Research on Spatial Characteristics and Influencing Factors of Urban Vitality at Multiple Scales Based on Multi-Source Data: A Case Study of Qingdao

Abstract

Urban vitality serves as an important indicator for evaluating the level of urban quality development and sustainability. In response to a series of urban challenges arising from rapid urban expansion, enhancing urban quality and fostering urban vitality have become key objectives in contemporary urban planning and development. This study summarizes the spatial distribution patterns of urban vitality at the street and neighborhood levels in the central area of Qingdao, and analyzes their spatial characteristics. A 5D built environment indicator system is constructed, and the effects of the built environment on urban vitality are explored using the Optimal Parameter Geographic Detector (OPGD) and the Multi-Scale Geographically Weighted Regression (MGWR) model. The aim is to propose strategies for enhancing spatial vitality at the street and neighborhood scales in central Qingdao, thereby providing references for the optimal allocation of urban spatial elements in urban regeneration and promoting sustainable urban development. The findings indicate the following: (1) At both the subdistrict and block levels, urban vitality in Qingdao exhibits significant spatial clustering, characterized by a pattern of “weak east-west, strong central, multi-center, cluster-structured,” with vitality cores closely aligned with urban commercial districts; (2) The interaction between the three factors of functional density, commercial facilities accessibility and public facilities accessibility and other factors constitutes the primary determinant influencing urban vitality intensity at both scales; (3) Commercial facilities accessibility and cultural and leisure facilities accessibility and building height exert a positive influence on urban vitality, whereas the resident population density appears to have an inhibitory effect. Additionally, factors such as building height, functional mixing degree and public facilities accessibility contribute positively to enhancing urban vitality at the block scale. (4) Future spatial planning should leverage the spillover effects of high-vitality areas, optimize population distribution, strengthen functional diversity, increase the density of metro stations and promote the coordinated development of the economy and culture.

1. Introduction

Cities serve as crucial platforms for advancing high-quality development, efficient governance and high-quality living in the new era. Within China’s ongoing pursuit of New-Type Urbanization, a national strategy explicitly prioritizing urban quality and connotative development over mere physical expansion [1], enhancing urban vitality and promoting intensive urban development emerge as central priorities. These objectives represent key focal points for the thorough implementation of urban renewal initiatives and the comprehensive practice of the people-centered philosophy underpinning New-Type Urbanization [2]. As of 2024, China’s urbanization rate has reached 67%. The acceleration of urbanization and the rapid population growth in central areas have significantly altered urban spatial structures. The rapid urbanization process and global population growth have led to an increase in the number of urban residents to 56% by 2021. Urban management has become one of the most severe development challenges facing the world in the 21st century. At the same time, it has caused inefficient and fragmented spatial functional layouts, manifested as the decline in old urban areas, uneven distribution of public services, traffic congestion, separation of living and working spaces and social stratification [3], ultimately disrupting the orderly distribution of urban vitality. Urban growth and the steady rise in living standards have propelled the advancement of urban vitality to the forefront of modern urban planning priorities [4]. In China, this is both driven by the intrinsic impetus of urban development and the national demand for sustainable and high-quality urban environments under the backdrop of new urbanization. Therefore, it is crucial to conduct in-depth research on strategies to enhance urban spatial vitality in order to achieve sustainable urban progress, improve residents’ well-being and strengthen the overall competitiveness of cities.

Urban vitality encompasses diverse and profound connotations. Jane Jacobs was the first to propose that the city’s vitality stems from the close integration of residents’ daily lives and living spaces, which creates rich diversity in life [5]. Kevin Lynch defined vitality as the degree to which an urban environment fosters essential human activities, ecological needs and capacity development, and regarded vitality as a key element for evaluating the form of space and social sustainability [6]. Gehl argued that the vitality of urban public spaces derives from the presence of people and their activities [7]. Jiang et al. conceptualized urban vitality as a city’s ability to foster a human-centered living environment, where the concentration of people gives qualities similar to those of living organisms, acting as the fundamental source of urban vitality [8]. Yang et al. indicated that urban vitality is mainly reflected in the appeal of public spaces to diverse groups of people [9]. Collectively, these perspectives emphasize the central role of human activities and their interaction with the environment as the main drivers of urban vitality. The Natural Resources and Planning Bureau of Qingdao has set forth development goals aimed at achieving comprehensive enhancement of regional functions and environmental quality, establishing a new engine for urban growth, and stimulating urban vitality. Therefore, examining the interplay between urban vitality and the built environment is critical for informing urban design and management approaches aimed at strengthening urban vitality.

The quantitative assessment of urban vitality has attracted significant interest across diverse fields such as urban planning, urban geography and the social sciences. Early methodologies relied on field surveys to collect data on individual activities, interaction behaviors and lived experiences to characterize urban vitality. However, these methods were limited by high costs, small sample sizes and insufficient spatiotemporal resolution, which restricted a thorough understanding of the dynamic features and developmental trajectories of urban vitality [10,11]. Subsequent studies have used point of interest (POI) data to represent spatial distribution patterns of urban functions and their impact on regional vitality [12], but POI data often lacks timeliness and objectivity, providing only a partial view of actual human activity levels and failing to accurately capture urban vitality [13]. The advancement of information technology has enabled more effective analyses of urban vitality. Compared to traditional survey methods, multi-source data offer finer-grained observations, broader coverage and richer spatiotemporal and semantic information, rendering them particularly suitable for characterizing urban vitality from a human perspective. Consequently, mobile phone signaling data, remote sensing data and online review data have been widely employed to measure urban vitality, facilitating detailed examinations of its spatiotemporal dynamics [14,15,16,17]. For instance, Levin et al. showed that satellite-based nighttime light data can accurately capture the dynamic patterns and intensity of urban population activities [18]. Kim pinpointed hotspots of online activity by analyzing Wi-Fi data, assessing the vibrancy of urban spaces through the lens of digital connectivity [19]. Cao et al. extracted spatial vitality at 24 intervals throughout the day from mobile phone signaling data, creating a vitality time series for urban blocks and interpreting urban vitality from a temporal dimension [17].

Research on the vitality of various urban areas has garnered significant attention [20]. In China, subdistricts serve as fundamental administrative units [21], and findings derived from this scale are more operational and scientifically robust, yielding critical insights to inform urban development and enhance the built environment. Employing the block as the unit of analysis accurately reflects urban vitality’s micro-scale spatial attributes and its connection to the built surroundings, thereby addressing the precision limitations associated with larger administrative regions and the over-arching perspectives lacking in smaller grid scales. Quantitative assessment of urban vitality at the subdistrict and block levels enhances the understanding of both overall and regional development, systematically revealing the hierarchical characteristics of urban vitality through macro patterns and micro mechanisms. Such insights can inform future urban activities, planning and construction initiatives [22,23].

The built environment of a city is recognized as a critical factor influencing urban vitality [24]. Researchers utilize a range of built environment indicators of different dimensions to examine urban vitality influences. Elements such as spatial configuration [25], development and construction [26] and supporting facilities [20] are recognized as closely associated with urban vitality and are frequently represented by one or a limited number of indicators. For example, Xuan et al. found that an increase in the density of road intersections can enhance the accessibility of a road network in a certain area, which to some extent can promote urban vitality [27]. Li et al. argued that higher development intensity and greater spatial compactness have a significant positive impact on urban vitality [28]. Lee pointed out that the accessibility of urban facilities is the most important environmental factor for enhancing urban livability [29].

Cervero et al. introduced a “3D” framework for evaluating the interplay between the constructed environment and transportation [30]. Ewing et al. expanded this model by incorporating destination accessibility and distance to transit, resulting in a more comprehensive “5D” built environment framework [31], which has become an important reference standard for quantitative research on the built environment on an international scale. In terms of analytical methods, related studies frequently utilize techniques such as ordinary least squares (OLS) [32], geographically weighted regression (GWR) [20] and spatial lag models (SLMs) [33]. For instance, Wang et al. [34] utilized multi-scale geographically weighted regression (GWR) to analyze the spatiotemporal driving factors affecting 24 h urban vitality in Beijing. Moreover, the use of Geodetector in related research is on the rise. However, the traditional Geodetector model is constrained by issues related to spatial scale, data discretization methods and the number of levels. To address these limitations, Song et al. developed a new optimal parameter geographic detector (OPGD) model [35], which mitigates the challenges associated with subjective data discretization and spatial scale, allowing for a more scientifically robust evaluation of the research subject [36].

A review of existing research indicates that current studies on the built environment and urban vitality predominantly examine their linear relationships, among which the application of OLS and GWR models is relatively common. In terms of research scales, investigations primarily focus on the street, block and community levels. However, analyses of urban vitality at the block and subdistrict levels utilizing the optimal parameter geographic detector (OPGD) model and multiscale geographically weighted regression (MGWR) model remain scarce. Based on the context outlined above, this research employs exploratory spatial data analysis techniques to investigate the distribution patterns and clustering tendencies of urban vitality within Qingdao City, focusing on the subdistrict and block levels. Additionally, the OPGD and spatial regression analysis methods are employed to uncover the interactive effects and spatial heterogeneity influencing urban vitality, thus providing targeted recommendations for enhancing vitality in sustainable urban development and renewal.

2. Research Area and Data

2.1. Research Area

Located along the coast of East China and adjacent to the Yellow Sea, Qingdao possesses abundant natural resources and a strategic geographic location, underpinned by a solid economic foundation and distinct regional advantages. As the economic hub of Shandong Province, a key sub-provincial city, and an internationally influential port city, Qingdao holds a significant socio-economic competitive edge within the region. The research area is defined as the central urban zone outlined in the Qingdao Spatial Planning (2021–2035), encompassing all of Shinan, Shibei and Licang Districts; most of Chengyang District; and portions of Laoshan and Huangdao Districts. It spans approximately 898 square kilometers and houses a resident population of 5.83 million, representing 57.9% of Qingdao’s total population. The central area constitutes the core of the city’s system, distinguished by a comprehensive range of facilities and a densely concentrated population. As the principal hub of vitality aggregation in Qingdao [37], it presents an optimal focus for investigating urban vitality measurement and the factors that influence it.

Since 1978, Qingdao has rapidly evolved from a seaside fishing village into a world-renowned international metropolis. The accelerated pace of urbanization has brought significant challenges to the city’s sustainable development, making it imperative for Qingdao to shift from the stage of rapid urbanization to one focused on improving urbanization quality. At this critical juncture of urban transformation, exploring the vitality characteristics at different urban scales is conducive to fostering local vibrancy and enhancing overall urban vitality. Therefore, this study adopts subdistricts and blocks as the basic units of analysis. The block serves as the fundamental functional unit of urban space, while the subdistrict represents the grassroots administrative unit within China’s hierarchical structure. This dual-level analysis of blocks and subdistricts offers a comprehensive perspective for urban governance and development. Block boundaries were derived from road network segmentation, then refined through integration of remote sensing imagery, hydrological features and natural landscape elements, resulting in 811 block units averaging 1.11 km² after the elimination of undersized blocks, while 66 subdistrict-level units, delineated according to China’s administrative division system, averaged 13.61 km² each (Figure 1).

2.2. Data Sources

Urban social vitality was derived from Baidu Heat Map information (https://map.baidu.com/; accessed on 16 December 2024). Prior studies have analyzed the characteristics of heat maps from Monday to Friday, revealing consistent patterns in residents’ travel behaviors, which also exhibit similarity on Saturdays and Sundays [38]. To avoid the interference of holidays and other special events on social vitality, this study selected two typical dates to measure social vitality. These two dates represent normal daily activities and are not affected by any major external factors that may cause data deviation. In this way, this study aims to obtain more accurate and consistent measurement data of social vitality and more clearly reflect the daily social interactions and activities. Consequently, the Baidu Heat Maps from 14 November 2024 (a weekday), and 17 November 2024 (a weekend), were selected as data sources, focusing on a time interval from 8 a.m. to 12 p.m., with data generated hourly. To represent social vitality, this study calculated the average heat density values at 34 distinct time points.

High-resolution nighttime light data from the Qingdao LuoJia1-01 remote sensing imagery (http://www.lmars.whu.edu.cn; accessed 6 January 2025) were employed to evaluate the economic vitality of the city [39]. The nearest-neighbor resampling technique was applied to the image, and the mean luminosity within the unit was calculated to represent economic vitality.

The assessment of urban cultural vitality is based on data obtained from the Amap Open Platform (https://lbs.amap.com/; accessed on 4 November 2024). This data encompasses significant cultural elements, including museums, exhibition centers, libraries, concert halls, theaters and art galleries. The density of Points of Interest (POIs) is employed as a metric to evaluate cultural vitality [40].

Drawing upon variables established in prior research [11,28,41,42,43] as well as the 5D framework of the urban built environment developed by Cervero [30] and Ewing [31], this study evaluates the urban built environment through five key dimensions (

Table 1. Built environment 5D detection index system.

Dimensionality	Variable	Calculation Method	Illustrate
	Resident population density (X1)	Distribution density of the resident population within the unit	Reflect the distribution characteristics of the resident population
Density	Building density (X2)	Ratio of the sum of built-up area within the unit to the unit area	Reflect the degree of building coverage
	Functional density (X3)	Total count of POIs within the unit to the unit area	Reflect the concentration degree of functional types
	Unit area (X4)	Area of the spatial unit	Reflect the unit area
Design	Building height (X5)	Mean value of the heights of all buildings in the unit	Reflect the three-dimensional form of buildings
	Greening rate (X6)	Mean value of the fractional vegetation cover (FVC) within the unit	Reflect the vegetation coverage
	Morphological compactness (X7)	Degree of compactness of the spatial form within the unit	Reflect the compactness of spatial form
Diversity	Functional mixing degree (X8)	Mixing degree of different types of POIs within the unit	Reflect the degree of functional diversity
	Metro accessibility (X9)	Linear distance from the spatial centroid of the block to the nearest metro station/Mean value of the kernel density of metro stations within the subdistrict	Reflect the convenience of metro transportation
Distance to Transit	Bus accessibility (X10)	Linear distance from the spatial centroid of the block to the nearest bus station/Mean value of the kernel density of bus stations within the subdistrict	Reflect the convenience of bus transportation
	Road network density (X11)	Ratio of the total length of the road network within the unit to the unit area	Reflect the density of road distribution
	Commercial facilities Accessibility (X12)	Ratio of the total quantity of commercial facilities within the unit to the unit area	Reflect the distribution density of commercial facilities
Destination Accessibility	Public facilities Accessibility (X13)	Ratio of the total quantity of public facilities within the unit to the unit area	Reflect the distribution density of public facilities
	Cultural and leisure facilities Accessibility (X14)	Ratio of the total quantity of cultural and leisure facilities within the unit to the unit area	Reflect the distribution density of cultural and leisure facilities

). Basic geographic data, including administrative boundaries and water areas, are sourced from the Resources and Environmental Science and Data Platform of the Chinese Academy of Sciences (https://www.resdc.cn/; accessed on 4 November 2024). Vector data for buildings and various Points of Interest (POIs), such as bus stops, subway stations and urban center points, are obtained from the Open Platform of AutoNavi Maps (https://lbs.amap.com/; accessed on 4 November 2024). Urban road network data is acquired from the vector layers provided by OpenStreetMap (OSM) (https://download.geofabrik.de/; accessed on 4 November 2024).

3. Research Methods

This research analyzes the distribution of urban vitality across multiple spatial scales and its relationship with the built environment. A comprehensive research framework is constructed, which includes data representation, indicator system formulation and quantitative evaluation (Figure 2). Initially, the social, economic and cultural vitality of Qingdao’s central area are evaluated at the subdistrict and block scales. After normalizing these three dimensions, their average is calculated to derive a composite overall vitality index. Secondly, a systematic analysis is conducted to examine the distribution patterns and agglomeration characteristics of vitality in central Qingdao. Thirdly, based on five dimensions, a built environment evaluation system with 14 specific indicators is developed to investigate the correlations between urban vitality levels at multiple scales and elements of the built environment. Finally, specific measures for improving the urban spatial structure are proposed based on a comparison with existing research.

3.1. Spatial Autocorrelation

Spatial autocorrelation is employed to assess whether significant relationships exist between the attributes of geographic locations and those of their neighboring areas [44], and it serves as a key method for identifying patterns of spatial heterogeneity and urban spatial structure [22]. Spatial autocorrelation analysis, which includes both global and local autocorrelation, is a fundamental statistical tool in spatial analysis [45]. Global spatial autocorrelation analysis helps identify the overall patterns in how phenomena are distributed across a region and measures the extent and type of spatial dependence throughout the entire study area. Moran’s I index, a widely used measure of spatial autocorrelation, is expressed as follows:

$$I={\displaystyle \frac{{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{w}_{ij}(x_i-\overline{x})(x_j-\overline{x})}}}{s^2{\displaystyle \sum _{i=1}^n{\displaystyle \sum _{j=1}^n{w}_{ij}}}}}$$

(1)

$$s^2={\displaystyle \frac{1}{n}}{\displaystyle \sum _{i=1}^n{\left(x_i-\overline{x}\right)}^2}$$

(2)

In the equation, n stands for the total count of units; x_i and x_j refer to the vitality values of spatial units i and j, respectively; $\overline{x}$ denotes the mean vitality value; and w_ij signifies the weight between i and j in the weight matrix, assuming a value of 1 when the spatial units are adjacent and 0 otherwise. Moran’s I statistic spans from −1 to 1. A positive value indicates an aggregation trend, a negative value indicates a dispersion trend, and the magnitude represents the intensity of the pattern; an absolute value of 0 indicates a random distribution.

Local spatial autocorrelation measures the strength of the association between a given spatial unit and its surrounding neighbors. This analysis is instrumental in detecting local instability across different spatial locations and uncovering spatial heterogeneity within the data. Local Moran’s I (LISA) measures spatial autocorrelation locally [46], with its formula given as follows:

$$I_i={\displaystyle \frac{n\left(x_i-\overline{x}\right){\displaystyle \sum _{j=1}^n\left(x_i-\overline{x}\right)}}{{\displaystyle \sum _{i=1}^n{\left(x_i-\overline{x}\right)}^2}}}$$

(3)

In the formula, w_ij represents the row-normalized form of the spatial weight matrix. Using Local Indicators of Spatial Association (LISA) metrics, four distinct categories of hotspot and coldspot areas emerge: low-value clustering regions (Low-Low, LL) and high-value clustering regions (High-High, HH), along with low-high (LH) and high-low (HL) spatial outlier clusters.

3.2. Optimal Parameter Geographic Detector (OPGD)

The Geodetector model is a spatial statistical approach grounded in variance analysis, designed to quantify the individual impacts of factors as well as their interaction effects on the dependent variable [47]. In this model, continuous variables are typically discretized and transformed into categorical variables. The OPGD model is an improved model based on the basis of the Geodetector model by solving for the optimal parameters [35]. This model achieves more precise spatial analysis by determining the optimal method for discretizing spatial data and the optimal number of breakpoints for all explanatory variables. It selects the discretization method yielding the highest q value by comparing q values derived from various discretization techniques. The alternative discretization methods employed in this study include the equal interval breakpoint method, natural interval breakpoint method, quantile interval breakpoint method and geometric interval breakpoint method. In this study, the analysis based on the OPGD model encompasses two components: factor detector and interaction detector.

3.2.1. Factor Detector

The factor detector within the geographical detector shows the comparative significance of explanatory variables via the q statistic [48,49,50]. This statistic compares the variance of observed values across the entire study area with the variance at the variable level, the q value for each explanatory variable is calculated as follows:

$$q=1-{\displaystyle \frac{{\displaystyle \sum _{h=1}^L{N}_h{\sigma }_h^2}}{N{\sigma }^2}}=1-{\displaystyle \frac{\mathrm{SSW}}{\mathrm{SST}}}$$

(4)

$$\mathrm{SSW}={\displaystyle \sum _{h=1}^L{N}_h{\sigma }_h^2}$$

(5)

$$\mathrm{SST}=N{\sigma }^2$$

(6)

In the formula, the explanatory power of the q factor ranges from 0 to 1. L denotes the number of classification factors, while N_h and N, respectively, represent category h and the number of blocks or subdistricts in the study area. ${\sigma }_h^2$ and ${\sigma }^2$ are the variances of classification h and the overall study area, respectively. SSW represents the overall within-class variance, SST indicates the total variance across the area [47].

3.2.2. Interaction Detector

The interaction detector evaluates the significance of interactions by examining the q values obtained from factor detectors, thereby identifying the combined effect of two overlapping spatial explanatory variables. This spatial interaction is conceptualized as the superimposition of these variables. Through analyzing the interaction’s q-value in relation to the individual factors’, the interaction detector delves into how the variables’ interplay impacts the dependent variable [51,52].

3.3. Multi-Scale Geographically Weighted Regression (MGWR)

Urban vitality is a complex and multi-dimensional indicator, resulting from the combined effects of various factors. When examining its influencing factors, it is important to consider the scale of independent variables to avoid model instability caused by uniform bandwidths, which may compromise the results. The MGWR model addresses the challenges of differing variable scales and bandwidths by employing an optimal bandwidth for each independent variable. This approach demonstrates superior performance compared to traditional geographically weighted regression models [53]. The model expression is as follows:

$$y_i={\beta }_h(x_i,y_i)+{\displaystyle \sum _{j=1}^k{\beta }_j(x_i,y_i)x_{ij}}+{ϵ}_i$$

(7)

In the formula, y_i represents a global variable; (x_i, y_i) denotes the centroid coordinates of block or subdistrict i; β_h(x_i, y_i) is the regression constant term; β_j(x_i, y_i) indicates the regression coefficient for point j; x_ij represents the value of the influencing factor; k is the number of influencing factors; and ϵ_i denotes the regression residual of the model.

4. Spatial Distribution of Urban Vitality

4.1. Spatial Distribution Pattern of Urban Vitality

Urban vitality in central Qingdao exhibits a clustered distribution pattern of “weak east–west, strong central, multi-enter, cluster–structured”, with significant spatial heterogeneity (Figure 3). From a comprehensive spatial perspective, areas with elevated vitality typically develop along major roads and near rail transit, forming a corridor along the southern coastline of Qingdao. Functionally, high-value areas of urban vitality are predominantly concentrated around municipal and district-level commercial and employment centers. At the subdistrict level, urban vitality exhibits distinct spatial patterns, with a marked decline from the central core towards the outskirts. Significant spatial differentiation is evident, with high-value areas primarily located on subdistricts such as Zhongshan Road, Taipingdang, Dunhua Road, Zhan Shan and Hong Kong Middle Road in the Shinan and Shibei Districts. In contrast, subdistricts located in the northeastern sector of the central area, especially within Chengyang District and the eastern part of Huangdao District, demonstrate comparatively lower levels of urban vitality. At the block level, a general trend of gradual decline in vitality is observed moving outward from the city’s central business district, with areas exhibiting strong comprehensive vitality predominantly forming a belt-like pattern along the central zones of major urban areas in Qingdao, while the eastern and western sections of Chengyang District display comparatively lower levels of comprehensive vitality.

The Getis-Ord Gi* analysis identified cold and hotspots within the study region, highlighting spatial patterns that align closely with the low-value and high-value zones identified at both subdistrict and block levels (Figure 4). Qingdao’s prominent vitality pole is centered on the central urban districts of Shinan and Shibei, which function as the economic core and cultural tourism hub, characterized by comprehensive urban infrastructure, convenient public transportation and high-density human activity zones. Additionally, smaller vitality poles have emerged in central business districts such as Chengyang and Huangdao. However, limited transportation accessibility in western Chengyang and northern Huangdao results in low functional densities, while underdeveloped supporting facilities constrain their long-term capacity to sustain population retention, consumption and residential stability.

4.2. Spatial Clustering Characteristics of Urban Vitality

The global Moran’s I was utilized to measure the spatial correlation in urban vitality at subdistrict and block levels within central Qingdao (Figure 5). The results demonstrate that both scales exceeded the significance threshold (p < 0.001), with Moran’s I indices of 0.564 at the subdistrict scale and 0.679 at the block scale, indicating significant spatial aggregation of urban vitality at both levels, with aggregation more pronounced at the block scale.

The spatial autocorrelation of urban vitality in Qingdao’s central districts was analyzed using the Local Moran’s I index, examining both subdistrict and block levels, and four distinct spatial patterns were identified (Figure 6), with an overall distribution characterized by a “strong center and weak periphery” pattern. At the block scale, several new “high-high” distribution areas emerged, including Chengyang Wanda Plaza, Huangdao Jiajia Yuan, Licang Wanda Plaza and Laoshan Jinshi Plaza. Specifically, 5.15% of areas at the subdistrict scale and 17.8% at the block scale fell into the “high-high” spatial state. These high urban vitality areas exerted a positive influence, elevating the vitality of surrounding regions. In terms of scale, only the “high-high” clusters in Shinan and Shibei Districts were relatively large, whereas the proportion of areas exhibiting the “low-low” clustering pattern increased to 27.6% at the subdistrict level and 48.6% at the block level, predominantly concentrated in the northern parts of Huangdao and Chengyang Districts, where diminished vitality values adversely affected adjacent blocks. This analysis demonstrates that the clustering of high and low vitality values markedly affects the vitality of neighboring areas. Overall, the proportion of areas with high vitality values at both scales is limited, indicating the necessity for further initiatives to stimulate new sources of vitality and enhance the overall dynamism of the city.

5. Analysis of Influencing Factors of Urban Vitality

5.1. Identification of the Key Influencing Factors of Urban Vitality

By employing the optimal parameter geographic detector for factor detection, the following results were obtained: Among all factors analyzed, only the correlation between morphological compactness and variations in subdistrict-scale vitality intensity failed to reach statistical significance at the 0.05 level, while all other factors exhibited statistical significance, indicating that a variety of variables influence the spatial distribution of urban vitality in Qingdao. Additionally, as illustrated by the q values (Figure 7), the significance of subdistrict-scale factors consistently exceeds that of block-scale factors. At the subdistrict scale, functional density (X3) exhibits the strongest correlation with the spatial distribution of urban vitality within the built environment. Other factors exerting a relatively large impact on subdistrict vitality, in descending order: commercial facilities accessibility (X12), public facilities accessibility (X13), road network density (X11), cultural and leisure facilities accessibility (X14), bus accessibility (X10) and metro accessibility (X9) comprise more than 50% of the spatial distribution of subdistrict vitality in the central area, thereby serving as the primary determinants of vitality intensity. At the block scale, functional density (X3) remains the factor most closely associated with the spatial distribution of urban vitality within the block environment. Other significant factors affecting block vitality, in order of importance, are public facilities accessibility (X13), commercial facilities accessibility (X12), road network density (X11) and cultural and leisure facilities accessibility (X14). These factors are pivotal in shaping the distribution of block vitality intensity. Morphological compactness (X7) demonstrates the lowest explanatory power at both the subdistrict and block scales.

Interaction detection allows for a deeper analysis of how driving factors interact and quantify their combined explanatory power. The q values for these interactions surpass those of individual factors (Figure 8), demonstrating that urban vitality intensity results from the synergistic effects of multiple influences. Upon analyzing factor interactions, two primary outcomes typically emerge: bi-factor enhancement and non-linear enhancement, and significant disparities exist in their interactions. Overall, the collaborative effect of morphological compactness (X7) with commercial facilities accessibility (X12) and functional density (X3) exerts the most pronounced impact on subdistrict-level urban vitality. At the block scale, the interaction between functional density (X3) and cultural and leisure facilities accessibility (X14) demonstrates the greatest influence, resulting in a stronger positive impact on urban vitality. Conversely, the combination of morphological compactness (X7) with resident population density (X1) and greening rate (X6) exhibits the lowest explanatory power at the block scale.

5.2. Exploration of the Spatial Relationship Between Urban Vitality and Influencing Factors

Prior to regression analysis, the selection of an appropriate spatial econometric model is essential. A multicollinearity assessment, employing the variance inflation factor (VIF), was conducted for urban vitality and 14 influencing factors within the central area of Qingdao. The results indicate VIF values exceeding 10 for building density, functional density and bus accessibility at the subdistrict scale, as well as for functional density at the block scale, thereby confirming the presence of multicollinearity among the variables. Following the removal of the aforementioned variables, the VIF test was repeated. The subsequent results indicated the absence of multicollinearity among the remaining factors. The details are presented in

Table 2. Covariance test results.

Modified Covariance Test (Subdistrict Level)		Modified Covariance Test (Block Level)
Variant	VIF Value	Variant	VIF Value
Resident population density (X1)	2.677	Resident population density (X1)	1.485
Unit area (X4)	1.728	Building density (X2)	1.878
Building height (X5)	3.849	Unit area (X4)	1.865
Greening rate (X6)	2.751	Building height (X5)	1.612
Morphological compactness (X7)	1.765	Greening rate (X6)	1.644
Functional mixing degree (X8)	3.270	Morphological compactness (X7)	1.457
Metro accessibility (X9)	5.188	Functional mixing degree (X8)	2.106
Road network density (X11)	7.663	Metro accessibility (X9)	1.683
Commercial facilities Accessibility (X12)	4.619	Bus accessibility (X10)	1.696
Public facilities Accessibility (X13)	9.130	Road network density (X11)	2.796
Cultural and leisure facilities Accessibility (X14)	4.053	Commercial facilities Accessibility (X12)	1.824
		Public facilities Accessibility (X13)	2.283
		Cultural and leisure facilities Accessibility (X14)	1.325

.

After passing the multicollinearity test, the ordinary least squares (OLS) method was used for analysis, and only the built environment variables with significant influence were retained, effectively enhancing the robustness of the model. Goodness-of-fit measures were then compared with those from the MGWR model (

Table 3. Model diagnostic coefficients.

Title 1	Subdistrict Level		Block Level
	OLS	MGWR	OLS	MGWR
R2	0.913	0.933	0.707	0.824
R2 adjusted	0.896	0.923	0.702	0.804
AICc	45.297	31.723	1331.317	1081.958

), which demonstrated the highest adjusted R² and strongest fit. Moreover, compared with the classic GWR model, the MGWR model considered the scale of the independent variables’ effects, providing support for exploring the spatial effects of various influencing factors. Therefore, the MGWR model was selected to study the spatial heterogeneity of driving factors in different regions (

Table 4. Descriptive statistics of MGWR regression coefficients for the urban vitality at the subdistrict scale.

Variable	Max	Median	Min	Average	Standard Deviation
Resident population density	0.413	0.208	−0.075	0.210	0.162
Building height	0.318	0.315	0.303	0.314	0.003
Commercial facility accessibility	0.415	0.411	0.407	0.411	0.002
Cultural and leisure facility accessibility	0.332	0.330	0.325	0.330	0.002

and

Table 5. Descriptive statistics of MGWR regression coefficients for the urban vitality at the block scale.

Variable	Max	Median	Min	Average	Standard Deviation
Resident population density	−0.058	−0.060	−0.070	−0.061	0.004
Building height	0.247	0.132	0.01132	0.123	0.066
Greening rate	0.020	−0.100	−0.239	−0.089	0.084
Functional mixing degree	0.098	0.091	0.074	0.089	0.007
Metro accessibility	0.145	−0.379	−1.870	−0.545	0.530
Road network density	0.240	0.136	−0.047	0.131	0.072
Commercial facility accessibility	0.347	0.198	0.015	0.174	0.103
Public facility accessibility	0.120	0.113	0.100	0.111	0.006
Cultural and leisure facility accessibility	0.647	0.102	−0.156	0.155	0.164

). The R² values are 0.933 at the subdistrict scale and 0.824 at the block scale, while the adjusted R² values are 0.923 and 0.804, respectively, indicating a significant association between the potential factors and urban vitality.

In terms of correlation strength, the MGWR model results are consistent with those obtained using the optimal parameter geographic detector method. At the subdistrict scale, the factors influencing comprehensive vitality, in descending order of impact, are commercial facilities accessibility, cultural and leisure facilities accessibility, building height and resident population density. At the block scale, the order of influence is metro accessibility, commercial facilities accessibility, cultural and leisure facilities accessibility, road network density, building height, public facilities accessibility, functional mixing degree, greening rate and resident population density. Commercial facilities accessibility and cultural and leisure facilities accessibility exert the greatest influence on comprehensive vitality at both the subdistrict and block scales, whereas resident population density and building height exhibit comparatively weaker effects. According to the standard deviation index, resident population density shows substantial spatial variation in its influence on subdistrict vitality, while the variability associated with commercial facilities accessibility and cultural and leisure facilities accessibility is minimal. For block vitality, metro accessibility demonstrates considerable spatial variability in its influence, whereas resident population density, public facilities accessibility and functional mixing degree display limited variability.

The model outputs were visualized and examined using ArcGIS10.8.1 (Figure 9 and Figure 10). At the macro level, four variables, namely the resident population density, building height, commercial facilities accessibility and cultural and leisure facilities accessibility, have a significant influence on the formation of urban vitality. With the exception of the resident population density, the other three variables exhibit a positive correlation with the urban vitality index at the subdistrict level. Commercial facilities accessibility showed the highest association with urban vitality, with the strongest correlation observed along Zhongshan Road and neighboring subdistricts in Shinan District, and reaching its lowest in Wangtai Town Subdistrict of Huangdao District and Xifu Town Subdistrict of Chengyang District, with spatial patterns indicating lower values in the western and eastern areas, higher values concentrated in the central zone and reduced levels in the peripheral surroundings. The cultural and leisure facilities accessibility exerted a stronger influence on urban vibrancy in eastern cities compared to western ones, with the eastern region showing a gradual increase in impact spreading from the east to the west, while the western region experienced a more subdued effect. Although the correlation between building height and urban vitality was weaker than that of cultural and leisure facilities accessibility, it was strongest along the eastern coastline, gradually decreasing laterally. The influence of resident population density on urban vitality is concentrated around the subdistricts of Shinan and Shibei Districts, exhibiting a radially declining pattern outward, while in Chengyang District, this effect gradually diminishes from west to east, with the weakest negative impact observed in the northwestern subdistricts.

At the micro-medium scale, urban vitality is significantly influenced by nine environmental factors: resident population density, building height, greening rate, functional mixing degree, metro accessibility, road network density, accessibility to commercial, public and cultural and leisure facilities. Within the dimensions of density and diversity, resident population density exhibits a slight negative association with urban vitality, with this correlation being strongest in the eastern neighborhoods and gradually diminishing toward the west, which suggests that, in densely populated areas, the resident population density has approached saturation. The functional mixing degree exhibits the strongest correlation with urban vitality along the eastern coastal neighborhoods, gradually diminishing toward the surrounding areas. Enhancing functional diversity can effectively improve regional vitality, with this effect being particularly pronounced in urban central areas.

Within the design dimension, building height is generally positively correlated with urban vitality, with the strongest associations observed in the blocks of Shinan District and the western and northeastern areas of Chengyang District. This correlation gradually weakens outward from these three focal areas, indicating that urban renewal in central districts can benefit from appropriately increasing building heights to accommodate a greater diversity of land uses. The relationship between greening rate and urban vitality is complex. In the northern neighborhoods, a negative correlation is observed, with the strongest adverse impact among all built environment indicators, indicating poorer conditions for green space development in urban fringe areas, while positive correlations are found in the western, central and southern neighborhoods, where the strength of this association gradually decreases from Shinan and Shibei Districts toward the surrounding areas.

Within the dimension of distance to transit, urban vitality exhibits the strongest correlation with metro accessibility, showing an overall negative relationship whereby proximity of a neighborhood center to a metro station enhances local vitality, and since metro stations are primarily concentrated in the central areas of each district, this effect is particularly pronounced in the peripheral zones of Huangdao and Chengyang Districts. An increase in road network density enhances regional accessibility and can, to some extent, stimulate the improvement in urban vitality; this effect is particularly pronounced in the northwestern Chengyang District and the western Huangdao District, while it is comparatively weaker in areas such as the scenic zones of southern Laoshan District.

In the dimension of destination accessibility, commercial facilities accessibility consistently shows a positive relationship with urban vitality, following a radial decline from the northeast toward the southwest. Likewise, public facilities accessibility influences urban vitality in a similar pattern, with values rising along the diagonal from southwest to northeast. This indicates that enhancing the commercial facilities accessibility and public facilities accessibility in the northeastern part of the city can significantly promote regional urban vitality. Cultural and leisure facilities accessibility exerts a distinctly polarized spatial effect on urban vitality. From a spatial distribution perspective, areas with high positive values exhibit a “one center, multiple clusters” pattern, with the core located in the northern neighborhoods of Huangdao District and multiple clusters distributed across the Laoshan District and the northeastern neighborhoods of Chengyang District. High-value negative areas are located in the northern neighborhoods of Chengyang District, which can be attributed to the area’s inadequate infrastructure and supporting service facilities, limiting its ability to attract populations and thereby enhance regional vitality.

6. Discussion

The concept of urban vitality, rooted in essential human activities, is complex and multidimensional, and its intensity cannot be fully measured or represented by relying on a single data source alone [54]. The integration of multi-source data enables urban planners and policymakers to conduct a more evaluation of urban vitality [55]. Combining Baidu heat map data, urban points of interest and night light images can more accurately depict the intensity of human activities within urban areas, thereby facilitating a deeper understanding of the spatial pattern of urban vitality distribution. Unlike earlier studies that relied on a single data source to analyze variations in urban vitality intensity, this research integrates multiple big datasets to quantitatively evaluate the spatial vitality of Qingdao’s central area, which provides a robust foundation for comprehensively exploring the multidimensional spatial driving factors influencing urban vitality.

Urban development requires the coordinated interaction of multiple spatial dimensions. The spatial variation in urban vitality is closely associated with diverse urban components across different spatial scales [56]. This study identifies that commercial facilities accessibility and cultural and leisure facilities accessibility have a significant influence on urban vitality, corroborating previous research [28]. This study, consistent with the findings of Li et al. [33] and Ta et al. [32], indicates that densely populated urban cores exhibit both positive and negative effects on urban vitality. At the block scale, a negative correlation between population density and urban vitality is observed, which aligns with the findings of Lu et al. [57]. Therefore, it is crucial to plan for mixed land use, appropriately infill available spaces to create more opportunities and reasonably allocate urban facilities. In addition, incorporating the concept of the 15 min city can enhance the accessibility of urban amenities, thereby stimulating the growth of active travel modes such as walking and cycling, achieving a compact city form, and promoting a win-win situation for urban vitality and sustainable transportation [58,59]. Future research should further explore the relationship between urban vitality planning and active travel, to understand how these can form a virtuous cycle and promote sustainable urban development.

Similarly, Xuan et al. observed that building height consistently exerts a positive influence on urban vitality, aligning with the results of the present research [27]. Excessive building heights may lead to a decline in residential comfort. Dense clusters of high-rise buildings are particularly prone to creating street canyon effects, which can hinder the dispersion of air pollutants [60] and intensify the urban heat island effect [61], resulting in adverse microclimatic conditions. These factors may ultimately have a negative impact on pedestrian comfort and outdoor activities, thereby indirectly diminishing urban vitality. Thus, increasing building heights to enhance urban vitality should be accompanied by reasonable planning to ensure sustainable improvements.

However, not all factors exert uniform effects across spatial scales. For instance, excessive functional diversity and randomness were found to impede urban vitality, in agreement with earlier findings [62,63]. This study reveals that the greening rate exhibits a negative influence on urban vitality, which is consistent with the findings of Ding et al. [64] and Li et al. [33]. Furthermore, metro accessibility exhibits an overall negative correlation with urban vitality at the block scale, suggesting that proximity to metro stations facilitates population aggregation and increased flow, as reported by previous studies [65]. These findings highlight the need for a comprehensive and multidimensional approach to enhancing urban vitality.

Multiple factors shape urban vitality, and effective urban development requires fostering synergy among these elements while recognizing how distinct components of the built environment uniquely affect a city’s vitality. According to regional functional zoning, the functional spatial organization of built environment factors requires optimization to achieve efficient resource allocation and to improve urban vitality. Informed by the findings of this study, targeted strategies for fostering urban vitality are proposed from multiple perspectives.

From the perspectives of functional diversity and spatial optimization, a diverse range of functional and commercial types is essential for stimulating economic activities and enhancing urban vitality. Research indicates that the interaction between morphological compactness and functional density significantly increases the vitality of cities at the subdistrict scale, while the combined effects of functional density and cultural and leisure facilities accessibility markedly improve the vitality of urban blocks. Therefore, the above and underground spaces can be utilized to develop multiple functions, enhance the construction of cultural and leisure facilities, adjust block spatial forms, and invigorate the city. Concurrently, land use functions can be diversified and enriched. Through the processes of renewal and transformation, the land use structure can be effectively adjusted. In areas where large-scale renewal is challenging, targeted, small-scale functional interventions can be implemented to meet diverse demands.

From the perspectives of built environment and infrastructure, a well-developed urban environment and transportation infrastructure form the foundation for attracting people and enhancing urban vitality, while increases in building height and density provide essential physical space for population aggregation. However, targeted management based on local conditions and the implementation of differentiated improvement strategies are necessary. For example, in city centers, spatial pressure can be reduced by appropriately reducing building density and increasing building height, combined with additional green spaces. The enhancement of public transportation systems facilitates convenient mobility, with the development of bus and subway stations creating efficient conditions for movement and gathering. The provision of diverse facilities establishes a robust basis for vibrant urban spaces, contributing significantly to the rational allocation of spatial resources and the promotion of urban vitality. At broader spatial scales, factors such as resident population density, building height, commercial facilities accessibility and cultural and leisure facilities accessibility have a substantial influence on urban vitality. Consequently, urban built environment development should carefully balance considerations across different spatial scales and characteristics to effectively meet the requirements for enhancing urban vitality at multiple levels.

From the perspectives of regional coordination and functional integration, future spatial planning should continue to leverage the radiative and driving functions of regional centers to optimize population distribution within the central area of Qingdao. Secondary business districts, predominantly located on the urban periphery, experience constrained population mobility and limited vitality diffusion due to natural boundaries, urban road networks and large-scale gated communities. Strategies such as enhancing functional diversity in the eastern sector of Chengyang District and increasing the density of subway stations may effectively promote population aggregation. By capitalizing on the economic influence of the Huangdao Business District, Chengyang Business District and the central business district (CBD), balanced development across economic centers in various directions within Qingdao’s central area can be advanced. Meanwhile, moderately opening enclosed blocks and introducing pedestrian-friendly connectivity channels can facilitate intra-block circulation, enhance spatial interactions, and boost overall urban vitality. Additionally, prioritizing improvements to cultural and leisure facilities in the southwestern Shinan District, Shibei District, eastern Chengyang District and Huangdao District is essential for fostering the coordinated development of economic and cultural functions within regional centers. The strategic integration of diverse functions—such as commercial enterprises, public amenities and creative cultural industries—can generate new nodes of vitality, establishing a dynamic cultural, creative and leisure business model that attracts population flows, enhances the city’s image and ultimately reinforces the vibrancy of these areas.

From a policy perspective, this study explores the main spatial drivers underlying the evolution of urban vitality based on spatial distribution differences at multiple scales. The findings can assist urban planning authorities in clarifying the spatial arrangement of functional zones and infrastructure, as well as formulating long-term regional development strategies. This, in turn, may help to mitigate the loss of urban vitality and address issues of uneven distribution. Recent policy shifts have moved away from large-scale land expansion, instead emphasizing urban regeneration and the improvement in existing urban structures, particularly in older neighborhoods and “urban villages” with insufficient public facilities [66]. The results of this study support these policy directions by highlighting the importance of infill development, functional diversity and improved accessibility in revitalizing urban spaces.

As a representative city of China’s rapid economic growth in recent years, Qingdao’s central urban vitality has also increased annually [28]. However, the analysis of vitality evolution reveals challenges such as environmental degradation, inefficient land use and the emergence of “ghost cities” characterized by excessive and underutilized housing stock. Urban managers should be alert to the risks of planning imbalance—over-concentration of resources in core areas may exacerbate spatial inequalities, while unrestrained expansion in peripheral zones may result in low land use efficiency and declining urban vitality. Therefore, adopting balanced and context-specific planning approaches is essential. Looking ahead, as urbanization in China transitions into a more stable and mature phase, planning and development should continue to adhere to people-oriented and sustainable principles, emphasize the quality of urban growth, address the needs of urban residents and ensure the sustained vitality of central urban areas.

7. Conclusions

Within the context of urban regeneration and district redevelopment initiatives, analyzing the spatial differentiation patterns and formation mechanisms of urban vitality from a multi-scale perspective holds significant theoretical value and practical significance for the assessment and revitalization of cities. This research innovatively employs multi-source data to provide urban planners and policymakers with an effective framework for evaluating urban vitality, offering a replicable model for other cities. By integrating big data, this study quantitatively assesses the economic, social and cultural vitality of central Qingdao and characterizes the spatial distribution and clustering patterns of urban vitality at both subdistrict and block scales. Through the development of a comprehensive 5D built environment indicator system, this study quantifies environmental impact factors and employs the OPGD and MGWR models to elucidate the mechanisms through which the built environment influences urban vitality. The findings provide evidence-based strategies for both micro-scale vitality optimization and macro-level urban quality enhancement, offering valuable references for future urban development initiatives. The principal conclusions are as follows:

(1) Urban vitality in central Qingdao displays a spatial structure characterized by a “weak east-west, strong central, multi-center, cluster-structured” pattern, with higher vitality predominantly concentrated in the core of the central area and diminishing unevenly in clustered patterns radiating outward from the business district core. The vitality pole located in the southern sector of the central area is well established, whereas smaller vitality poles in the northern and southeastern regions are still developing and require further efforts to enhance their attractiveness.

(2) The distribution of urban vitality exhibits significant spatial autocorrelation. At the subdistrict scale, high-high vitality clusters are concentrated within the central urban districts of Shinan and Shibei, gradually extending from the core toward peripheral areas, and low-low vitality clusters are primarily located distributed in the northern and southwestern parts of Huangdao District, Chengyang District and the northern section of Licang District. At the block scale, new patterns of high-high and low-low clusters have emerged, building upon the spatial distribution observed at the subdistrict scale.

(3) Urban vitality aggregation arises from the combined effects of multiple factors, with interactions between any two factors exerting a significantly greater influence than individual factors alone. Notably, the interaction between the three factors of functional density, commercial facilities accessibility and public facilities accessibility and other factors constitutes the primary determinant influencing urban vitality intensity at both scales. The interaction between building density and morphological compactness exhibits the least influence at the subdistrict scale, while at the block scale, the interaction between morphological compactness, resident population density and greening rate exerts the lowest impact.

(4) The impact of individual factors on urban vitality varies considerably across different spatial variability. The indicators most positively correlated with urban vitality are commercial facilities accessibility and cultural and leisure facilities accessibility. A moderate increase in building height promotes urban vitality, while resident population density exerts both positive and negative effects on urban vitality at the subdistrict scale, while it has a predominantly negative impact on urban vitality at the block scale. In addition to the aforementioned factors, other dimensions such as density, design and distance to transit positively contribute to urban vitality at the block scale, playing a significant role in its enhancement. Among these, commercial facilities accessibility exerts a stronger positive effect on urban vitality, whereas metro accessibility is generally negatively correlated with urban vitality, exhibiting a spatial pattern that decreases from the urban periphery toward the central areas of each district.

Several limitations of this study warrant acknowledgment. Baidu Heat Map data predominantly reflects the activity scope of middle-aged and young populations. However, it inadequately represents the activity spaces of groups such as juveniles and the elderly, who use mobile phones less frequently. This study conducted a cross-sectional comparison of urban vitality at the subdistrict and block scales but did not explore the 24 h diurnal variation trends and their influencing factors through a multi-stage longitudinal approach. Future research should leverage urban vitality and built environment data across multiple time periods to analyze the spatiotemporal evolution of urban vitality more comprehensively. At the same time, more days and data from different seasons should be collected to avoid the limitations caused by seasonal differences and insufficient days. Moreover, employing nighttime light data as a proxy for economic vitality is limited by the spatial scale of the study units. Subsequent studies could enhance economic vitality assessment by integrating additional indicators, such as the density of operating restaurants and the number of commercial institutions, to provide a more comprehensive representation of regional economic dynamics [57,58]. In terms of dimensionality, this study included indicators reflecting human environmental perception, such as greening rate; however, human perception of environmental emotions is multifaceted. Future research should incorporate additional elements, including emotional responses to different environments, into the indicator framework to more fully capture the influence of human perception on block-level vitality. Furthermore, conducting longitudinal monitoring of urban vitality to capture its dynamic changes over time warrants further investigation.