A city center spatio-temporal correlation evolution identification method for urban agglomeration

Abstract

The application discloses a city center spatio-temporal correlation evolution identification method for urban agglomeration, belongs to the technical field of geographic information science, spatial network analysis and land space planning auxiliary decision-making, and comprises the following steps: (1) acquiring urban agglomeration boundaries, urban administrative divisions, night light remote sensing, interest points, road networks, population grids and economic grids and other multi-source data and performing pretreatment, fusing night light, interest point kernel density and road network kernel density data through wavelet transform to obtain multi-source fusion grid data; (2) identifying the spatial range of the city center by using Moran index analysis, natural break classification and centroid point superposition screening, and calculating a city center comprehensive index; (3) constructing an improved gravity model based on the city center comprehensive index and spatial distance, measuring the correlation strength of the city center, constructing a city center correlation graph structure of the urban agglomeration, calculating the node centrality, and identifying the spatio-temporal correlation evolution characteristics of the city center of the urban agglomeration. The application can improve the accuracy of city center spatial range identification, and finely depict the spatial correlation relationship and evolution characteristics of the city center.

Description

Technical Field

[0001] This invention relates to the fields of geographic information science, urban geography, spatial network analysis, remote sensing and geospatial big data processing, and land spatial planning auxiliary decision-making. More specifically, it relates to a method for measuring and identifying the spatial correlation strength of urban centers in the spatiotemporal correlation evolution of urban centers in urban agglomerations, based on the comprehensive index of urban centers, an improved gravity model, and graph structure analysis. Background Technology

[0002] Urban agglomerations are an important spatial organization form characterized by the high concentration and flow of population, industry, transportation, and information, and also a crucial vehicle for promoting coordinated regional development and new urbanization. Urban centers, as the areas within cities where population, economic, and public service activities are most concentrated, are important nodes for cities to participate in regional connections and the flow of factors. Accurately identifying the spatial relationships and evolutionary characteristics among urban centers within urban agglomerations is of great significance for revealing the spatial structure of urban agglomerations, optimizing regional resource allocation, and supporting national spatial planning.

[0003] Existing studies on spatial connections within urban agglomerations typically employ gravity models, accessibility models, socioeconomic linkage models, or complex network analysis methods to measure population flows, economic ties, transportation links, and spatial interactions between cities. Among these, gravity models are widely used for analyzing the strength of inter-city connections and regional spatial interactions due to their clear structure, well-defined parameters, and broad applicability. These methods generally use city population size, economic output, or overall development level as urban quality parameters, and the distance between cities as spatial impedance parameters, thereby calculating the strength of interactions between cities. Complex network analysis methods, on the other hand, typically abstract cities as nodes and connections between cities as edges, revealing regional spatial structural characteristics through indicators such as network density, centrality, and connectivity.

[0004] However, existing research still has certain limitations. On the one hand, most studies use the geometric centroid or administrative center of a city's administrative district as a spatial node, which makes it difficult to accurately reflect the true agglomeration center of population and economic activity within the city. On the other hand, city size is usually characterized by single indicators such as population or GDP, which makes it difficult to comprehensively reflect the overall development level of the city center. In addition, existing methods mostly focus on spatial connection analysis at the city scale, paying insufficient attention to the relationships between city centers, making it difficult to finely characterize the spatial organization characteristics of the central system within a city cluster. At the same time, there are relatively few studies on the changes in the strength of connections, network structure, and nodal status of city centers at different time stages, lacking a systematic identification method for the spatiotemporal evolution of city clusters.

[0005] Therefore, there is an urgent need to propose a method for identifying the spatiotemporal correlation evolution of urban centers in urban agglomerations. This method takes the actual functional centers of cities as the research object, comprehensively considers the characteristics of multi-source data such as population and economy, constructs an urban center correlation network, and combines time series analysis to identify the correlation strength and evolution law between urban centers. This will provide more scientific technical support for optimizing the spatial structure of urban agglomerations and promoting regional coordinated development. Summary of the Invention

[0006] This invention provides a method for identifying the spatiotemporal correlation evolution of urban centers in urban agglomerations, addressing issues such as inaccurate representation of urban spatial nodes, insufficient characterization of urban center correlations, and inadequate evolution identification in existing urban agglomeration spatial correlation identification methods. This method takes the spatial extent of urban centers as the research object, combining a comprehensive urban center index, an improved gravity model, and graph structure analysis to achieve the calculation of the spatial correlation strength and spatiotemporal evolution identification of urban centers within urban agglomerations.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A method for identifying the spatiotemporal correlation evolution of urban centers in urban agglomerations includes:

[0009] 1) Acquire basic data of the study area and perform preprocessing; wherein, the basic data of the study area includes urban agglomeration boundary data, urban administrative boundary data, nighttime light remote sensing data, point of interest data, road network data, population raster data and economic raster data;

[0010] 2) Identify and extract the spatial extent of the city center based on nighttime light remote sensing data, point-of-interest data, and road network data;

[0011] 3) Calculate the comprehensive index of the city center based on population raster data and economic raster data, which is used to characterize the population agglomeration capacity, economic development level and comprehensive functional level of the city center;

[0012] 4) Construct an improved gravity model based on the comprehensive index of urban centers and the spatial distance between urban centers, and calculate the correlation strength between different urban centers within the urban agglomeration;

[0013] 5) Construct a city center association graph structure based on the association strength between city centers, and calculate the centrality index of city center nodes to identify core nodes, secondary nodes and edge nodes within the city cluster;

[0014] 6) Identify the spatiotemporal correlation evolution characteristics of urban centers in urban agglomerations based on the comprehensive index of urban centers, correlation strength, correlation graph structure, and node centrality results for multiple periods.

[0015] This invention identifies and extracts the spatial range of urban centers through multi-source spatial data, constructs an improved gravity model using the comprehensive index of urban centers and the distance between centers, and calculates the correlation strength of urban centers. It also combines graph structure and centrality index to express the spatial organization relationship within urban agglomerations, thereby improving the accuracy of urban agglomeration spatial correlation identification and the systematic nature of spatiotemporal evolution analysis, and providing technical support for urban agglomeration spatial structure optimization and regional coordinated development. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely embodiments of the present invention, and those skilled in the art can obtain other drawings based on the provided drawings without creative effort.

[0017] Figure 1 A flowchart of a method for identifying the spatiotemporal correlation evolution of urban centers in urban agglomerations provided by the present invention;

[0018] Figure 2 This is a schematic diagram of multi-source basic data for urban center identification provided by the present invention;

[0019] Figure 3 This is a schematic diagram of kernel density analysis of point-of-interest data and road network data provided by the present invention;

[0020] Figure 4 This is a schematic diagram of the nighttime light data normalization process provided by the present invention;

[0021] Figure 5 This is a schematic diagram of the wavelet transform fusion result of multi-source spatial data provided by the present invention;

[0022] Figure 6 This is a schematic diagram of the urban center spatial range identification results provided by the present invention;

[0023] Figure 7 This is a schematic diagram illustrating the calculation of the comprehensive index of urban centers provided by the present invention;

[0024] Figure 8 This is a schematic diagram of the urban cluster urban center association strength measurement and graph structure construction provided by the present invention;

[0025] Figure 9 This is a schematic diagram of the spatiotemporal correlation evolution identification results of urban centers in urban agglomerations provided by the present invention; Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0027] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. Example Description

[0028] As shown in Figure 1, this embodiment of the invention provides a method for identifying the spatiotemporal correlation evolution of urban centers in urban agglomerations. This method takes urban agglomerations as the research object and the identified urban centers as spatial correlation calculation nodes. It comprehensively utilizes nighttime light remote sensing data, point-of-interest data, road network data, population raster data, and economic raster data to sequentially complete urban center identification, urban center comprehensive index calculation, urban center correlation strength measurement, urban agglomeration urban center correlation graph structure construction, centrality calculation, and multi-period spatiotemporal correlation evolution identification.

[0029] In this embodiment, prefecture-level cities and above and national-level urban agglomerations across the country are used as the application objects, and multi-source spatial data from multiple periods are selected for analysis. The multiple periods include 2013, 2018, and 2023. The urban agglomerations include optimized and upgraded urban agglomerations, growing and expanding urban agglomerations, and nurturing and developing urban agglomerations, and can also be adjusted to other research areas and other time series according to actual application needs.

[0030] This embodiment mainly includes the following steps:

[0031] Step 1: Acquisition and preprocessing of basic data for the study area.

[0032] The study area requires the acquisition of basic data, including urban agglomeration boundary data, urban administrative division boundary data, nighttime light remote sensing data, points of interest (POI) data, road network data, population raster data, and economic raster data. Specifically, urban agglomeration boundary data is used to define the analysis scope of the urban agglomeration; urban administrative division boundary data is used to determine urban affiliation; nighttime light remote sensing data reflects the intensity of urban human activity and the degree of construction activity concentration; POI data reflects the spatial distribution of urban functional elements such as commerce, offices, public services, and transportation facilities; road network data reflects urban transportation connectivity and spatial accessibility; population raster data reflects population spatial concentration; and economic raster data reflects the spatial distribution of economic activities.

[0033] The above data undergoes preprocessing, including coordinate system unification, spatial clipping, data cleaning, outlier removal, topology checking, raster resampling, and spatial matching. For point-of-interest (POI) data, categories weakly associated with urban center functions are removed, while types closely related to urban center activities, such as commercial services, catering and accommodation, office and business, finance and insurance, culture and education, healthcare, and transportation facilities, are retained. For road network data, topology checks and repairs are performed, removing duplicate, broken, and obviously abnormal line segments, while retaining road types relevant to urban center spatial structure identification. For nighttime light data, atmospheric correction, negative value removal, maximum value processing, projection transformation, and clipping are performed. For population and economic raster data, a unified projection coordinate system and spatial resolution are established, and the data is clipped according to urban cluster boundaries.

[0034] After the above processing, multi-source spatial data with consistent spatial reference, matching temporal caliber, and unified spatial scale are obtained, providing a data foundation for subsequent identification of urban centers and measurement of the correlation strength of urban agglomerations.

[0035] Step 2: Identify and construct multi-source basic data for urban centers.

[0036] As shown in Figure 2, nighttime light remote sensing data, point-of-interest (POI) data, and road network data are used as the basic data for urban center identification. Nighttime light remote sensing data is used to characterize urban construction intensity and nighttime activity intensity; POI data is used to characterize the degree of agglomeration of urban functional facilities; and road network data is used to characterize urban traffic accessibility and road network density.

[0037] Kernel density analysis was performed on both point-of-interest (POI) data and road network data to obtain POI kernel density raster data and road network kernel density raster data, respectively. Kernel density analysis is used to transform discrete point or linear spatial features into continuous density surfaces to reflect the spatial clustering of urban functional and transportation elements. The kernel density calculation method is shown in Equation 1:

[0038]

[0039] In the formula, Indicates position The kernel density estimate at that location; Indicates the number of elements within the neighborhood; Indicates the search radius or bandwidth; Represents the kernel function; Indicates position With elements The spatial distance between them.

[0040] As shown in Figure 3, by performing kernel density analysis on the point of interest data and road network data respectively, we can obtain the spatial clustering surface of the point of interest and the spatial clustering surface of the road network, which can be used for subsequent multi-source data fusion.

[0041] Step 3: Normalization processing of nighttime light data, point of interest kernel density data, and road network kernel density data.

[0042] Because nighttime light data, point-of-interest kernel density data, and road network kernel density data differ in numerical magnitude, data units, and spatial representation, they cannot be directly overlaid and fused. Therefore, it is necessary to normalize the three types of data separately, mapping them to a unified numerical range.

[0043] like Figure 4 As shown, the normalization process uses the deviation standardization method, and the calculation formula is:

[0044]

[0045] In the formula, Represents the original data value; This represents the normalized data value; This represents the minimum value in this type of data; This represents the maximum value in this type of data.

[0046] Normalization eliminates dimensional differences between different data sources, making nighttime light data, point-of-interest kernel density data, and road network kernel density data comparable, and providing a unified data foundation for subsequent wavelet transform fusion.

[0047] Step 4: Wavelet transform fusion of multi-source spatial data.

[0048] like Figure 5 As shown, normalized nighttime light data, point-of-interest kernel density data, and road network kernel density data are fused using multi-source spatial data. In this embodiment, a two-dimensional wavelet transform method based on the Daubechies 4 (DB4) wavelet basis is preferably used for fusion.

[0049] Specifically, wavelet decomposition is performed on the normalized nighttime light raster data, interest point kernel density raster data, and road network kernel density raster data to obtain low-frequency approximate information representing the overall distribution trend of urban spatial activities, and high-frequency detail information representing local clustering characteristics and spatial boundary differences. The wavelet coefficients of the three types of raster data at their respective scales are fused, and wavelet reconstruction is performed based on the fused wavelet coefficients to obtain multi-source fused raster data. The spatial reference, spatial extent, and spatial resolution of the multi-source fused raster data are consistent with the input raster data. The two-dimensional wavelet transform can be expressed as:

[0050]

[0051] In the formula, This represents the two-dimensional wavelet transform result based on the DB4 wavelet basis; Indicates the scale factor; and These represent the wavelet basis functions at... direction and Translation parameters in the direction; Indicates input raster data; This represents the DB4 wavelet basis function.

[0052] By using the above fusion method, while preserving the macroscopic distribution characteristics of nighttime light data, the spatial clustering characteristics of points of interest and the linear structural characteristics of the road network are comprehensively expressed, thereby forming multi-source fused spatial data that can reflect the intensity of urban construction activities, the degree of agglomeration of urban functional facilities, and the accessibility of road traffic.

[0053] Step 5: Identification of the spatial scope of the city center.

[0054] like Figure 6 As shown, the Moran index and natural discontinuity classification are performed on the multi-source fused raster data obtained in step 4, and the spatial range of the city center is obtained by centroid point overlay screening.

[0055] Specifically, a global spatial autocorrelation analysis is first performed on the multi-source fused raster data to calculate the global Moran's index, which is used to determine whether the multi-source fused data exhibits significant spatial clustering characteristics throughout the study area. The formula for calculating the global Moran's index is:

[0056]

[0057] In the formula, This represents the global Moran exponent, with a value range of [-1, 1]. Indicates the total number of spatial units; and Representing spatial units and spatial units Observed values; This represents the average value of all spatial unit observations; Representing spatial units With spatial units The spatial weight matrix elements between them. When When the value is greater than 0, it indicates that the multi-source fusion data has positive spatial autocorrelation, and spatial units with similar values ​​are clustered together; when... When <0, it indicates that the multi-source fusion data has negative spatial autocorrelation, and the spatial units with large numerical differences are discretely distributed; when When the value is 0, it indicates that the multi-source fusion data is approximately randomly distributed. Further significance testing of the calculation results is performed using the Z-standardized value and p-value. When p < 0.05, the spatial autocorrelation result is considered statistically significant.

[0058] Based on significant global spatial autocorrelation, local spatial autocorrelation analysis is performed on multi-source fused raster data to calculate the local Moran index, which is used to identify the local correlation characteristics between each spatial cell and its surrounding spatial cells. The formula for calculating the local Moran index is:

[0059]

[0060] In the formula, Representing spatial units The local Moran index; Represents the variance of all spatial unit observations; the meanings of the other symbols are the same as in the formula. same.

[0061] Based on the local Moran's index and its significance test results, local spatial association types are classified into high-value-high-value clustering, low-value-low-value clustering, high-value-low-value dispersion, and low-value-high-value dispersion. Among these, high-value-high-value clustering indicates that high-value spatial units are surrounded by other high-value spatial units, reflecting hotspot areas where urban construction activities, functional facilities, and transportation elements are continuously concentrated. Therefore, high-value-high-value clustering units that meet the significance requirements are extracted and transformed into high-value-high-value clustering vector surfaces.

[0062] While performing local spatial autocorrelation analysis, the multi-source fused raster data obtained in step 4 is graded for natural discontinuities. Natural discontinuity grading is achieved by adjusting the grading discontinuities between different grades to minimize the sum of squared deviations within the same grade and maximize the differences between different grades. The calculation formula is as follows:

[0063]

[0064] In the formula, This represents the sum of squared deviations within each group for all grades. Indicates the total number of levels; Indicates the first Number of spatial units within each level; Indicates the first Within each level Observations of each spatial unit; Indicates the first The average value of all spatial unit observations within each level. By optimizing the hierarchical breakpoints, The minimum value is reached, thus obtaining a classification result with smaller intra-class differences and larger inter-class differences.

[0065] In this embodiment, the multi-source fused raster data is preferably divided into 5 levels. The highest-level spatial unit in the natural discontinuity classification result is extracted, and the highest-level spatial unit is converted into a natural discontinuity highest-level vector surface. Local Moran index analysis and natural discontinuity classification are carried out simultaneously and in parallel, respectively obtaining high-value-high-value clustering vector surfaces and natural discontinuity highest-level vector surfaces. The two analysis processes are independent of each other and there is no sequential dependency.

[0066] Furthermore, the highest-level vector surface of natural discontinuity is converted into the corresponding geometric centroid, and the multi-source fusion index value and spatial coordinate information of each spatial unit are retained. Using the geometric centroid as the basis for spatial screening, high-value-high-value clustering vector surfaces are overlaid and screened, retaining high-value-high-value clustering surface units containing the geometric centroid, and excluding surface units that only have spatial clustering characteristics but whose values ​​do not reach the highest level of natural discontinuity. Finally, the urban center spatial range that simultaneously meets the two conditions of the highest numerical level and high-value spatial clustering is obtained.

[0067] Natural discontinuity grading can identify spatial units with values ​​at the highest level, but may include scattered high-value areas lacking spatial continuity; local Moran's index can identify spatially continuous high-value clusters, but cannot guarantee that the identified units reach the highest level. By using the centroid overlay screening method described above, the two types of identification results can be bidirectionally coupled and verified, improving the accuracy of identifying the spatial extent of urban centers.

[0068] When only one urban center spatial range is identified for the same city, that urban center is designated as the main urban center; when two or more separate urban center spatial ranges are identified for the same city, the main center and sub-centers are divided according to the size of each urban center spatial range, with the urban center with the largest area designated as the main center and the remaining urban centers designated as sub-centers.

[0069] Step 6: Calculation of the comprehensive index of the city center.

[0070] like Figure 7 As shown, based on the urban center spatial range obtained in step 5, the comprehensive urban center index for each urban center is calculated using population raster data and economic raster data at different target periods. In this embodiment, the target periods include 2013, 2018, and 2023, but other time points can be selected according to actual application needs.

[0071] Since population raster data and economic raster data differ in units, dimensions, and numerical ranges, the population raster data and economic raster data for each target period are first normalized to map the two types of data to a unified numerical range, thereby eliminating the influence of different data dimensions and orders of magnitude on the calculation results.

[0072] After normalization, wavelet transform is used to fuse and denoise population raster data and economic raster data for the same target period, obtaining population-economic fused spatial data for the corresponding target period. This fusion process does not involve a simple arithmetic average or fixed-weighting of the population and economic data; instead, it uses wavelet transform to extract the overall trend and local differences between the two types of data, reducing the impact of outliers and local noise, and expressing the spatial coupling relationship between population agglomeration and economic activity.

[0073] Furthermore, the urban center spatial range identified in step 5 is used as a spatial trimming and zoning statistical unit. Population-economic integration spatial data for each target period are trimmed and statistically analyzed to spatially match the population and economic integration data with the urban center area, eliminating interference from data from non-central urban areas. Using urban agglomerations as independent analysis units, a comprehensive urban center index is calculated for each main and sub-center within the urban agglomeration. A higher comprehensive urban center index indicates a higher level of population concentration, economic development, and comprehensive functional capacity within the corresponding urban center area. Each comprehensive urban center index is assigned to a corresponding urban center spatial node for subsequent urban center correlation strength calculation and graph structure analysis.

[0074] Step 7: Calculation of the correlation strength between the city center and the city center.

[0075] like Figure 8 As shown, the comprehensive index of urban centers calculated in step 6 is used as the urban center size parameter, and the spatial distance between urban centers is used as the distance impedance parameter to construct an improved gravity model and calculate the correlation strength between different urban centers within the same urban agglomeration.

[0076] This invention uses the real functional center of the city identified in step 5 to replace the geometric centroid of the city's administrative region as a spatial node, thereby reducing the spatial deviation between the geometric center of the administrative region and the actual population and economic agglomeration center.

[0077] The calculation formula for the improved gravitational model is as follows:

[0078]

[0079] In the formula, Indicates the city center With the city center The strength of the correlation between them; and Representing the city center and city center The comprehensive index of urban centers; Indicates the city center With the city center The straight-line distance between them; denoted as the gravitational constant, which is preferably 1 in this embodiment.

[0080] The higher the comprehensive index of city centers and the smaller the straight-line distance between two city centers, the stronger the connection between them; conversely, the lower the comprehensive index of city centers and the larger the straight-line distance between two city centers, the weaker the connection between them.

[0081] The correlation strength between candidate city centers within the same urban agglomeration is calculated for each target period, forming city center correlation strength data for the corresponding period. Spatial connections between city centers are displayed hierarchically based on the correlation strength values, where the size of the city center node represents the city center comprehensive index, and the width of the lines connecting nodes represents the city center correlation strength.

[0082] Step 8: Construction of the urban center association graph structure of the urban agglomeration.

[0083] Based on the main and secondary city centers determined in step 5 and the city center association strength calculated in step 7, each city cluster is abstracted into an independent undirected spatial topology network. The undirected spatial topology network uses the main and secondary city centers as nodes, the spatial connections between city centers as edges, and the association strength between corresponding city centers as the edge attribute.

[0084] The city center connection graph structure of the urban agglomeration is constructed according to a hierarchical priority logic. Specifically, the connection relationships between city centers within a single city are constructed first. For a city containing one main center and one or more sub-centers, each sub-center is connected to the main center of that city, forming a main-sub-center connection structure within the city area. For a city containing only one city center, that city center is treated as the main center node, and no internal connections are constructed within the city area. After completing the connections between the main and sub-centers within the city area, the connection relationships between city centers of different cities are then constructed.

[0085] Using the main centers of each city as intercity spatial connection nodes, the intercity connection edges between different city main centers are determined based on the city center connection strength calculated in step 7. Intercity main center-to-main-center connections are retained according to the magnitude of the connection strength or a preset connection strength threshold, thus forming an intercity main center-to-main-center connection structure. By first establishing main center-sub-center connections within a city and then establishing intercity main center-to-main-center connections, the hierarchical relationships within a single city can be preserved, while simultaneously expressing the spatial connections between different cities, forming a complete urban cluster city center connection graph structure.

[0086] The node positions in the graph structure are represented by the actual spatial locations of the corresponding city centers; the node sizes are represented by the comprehensive index of the corresponding city centers; and the edge widths are represented by the association strength of the corresponding city centers. This achieves a comprehensive expression of the scale, hierarchical relationships, and spatial connections of city centers within a city cluster.

[0087] Step 9: Calculate the centrality of the city center node.

[0088] Based on the urban center association graph structure constructed in step 8, the absolute centrality and relative centrality of each urban center node are calculated to evaluate the connectivity and hub status of the urban center in the urban agglomeration association network.

[0089] Absolute centrality is used to represent the number of direct connections between a city's central node and other nodes. Its calculation formula is:

[0090]

[0091] In the formula, Represents the city center node absolute centrality; Indicates the connection with the city center node The number of directly connected edges.

[0092] The greater the absolute centrality, the more direct connections the city's central node has with other nodes within the city cluster, and the stronger its direct radiation and communication capabilities; the smaller the absolute centrality, the fewer direct connections the node has in the city cluster's network.

[0093] Since the number of urban center nodes varies across different urban agglomerations, it is difficult to make a horizontal comparison of the node status of different urban agglomerations using only absolute centrality. Therefore, the absolute centrality is standardized to obtain the relative centrality, which is calculated using the following formula:

[0094]

[0095] In the formula, Represents the city center node The relative centrality; Represents the city center node absolute centrality; This represents the total number of nodes in the corresponding city cluster association graph structure.

[0096] The relative centrality ranges from [0, 1]. A relative centrality closer to 1 indicates stronger connectivity and a more prominent hub status for the corresponding city center within the urban agglomeration's network; a relative centrality closer to 0 indicates relatively weaker network connections for the corresponding city center. The centrality level of nodes within the urban agglomeration is evaluated based on the city center comprehensive index, association strength, absolute centrality, and relative centrality of each city center node. Nodes with higher comprehensive indices and relative centralities correspond to core nodes of the urban agglomeration; nodes directly connected to core nodes and possessing a certain degree of association ability correspond to secondary nodes; and nodes with relatively lower comprehensive indices and relative centralities correspond to peripheral nodes.

[0097] Step 10: Identification of the spatiotemporal correlation evolution of urban centers in urban agglomerations.

[0098] like Figure 9 As shown, spatial matching and comparative analysis are performed on the comprehensive index of urban centers, the intensity of urban center association, the structure of urban center association graph, absolute centrality and relative centrality for different target periods to identify the spatiotemporal association evolution characteristics of urban centers in urban agglomerations.

[0099] Specifically, based on city name, city affiliation, primary and secondary center level, and spatial location, the corresponding city center nodes in 2013, 2018, and 2023 are linked to establish the correspondence between city center nodes in different target periods.

[0100] By comparing the comprehensive index of the same city center in different target periods, we can identify the enhancement, weakening, or relative stability of the population-economic comprehensive functional level of the city center; by comparing the correlation strength between pairs of the same city centers in different target periods, we can identify the enhancement, weakening, addition, or disappearance of spatial connections between city centers; by comparing the graph structure in different target periods, we can identify the changes in the number of network connections, spatial extension direction, and overall connectivity pattern of the city cluster; by comparing the absolute centrality and relative centrality in different target periods, we can identify the changes in the node hub status and radiation capacity of the city center.

[0101] For urban agglomerations that are being optimized and upgraded, the focus is on identifying whether the status of the original core nodes remains stable, whether the strength of the connections between core nodes has increased, and whether the peripheral nodes are gradually integrating into the core network.

[0102] For growing urban agglomerations, the focus should be on identifying the expansion of the radiation range of the core cities in the region, the improvement of the comprehensive functional level of secondary nodes, and the development process of axial connections and multi-node networks within the urban agglomeration.

[0103] For the cultivation and development of urban agglomerations, the focus is on identifying the leading capabilities of single or dual core nodes, the process of peripheral nodes connecting to the urban agglomeration network, and the characteristics of the urban agglomeration's interconnected structure evolving from relatively sparse to continuous and networked.

[0104] Finally, the comprehensive index of the city center is represented by the size of the city center node, the node level is represented by the main center, the sub-center and different node levels, the correlation strength is represented by the width of the connection line, and the results of multiple target periods are displayed side by side to obtain the spatiotemporal correlation evolution identification results of the city center of the urban agglomeration.

[0105] Through the above steps, this invention combines the identification of the spatial scope of the city center, the calculation of the comprehensive functional level of population and economy, the calculation of the correlation strength of the city center, the construction of graph structure, and the evaluation of node centrality. It can replace the geometric centroid of the traditional administrative region with the real functional center of the city, and characterize the spatial correlation of the city cluster from two levels: the main and secondary centers within the city and the main centers between cities, thereby improving the accuracy of the calculation of the spatial correlation strength of the city cluster and the identification of its spatiotemporal evolution.

[0106] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any adjustments, substitutions, or improvements made by those skilled in the art to the research area, target period, data source, spatial resolution, number of levels, and graph structure connection method without departing from the technical principles and concept of the present invention should fall within the scope of protection of the present invention.

Claims

1. A method for identifying the spatiotemporal correlation evolution of urban centers in urban agglomerations, characterized in that, include: S1: Acquire basic data of the study area and perform preprocessing; wherein, the basic data of the study area includes urban agglomeration boundary data, urban administrative division boundary data, nighttime light remote sensing data, point of interest data, road network data, population raster data and economic raster data; S2: Construct multi-source basic data for urban center identification based on nighttime light remote sensing data, point of interest data, and road network data, and perform kernel density analysis on point of interest data and road network data respectively to obtain point of interest kernel density raster data and road network kernel density raster data. S3: Normalize the nighttime light remote sensing data, point of interest kernel density raster data, and road network kernel density raster data to obtain normalized multi-source raster data; S4: Perform two-dimensional wavelet transform fusion based on DB4 wavelet basis on the normalized multi-source raster data to obtain multi-source fused raster data; S5: The Moran index is calculated and natural discontinuity is classified for the multi-source fused raster data respectively, and the spatial range of the city center is obtained by centroid point overlay screening. S6: Calculate the comprehensive index of the city center based on population grid data and economic grid data, which is used to characterize the population concentration level, economic development level and comprehensive functional level of the city center; S7: Construct an improved gravity model based on the comprehensive index of urban centers and the spatial distance between urban centers to calculate the correlation strength between different urban centers within the same urban agglomeration; S8: Construct the urban center association diagram structure of the urban agglomeration based on the association strength between the main urban center, sub-centers, and urban centers; S9: Calculate the absolute centrality and relative centrality of urban center nodes based on the urban center association graph structure of urban agglomerations; S10: Identify the spatiotemporal correlation evolution characteristics of urban centers in urban agglomerations based on the comprehensive index of urban centers, correlation strength, correlation graph structure, absolute centrality, and relative centrality over multiple periods.

2. The method for identifying the spatiotemporal correlation evolution of urban centers in urban agglomerations according to claim 1, characterized in that: In steps S4 to S5, the normalized nighttime light raster data, interest point kernel density raster data, and road network kernel density raster data are subjected to two-dimensional wavelet decomposition, and the wavelet coefficients are fused and reconstructed to obtain multi-source fused raster data. Moran index analysis and natural discontinuity grading are used to perform dual-path identification on the multi-source fused raster data, and the high-value-high-value clustering vector surface is overlaid and filtered by the centroid of the highest-level natural discontinuity vector surface to obtain the spatial range of the city center.

3. A method for identifying the spatiotemporal correlation evolution of urban centers in urban agglomerations according to claim 1 or 2, characterized in that: In steps S6 to S10, the spatial extent of the city center is spatially overlaid with population raster data and economic raster data to calculate the comprehensive index of the city center; an improved gravity model is constructed based on the comprehensive index of the city center and the spatial distance between city centers to calculate the correlation strength of the city centers; the correlation graph structure of the city center of the city cluster is constructed according to the rules of connecting the main center and the sub-center within the city area and connecting different main centers of cities within the city cluster according to the correlation strength, and then the absolute centrality and relative centrality of the city center nodes are calculated, and the spatiotemporal correlation evolution characteristics of the city center of the city cluster are identified based on the comprehensive index, correlation strength, correlation graph structure and centrality changes over multiple periods.

Images