Freight Big Data-Based Dual-Scale Study of Economic Spatial Organization and Planning Responses in Hubei Province

Abstract

Using truck GPS trajectory data, this study measures the intensity of economic spatial linkages in Hubei Province at both administrative and cross-administrative scales and examines the hierarchical structure and spatial pattern of its urban economic network. By comparing the results with existing regional plans, the study provides empirical support for regional coordination and spatial planning. Network centrality analysis, linkage intensity measurement, and community detection algorithms are integrated to construct a topological model of the urban economic network from three dimensions: urban node hierarchy, inter-city linkage intensity, and urban cluster structure. To overcome administrative boundary constraints, a 5 km × 5 km grid-based approach is applied to identify functionally connected urban economic communities. The results indicate that Hubei Province’s urban economic network exhibits a highly dominant core accompanied by multiple secondary supporting centers. While the Wuhan Metropolitan Area demonstrates high economic activity, internal horizontal linkages remain relatively weak, and the roles of Yichang and Xiangyang as regional sub-centers require further strengthening. Grid-based analysis further reveals pronounced cross-administrative economic linkages. Accordingly, this study suggests strengthening support for regional sub-centers and promoting better alignment between administrative space and functional space within the spatial planning system, with enhanced cross-regional coordination.

1. Introduction

Implementing regional coordinated development is one of the key measures for promoting high-quality development and is of great significance for sustaining healthy and stable economic growth in China. Coordinated development among multiple central cities within a region can enhance agglomeration effects and economies of scale across a broader territorial space [1]. With the accelerating processes of urbanization and regional integration, cities are no longer isolated administrative units; rather, they have become interconnected through flows of population, industry, capital, information, and freight, forming large-scale regional linkage systems and gradually evolving into interdependent and interactive urban agglomerations [2]. As important spatial carriers of regional division of labor, optimized factor allocation, and innovation diffusion, urban agglomerations play an increasingly significant role in regional economic cooperation and global competition [3]. As a complex socio-economic phenomenon, cross-regional flows and exchanges of various factors within urban agglomerations profoundly shape economic interactions and their network structures, prompting a re-examination of traditional static spatial delineations and their correspondence with actual interaction patterns and spatial organization [4]. In this context, scientifically identifying the internal structure of regional economic linkages and cross-regional flow characteristics, and on this basis rationally delineating urban agglomerations and economic clusters, has become a fundamental task for achieving coordinated regional development [5].

In the context of deepening regional integration and increasingly intensive factor mobility, regional spatial organization can no longer be adequately explained solely through traditional territorial units and administrative hierarchies. Relevant theoretical perspectives can generally be categorized into two approaches: the “space of place” and the “space of flows.” The former, grounded in Christaller’s central place theory [6], emphasizes hierarchical urban systems and center–hinterland relationships. The latter, proposed by Castells, highlights cross-city functional linkages, network structures, and relational connections formed through flows of population, capital, information, and goods [7]. Building upon these perspectives, Taylor’s central flow theory further argues that both space of place and space of flows coexist within regions: the former manifests as hierarchical urban systems and their hinterland relations, while the latter reflects external linkages that transcend traditional urban hinterlands [8]. Central flow theory incorporates core ideas from the urban network paradigm without entirely rejecting central place theory, instead emphasizing the interpretation of place-based structures through the lens of functional linkages [1]. On this theoretical basis, the understanding of regional boundaries has also evolved from a purely institutional delineation toward function-based identification grounded in spatial interactions. Administrative boundaries are institutional constructs formed through governance, statistical, and managerial needs in a top–down manner, whereas functional boundaries emerge bottom–up from actual spatial interactions [9]. While the two often overlap at the national scale, spatial mismatches are more likely to occur at regional and urban scales [4]. In light of this perspective, this study adopts a dual-scale analytical framework that juxtaposes administrative space and functional space. The administrative scale is employed to identify urban economic spatial structures within existing governance units, while the cross-administrative scale is used to delineate functional economic communities organized by freight flows. Through this dual-scale approach, the study aims to provide a more comprehensive understanding of regional economic spatial organization.

Studies on regional economic spatial structure typically take existing governance units, such as prefectural-level cities and counties/districts, as the basic units of analysis. By applying urban network analysis methods, cities are conceptualized as network nodes, inter-city relationships as network edges, and urban clusters are subsequently delineated to identify internal economic linkage patterns and spatial organizational characteristics within regions. On this basis, urban network research at the administrative scale has gradually developed a relatively mature “point–line–area” analytical framework. First, node-level analysis primarily relies on centrality measures—such as degree centrality, closeness centrality, and betweenness centrality—to identify the hierarchical position and functional roles of cities within the network, thereby revealing their relative positions in processes of resource agglomeration, diffusion, and control [10]. Second, edge-level analysis focuses on measuring the intensity, directionality, and density of inter-city linkages. Methods such as the Gravity Model and social network analysis are widely employed to characterize patterns of regional interaction and structural features of economic linkage networks [2,11]. Third, area-level analysis introduces community detection algorithms and related approaches to identify regions of high internal connectivity within the network, thereby delineating urban clusters at different hierarchical levels [12]. Overall, urban network studies at the administrative scale provide an important methodological foundation for understanding hierarchical urban systems, regional linkage structures, and cluster configurations within existing governance units.

With the rapid development of urban agglomerations and metropolitan areas, flows of population, goods, and information increasingly transcend administrative boundaries [4]. Research focusing on functional boundaries and cross-administrative spatial delineation has continued to deepen, promoting regional spatial analysis beyond existing administrative units toward the identification of actual interaction spaces. Yu et al. [13] introduced the concept of “cross-city communities,” demonstrating that interaction intensity between adjacent cities may, in some cases, exceed intra-city interactions, and arguing that such cross-boundary communities should be identified based on functional linkages to provide new spatial references for economic development and urban governance. Liu et al. [14] similarly suggested that metropolitan economic regions should be delineated according to the spatial scope of economic activities in order to reconcile administrative divisions with the trend toward economic integration. Building on this perspective, recent studies have employed fine-grained spatial units—such as grid cells, Voronoi polygons, and hexagonal tessellations—combined with flow data derived from truck trajectories, mobile phone signaling, passenger travel records, and social media activity. Using spatial interaction networks and community detection algorithms, these studies have identified cross-administrative functional communities, activity ranges, and their internal organizational structures [13,15,16,17]. Notably, although these studies differ in data types and methodological approaches, their findings consistently indicate that highly connected urban areas identified as communities within network space tend to exhibit strong geographical clustering. Such geographically cohesive regions are widely interpreted as manifestations of spatial proximity effects, whereby interaction intensity between regions typically decreases as geographical distance increases [9].

Although existing research has progressively shifted regional economic network analysis from “static interactions” toward the identification of “dynamic network structures” based on factor flows, and studies of urban spatial structure are increasingly oriented toward multi-scale relational and functional network analysis [18,19], several limitations remain in the comprehensive identification of regional economic spatial organization. First, urban network studies conducted at the administrative scale and functional space identification across administrative boundaries are often undertaken separately, lacking a unified dual-scale analytical framework. As a result, it is difficult to simultaneously capture the urban network structure within existing governance units and the functional spatial structures shaped by actual factor flows. Second, although some scholars have employed big data sources—such as mobile phone signaling data, social media check-ins, and passenger transport schedules—to analyze flows of people and information, these datasets primarily reflect patterns of social interaction and commuting behavior rather than real industrial and economic linkages. Applications based on freight data, which more directly represent the operation of the real economy through freight flows and the flow of goods, remain relatively limited. Third, in measuring economic linkages, many existing studies still rely heavily on macro-level statistical indicators such as population and gross domestic product (GDP) to construct Gravity Models or related measurement frameworks. While such approaches effectively characterize potential linkages between cities, they pay insufficient attention to the characteristics of actual factor flows. Therefore, developing a unified analytical framework that integrates administrative-scale and cross-administrative-scale analysis, and incorporating freight flow data that better represent the operational processes of the real economy, remains an important direction for further research. Such an approach would enable the identification of regional economic spatial organization from the dual perspectives of potential linkages and actual flows.

At present, regional economic development in Hubei Province exhibits a pronounced imbalance. At the provincial level, the economic output of the Wuhan Metropolitan Area reached 3.6 trillion yuan in 2024, accounting for more than 60% of the total economic output of Hubei Province. Within the Wuhan Metropolitan Area, Wuhan alone generates a GDP exceeding the combined total of all other prefecture-level cities, indicating a highly uneven regional development pattern. On this basis, this study takes Hubei Province as the case study and focuses on the core issue of provincial economic spatial organization by constructing a dual-scale analytical framework that integrates the administrative scale and the grid scale. At the administrative scale, truck GPS trajectory data are employed to build a “point–line–area” urban network analytical framework. From the perspective of freight flows, urban nodes at the county and district levels are classified hierarchically, inter-city economic linkage intensity is measured, and urban clusters characterized by dense economic interactions are identified. On this basis, an urban economic network topology model of Hubei Province is constructed to reveal the hierarchical structure, linkage configuration, and spatial distribution patterns of the provincial economic network. At the grid scale, this study applies a 5 km × 5 km grid-based modeling approach in combination with the Louvain community detection algorithm to identify urban economic communities beyond administrative boundary constraints. This enables the delineation of the actual spatial scope of economic activities and their internal functional linkage structures. Furthermore, the results are analyzed in relation to the existing regional planning framework of Hubei Province through a coupling analysis. Specifically, this study seeks to address three research questions: (1) At the administrative scale, what hierarchical patterns, linkage configurations, and cluster structures characterize the urban economic network of Hubei Province? (2) At the cross-administrative scale, what are the spatial extent and internal structural features of functional economic communities organized by freight flows? (3) How do the dual-scale identification results correspond to the existing regional planning framework of Hubei Province, and what implications can be derived for planning optimization?

2. Data and Methods

2.1. Study Area

This study takes the entire territory of Hubei Province as the study area, encompassing 12 prefecture-level cities, one autonomous prefecture, and four province-administered county-level administrative units. The prefecture-level cities include Wuhan, Huangshi, Huanggang, Ezhou, Xiaogan, Xianning, Suizhou, Jingmen, Jingzhou, Shiyan, Yichang, and Xiangyang, along with the Enshi Tujia and Miao Autonomous Prefecture. The four province-administered county-level units are Xiantao, Qianjiang, Tianmen, and the Shennongjia Forestry District.

In the administrative-scale analysis, all counties, districts, and county-level cities within the jurisdiction of each city in the study area are treated as independent urban nodes, resulting in a total of 87 spatial units (Figure 1). These units are used to characterize the economic network structure among administrative units within Hubei Province. Furthermore, to overcome the limitations imposed by administrative boundaries on spatial identification, this study constructs a standardized 5 km × 5 km grid system [20] and adopts it as the basic spatial unit for identifying functional economic communities at the grid scale. This approach enables the detection and analysis of cross-administrative economic linkage patterns. Considering the spatial accuracy of truck GPS trajectory data, the need to capture sub-city functional structures, and the balance between analytical resolution and computational feasibility, this resolution was selected as the unified unit of analysis. The study area was divided into 7878 grid cells in total, among which 6352 contained valid freight flow records. Grid cells intersecting the provincial boundary were retained in the analysis.

2.2. Research Framework

This study draws upon established methodologies in urban network analysis and regional spatial structure research to examine the economic spatial pattern of Hubei Province from both the administrative and cross-administrative scales (Figure 2). At the data processing stage, inter-city and inter-grid origin–destination (OD) linkages are first extracted from truck GPS trajectory data, providing the foundational dataset for subsequent network analysis based on freight flows. At the administrative scale, the analysis is structured around a “point–line–area” framework. First, urban nodes are hierarchically classified using degree centrality, alter-based centrality, and alter-based power, and a composite node classification is calculated to identify the hierarchical characteristics of the urban network in Hubei Province. Second, by integrating the Gravity Model and the freight linkage intensity index, inter-city economic linkage intensity is measured from the dual perspectives of potential linkages and actual flows, and a Composite Linkage Intensity classification is conducted to reveal the spatial characteristics of economic linkage strength at the provincial scale. Third, a community detection algorithm is employed to identify regions of dense internal connections within the urban network, thereby delineating urban cluster structures. On this basis, an urban network topology model of Hubei Province is constructed to present the overall configuration of provincial economic linkages. At the grid scale, 5 km × 5 km grid-based modeling is used to delineate the boundaries of urban economic communities and to identify cross-administrative economic spatial structures in Hubei Province. The results are subsequently subjected to a coupling analysis with the existing regional planning framework of Hubei Province, providing empirical support for regional coordinated development and territorial spatial planning optimization. All data processing, spatial analysis, and visualization procedures are implemented using the Python 3.11 programming language and the ArcGIS 10.8 Desktop software suite.

2.3. Research Data

In recent years, with advances in big data acquisition and analytical techniques, an increasing number of studies have employed mobile phone signaling data and taxi positioning data to identify population and passenger flows for urban network analysis [19,21,22]. However, the core of regional integration lies in the deep integration of industrial and supply chains. Inter-city freight linkages represent interactions among cities in production, transportation, and consumption activities, and directly map industrial linkages and material exchanges [23]. Therefore, freight data are strongly correlated with inter-city economic linkages. In Hubei Province, freight transportation is dominated by road-based modes, and thus truck GPS trajectory data can largely reflect the spatial pattern of the flow of goods among cities within the province. Truck GPS trajectory data refer to vehicle movement records collected by onboard GPS navigation systems installed in freight trucks. These data contain multidimensional information, including travel time, spatial location, speed, and vehicle identifiers, and are characterized by high positional accuracy and spatiotemporal continuity. As such, truck GPS trajectory data are capable of accurately capturing actual freight interactions between cities [24]. According to the Regulations on Dynamic Supervision and Administration of Road Transport Vehicles, heavy freight trucks engaged in road transport (with a total weight of 12 tons or above) are required to install standardized satellite positioning devices and connect to regulatory platforms. Therefore, the truck GPS trajectory dataset used in this study is supported by a relatively robust sample base.

Existing studies have demonstrated that truck GPS trajectory data have been widely applied in freight flow and volume estimation [25,26], transportation behavior identification [27], urban network structure analysis [28], and trajectory-based energy consumption and emission estimation [29], thereby providing new empirical data sources for understanding inter-city freight flows activities and economic linkages. However, current research has primarily focused on methodological issues related to the mining and processing of truck GPS trajectory data [30] and on descriptive profiling of freight activity characteristics [31], while relatively few studies have extended their application to the analysis of inter-city economic linkages.

Prior to analysis, the freight trajectory data were cleaned and screened, including the removal of invalid positioning records, duplicate entries, and spatiotemporally inconsistent records. Records with coordinates outside the administrative boundary of Hubei Province were excluded, and vehicle samples were identified and de-duplicated based on unique vehicle identifiers. After cleaning, only truck samples forming valid trajectory chains within Hubei Province were retained. The final dataset includes an average of approximately 168,000 monitored heavy freight trucks per day within the province. The vehicle types mainly consist of heavy semi-trailer tractors, heavy dump trucks, and other heavy-duty freight vehicles. Onboard satellite positioning devices upload data approximately every 30 s, including vehicle ID, vehicle type, spatiotemporal location information, and operational status. The total data volume is approximately 1143 GB, and key data fields are presented in

Table 1. Overview of Truck GPS Trajectory Data.

Data Field	Example
Vehicle ID	10468704
Longitude	114.06114°
Latitude	30.6179°
GPS Timestamp	15 March 2024 16:54:48
Truck Type	Heavy Semi-Trailer Tractor
Mileage	30,820 km

. The selected data cover one week in March, June, September, and December 2024, representing different quarters of the year and capturing the regular characteristics of provincial freight flows. A rule-based identification approach based on travel distance and dwell time is employed to extract truck stop points and trip origin–destination (OD) information. The specific procedures are as follows.

• Stop point identification: For each truck, trajectory records within the same day are first sorted by timestamp. The time difference between two consecutive records is defined as the dwell time, while the mileage difference represents the travel distance. Records in which the mileage remains unchanged are used to identify candidate stop points, with the first and last records of each stationary segment extracted as potential stop locations.
• Removal of short-duration stops: To exclude brief stops caused by traffic signals, short breaks, or refueling, candidate stop points with dwell times shorter than a predefined time threshold are removed.
• Merging of consecutive stops: During loading and unloading operations, trucks may move short distances within freight yards. Therefore, candidate stop points separated by travel distances smaller than a predefined distance threshold are merged and treated as a single continuous stop.
• Generation of trip OD records: Based on the final set of identified stop points, OD records are extracted for each vehicle. Each OD record includes the geographic coordinates of the origin and destination, departure and arrival times, travel distance, and travel duration, from which average travel speed can be calculated.

To improve the robustness of stop point identification, this study first conducted a statistical analysis of the dwell duration and displacement distance of candidate stop points to identify concentration intervals associated with short-duration stops and small-range movements. On this basis, threshold values were calibrated by considering the typical duration of freight loading and unloading activities, waiting operations, and the spatial scale of activities within large logistics hubs. Balancing the completeness of stop detection with the control of misclassification, this study ultimately adopted 800 m and 15 min as the thresholds for identifying truck stop points [32]. Using the Python programming language, origin (O) and destination (D) points were extracted from the truck GPS trajectory data and organized into an OD pair table, yielding a total of 396,460 valid OD records.

Based on the OD data described above, freight linkage statistics are conducted at both the administrative scale and the grid scale. At the administrative scale, OD link data are aggregated according to administrative boundaries to obtain the freight linkage volume among prefectural-level cities in Hubei Province (Figure 3), providing the computational basis for subsequent urban network analysis. At the grid scale, this study further adopts a 5 km × 5 km grid-based modeling approach to partition the entire province into spatial units, with each grid cell treated as an independent analytical unit. Grid cells are defined as network nodes, while inter-grid OD flows extracted from freight trajectories are used as network edges. Freight flow volumes between different grid cells are calculated, and a directed edge is established whenever freight movement occurs between two grid cells, with the frequency of freight trips assigned as the edge weight. In this manner, a grid-based weighted and directed freight linkage network is constructed to characterize the economic spatial organization shaped by freight flows across administrative boundaries.

In addition, the socio-economic statistical data used in this study are primarily obtained from the Hubei Statistical Yearbook and statistical yearbooks of individual prefecture-level cities, as well as publicly released data from local statistical authorities. These data cover both prefecture-level and county-level administrative units in Hubei Province and include resident population figures (based on the Seventh National Population Census) and gross domestic product (GDP), which are used for measuring economic linkages and interpreting analytical results.

2.4. Research Methods

2.4.1. Centrality Measures

Centrality analysis is an effective quantitative approach for identifying and evaluating nodes in urban network studies, providing essential support for examining hierarchical structures and interaction patterns within urban networks. Considering the characteristics of the OD flow data, this study employs three centrality indicators: degree centrality [33], alter-based centrality, and alter-based power [34,35]. Centrality indicators reflect the unified processes of resource agglomeration and diffusion, such that cities with higher centrality tend to exhibit stronger capacities for both accumulation and outward dissemination of resources [36]. Alter-based centrality measures the combined ability of a city to attract and disperse resources and essentially represents the level of a network city’s participation in resource exchanges. Compared with degree centrality, alter-based centrality more effectively captures a city’s degree of involvement in regional economic activities, resulting in substantial differences between economically active and inactive cities. The alter-based power indicator represents a city’s influence in resource exchange processes and is closely associated with its position and functional role within the urban network. Alter-based power is less sensitive to absolute flow volumes; instead, it captures the capacity to control resources, reflecting the extent to which core cities exert influence over other interconnected cities in the network.

The calculation methods for the three indicators are as follows:

$$N_j={\sum }_i{R}_{ij}$$

(1)

$${RC}_j={\sum }_j{R}_{ij}\times {DC}_j$$

(2)

$${RP}_i={\sum }_j{\displaystyle \frac{R_{ij}}{{DC}_j}}$$

(3)

where $N_j$ denotes degree centrality, $R_{ij}$ represents the freight linkage volume between city $i$ and city $j$; ${RC}_j$ denotes alter-based centrality, ${DC}_j$ represents the degree centrality of city $j$; and ${RP}_i$ denotes alter-based power.

2.4.2. Measurement of Economic Linkage Intensity

In studies of urban networks and regional economic spatial linkages, both domestically and internationally, gravity models have been widely employed to quantify potential inter-city economic linkages using variables such as economic output, population size, and spatial distance [11,37]. This approach effectively reveals the spatial interaction and radiation effects of economic factors within urban systems at the macro scale. Such results tend to represent potential inter-city linkages rather than accurately capturing the characteristics of actual factor flows.

With advances in big data technologies, measurement approaches based on observed flow data—such as transportation and freight flows—have increasingly been adopted to assess inter-city linkages. These flow-based methods provide a more direct representation of actual economic interactions between cities and offer significant advantages in terms of spatiotemporal resolution and empirical interpretability. Nevertheless, such data primarily describe the manifested layer of factor movements, while the underlying structural economic relationships involving industries, capital, and information flows are not fully captured.

Taking into account the respective strengths and limitations of theoretical models and empirical flow data, this study integrates a gravity model with a freight linkage–based intensity index to measure and classify inter-city economic linkages in Hubei Province from two complementary dimensions: potential linkages and actual flows. The specific methods are described as follows.

On the one hand, an improved gravity model is applied to estimate the idealized economic linkage intensity based on urban scale (Equation (4)):

$${GR}_{ij}={\displaystyle \frac{\sqrt{P_i{G}_i}\sqrt{P_j{G}_j}}{{d_{ij}}^2}}$$

(4)

In Equation (4), ${GR}_{ij}$ denotes the gravity-based economic linkage intensity between cities $i$ and $j$; $P_i$ and $P_j$ represent the resident population of the two cities, respectively; $G_i$ and $G_j$ denote the GDP of cities $i$ and $j$ respectively; $d_{ij}$ represents the geographical distance between cities $i$ and $j$.

On the other hand, a freight linkage intensity index is constructed. Given that inter-city freight interactions represent a direct manifestation of regional economic collaboration, this study develops a freight linkage intensity index (Equation (5)):

$${EC}_{ij}={\displaystyle \frac{{R_{ij}}^2}{{d_{ij}}^k\left(R_i+R_j\right)}}$$

(5)

In Equation (5), ${EC}_{ij}$ denotes the freight linkage intensity index between cities $i$ and $j$; $R_{ij}$ represents the freight linkage volume between cities $i$ and $j$; $R_i$ and $R_j$ denote the total freight trip generation of cities $i$ and $j$, respectively; $d_{ij}$ represents the geographical distance between cities $i$ and $j$; and $k$ is a distance decay adjustment parameter that represents the degree to which inter-city economic linkages decay with increasing spatial distance. In this study, k is set to 1.

Compared with traditional models, the construction of this index is primarily grounded in three theoretical considerations. First, drawing on urban scaling laws [38] and network externality theory [39], outputs and factor flows within urban systems often exhibit superlinear growth characteristics. Large-scale freight interactions imply deeply embedded supply-chain networks whose economic value demonstrates a multiplier effect. Accordingly, a squared term of the freight linkage volume $R_{ij}$ is introduced to compensate for the limitations of linear measurements. Second, drawing on the concept of network coupling [40,41], the term $R_i+R_j$ is introduced in the denominator for scale normalization. This adjustment is intended to mitigate the “scale masking effect” caused by the disproportionately large size of major cities and to emphasize the relative functional dependence between urban nodes. Finally, based on spatial interaction theory [42] and the iceberg transportation cost model [43], the distance term ${d_{ij}}^k$ is retained to capture the spatial constraints imposed by transportation costs and information frictions on economic linkages. This specification is consistent with the spatial distance-decay principle articulated in Tobler’s First Law of Geography [44]. In this study, k is set to 1 based on the following considerations. First, Hubei Province spans a considerable east–west distance, and inter-city distances vary substantially. An excessively strong distance decay effect would impose disproportionate penalties on geographically distant cities such as Yichang and Shiyan. Second, under a well-developed expressway network, road-based freight transport is more strongly influenced by accessibility conditions and time costs, and increases in geographical distance do not necessarily translate into proportional increases in transport costs. Therefore, a relatively moderate distance decay assumption is adopted in this study. By avoiding the subjectivity inherent in parameter selection in traditional models, this approach enables a more accurate representation of the actual intensity and structural characteristics of inter-city economic linkages from a space-of-flows perspective.

2.4.3. Community Detection Algorithm

This study employs community detection algorithms to delineate urban economic clusters and urban economic communities in Hubei Province. Community detection algorithm was initially proposed by Newman based on a greedy optimization approach [45]. By applying clustering techniques to complex networks, this method identifies densely connected subgroups, or communities, within a network. Owing to its effectiveness in revealing internal structural organization, community detection has become a widely used approach in both domestic and international studies of urban network structures [46,47].

A community refers to a subnetwork within a larger network in which connections among internal nodes are relatively dense, while connections with external nodes are comparatively sparse [48]. A key indicator for assessing the significance of community structure in a network is modularity, originally proposed by Newman et al. [45]. Through iterative optimization, groups that yield the maximum modularity value $Q$ are successively merged, until the entire network is aggregated into a single group. The hierarchical level corresponding to the maximum $Q$ value is then selected as the final community structure. Among various community detection approaches, this study adopts the Louvain algorithm, which was proposed by Blondel et al. [49]. Owing to its high computational efficiency and suitability for large-scale weighted networks, the Louvain algorithm has been widely applied in complex network analysis. The modularity $Q$ is calculated as shown in Equation (6):

$$Q={\displaystyle \frac{1}{2m}}{\sum }_{i,j}\left\{A_{ij}-{\displaystyle \frac{k_i{k}_j}{2m}}\right\}\delta \left(C_i,C_j\right)$$

(6)

In Equation (6), $A_{ij}$ denotes the weight of the edge connecting nodes $i$ and $j$; $k_i$ and $k_j$ represent the sum of the weights of all edges connected to nodes $i$ and $j$, respectively; $\delta \left(C_i,C_j\right)$ is an indicator function that equals 1 if nodes $i$ and $j$ belong to the same community and 0 otherwise; and $m$ denotes the maximum possible number of connections in the network. The modularity value $Q$ ranges between 0 and 1, with larger values indicating denser connections within communities and a more pronounced community structure.

2.4.4. Construction of the Urban Network Topology Model

Exploring the topological structure of urban agglomerations is a central topic in urban geography and regional planning research [50]. This approach has been widely applied to the identification of regional centers [51] and to the analysis of urban network structures [52]. Drawing on analytical perspectives from urban network topology, this study constructs a topological model of the economic network of Hubei Province, with the aim of revealing its overall structural characteristics and spatial connectivity patterns.

To systematically present the overall structure of the urban economic network of Hubei Province, this study integrates the results of centrality-based classification, inter-city economic linkage intensity, and urban economic cluster delineation to construct a provincial-scale urban economic network topology model based on a “point–line–area” framework. Within this framework, “points” are defined through comprehensive centrality analysis to identify core nodes at different hierarchical levels; “lines” are derived from linkage intensity analysis to extract inter-city connections of varying strengths, representing the degree of association between cities; and “areas” are delineated using community detection algorithms to reveal the aggregation characteristics of regional clusters. The specific construction process is described as follows:

(1)

Extraction of urban node hierarchy

All cities are treated as network nodes. Based on the comprehensive node classification results, cities ranked within the top three hierarchical levels are selected as the primary structural nodes, and municipal districts under the same jurisdiction are merged. Different symbols or graphical representations are then used to denote node hierarchies, thereby reflecting their relative importance and influence within the network.

(2)

Construction of the county-level economic linkage network

Economic linkages between urban nodes are regarded as network edges. Based on the integrated economic linkage intensity results, the dominant linkage path between each city and its primary connected city is extracted and incorporated as a connecting edge in the topological model. Variations in line width are used to represent differences in linkage intensity, providing an intuitive visualization of the strength of inter-city economic connections.

(3)

Construction of the prefecture-level economic linkage network

On this basis, the integrated economic linkage intensity between prefecture-level cities is further aggregated and calculated. For each prefecture-level city, its primary linkage counterpart is identified to construct a prefecture-level economic linkage network. Similarly, differences in line width are employed to illustrate variations in linkage intensity, thereby revealing dominant economic linkage relationships among prefecture-level cities.

(4)

Overlay of urban cluster spatial structure

Urban economic clusters identified using the Louvain community detection algorithm are overlaid as the areal layer of the model to delineate regional aggregation units characterized by dense economic linkages. Different urban clusters are distinguished by boundary lines, reflecting the functional affiliation and aggregation relationships among cities within the network.

This model provides an intuitive representation of the hierarchical structure of the provincial urban system, the configuration of inter-city economic linkages, and the internal spatial patterns of regional clusters, thereby offering a structured spatial cognition framework for subsequent urban network analysis, regional planning evaluation, and policy formulation.

3. Results and Analysis

3.1. Hierarchical Characteristics of Urban Nodes

This study calculates the values of degree centrality, alter-based centrality, and alter-based power for each city- and county-level spatial unit in Hubei Province. The computed results are then classified using the natural breaks (Jenks) classification method. Following existing studies, urban nodes within the Hubei urban network are categorized into four hierarchical levels [11].

3.1.1. Degree Centrality Classification

The classification results of city- and county-level spatial units in Hubei Province based on degree centrality are shown in Figure 4. A total of five nodes are classified into the first hierarchical level, including the central urban area of Wuhan, Dongxihu District, Caidian District, Jiangxia District, and Huangpi District. Among them, the central urban area of Wuhan exhibits the highest degree centrality value, followed by Dongxihu District.

From the perspective of degree centrality, the urban economic network of Hubei Province exhibits a core-dominated structure supported by multiple secondary nodes. At the provincial scale of Hubei, all five first-tier nodes are located within the administrative area of Wuhan, indicating that Wuhan occupies a prominent hub position in the provincial economic network and exhibits the strongest capacity for linkage agglomeration and network control. Other cities with relatively high degree centrality values include Jingshan, the urban districts of Xiangyang, and Zhongxiang, none of which are situated within the Wuhan Metropolitan Area. This pattern indicates that although Wuhan occupies the highest central position in the provincial urban network, its radiation and spillover effects on surrounding cities remain limited and have not reached an optimal level. The urban districts of Xiangyang, Yichang, Xian’an District, among others, are classified as second-level nodes and exhibit relatively high degree centrality compared with their neighboring cities, suggesting that they possess certain characteristics of regional centers. However, compared with the high level of concentration in Wuhan, these nodes still exhibit a substantial gap in network hierarchy. This indicates that the provincial economic linkage structure continues to display pronounced polarization characteristics. Wuhan’s capacity to organize and concentrate inter-city linkages far exceeds that of other cities, while the development of polycentric coordination across the province remains relatively limited.

At the scale of the Wuhan municipality, the urban network demonstrates a pronounced polycentric pattern: with the exception of Hannan District, degree centrality values across districts are relatively similar. This indicates that Wuhan is not dominated by a single central district; rather, it exhibits a network configuration supported by multiple strongly connected and functionally complementary nodes. When each administrative district of Wuhan is treated as an independent node, the results show that Dongxihu District, Caidian District, and Jiangxia District exhibit relatively high degree centrality values, whereas central districts such as Jianghan District rank lower. This result is broadly consistent with the spatial distribution of economic development zones, high-tech industrial parks, and modern logistics functional areas in Wuhan, indicating that the most active areas of economic linkages within the municipality have gradually expanded from the traditional central urban districts toward peripheral functional new zones.

3.1.2. Alter-Based Centrality Classification

The results of the alter-based centrality analysis are presented in Figure 5. Nodes classified into the first hierarchical level include the central urban area of Wuhan, Dongxihu District, Jiangxia District, Caidian District, and Huangpi District, which are similar to the results obtained from the degree centrality classification. Among these nodes, the central urban area of Wuhan exhibits the highest alter-based centrality value, followed by Dongxihu District.

Compared with the urban node classification based on degree centrality, the nodes classified into the first hierarchical level under alter-based centrality are identical. However, the number of second-tier nodes is significantly reduced, indicating that when the roles of nodes in transmitting and transforming linkages within the network are taken into account, high-level nodes in the economic network of Hubei Province become further concentrated in the core region. Except for Zhongxiang, all second-level nodes belong to the Wuhan Metropolitan Area, with Jingshan included as an observer city. This pattern indicates that economically active areas in Hubei Province are largely concentrated within the Wuhan Metropolitan Area. Within the Wuhan Metropolitan Area, cities located farther from Wuhan—such as Qianjiang and Xian’an District of Xianning—are classified as third-level nodes, whereas cities closer to Wuhan, including Hanchuan, Xiaonan District, and Yingcheng, are promoted to the second level. Notably, Hanchuan exhibits a higher alter-based centrality value than Xiaonan District, which serves as the administrative seat of Xiaogan. This indicates that node hierarchy is not determined solely by administrative rank but is more strongly influenced by spatial proximity to Wuhan and the intensity of economic linkages with it.

At the provincial scale, the urban network of Hubei Province exhibits a concentric diffusion structure centered on the central urban area of Wuhan, with urban hierarchy declining as spatial distance from the core increases. This pattern reflects a typical core–periphery structure. Most urban nodes located outside the Wuhan Metropolitan Area are classified into the third and fourth hierarchical levels. Notably, the urban districts of Xiangyang and Yichang—designated as regional central cities in the Hubei Territorial Spatial Plan (2021–2035) and the Outline of the 14th Five-Year Plan of Hubei Province—also exhibit lower alter-based centrality values than many counties and county-level cities within the Wuhan Metropolitan Area. This finding suggests that current economic linkages in Hubei Province remain highly dependent on geographic proximity, and that an efficient, network-based functional transmission mechanism capable of overcoming spatial distance has yet to be fully established.

From a regional perspective, the Wuhan Metropolitan Area currently exhibits a single-tier system characterized by Wuhan as the absolute core. The relatively active economic performance of surrounding cities is largely attributable to spillover effects generated by the excessive concentration of economic activities and industries in Wuhan, rather than to endogenous growth driven by inter-city synergy. In contrast, the agglomeration and coordinated development features of the Xiangyang Metropolitan Area and the Yijingjing Metropolitan Area remain insufficiently evident. The Wuhan Metropolitan Area Development Plan, officially approved in 2023, explicitly proposes strengthening Wuhan’s radiating role to promote the joint development of surrounding cities, accelerating infrastructure interconnectivity, fostering specialized industrial division of labor and collaboration, advancing the co-construction and sharing of public services, and establishing coordinated mechanisms for joint risk prevention and control, thereby improving the institutional framework for metropolitan integration. However, from the perspective of the actual economic hierarchical structure, the construction of a modern, networked metropolitan area remains a long-term and challenging task.

3.1.3. Alter-Based Power Classification

The classification results of urban nodes based on the alter-based power indicator are shown in Figure 6. Six nodes are classified into the first hierarchical level, including the central urban area of Wuhan, Dongxihu District, Caidian District, the urban districts of Xiangyang, Enshi City, and the urban districts of Yichang. Nodes classified into the second hierarchical level are more numerous and are mainly composed of prefecture-level Party committee and government seats, exhibiting an approximately even spatial distribution across the province.

Based on the hierarchical classification of urban nodes derived from the alter-based power indicator, the urban network of Hubei Province already exhibits pronounced polycentric characteristics. Unlike degree centrality and alter-based centrality, which primarily reflect the level of node activity and linkage transmission capacity, alter-based power places greater emphasis on a node’s ability to control surrounding resource exchanges and organize network connections. Accordingly, although Wuhan continues to dominate in terms of the overall volume of economic linkages and network activity at the provincial scale, multiple regional centers with relatively independent control capacity have emerged in Hubei Province in terms of linkage governance and regional organization. Yichang and Xiangyang, which are classified as first-level nodes, serve as the core cities of the “two wings” within the regional economic development pattern of “one core leading with two wings driving,” as proposed in the Outline of the 14th Five-Year Plan of Hubei Province. Among them, the urban districts of Xiangyang display the highest alter-based power value, exceeding those of all districts in Wuhan. In contrast, the urban districts of Yichang exhibit the lowest alter-based power value among first-level nodes, indicating a comparatively weaker capacity to dominate resource exchanges with surrounding cities relative to other first-level urban nodes.

It is noteworthy that Enshi City is classified into the third and fourth hierarchical levels in the degree centrality and alter-based centrality classifications, respectively, while it is elevated to the first hierarchical level in the alter-based power classification. This indicates that although the overall scale of Enshi’s inter-city economic activities is relatively small and its level of participation in resource exchanges within the surrounding urban network is extremely low, it exerts a disproportionately strong controlling influence over neighboring urban nodes. Its dominance over cities within its hinterland even exceeds that of Yichang, the regional center of southwestern Hubei, thereby forming a “gateway node” in the southwestern Hubei region. This pattern is broadly consistent with the population distribution and resource endowment of Enshi City and the Enshi Autonomous Prefecture.

3.1.4. Comprehensive Classification of Urban Nodes

Based on the hierarchical levels of urban nodes derived from degree centrality, alter-based centrality, and alter-based power, this study assigns scores to each urban node accordingly. The average of the three scores is then calculated to obtain a composite score for each city, resulting in a comprehensive classification of urban nodes in Hubei Province (Figure 7). All first-level urban nodes are concentrated within the administrative area of Wuhan, reflecting Wuhan’s absolute core position in the provincial economic network. The second hierarchical level comprises 13 urban nodes, including the urban districts of Xiangyang, Xinzhou District of Wuhan, Jingshan, Zhongxiang, Daye, Xiaonan District of Xiaogan, Huarong District of Ezhou, Macheng, Tianmen, the urban districts of Yichang, Dangyang, Hanchuan, and Shashi District of Jingzhou. This distribution indicates a locally polycentric pattern.

Overall, the hierarchical classification of urban economic network nodes in Hubei Province exhibits a core-dominated structure supported by multiple secondary nodes. As the core and leading city of provincial economic development, Wuhan demonstrates a pronounced polycentric pattern within its internal economic linkage network: with the exception of Hannan District and Xinzhou District, all districts of Wuhan are classified into the first hierarchical level. At the provincial scale, second-level nodes are mainly distributed across the Jianghan Plain, forming a spatial pattern characterized by “concentration in the central and eastern regions and dispersion in the western region.” This configuration indicates that regional economic development in Hubei Province has gradually evolved toward a networked structure driven by a primary core and supported by multiple nodes, which is of significant importance for improving the provincial spatial organization system and promoting regional coordinated development. Moreover, these nodes are predominantly prefecture-level administrative seats or economically important county-level cities, possessing strong inter-regional linkage capacities and functioning as key secondary hubs for receiving and transmitting economic flows within the provincial network. Notably, several county-level cities—such as Hanchuan and Daye—are classified into the second hierarchical level, suggesting that within the provincial economic network, industrial and logistics linkages have, to some extent, surpassed administrative hierarchy in shaping urban economic status.

3.2. Economic Linkage Intensity Analysis

This study applies two approaches—the gravity model and a freight linkage-based intensity index—to calculate the economic linkage intensity among 87 independent urban nodes across Hubei Province. The resulting linkage intensities are classified into hierarchical levels using the geometric interval classification method and visualized accordingly (Figure 8 and Figure 9). Furthermore, the linkage intensity results are standardized and adjusted to obtain a composite analytical outcome that more accurately reflects the actual pattern of economic linkages.

3.2.1. Linkage Intensity Analysis Based on the Gravity Model

The classification results of linkage intensity based on the gravity model indicate that the economic linkage network of Hubei Province exhibits a pronounced radial structure (Figure 8). Overall, the top five inter-city linkages in terms of intensity, ranked from highest to lowest, are as follows: the central urban area of Wuhan–Dongxihu District, the central urban area of Wuhan–Caidian District, the central urban area of Wuhan–Jiangxia District, the central urban area of Wuhan–Huangpi District, and the central urban area of Wuhan–Hanchuan City. Among the top 20 strongest linkages, 13 involve the central urban area of Wuhan, indicating that the central urban area of Wuhan (including the Optics Valley High-Tech Development Zone) occupies an overwhelmingly dominant position in the provincial economic network. This dominance can be attributed to the concentration of population and industrial resources in the central urban area of Wuhan, which generates strong attraction and agglomeration effects on surrounding urban nodes.

Beyond Wuhan’s role as the primary hub in the economic network, distinctive linkage patterns are also observed among other cities in Hubei Province. For example, Tianmen, Qianjiang, and Xiantao exhibit relatively dense economic linkages with one another, forming a localized radial structure. Similarly, cities within the Yichang-centered and Xiangyang-centered urban agglomerations display relatively strong inter-city linkages, forming localized outward-radiating structures centered on Yichang and Xiangyang, respectively. These patterns suggest that urban agglomerations centered on Yichang and Xiangyang have begun to take shape.

3.2.2. Linkage Intensity Analysis Based on Freight Linkage Volume

The results of economic linkage intensity measurement based on freight linkage volume are presented in Figure 9. Ranked in descending order of linkage intensity index values, the top five inter-city linkages are Huangshi urban districts–Daye, the central urban area of Wuhan–Jiangxia District, Dongbao District–Duodao District of Jingmen, the central urban area of Wuhan–Dongxihu District, and Dongxihu District–Caidian District of Wuhan. Among the top 20 strongest linkages, seven involve Wuhan, while the remaining strong linkages are distributed across multiple regional nodes throughout the province. Compared with the gravity model–based results, Wuhan continues to occupy a dominant position; however, the distribution of freight-based linkage intensity is more dispersed, indicating a tendency toward a multi-core development pattern in the provincial economic network.

From the perspective of freight interactions, the urban economic network of Hubei Province exhibits a polycentric structure dominated by Wuhan, with Yichang and Xiangyang serving as secondary centers. Compared with the gravity model–based results, the freight linkage intensity index reveals a more balanced pattern of inter-city economic linkages. Within the Wuhan Metropolitan Area, interactions among Wuhan’s districts and surrounding counties no longer display a strictly radial structure centered on the central urban area but instead form a network-like configuration with relatively high linkage intensity. This indicates that, in the actual process of freight flows, provincial economic linkages do not rely entirely on Wuhan as a single core but instead form a more networked interaction structure among multiple regional nodes. Although the Wuhan Metropolitan Area as a whole still exhibits a centripetal tendency centered on Wuhan, horizontal linkages among cities within the metropolitan area have been significantly strengthened.

In addition, the areas surrounding Yichang and Xiangyang display typical radial linkage patterns, forming regionally radiating structures centered on Yichang and Xiangyang, respectively. This reflects the emergence of these two cities as secondary cores within the provincial economic network. Notably, Shiyan and several western cities within the Xiangyang region (including Yunyang District–Shiyan–Danjiangkou–Laohekou–Gucheng) exhibit relatively low overall levels of economic activity and weak external linkages with other cities. Nevertheless, their internal connections are relatively strong, forming relatively independent local network units.

3.2.3. Composite Linkage Intensity Classification

To enhance the accuracy and comparability of economic linkage measurements between urban nodes, this study performs a comprehensive classification by integrating the results derived from the gravity model and the freight linkage intensity index. First, the potential economic linkage intensity obtained from the gravity model and the actual economic linkage intensity based on freight flows are standardized to eliminate differences in measurement scales. Second, the results of the two approaches are averaged to derive a composite linkage intensity between urban nodes in Hubei Province. Based on this composite measure, the inter-city linkage intensities are reclassified into hierarchical levels.

The correspondence between urban nodes in the top three hierarchical levels and their primary linked cities is presented in

Table 2. Primary linkages of urban nodes.

Urban Nodes	Primary Linked Urban Node	Composite Linkage Intensity	Urban Nodes	Primary Linked Urban Node	Composite Linkage Intensity
Central Urban Area of Wuhan	Dongxihu District	0.765331	Urban Districts of Yichang	Zhijiang City	0.119619
Dongxihu District	Central Urban Area of Wuhan	0.765331	Zhijiang City	Urban Districts of Yichang	0.119619
Jiangxia District	Central Urban Area of Wuhan	0.721134	Yingcheng City	Jingshan City	0.111736
Huangpi District	Central Urban Area of Wuhan	0.512278	Zaoyang City	Urban Districts of Xiangyang	0.107126
Caidian District	Central Urban Area of Wuhan	0.385399	Suixian County	Zaoyang City	0.104873
Daye City	Urban Districts of Huangshi	0.298973	Hannan District	Caidian District	0.095519
Urban Districts of Huangshi	Daye City	0.298973	Xian’an District	Tongshan County	0.085641
Xinzhou District	Macheng City	0.267004	Yidu City	Songzi City	0.077235
Macheng City	Xinzhou District	0.267004	Huarong District	Echeng District	0.077159
Urban Districts of Shiyan	Yunyang District	0.214275	Dangyang City	Urban Districts of Yichang	0.065529
Yunyang District	Urban Districts of Shiyan	0.214275	Shayang County	Dongbao District	0.062915
Echeng District	Huangzhou District	0.204213	Changyang Tujia Autonomous County	Yidu City	0.060759
Huangzhou District	Echeng District	0.204213	Gongan County	Songzi City	0.05822
Dongbao District	Duodao District	0.199984	Hanchuan City	Dongxihu District	0.055558
Duodao District	Dongbao District	0.199984	Enshi City	Lichuan City	0.053455
Jingshan City	Tianmen City	0.191665	Xiantao City	Jingshan City	0.046094
Tianmen City	Jingshan City	0.191665	Xiaonan District	Dongxihu District	0.045674
Urban Districts of Xiangyang	Yicheng City	0.178546	Qianjiang City	Shayang County	0.038571
Yicheng City	Urban Districts of Xiangyang	0.178546	Chibi City	Jiayu County	0.03251
Shashi District	Jingzhou District	0.152833	Jiayu County	Chibi City	0.03251
Danjiangkou City	Urban Districts of Shiyan	0.132848	Xishui County	Huangzhou District	0.0293
Zhongxiang City	Dongbao District	0.120354	Urban Districts of Yichang	Zhijiang City	0.119619

.

3.3. Urban Economic Cluster Delineation

To identify the spatial linkage characteristics of the urban economic network in Hubei Province, this study applies the Louvain community detection algorithm to delineate urban clusters based on composite economic linkage intensity. The algorithm iteratively optimizes modularity to obtain an optimal partition, with higher modularity values indicating denser internal connections and weaker external linkages within clusters. When the modularity value converges and becomes stable, the primary and secondary hierarchical cluster structures of the urban network can be determined. Such multi-level clustering facilitates a more in-depth examination of the internal economic interactions within each cluster.

3.3.1. Major Urban Economic Clusters

The results of the Louvain community detection algorithm indicate that the urban economic network of Hubei Province can be divided into six major urban economic clusters (Figure 10): the Wuhan–Macheng cluster, the Xiaogan cluster, the Xianning cluster, the Ezhou cluster, the Yichang cluster, and the Xiangyang–Shiyan–Suizhou cluster. Overall, this clustering pattern is broadly consistent with Hubei Province’s regional coordinated development strategy of “one core with two wings.” However, cities within the Wuhan Metropolitan Area do not form a single unified cluster; instead, they are subdivided into four distinct urban clusters. Only Macheng City and Hong’an County in Huanggang are grouped together with Wuhan, indicating that economic linkages within the Wuhan Metropolitan Area remain relatively fragmented. The Yichang cluster encompasses not only most areas of Yichang, Jingmen, Jingzhou, and Enshi, but also includes Shennongjia Forestry District and Qianjiang City. The Xiangyang–Shiyan–Suizhou cluster covers the Shiyan, Xiangyang, and Suizhou regions, suggesting that the actual structure of economic linkages deviates to some extent from the existing “Xiang–Shi–Sui–Shen” regional delineation.

From a spatial perspective, the boundaries of most clusters generally coincide with prefecture-level administrative boundaries. This is particularly evident in the Xianning cluster, whose boundaries fully overlap with the administrative area of Xianning and which exhibits the highest modularity value, reflecting the strongest internal economic linkages. Meanwhile, several county-level cities—such as Jingshan, Guangshui, and Honghu—are assigned to clusters that do not correspond to the dominant cluster of their respective prefecture-level cities, indicating a mismatch between administrative boundaries and functional economic clusters. In particular, Guangshui City has stronger economic ties with the Xiaogan region due to its long-term administrative affiliation with Xiaogan prior to the establishment of Suizhou as a prefecture-level city in 2000. Jingshan City, owing to its geographic proximity to the Wuhan Metropolitan Area, is classified into the Xiaogan cluster. These findings indicate that the formation of urban economic clusters is influenced not only by current administrative affiliations, but also closely associated with historical administrative linkages, spatial proximity conditions, and the intensity of actual economic interactions.

3.3.2. Secondary Urban Economic Clusters

Based on the six major urban economic clusters, a further subdivision yields sixteen secondary urban economic clusters (Figure 11). Overall, the Xianning cluster exhibits highly cohesive internal economic linkages and is therefore not further divided into secondary clusters. The Xiangyang–Shiyan–Suizhou cluster, the Xiaogan cluster, and the Wuhan–Macheng cluster are each subdivided into two secondary clusters, while the Ezhou cluster is divided into three secondary clusters. In contrast, the Yichang cluster displays substantial internal heterogeneity and is further divided into six secondary clusters.

From a spatial perspective, the delineation of secondary clusters in western Hubei largely corresponds to administrative boundaries. For instance, the clusters centered on Shiyan and Xiangyang generally coincide with their respective prefecture-level administrative jurisdictions, indicating relatively strong internal economic linkages. In contrast, the economic linkage patterns in central and eastern Hubei are more complex. Although Wuhan functions as the provincial core, its administrative districts are not entirely integrated into a single cluster. Xinzhou District forms a cross-boundary cluster together with Macheng and Hong’an in Huanggang, reflecting a certain degree of fragmentation in economic linkages within the Wuhan Metropolitan Area and suggesting that the spillover effects of the core city have not yet been fully realized. Regions such as Huanggang and Huangshi exhibit patterns characterized by the coexistence of multiple clusters and relatively weak inter-cluster linkages. From the perspective of the road freight–based urban network, these areas have yet to form stable urban agglomeration structures, further highlighting the tendency for economic linkages to transcend administrative boundaries in network terms.

Overall, the regional economic network of Hubei Province is characterized by strong internal linkages within clusters and weak linkages across clusters. While economic activities exhibit pronounced cluster-based spatial agglomeration, insufficient inter-cluster connectivity constrains, to some extent, the coordinated and integrated development of the provincial economic network.

3.4. Construction of the Urban Economic Network Topology Model of Hubei Province

In this study, cities are treated as network nodes, inter-city economic linkages as network edges, and urban economic clusters identified through community detection as the areal layer of the network. Based on a “point–line–area” framework, a topological model of the urban economic network is constructed to characterize the hierarchical relationships and spatial organization of the urban system in Hubei Province (Figure 12, where prefecture-level cities are highlighted in bold red). Specifically, the top three tiers of urban nodes identified through the comprehensive node classification are selected as the primary structural nodes, with municipal districts merged to reflect their relative influence within the network. Based on the composite economic linkage intensity, the dominant linkage between each city and its primary linked city is extracted and represented by edges with varying line widths to depict linkage strength, thereby constructing a county-level economic linkage network. These linkages are then aggregated to form the prefecture-level economic network structure. Finally, urban economic clusters identified using the Louvain community detection algorithm are overlaid as the areal layer of the network to reveal functionally cohesive spatial aggregation areas characterized by dense economic linkages.

From an overall structural perspective, the urban economic network of Hubei Province is characterized by single-core dominance and cluster-based organization. Wuhan occupies an absolute core position in the network and is the only first-tier city, exhibiting the strongest economic linkage intensity and the widest spatial influence at both prefecture-level and county-level scales. Provincial economic interactions are organized around several urban economic clusters, within which internal linkages are relatively strong, while inter-cluster connections remain limited. This indicates that a highly balanced polycentric network has yet to emerge in Hubei Province.

In terms of hierarchical distribution, the topological model clearly reveals a multi-level structure nested between prefecture-level and county-level cities. As the sole first-tier city, Wuhan not only maintains first-tier economic linkages with multiple prefecture-level cities but also directly attracts primary linkages from several county-level cities, demonstrating strong agglomeration and control capacity. Most prefecture-level cities—including Xiangyang, Yichang, Jingzhou, Xiaogan, Ezhou, and Huangshi—constitute second-tier nodes in the provincial network, functioning as regional hubs for the aggregation and redistribution of economic linkages. However, notable hierarchical mismatches are observed: some prefecture-level cities (e.g., Huanggang, Suizhou, and Shennongjia Forestry District) do not rank among the top three tiers, whereas several county-level cities (e.g., Macheng, Daye, Zhongxiang, and Jingshan) exhibit higher network positions than their administratively affiliated prefecture-level cities. This suggests that a city’s position within the provincial economic network is shaped less by administrative status than by its industrial base and the strength of its economic linkages.

From the perspective of economic connections, the topological model displays a pronounced vertical dependency pattern and relatively weak horizontal interactions. For most county-level cities, primary economic linkages are directed toward their corresponding prefecture-level cities, reflecting a typical bottom–up dependency structure. At the prefecture level, certain inter-city connections—such as Xiangyang–Shiyan, Yichang–Jingzhou–Jingmen, and Ezhou–Huangshi—are evident, but their overall intensity remains weaker than that of linkages between Wuhan and selected prefecture-level cities. A stable and tightly connected horizontal network among prefecture-level cities has therefore not yet formed. Within the central region, Jingmen and Jingzhou exhibit partial hub-like characteristics, while the linkage between Ezhou and Huangshi reaches the first-tier intensity level. In contrast, several county-level cities (e.g., Gongan, Changyang Tujia Autonomous County, and Yidu) display clear peripheral characteristics, with relatively weak economic linkages to other cities.

At the level of urban economic clusters, inter-cluster linkages are generally limited, indicating insufficient cross-cluster coordination. The Wuhan–Macheng cluster and the Xiaogan cluster exhibit the strongest inter-cluster connections, forming a network involving multiple cities. By contrast, the Ezhou cluster remains relatively isolated within the provincial network, lacking direct connections with other clusters. Most remaining clusters are connected only through single linkage pathways, resulting in weak overall inter-cluster integration. Notably, although Yichang is designated as a regional sub-center in provincial planning, its cluster does not exhibit strong cross-regional spillover effects in the topological structure; instead, its external connections rely primarily on Jingmen as an intermediary node. Furthermore, urban economic clusters display clear cross-administrative characteristics—for example, Jingshan is incorporated into the Xiaogan cluster and maintains relatively strong economic linkages with multiple cities within the cluster.

3.5. Urban Economic Community Identification

Based on grid-based modeling and community detection, this study constructs a freight-based economic linkage network for Hubei Province using 5 km × 5 km grid cells. The Louvain community detection algorithm is applied to identify network communities through iterative modularity optimization, yielding partitions with maximized modularity. This approach produces a classification of urban economic communities grounded in actual economic linkages, thereby revealing spatial organization patterns that transcend administrative boundaries. The visualization results (Figure 13) identify ten major urban economic communities. Their boundaries differ substantially from prefecture-level administrative divisions but show a high degree of consistency with county-level boundaries, indicating that cities in Hubei Province primarily participate in provincial economic activities at the county scale.

From an overall perspective, urban economic communities in Hubei Province exhibit pronounced cross-boundary reconfiguration characteristics. On the one hand, some communities clearly transcend existing prefectural-level administrative boundaries, forming continuous functional spaces composed of multiple adjacent areas. For example, Community 5 (Wuhan–Hanchuan), Community 8 (Northern Huanggang–Xinzhou–Huangpi), and Community 9 (Eastern Huanggang–Huangshi–Ezhou) each span two or more prefectural-level cities, indicating that actual economic linkages have formed relatively stable cross-boundary organizational units in several regions. On the other hand, such cross-boundary integration does not occur in a random or unstructured manner; rather, it follows patterns of spatial proximity, transportation corridors, and industrial linkages, demonstrating clear directionality and selectivity. This suggests that cross-administrative urban economic communities in Hubei Province do not simply represent a mechanical break from administrative boundaries, but instead constitute functional spatial units more closely aligned with the underlying logic of economic operations shaped by sustained freight flows.

The identified urban economic communities also exhibit evident concentric differentiation in spatial scale: communities located closer to Wuhan tend to be smaller in size, whereas those farther from Wuhan cover broader spatial areas. This pattern indicates that freight linkages in the vicinity of Wuhan are denser and more multidirectional, and that economic spatial organization in these areas has evolved toward a relatively high degree of subdivision and networkization, thereby facilitating the formation of multiple small-scale functional communities. In contrast, peripheral regions generally display weaker inter-node linkages and lower internal heterogeneity, making it more likely for communities to merge into larger spatial units. Essentially, this phenomenon reflects a differentiated provincial economic spatial structure characterized by a declining gradient from the Wuhan core toward the periphery.

Further examination reveals that cross-administrative urban economic communities in Hubei Province can be categorized into three typical types:

(1)

Core city spillover-driven communities.

Represented by Community 5 (Wuhan–Hanchuan) and Community 8 (Northern Huanggang–Xinzhou–Huangpi), and also including Community 1 (Xianning–Eastern Honghu), Community 3 (Xiaogan–Tianmen–Xiantao–Guangshui–Jingshan), and Community 9 (Eastern Huanggang–Huangshi–Ezhou), this type typically relies on a core city or high-hierarchy node and expands continuously into adjacent areas, exhibiting strong cross-boundary integration characteristics.

(2)

Regionally coordinated integration communities.

Including Community 4 (Jingmen–Northern Jingzhou–Qianjiang–Dangyang–Zhijiang) and Community 7 (Western Suizhou–Zaoyang), these communities are not entirely dependent on a single core city but are instead formed through relatively strong horizontal linkages among multiple neighboring nodes, reflecting a certain degree of regional coordination and integration. Although some of these areas are administratively designated as part of the Wuhan Metropolitan Area, differences in spatial proximity and actual linkage directions with Wuhan have resulted in relatively independent economic communities.

(3)

Transport–geography constrained communities.

Including Community 2 (Wufeng–Yidu–Songzi–Gongan–Shishou), Community 6 (Central–Western Yichang–Enshi–Shennongjia), and Community 10 (Xiangyang–Shiyan), this type is spatially influenced by topography and transportation conditions, particularly in southwestern Hubei and mountainous areas. Communities 6 and 10, located in mountainous regions, tend to form relatively cohesive units at larger spatial scales due to terrain constraints and limited accessibility. Community 2 forms an elongated belt along provincial highways S322, S242, and S325, spanning Yichang, Jingzhou, and Enshi, highlighting the guiding role of transportation corridors in shaping economic linkages.

Meanwhile, significant differences exist among communities in terms of internal economic cohesion and network integration. For instance, Community 1 (Xianning–Eastern Honghu) exhibits the highest modularity value, indicating the strongest internal economic cohesion. In contrast, Community 10 (Xiangyang–Shiyan), despite covering a relatively large area, has the lowest modularity, suggesting that the middle and upper reaches of the Han River have not yet formed a highly cohesive and integrated cross-boundary functional space; instead, its community structure reflects relative aggregation among geographically adjacent areas rather than a mature and stable high-intensity economic community.

4. Discussion

4.1. Structural Characteristics of the Urban Economic Network at the Administrative Scale

By integrating the results of urban node classification, economic linkage intensity measurement, cluster delineation, and topological modeling, this study finds that the urban economic network of Hubei Province overall exhibits a core-dominated structure supported by multiple nodes.

At the provincial scale, the regional development strategy of “one core with two wings” has begun to manifest in the economic network structure. According to the Hubei Territorial Spatial Plan (2021–2035), Wuhan is designated as the provincial core city, while Xiangyang and Yichang serve as sub-centers, jointly forming the spatial framework of “one core with two wings” (Figure 14). The results of this study indicate that both Xiangyang and Yichang have developed independent urban economic clusters centered on themselves, with their network hierarchies significantly higher than those of other nodes within the same regions. The primary economic linkages of surrounding cities—such as Suizhou, western Jingmen, western Jingzhou, Shiyan, Enshi, and the Shennongjia Forestry District—are largely oriented toward these two cities. This suggests that Xiangyang and Yichang are gradually assuming their roles as regional sub-centers, although their spatial influence and radiating capacity remain to be further strengthened.

From a regional development perspective, inter-city economic linkages in Hubei Province exhibit pronounced spatial imbalance. Economic connections within the “1 + 8” Wuhan Metropolitan Area are relatively strong, whereas linkages with central and western regions are considerably weaker. The radiating and driving roles of Xiangyang and Yichang within their respective regions have not yet been fully realized. As the core area under the “one core leading” strategy, the Wuhan Metropolitan Area demonstrates the highest level of economic activity; however, horizontal linkages among its constituent cities remain relatively weak. Economic activities are highly concentrated in Wuhan, and with the exception of geographically proximate cities such as Hanchuan, most member cities exhibit predominantly one-directional dependency on Wuhan rather than forming an effective multi-node collaborative network. This pattern indicates that internal coordination mechanisms within the Wuhan Metropolitan Area require further strengthening, and that a more interactive economic linkage structure is urgently needed to alleviate the structural imbalance characterized by an overly dominant core and dependent peripheries.

From a spatial planning perspective, a certain degree of misalignment remains between the structure of regional economic linkages and existing administrative divisions. Community detection results indicate that the Wuhan Metropolitan Area is subdivided into several relatively independent urban economic clusters, while Jingmen and Jingzhou exhibit pronounced east–west differentiation, with parts of their eastern areas already integrated into the Wuhan Metropolitan Area. In addition, the Shennongjia Forestry District demonstrates stronger economic linkages with Yichang. These findings suggest that the overall spatial organization of Hubei Province generally aligns with the strategic framework of “one core with two wings,” yet further optimization is needed in terms of internal hierarchical coordination and inter-cluster connectivity. Future spatial planning should strengthen connectivity between the core city and regional sub-centers, thereby promoting functional coordination and the integrated development of the provincial economic network.

4.2. Characteristics of Urban Economic Communities at the Cross-Administrative Scale

By comparing the identified urban economic communities with the “one core with two wings” regional development framework of Hubei Province (Figure 15), a high degree of consistency is observed between the two in terms of their overall spatial configuration. Both delineate regional structures centered on Wuhan, Xiangyang, and Yichang. Wuhan occupies the core position within the provincial network, exhibiting a pronounced primary-center role, while Xiangyang and Yichang function as secondary centers supporting the northwestern and southwestern wings of the province, respectively. Together, they constitute key pillars of Hubei’s emerging multi-core regional structure.

The coupling analysis indicates that although Hubei Province’s current regional planning has achieved certain progress in implementation, discrepancies remain between actual economic linkages and planned spatial structures. These discrepancies are mainly reflected in the following aspect:

(1)

The internal evolution of the Wuhan Metropolitan Area has not yet reached the level of coordination envisioned in the planning framework.

The plan positions the Wuhan Metropolitan Area as the core region under the “one core leading” strategy, aiming to establish an integrated spatial structure centered on Wuhan with coordinated development among member cities. However, the urban economic community delineation reveals that the Wuhan Metropolitan Area has not formed a tightly integrated whole. Instead, it exhibits a multi-area configuration composed of the “Wuhan–Hanchuan” core community alongside several eastern, southern, western, and northern communities. Nodes such as Xian’an District and Ezhou play pronounced gateway roles in resource exchange and factor flows, while some cities in the Xiaogan area maintain predominantly one-directional dependency on Wuhan, with limited horizontal interaction. This pattern suggests that economic linkages within the Wuhan Metropolitan Area remain fragmented, with Wuhan functioning as the sole dominant hub and insufficient horizontal connectivity and coordination among other nodes. Consequently, a gap persists between actual economic interactions and the planning objectives of internal coordination and functional complementarity. Future efforts should therefore place greater emphasis on strengthening horizontal linkages and functional integration among cities surrounding Wuhan through enhanced metropolitan integration and cross-jurisdictional cooperation mechanisms.

(2)

Mismatches between actual economic linkage structures and planned regional affiliations in certain areas.

The results of urban economic community identification indicate that economic linkages in Hubei Province exhibit pronounced cross-administrative characteristics, with functional inter-city relationships partially transcending the administrative divisions defined in regional planning. In several cases, the primary directions of economic interaction for certain counties and cities are inconsistent with their designated regional affiliations under the current planning framework. For example, although Qianjiang is classified as part of the “1 + 8” Wuhan Metropolitan Area, its freight flows are more strongly integrated into the central–western Hubei economic network formed by Yichang, Jingmen, and Jingzhou. Similarly, Guangshui City in Suizhou and Jingshan City in Jingmen, which are designated as observer cities of the Wuhan Metropolitan Area, exhibit economic linkage structures that are more closely aligned with cities in the western part of the metropolitan area. In addition, although the Shennongjia Forestry District and Baokang County are included in the “Xiangyang–Shiyan–Suizhou–Shennongjia” region in planning documents, their economic linkages are more oriented toward the Yichang–Enshi corridor. These patterns reveal clear discrepancies between the actual economic networks of certain cities or counties and their planned regional affiliations, highlighting the tendency for functional economic spaces to transcend administrative boundaries in practice. Overall, while maintaining the overarching framework of the “one core with two wings” strategy, regional planning should more closely incorporate empirically observed economic linkage networks and further refine functional zoning at the county scale.

4.3. Dual-Scale Characteristics of Provincial Economic Spatial Organization in Hubei Province

Existing studies on urban networks at the administrative scale generally indicate that regional economic linkages tend to exhibit a hierarchical structure dominated by high-level central cities and supported by secondary centers [2,18]. In contrast, research on functional boundaries and cross-city communities suggests that actual interaction spaces do not always conform to administrative boundaries but may form cross-boundary functional units shaped by spatial proximity, transportation connectivity, and factor flows [15,19]. The findings of this study are broadly consistent with these two strands of research. On the one hand, the economic spatial structure of Hubei Province displays a clearly defined network configuration in which Wuhan functions as a dominant core, while Xiangyang and Yichang emerge as regional secondary centers. This aligns with the general conclusion in the literature that regional economic networks are characterized by core dominance supported by secondary nodes. On the other hand, the urban economic communities identified in this study transcend certain existing administrative boundaries, indicating that actual economic activities in Hubei are not entirely confined to institutional space, but instead reorganize across boundaries along directions defined by spatial proximity and transportation linkages. This observation echoes findings from functional boundary and cross-city community research. At the same time, some differences are evident. Compared with the more pronounced polycentric coordination patterns observed in regions such as the Yangtze River Economic Belt and the Pearl River Delta [5], the provincial economic spatial organization of Hubei—although exhibiting a “one core with multiple nodes” network structure—remains predominantly dominated by a single core centered on Wuhan. The secondary center roles of Xiangyang and Yichang have not yet fully matured. This suggests that the level of network balance and regional coordination in Hubei Province remains weaker than that of more developed urban agglomeration regions.

By integrating the results from both administrative-scale and cross-administrative-scale analyses, it becomes evident that the provincial economic spatial organization of Hubei exhibits pronounced dual-scale characteristics. At the administrative scale, it presents a relatively stable hierarchical network structure. At the cross-administrative scale, it manifests as economic communities that transcend administrative boundaries. The former primarily reflects urban hierarchy, linkage structures, and regional center systems within institutional space, corresponding to the hierarchical order emphasized in space of place theory [6]. The latter reveals cross-boundary economic communities and functional boundaries shaped by actual freight flows, aligning more closely with the factor-linkage structures emphasized in space of flows theory [7]. Together, these two dimensions reflect the coexistence and interaction of space of place and space of flows as articulated in central flow theory [8]. This dual-scale perspective provides empirical evidence for understanding the interaction between institutional space and functional space at the provincial scale and further reveals the complexity and dynamic nature of economic spatial organization in Hubei Province.

5. Conclusions and Implications

5.1. Provincial Economic Spatial Optimization and Planning Recommendations

5.1.1. Optimizing the Provincial Economic Spatial Structure: Building a Polycentric and Coordinated Provincial Economic Linkage Network

A synthesis of this study’s analysis of the urban economic network structure and functional community characteristics in Hubei Province indicates that the provincial economic spatial organization has generally formed a basic configuration centered on Wuhan, with Xiangyang and Yichang serving as regional sub-centers. However, several issues remain, including the overwhelming dominance of Wuhan, the high degree of dependence of surrounding cities on Wuhan, and the insufficient radiating and driving effects of the secondary centers. Going forward, within the overarching framework of “one core with two wings,” it is necessary to promote a transition from a single-core radial pattern toward a polycentric and coordinated network structure.

First, efforts should be made to strengthen the development of the Wuhan Metropolitan Area and enhance internal connectivity and coordinated growth within the metropolitan region. While maintaining Wuhan’s leading role as the core city, greater emphasis should be placed on reinforcing horizontal linkages among node cities within the metropolitan area, thereby facilitating broader circulation and reorganization of factors. Through the improvement of inter-city transportation networks, the strengthening of industrial chain collaboration, and the joint development of logistics nodes, a networked economic linkage structure centered on Wuhan and supported by multiple interconnected nodes can gradually be established, enabling the Wuhan Metropolitan Area to better drive the overall development of eastern Hubei.

Second, greater emphasis should be placed on leveraging the radiating and driving roles of Yichang and Xiangyang as provincial sub-centers. In the Xiangyang–Shiyan–Suizhou region, priority should be given to removing bottlenecks along the Shiyan–Xiangyang–Suizhou corridor, strengthening economic linkages between Xiangyang and Shiyan, and fostering an urban economic belt along the Han River. In the Yichang-centered urban agglomeration, Yichang’s role as a strategic gateway connecting central and western China should be reinforced, the development of the Yichang–Jingzhou–Jingmen–Enshi urban agglomeration should be accelerated, and economic linkages with other cities in Hubei—particularly Wuhan—should be strengthened. In addition, consideration could be given to upgrading the functional status of central Hubei cities such as Jingmen and Zhongxiang, cultivating new growth poles in the provincial core area, and ultimately forming a polycentric, coordinated provincial spatial network.

5.1.2. Reorienting Regional Economic Planning in Hubei Province: Strengthening the Alignment Between Administrative Space and Functional Space

A synthesis of the analyses of the urban economic network structure and functional communities in Hubei Province indicates that, although the current provincial spatial pattern broadly aligns with the strategic vision of “one core with two wings and province-wide coordination,” notable gaps remain between actual economic linkage networks and planning expectations. To accommodate the ongoing trends of networked and polycentric economic development, as well as cross-boundary reconfiguration of economic activities, regional economic planning in Hubei Province requires further optimization in terms of spatial delineation, functional positioning, and coordination mechanisms, in order to enhance the alignment between administrative space and functional space.

First, planning concepts should be upgraded. While respecting existing administrative divisions and the overall planning framework, economic spaces formed through factor flows should be incorporated into the processes of plan formulation and implementation evaluation. Second, the functional hierarchy of provincial space should be further refined by constructing a multi-level network structure in which multiple centers, economic linkage corridors, and cross-boundary communities are mutually integrated. Finally, cross-regional coordination mechanisms for planning and governance should be established. By improving the provincial spatial governance system and promoting infrastructure interconnectivity, coordinated industrial chain development, and the co-construction and sharing of public services, closer alignment can be achieved between administrative boundaries and functional economic spaces. Such efforts would contribute to the formation of an open, balanced, and efficient provincial economic spatial structure.

5.1.3. Strengthening Implementation Mechanisms: Promoting Policy Execution Through Key Corridors and Cross-Boundary Collaborative Units

(1) Single-core upgrading: enhancing the carrying capacity of Wuhan’s outer zones

To alleviate the excessive concentration of population, land use, and logistics activities in Wuhan’s central urban area, logistics hubs should be gradually relocated toward the outer ring zones (the Fourth Ring Road and the outer ring). By developing a logistics buffer belt around Wuhan—centered on Dongxihu, Jiangxia, and Huangpi—a “logistics defensive ring” can be formed to redistribute freight flows and effectively reduce transportation and logistics pressures in the urban core.

(2) Densifying three strategic corridors: constructing a multi-dimensional provincial logistics backbone system

• Han River economic corridor.

To address weak linkages between Xiangyang and Shiyan, digital logistics platforms should be leveraged to integrate automobile parts logistics demand in both cities. The introduction of “Xiangyang–Shiyan circular freight services” or shared dedicated freight lines would reduce empty-haul rates and substantively promote the integration of the industrial corridor.

• Yangtze River integrated logistics corridor.

Relying on the Yangtze River as a major inland waterway, intermodal water–road transportation along the Yichang–Jingzhou–Wuhan axis should be strengthened to improve the circulation efficiency of bulk commodities and reinforce the role of the river corridor as a provincial logistics backbone.

• Southern Hubei east–west transversal corridor.

To overcome the relative isolation of Xianning, priority should be given to the planning and construction of high-standard east–west highways linking Xianning with Jingzhou and Huangshi. This would help break the current pattern of strong north–south but weak east–west connectivity and enhance horizontal linkages across southern Hubei.

(3) Cross-boundary collaboration: establishing functional cooperation demonstration zones across administrative boundaries

• Hanchuan–Dongxihu–Caidian metropolitan integration pilot zone.

Recognizing Hanchuan’s functional integration into the Wuhan Metropolitan Area, unified planning should be implemented for industrial land allocation and transport network connectivity, reducing traffic fragmentation and institutional barriers induced by administrative boundaries.

• Jingshan–Tianmen Jianghan cooperation zone.

Given the strong freight interactions between Jingshan and Tianmen, coordinated development of upstream–downstream linkages in machinery manufacturing and textile and apparel industries should be encouraged through the joint development of cross-regional industrial parks.

• Yidu–Songzi–Gongan boundary cooperation zone.

For the economic cluster formed along provincial highways by Yidu, Songzi, and Gongan, agricultural cold-chain logistics centers are recommended at inter-county boundary areas to facilitate the aggregation and circulation of specialty agricultural products from southwestern Hubei, thereby creating regional agri-logistics hubs.

5.2. Conclusions

Based on truck GPS trajectory data, this study quantitatively measures inter-city economic linkages among urban nodes in Hubei Province and systematically reveals the topological characteristics of the provincial urban economic network from three dimensions: points (urban node hierarchy), lines (inter-city linkage intensity), and areas (urban cluster structure). In addition, a 5 km × 5 km grid-based modeling approach is employed to construct a cross-administrative economic linkage network, overcoming the limitations of traditional analyses based on administrative units and identifying the spatial distribution patterns of urban economic communities in Hubei Province.

The findings indicate the following: (1) The urban economic network of Hubei Province currently exhibits an overall monocentric structure with localized polycentric features. Although the Wuhan Metropolitan Area shows high economic activity, internal horizontal linkages remain relatively weak and inter-city coordination requires further improvement. The development of Yichang and Xiangyang as regional sub-centers also needs to be strengthened. (2) Grid-based identification of urban economic communities reveals pronounced cross-administrative characteristics of economic linkages in Hubei Province, indicating persistent discrepancies between actual economic interactions and existing regional planning frameworks. Accordingly, this study proposes two directions for optimizing economic spatial planning in Hubei Province. First, policy support and resource allocation should be directed toward strengthening Yichang and Xiangyang as provincial sub-centers, thereby fostering a multi-core support structure. Second, it is recommended that Hubei Province promote the alignment between administrative space and functional space within its planning system, construct a multi-level network structure characterized by multiple centers, economic linkage corridors, and cross-boundary communities, and strengthen cross-regional coordination mechanisms.

This study demonstrates that network analysis based on freight flow data can more accurately capture inter-city economic linkage patterns, providing both data support and methodological innovation for provincial spatial planning. Nevertheless, several limitations remain. First, the findings represent only a single temporal snapshot of the evolving process of urban economic linkages. Changes in industrial spatial organization, regional policies, and external shocks may lead to dynamic adjustments in factor flows and inter-city economic relationships, thereby reshaping urban community structures and spatial configurations. Second, the truck GPS trajectory data used in this study primarily reflect road-based freight activities and do not incorporate other transport modes such as rail or waterway freight. Future research should therefore adopt a longitudinal perspective, constructing dynamic economic linkage networks based on multi-period freight data to examine the evolutionary trajectories and driving mechanisms of urban communities, as well as the long-term dynamics of urban networks and their implications for regional economic development. Future research could integrate rail and waterway freight data to construct a multi-modal freight linkage network, thereby providing a more comprehensive understanding of regional economic spatial organization. In addition, freight data could be integrated with multi-source datasets—such as mobile phone signaling data, firm linkage data, and internet information flows—to develop multi-factor urban functional linkage models. Such approaches would enable multi-factor, multi-scale analyses of economic network evolution and provide a more systematic and fine-grained understanding of polycentric spatial structures and functional dynamics within urban agglomerations.