the scope of HH clusters by optimizing their public service structure. The key in this stage is that policy intervention unblocks the transmission path from integration to prosperity, transforming the outcomes of urban–rural integration into a tangible welfare improvement for residents.

From 2021 to 2025 is a period of structural optimization with diminishing marginal effects, where conversion efficiency encounters obstacles. After the hardware layout of public service facilities approaches saturation, the promoting effect of urban–rural integration on common prosperity begins to rely on software upgrades, such as improvements in service quality and the refinement of equalization mechanisms. However, due to limited fiscal capacity in some non-Pearl River Delta areas, it is difficult to continuously invest resources in software optimization, leading to diminishing marginal returns from urban–rural integration and a stabilization of the clustering patterns. This phenomenon indicates that a model of integration relying solely on spatial expansion and facility increases cannot sustainably drive common prosperity, and a shift to a quality and efficiency-oriented, connotative development model is necessary.

Based on the spatial clustering results of urban–rural integration and common prosperity, it can be further observed that the impact of urban–rural integrated development on common prosperity does not manifest immediately, nor does a simple linear correspondence exist. Instead, it exhibits distinct characteristics of being staged, lagging, and being subject to marginal effects.

First, the achievement of common prosperity exhibits significant lag characteristics. The results showing the basic stability and insignificant changes in the clustering patterns of urban–rural integration and common prosperity from 2013 to 2017 indicate that, in its early stages, urban–rural integration is more reflected in spatial form adjustments and accelerated factor mobility. Its direct effects are mainly concentrated on land use changes and enhanced economic activities, while core aspects of common prosperity such as the improvement in the public service system and the enhancement in quality of life are not yet fully manifested. This sequential characteristic of integration first, prosperity later results in some areas still exhibiting the HL clustering pattern, where the improvement in common prosperity lags behind even as the level of urban–rural integration rises. Second, the significant changes in clustering types from 2017 to 2021 reflect the concentrated release phase of the impact of urban–rural integration on common prosperity. Combined with the substantial increase in the POI diversity index during this stage, it can be inferred that as the degree of urban–rural integration continuously accumulates, its promoting effect on the structure of public service provision, facility allocation, and improvements in residents’ welfare gradually becomes apparent. This drives a noticeable leap in the level of common prosperity within a relatively short period. The simultaneous expansion of HH and LH clusters during this stage reflects the phased amplification effect of converting urban–rural integration outcomes into common prosperity. Finally, the clustering pattern tends to stabilize from 2021 to 2025, reflecting a certain degree of diminishing marginal effects in the impact of urban–rural integration on common prosperity. Following the rapid improvement in infrastructure and the public service system in the earlier period, new investments in urban–rural integration contribute more to structural optimization rather than quantitative expansion in the improvement in common prosperity. Their potential for enhancing indicators such as POI diversity becomes relatively limited, leading to a slowdown in the growth rate of the common prosperity level and a gradual solidification of the spatial pattern. This result indicates that relying solely on the scale expansion of urban–rural integration can no longer sustainably drive a significant improvement in common prosperity.

Overall, the impact of urban–rural integration development on common prosperity exhibits typical characteristics such as temporal lag, phased concentration, and diminishing marginal effects. Its actual outcomes are profoundly constrained by the regional development foundation and structural conditions. This finding not only explains the differentiated patterns observed in the spatial clustering of urban–rural integration and common prosperity, but also provides an important basis for subsequent efforts to promote the synergistic deepening of urban–rural integration and common prosperity from the perspectives of mechanisms and policies.

4. Discussion

Based on land use data, NTL data, and POI facility data, this study systematically analyzes the level of urban–rural integration development and common prosperity from 2013 to 2025 across multiple dimensions, including urban–rural spatial evolution, economic activity intensity, and public service diversity. This study finds that urban–rural integration significantly promoted urban spatial expansion and overall economic level improvement during the study period, but the extent of development varied markedly among different regions.

From the perspective of urban–rural integration development research, the existing literature often reveals the impact of urban–rural integration on economic growth and spatial pattern evolution from the perspectives of factor mobility, spatial structure adjustment, and regional coordinated development. It is generally believed that urban–rural integration can enhance the overall regional development level by promoting the rational allocation of population, industrial, and land factors [52,53]. The findings of this study are consistent with the above research in terms of the overall trend, indicating that urban–rural integration significantly drives urban spatial expansion and enhances economic activities during the study period [54]. By integrating land use and NTL data, this study characterizes the outcomes of urban–rural integration from both spatial form and economic activity perspectives, further revealing significant differences in the advancement speed and spatial manifestations of urban–rural integration across different regions, thereby providing more spatially heterogeneous empirical evidence for urban–rural integration research [6]. From the perspective of common prosperity research, existing studies often focus on dimensions such as income distribution, public service provision, and equality of opportunity [55], emphasizing the structural disparities in common prosperity among regions and groups. It is widely found that developed regions and central cities possess obvious advantages in achieving common prosperity [56,57]. This study characterizes the level of common prosperity based on the POI facility diversity index, and its spatial distribution characteristics show high consistency with existing research conclusions, namely that the level of common prosperity tends to cluster in core cities and central areas [58]. However, by spatially measuring common prosperity based on the structure of public services and living convenience, this study supplements the relative inadequacy in characterizing the outcome structure of common prosperity in existing research.

The existing literature generally recognizes urban–rural integration as an important pathway to promote common prosperity. However, most studies employ linear regression or policy effect evaluation methods, implying the assumption that urban–rural integration has a synchronous and positive effect on common prosperity [9,59]. However, such research has neither revealed the spatial–temporal heterogeneity of their effects, nor has it answered the core question of how does urban–rural integration translate into common prosperity. The incremental value and innovative insights of this study, compared to previous work, are mainly reflected in the following points. First, this study breaks through the single perspective of factor mobility and builds a trinity analytical framework of space, economy and society. Previous studies mostly focus on the impact of urban–rural factor mobility on income disparity, essentially remaining at a single-dimensional economic analysis [14,20]. This study innovatively integrates multi-source data including land use, NTL, and POI to characterize the coupling relationship between urban–rural integration and common prosperity from three dimensions: spatial form evolution, economic activity intensity, and equalization of public services. It is the first to incorporate spatial structure optimization and universal accessibility of public services into the mechanism through which urban–rural integration promotes common prosperity, addressing the deficiency of insufficient attention to the social dimension in existing research. Moreover, we reveal the nonlinear interaction patterns, correcting the traditional perception that urban–rural integration necessarily promotes common prosperity. Existing studies often assume a positive linear correlation between urban–rural integration and common prosperity [6,7]. However, through long-term time-series spatial clustering analysis, this study finds that the impact of urban–rural integration on common prosperity exhibits three major characteristics: regional heterogeneity, temporal lag, and diminishing marginal effects. This finding indicates that urban–rural integration does not necessarily or synchronously translate into common prosperity; its effectiveness depends on regional development foundations and public service supply capacity. This conclusion breaks through the idealized perception of urban–rural integration policy outcomes and provides a new explanation for understanding the real-world dilemma of high integration levels but low prosperity levels in some regions. Finally, this study proposes a policy orientation of differentiated by region and by stage, moving beyond the mere verification of spatial inequality issues. Previous studies have mostly stopped at verifying the current state of urban–rural and regional development inequalities [5,35], but have failed to propose targeted optimization paths. This study distinguishes four types of regions based on clustering results, such as high integration–high prosperity (HH) and high integration–low prosperity (HL), and clarifies the core constraints of each region. Core Pearl River Delta areas need to address the issue of diminishing marginal effects by optimizing public service structure rather than expanding scale to promote common prosperity. Non-Pearl River Delta regions need to overcome the problem of lagging integration conversion by strengthening urban–rural industrial synergy and infrastructure connectivity to accelerate the transformation of integration outcomes into resident welfare. This approach of targeted policies for different categories provides actionable policy references for promoting the synergistic development of urban–rural integration and common prosperity. The findings of this paper extend this understanding to some extent. On one hand, the expansion of HH clusters in the Pearl River Delta region confirms that urban–rural integration can effectively promote common prosperity under specific conditions. On the other hand, the existence of cluster types such as HL and LH indicates that the effect of urban–rural integration on common prosperity has significant regional heterogeneity and stages. It does not translate into an increase in common prosperity levels immediately and proportionally in all regions. Further combining the temporal evolution results, this study reveals the characteristics of lagging common prosperity and diminishing marginal effects, which are relatively less systematically discussed in existing research [60]. Some research pays attention to the stage differences in the implementation effects of urban–rural integration policies, but lacks explanations for this phenomenon from the perspectives of spatial structure and outcome transformation [61,62]. Using long-term time-series data and spatial clustering analysis, this study shows that the initial stage of urban–rural integration primarily reflects factor and spatial integration. Its promoting effect on common prosperity often concentrates in the mid-term and shows a trend of diminishing marginal effects in the later stage, thus providing new empirical support for understanding the nonlinear relationship between urban–rural integration and common prosperity [63].

By placing urban–rural integrated development and common prosperity goals within a unified analysis framework and systematically examining the actual effect of urban–rural integrated development on common prosperity from the perspectives of structure and spatial correlation, this study breaks through the analysis paradigm in existing research that relatively separates the two or only tests linear relationships, thereby providing a new research perspective for understanding the complex relationship between urban–rural integration and common prosperity. This study, through multi-source data analysis, reveals that the impact of urban–rural integration development on common prosperity exhibits significant characteristics of being phased, lagging, and exhibiting diminishing marginal effects. It indicates that urban–rural integration does not necessarily and immediately translate into an increase in common prosperity levels. Its effect is profoundly constrained by regional development foundations and structural conditions. This provides targeted empirical evidence for promoting the coordinated development of urban–rural integration and common prosperity by region and by stage.

While this study systematically explores the relationship between urban–rural integration development and common prosperity from the perspectives of multi-source data and spatial analysis, there are still aspects that require further refinement. Although this study systematically explores the relationship between urban–rural integrated development and common prosperity from the perspectives of multi-source data and spatial analysis, there are still some aspects that require further improvement. This study adopts the POI facility diversity index to represent the level of common prosperity, with its core basis being that the accessibility of public services and the convenience of life are important dimensions of common prosperity. However, there remains a certain deviation between this indicator and the core connotation of common prosperity. On the one hand, POI data can only reflect the spatial distribution and quantity supply of public service facilities; it cannot capture demand-side characteristics such as residents’ actual usage frequency and satisfaction with service quality. On the other hand, there is also bias in the type of coverage of POI data; in the Amap and Baidu POI data used in this study, the collection coverage for commercial and transportation facilities is relatively high, while public welfare facilities such as elderly care, childcare, and cultural services are often not fully included, especially in remote rural areas. This may lead to insufficient recording of facilities closely related to people’s well-being, thereby potentially underestimating the level of common prosperity in rural areas. Furthermore, the core connotation of common prosperity includes key dimensions such as income distribution equity, improvement of social security, and enhancement in subjective well-being. Due to limited access to high-spatial-resolution microdata, this study does not incorporate direct measurement indicators like household income, the gap in property income, pension insurance coverage, and residents’ sense of happiness. This limitation further confines the research findings to a surface-level analysis of spatial equilibrium of public services, making it difficult to deeply reveal the mechanism through which urban–rural integration affects residents’ actual income growth and the enhancement in their welfare perception. Finally, it needs to be clarified that this study primarily adopts descriptive analysis and spatial clustering methods, aiming to reveal the spatial correlation characteristics between urban–rural integration and common prosperity, and has not conducted rigorous empirical identification of the causal relationship and underlying mechanism between the two. This research design is based on the core positioning of this paper, which focuses on typical fact identification and phenomenon description. The core objective of this study is to characterize the spatiotemporal coupling features between urban–rural integration and common prosperity, and to reveal the phased and heterogeneous patterns of their association, which belongs to the basic research stage of phenomenon identification and mechanism exploration. Correlation analysis can clearly present the spatial distribution and temporal evolution characteristics of both, providing typical facts support for subsequent causal analysis. Future research can introduce spatial econometric models, panel regressions, or quasi-natural experiment methods to further test the impact pathways and internal mechanisms of urban–rural integration on common prosperity, thereby deepening the understanding of the synergistic advancement path between urban–rural integration and common prosperity from multiple dimensions.

5. Conclusions

Based on land use data, NTL data, and POI data from 2013 to 2025, this study comprehensively applies spatial analysis and deep learning methods to systematically study the spatiotemporal evolution characteristics and mutual relationships of urban–rural integrated development and common prosperity levels. This study characterizes the actual results of urban–rural integration and common prosperity from multiple dimensions such as urban–rural spatial form evolution, economic activity intensity, and public service diversity, and identifies the coupling types and evolution processes of the two in different regions and different periods through spatial clustering. This study finds that urban–rural integration significantly promotes urban spatial expansion and the improvement in overall economic activity levels during the study period, but its development results show obvious spatial imbalance. The common prosperity level generally improves, but it highly concentrates in the Pearl River Delta and city–county center areas. Further analysis finds that the promotion effect of urban–rural integrated development on common prosperity possesses significant characteristics of regional heterogeneity, stages, and time lags. It manifests as synergistic improvement in core city areas, while insufficient conversion efficiency and structural differentiation phenomena exist in some non-core areas. The above findings indicate that urban–rural integration does not necessarily and immediately translate into an improvement in the common prosperity level, and factors such as regional development foundation, industrial structure, and public service supply capability profoundly constrain its effect.

This study provides empirical evidence for understanding the complex relationship between urban–rural integration and common prosperity, holding significant practical implications. It aids in examining the actual effects of urban–rural integration policies from spatial and structural perspectives, preventing the simplistic equation of urban–rural integration with the achievement of common prosperity. Furthermore, it offers decision-making references for optimizing the development path of urban–rural integration by region and by stage, and for improving the efficiency of converting integration outcomes into common prosperity. This contributes certain policy insights for promoting the synergistic advancement of urban–rural integration and common prosperity.