Re-Examining Urban Vitality through Jane Jacobs’ Criteria Using GIS-sDNA: The Case of Qingdao, China

Abstract

This study focuses on the assessment of historic city vitality to address increasingly fragmented urban patterns and to prevent the decline of livability in older urban areas. In 1961, Jane Jacobs theorized urban vitality and found the main conditions that were required for the promotion of life in cities: diversity of land use, small block sizes, diversity of buildings with varied characteristics and ages, density of people and buildings, accessibility for all people without depending on private transport, and distance to border elements. Jacobs’ criteria for urban vitality has had an indisputable influence on urban researchers and planners especially in the Anglo-American context. This perspective has influenced the development of New Urbanism and similar planning policies, such as neo-traditional communities and transit oriented development, yet her theories have to be more substantiated in Asia’s developing cities, especially in China’s historic cities. In order to verify the significance of Jacobs’ urban vitality theory in Chinese historic cities, we develop a composite measure of 16 variables of built environment, and we test it using GIS-sDNA in a historic city with an aging population and low-income in Qingdao. A systematic approach to urban spatial analysis allows us to provide a detailed spatial interpretation of a historic city form. The results emphasize that historic cities vitality, far from being homogeneous, followed a multi-centered distribution pattern, which is related to the previous European planning of the region, where a grid-type pattern was more likely to disperse urban vitality. The results can serve as a useful framework for studying the livability and vitality of different areas of the city in different geographical contexts.

1. Introduction

The world is experiencing rapid urbanization, especially in developing countries, and while urbanization has significantly improved living standards, it has also brought about unprecedented urban problems, such as land Fiscal-driven urbanization, hollowing out of old urban areas, and “ghost cities” [1]. In this context, historic cities generally face double decay of social vitality and physical morphology, such as over-tourism, gentrification, aging and impoverishment of aborigines, aging infrastructure, and a deteriorating living environment [2,3,4]. Historic city vitality is an isomorphism of the urban morphology and the social activities behind it [5], determined by the spatial distribution pattern of the elements of the historic city form [6,7]. The unique urban scenes of the historic city have often undergone centuries of shaping and are significantly different from what is commonly thought of as vibrant urban areas [8,9]. With rapid urbanization, the original functional types and spatial patterns of the historic city can no longer meet the demands of modern life and require necessary adjustments to meet the needs of residents’ lives and the needs of regional development. Numerous studies have shown that the distribution of urban vitality can be significantly influenced by the planning of urban form elements [10,11,12,13,14]. Therefore, in different urban contexts, how to evaluate the vitality of historical cities more rational and how to propose targeted strategies to enhance them has become a major research problem.

“The Death and Life of Great American Cities” (Jacobs, 1961) provided a landmark for urban planning theory with abiding influences [15]. This perspective is considered the benchmark for understanding how neighborhood vitality works today [16]. The mix of building functions and densities that meet the needs of citizens is fundamental to the urban vitality [17]. Her criteria have been recurrently rediscovered and revisited to respond to different needs and from different contexts and are still relevant in the present day. The outbreak of the COVID-19 has once again led to a consideration of Jacobs’ theory and the revival of urban vitality. Observing the changes of the epidemic, it can be seen that traditional unit communities within historical cities have better epidemic prevention effects due to their relatively open boundaries and the diversity of functions within the community, carrying compounded functions within the residential units, than, for example, new commercial housing communities on the outskirts of large cities. Cities and other public institutions around the world are also explicitly or implicitly using her principles as part of planning strategies for the revitalization of local historic cities [18]. As a result, her theories have also received increasing attention from urban researchers around the world, who use her general view of urban vitality as a framework for reflecting on the success or failure of urban planning in ensuring urban quality [19]. The current urban vitality study aims to validate her ideas by assessing the degree of mixing and spatial distribution of buildings and facilities in the city [20,21,22,23,24].

In terms of the analytical approach to urban vitality, Jacobs argues that the city is an ever-expanding network of mobility, with population expansion, size, and traffic congestion problems expanding outward as the city itself becomes saturated [15]. Urban space is a collection of information flows and networks, and a growing number of studies prove that many urban phenomena are essentially network phenomena, such as the hierarchical scale, resilience, agglomeration, or scale efficiency of cities [25,26]. In network analysis methods, spatial information plays a leading role, containing location data, allowing us to study the physical underpinnings that govern and shape the flow of human information in urban spaces. [27]. More recently, the centrality assessment model and space syntax analysis have been used to evaluate the structural properties of street networks in an urban system. Street centrality indices representing closeness, betweenness, and straightness capture the skeleton of the urban system; these factors shape economic activities and land use intensity [28,29]. A research team from Cardiff University, UK, developed the spatial design network analysis (sDNA) tool [30], which proposes a line segment model that is based on the spatial syntax axis model, while the topological relationships between streets can be weighted with information such as street length and geometric angular distance to reflect real geospatial information, while compensating for the failure of spatial syntax to capture physical isolation and network efficiency, especially the navigation difficulties and psychological barriers of pedestrians [31]. Jacobs’ vitality criteria are directly mirrored in sDNA, which incorporates four important indexes (density, connectivity, closeness, and betweenness) that are hypothesized to affect urban vitality in an urban system. In China, driven by both industrialization and urbanization, numerous satellite towns, industrial zones, commercial centers, and residential areas have emerged at the historic city fringe and pose significant challenges to sustainable urban development [32]. Until recently, empirical evidence in developing countries has indicated that the dynamic process and determinants of urban vitality can be different from the case studies in the United States or Europe due to their dissimilar historic urban morphology and territorial spatial planning policies. A lot of historic cities launched large-scale street planning projects and replicated the modern style of broad and grid roads from the US, thereby reducing their urban vitality [33].

The following questions are posed in this study: (1) What are the implications of Jacobs’ criteria for urban vitality in the context of historic cities? (2) How to construct the evaluation index system of “JANE Index” of Asian historical cities based on Jacobs’ criteria? 3) How to visualize and analyze the vitality of Qingdao’s historical city through GIS-sDNA and other related methods, and propose targeted vitality regeneration strategies? Therefore, this study aims to enrich the existing empirical studies by assessing the level of vitality that is exhibited by each neighborhood of a historic city and formulate strategies to enhance urban vitality.

2. Jacobs’ Criteria of Urban Vitality

Jacobs defines urban vitality as “the production of a diverse urban life consisting of human activities and living places” [15]. She proposed four main conditions for the creation of urban vitality: diversity of land use, small block sizes, diversity of buildings with varied characteristics and ages, and density of people and buildings, as well as two secondary conditions: accessibility without reliance on private transportation and distance from boundary elements [20].

Land use diversity, similar to architectural diversity of different characteristics and ages, is seen as paramount to maintaining urban vitality. She believes that the mix of functions creates spatial differentiation and increases the efficiency of the use of urban facilities, such as reducing commuter traffic and increasing the potential for public transportation use in a high-density environment [34,35]. On the other hand, building diversity also reflects the diverse mix of socio-economic groups [36]. Jacobs paid close attention to the “old building needs” of urban neighborhoods and argued that the healthiest areas of the city need not only old buildings, but also new buildings that are interspersed with old buildings. She noted that old buildings and new buildings require different levels of economic gain, i.e., the economic return that is generated in the building, and that new businesses tend to naturally emerge in buildings with low economic overhead. Th diversity index measures the degree of concentration, mix, and proportional relationships by measuring the number of different building types at a given scale [37]. The diversity index makes a morphological description, which calculates the number of buildings, number of functional types, and evenness in a site unit, quantifying the degree of spatial diversity [38]. When urban facilities show a high number and type with even distribution in space, this pattern is positively correlated with urban vitality.

Small-scale street blocks are based on the need for social spaces, which helps to create connections between people in specific areas of the city. This index is used to measure the connectivity characteristics of the street network, such as the number of intersections, the average distance between intersections, and other variables [39]. Studies have shown that potential traffic routes can be increased if the distance between intersections in a street network is shorter and the number of intersections is higher at the same scale. However, too short intersection distances with too many intersections can lead to a larger road network footprint and adversely affect the layout of other facilities [40]. A reasonable proportional and quantitative distribution of small-scale streets is positively correlated with urban vitality. Connectivity is only possible when streets vary in size and each size has a reasonable proportion of streets.

The influence of building and population density on vitality builds on the above indexes and emphasizes the degree of aggregation of population and activity places at a certain spatial scale [41,42]. The higher the building density, the higher the pedestrian traffic on the street. However, dense land occupation should not be confused with the construction of large apartment buildings, and areas with high housing density are often seen as a disadvantage. The presence of public services, commercial facilities, etc., is related to Jacobs’ concept of the “eye on the street,” which means that the constant presence of people creates a “natural surveillance” system that has also been shown to improve safety [43]. Compared to large public spaces, small infrastructures can be considered as potential social places. The POI kernel density calculation method has proven to be a valid method for vibrancy evaluation, for example for recreational facilities, small restaurant facilities, green infrastructures, and other types of calculations [7,44,45].

In terms of accessibility as one of the two supplementary conditions for urban vitality, in contrast to automobile-driven urban planning, studies have generally referred either to the supply of public transportation on the one hand, or to walkability conditions on the other [46]. In terms of walkability, sDNA’s closeness index measures walkability within a specified radius. Boundary vacuum refers to the segregated forms of urban existence, in the form of barrier surfaces or lines, including functionally homogeneous urban spaces (i.e., large parks), natural elements that could act as barriers (i.e., rivers), and ground-level heavy transportation infrastructures (i.e., railways). Jacobs argued that these elements adversely affect urban vitality through man-made impermeable boundaries [22,46].

These six criteria that are necessary for the generation of urban vitality, as proposed by Jacobs, form the basis of the theoretical approach that was analyzed in this paper. Jacobs’ influence on urban theory is most evident in Anglo-Saxon contexts. Her work has also inspired European, South American, and Asian urban studies on how to respond to urban development issues, such as the case studies of Barcelona, Cyprus, Santiago, and Seoul [20,21,46,47]. Revisiting Jacobs’ theory of urban vitality is necessary because her focus on the incremental revitalization of already urbanized areas, rather than the one-time redevelopment of master plans [36] is consistent with the focus of current considerations in developing countries to address the low vitality of historic cities that are brought about by rapid urbanization. Although most applied studies that are based on Jacobs’ original evaluation indexes have been conducted for various purposes, the means of handling the evaluation indexes are basically the same, usually transforming them into multiple morphological variables, adaptively weighting the set of values, and overlaying them into a comprehensive urban vitality index.

3. Methods

3.1. Study Area

This study focuses on the historic city of Qingdao, China (Figure 1). Qingdao city is located in the southeastern of Shandong peninsula (latitude 35°35′–37°09′ N, longitude 119°30′–121°00′ E) in eastern China, with a total area of about 28 km². The historic city area includes 14 historic conservation areas with a total area of 13.041 km² and a core protection area of 6.394 km², including 549 cultural relics protection units at all levels, 309 historical buildings, 1694 traditional style buildings. and 39 industrial heritage sites within the city area. The average population density is 8059 people/km². The distribution of landscape elements in each historic conservation area varies greatly, which is related to the adjustment of urban form planning and industrial structure in different historical stages. Until the end of the 19th century, Qingdao was a rural area with agriculture and fishing as the mainstay. In 1898, German colonialists forcibly leased Jiaozhou Bay, and Qingdao then entered colonial city stage. Urban planning of Qingdao was dominated by German colonialists, and the layout of city conformed to the topography and climate of the seaside hills. The urban residential area was divided into European and Chinese areas. The European area was spread along the southern coastal flat area, and the nature of the land was mainly for European residence, leisure, and entertainment, with outstanding landscape conditions. The Chinese area was located in the northern area of the city, including Dabaodao district, Taidong town, and Taisi town. Based on the unequal colonial zoning pattern of German-occupied Qingdao, there are huge differences between the European and Chinese districts in terms of urban location, plot scale, road width, building density, green landscape, and other aspects of the living environment. In 1902, the population density of the European areas was only 19.1 people/ha, and building density was only 20–25%; the population density of the Chinese areas was 417 people/ha, and building density was over 75%. In terms of green space, the European area accounted for 72.6% of the city’s greenery, while there was no public green space within the Dabaodao district. Urban planning and construction in Qingdao during the German occupation period (1897–1914) established the spatial morphological diversity of Qingdao’s historical urban area, resulting in six typical morphological types (Figure 2): (a) south of Guanhai Mountain (former European area), (b) south of Zhongshan Road (former European area), (c) Guanhai Mountain (former European area), (d) Dabaodao (former Chinese merchant settlement, (e) Taidong Town (former Chinese laborers settlement), and (f) Taisi Town (former Chinese poor settlement).

We choose Qingdao as a case study of the application of Jacobs’ theory to Asian cities because Asia, as the origin of human civilization and the most populous continent in the world, has accumulated a profound urban culture during the course of history, and its urban morphological development has the following characteristics: 1. High-density concentration of population and buildings, accompanied by a long-standing phenomenon of large differences in scale, class, and form in the central area. 2. Since 1897, Qingdao, as a modern colonial city, has experienced colonial urban planning led by Germany and Japan, and its urban development has both a technologically advanced industrial and trade base and a unique urban regional culture, with distinctive power in terms of street texture, neighborhood scale, green space ratio, and it has distinctive power and hierarchical differences in terms of street texture, neighborhood scale, green space ratio, building density, etc. It conforms to the general characteristics of Asian historical urban form while maintaining a unique regional culture.

3.2. Variables and Data Sources

This paper is based on Jacobs’ theory of urban vitality, focusing on built environment characteristics, i.e., street form, land use, and human activity factors, adapted to the urban context of Qingdao, China, and translates Jacobs’ theory into the general empirical model in Equation (1). Other variables were combined based on recent literature exploring vitality and its drivers in the Jacobs framework. The result is the adjusted JANE index, which includes 16 indicators that were compiled from different data sources. The model is based on the JANE index of urban vitality that was first constructed by Xavier Delclòs-Alió (2018).

Table 1. Conditions and indicators that were used in the analysis.

Jacobs Criteria	Indicators	Calculation Method	Formula	Number
1. Building function diversity ($Div$)	01. Land use mixture ($D_1$)	Shannon–Wiener index of building function.	$D_1=-{\displaystyle {\displaystyle \sum }_{i=1}^n}(p_i·\ln p_i)$	(2)
	02. Residential and non-residential ratio ($D_2)$	Residential and non-residential balance (0–1).	$D_2=1-\left|\frac{Re{s}_i-NonRe{s}_i}{Re{s}_i+NonRe{s}_i}\right|$	(3)
	03. Commercial and facility ratio ($D_3$)	Commercial facility and public facility mix (0–1).	$D_3=1-\left|\frac{ComFa{c}_i-PubFa{c}_i}{ComFa{c}_i+PubFa{c}_i}\right|$	(4)
2. Road scale ($Sca$)	04. Block size ($S_1$)	Area of the traffic analysis zone	$S_1=S_i$	(5)
	05. Road centerline density ($S_2$)	Total length of road centerline in TAZ unit.	$S_2=\frac{L_i }{S_i }$	(6)
	06. Road centerline connectivity ($S_3$)	Number of interconnections per road in TAZ unit.	$S_3=K_i$	(7)
	07. Road centerline betweenness ($S_4$)	Sum of geodesics that pass through a street in TAZ unit.	$S_4={\displaystyle {\displaystyle \sum }_{y\in N}}{\displaystyle {\displaystyle \sum }_{z\in R_y}}OD\left(y,z,x\right)\frac{P\left(z\right)}{links\left(y\right)}$	(8)
3. Building age diversity ($Di{v}_A$)	08. Building age of construction ($D{A}_1$)	Average building age of construction.
	09. Diversity of building age of construction ($D{A}_2$)	Diversity of average building age of construction (0–1).
4. Concentricity ($Con$)	10. Buildings density ($C_1$)	Building footprint in TAZ unit.	$C_1=\frac{M_i}{S_N}$	(9)
	11. Floor area density ($C_2$)	Total building area in TAZ unit.	$C_2=\frac{Q_i}{S_N}$	(10)
	12. People density ($C_3$)	Number of population in TAZ unit.	$C_3=\frac{N_i}{S_i}$	(11)
	13. Public facility density ($C_4$)	Kernel density values for public service facilities.	$C_4={\displaystyle {\displaystyle \sum }_{i=1}^n}\frac{1}{r^2}k\left(\frac{x-x_i}{r}\right)$	(12)
5. Accessibility ($Acc$)	14. Road centerline closeness ($A_1$)	Network quantity penalized for distance: for all streets in TAZ unit.	$A_1={\displaystyle {\displaystyle \sum }_{y\in Rx}}\frac{W\left(y\right)P\left(y\right)}{d_M\left(x,y\right)}$	(13)
	15. Distance to public transportation ($A_2$)	Nearest distance of TAZ unit to adjacent public transportation facilities.	$A_2=D_0-Min\left(D_{ip}\right)$	(14)
6. Boundary vacuum ($Bou$)	16. Distance to boundary vacuum ($DB$)	Nearest distance of TAZ unit to adjacent boundary vacuum elements.	$DB=D_0-Min\left(D_{ib}\right)$	(15)

Source: compiled by the authors.

details the variables that were used to construct these six dimensions and their respective data sources. Most of the data were obtained from official sources, while the rest were collected from spatial analysis by the research team using a geographic data system.

$$JANEIndex=f\left(Div,Sca,{Div}_A,Con,Acc,Bou\right)$$

(1)

Div = Function diversity, (land use mixture, residential and non-residential ratio, commercial and facility ratio)

Sca = Road scale, (block size, road length density, connectivity and betweenness)

Div_A = Building age diversity, (average building age, building age diversity)

Con = Concentricity, (building and floor area density, public facility density, people density)

Acc = Accessibility, (road closeness, distance to public transportation)

Bou = Distance to large single-use buildings, large parks, surface large roadways, parking areas, and empty lots

Building function diversity (Div.) is defined as the presence of land use mixture (D₁), residential and non-residential ratio (D₂) and commercial and facility ratio (D₃). In this study, based on the calculation method of information entropy, the Shannon–Weaver diversity index was used to calculate the land use diversity of spatial units using the main 11 land use function POIs (residential, commercial, work-related, recreational, and others) that were extracted from Amap, where n represents richness and p_i with proportional richness. When all building types in the dataset are the same, all p_i values are 1/n. The larger the difference in building type richness, the larger the weighted geometric mean ln p_i, and the smaller the corresponding D₁ value. If practically all the richness values are concentrated in one building type and the other types are very rare (even if they are numerous), the D₁ value is close to 0. When there is only one type in space unit, the D₁ value is equal to 0 [48,49]. The D₂ index was created with the expression as in the table, where Res_i refers to residential uses and NonRes_i to non-residential uses. Both indices take values from 0 to 1.

For road scale (Sca), the indexes included block size (S₁), road density (S₂), connectivity (S₃), and betweenness (S₄). The street block size was first calculated by GIS, and then the road density, connection value, and penetration were calculated using sDNA with a radius of 15 min walking distance (1500 m). Where density indicates the length ($L_i$) of a road within specified radius. The connectivity value indicates the sum of the street and all other connected streets ($K_i$), and the node with a higher connectivity value is the center of the street network. Usually the streets with more intersections are more connected and the ends are less connected. Betweenness is a measure of centrality that is based on the shortest path, which indicates the likelihood and frequency of a person passing a node in a street network movement [31,40]. Since people prefer to follow straight paths and angular distances in daily life because they are easier to remember and faster on average, this paper emphasizes the importance of angular distance (the distance that is measured as angular change) and prefers to use the shortest angular path rather than Euclidean distance. Then, we use the angle-weighted TPBtA metric from the sDNA theoretical framework for the measurement with the expression as in the table where N is the set of streets, R_y is the set of streets within the proposed radius from street y, and OD denotes the ratio of beginning to end in the penetration, between the interval (0, 1). z and y denote the two endpoints of the penetration path, and x denotes the measurement point of the street penetration degree. TPBtA measures the “through” potential of a street, which is positively correlated with house prices and rents, traffic flow, population density, and commuting flows in the street network [50].

As for building age diversity (Div_A), Jacobs highlighted that cities need to ensure that buildings with different characteristics are present in order to guarantee a certain degree of socioeconomic diversity. “Plain, ordinary, low-value old buildings” are critically needed environments for small, entrepreneurial businesses and healthy districts and cities” [36,51]. With this intention, we first included building age of construction (DA₁) and diversity of building age of construction (DA₂), which correspond to the original concept of Jacobs. In Qingdao historic city, strict restrictions on the preservation of historic conservation areas, as well as phenomena such as tourism and gentrification, can clearly affect building age diversity. Too many old buildings can reduce the diversity of old and new buildings, thus affecting the self-organization of historic urban renewal and discouraging the generation of urban vitality.

In order to study the concentration criteria (Con), we have considered building density (C₁), floor area ratio (C₂), population density (C₃), and public service facilities density (C₄). Building density is the ratio of building projection area ($M_i$) to building land area ($S_N$), reflecting the open space rate per unit space and the density of building coverage. The higher building density reflects the higher degree of intensive use of urban space. The floor area ratio is a concentrated reflection of the development intensity of urban space, and a higher development intensity means a higher degree of land use. The population density was calculated from Landsat Enhanced Thematic Mapper (ETM) satellite images with a 100 m spatial resolution provided by Worldpop. Public service facilities density is calculated by kernel density tool of ArcGIS. The principle is to define a threshold range (a circle of radius r) that is centered on the location of selected elements, and the kernel density value reaches a maximum at each core element n_i and decreases as it moves away from n_i until it drops to 0 when the distance from n_i reaches the threshold r. Due to the properties of buildings that are distributed along roads, this method is suitable for analyzing the concentration characteristics of infrastructure points in urban networks [44,52]. In this study, kernel density is considered as a cluster characterizing the concentration, and the peak area indicates the area of vitality hotspots, and the expression is shown in the table, where k is kernel function, r is distance decay threshold, n is the number of elements in threshold range, and (x-x_i) is the distance from point x to sample x_i. The size of the bandwidth has a significant impact on the accuracy of the analysis results [53]. In order to accurately reflect the concentration characteristics and to consider the discrete degree of building distribution and its average influence range, the bandwidth is set to 1 km in this paper.

Accessibility (Acc) is defined by the closeness of the road network (A₁) and pedestrian accessibility of transportation facilities (A₂). Closeness calculates the accessibility of the spatial structure of the street network through sDNA, which measures the difficulty of directing each street within a given radius to all possible destinations. A path with high closeness usually has high accessibility and it is easier to reach a further place, so that people can more easily reach the space from the surrounding areas for activities and social interactions, thus stimulating space vitality [40]. In this paper, the angle-weighted NQPDA model in the sDNA theoretical framework is used as the expression of closeness, as shown in the table, where d_M (x, y) denotes the shortest angular distance from street x to y based on the metric system. Rx denotes the set of polylines starting from street x within the proposed radius. W(y) denotes the weight of polyline y. P(y) denotes the ratio of any polyline y within the radius, when in discrete space, P(y) = 1 if the point is within the search radius, otherwise P(y) = 0; in continuous space, it is determined according to the ratio of radius to the length of segment, 0 ≤ P(y) ≤ 1. NQPDA can reflect street accessibility and mobility potential and is closely related to diverse land uses [54]. Pedestrian accessibility of transportation facilities measures the convenience of pedestrian access to public transportation services and is mainly related to the distribution of public transportation facilities (e.g., bus stops, subway stations, etc.) on a spatial scale. If the public transportation stations are distributed within the scale that is comfortable and reachable by walking, the higher people’s willingness to choose public transportation; otherwise, the willingness will decrease. In this paper, when discussing pedestrian accessibility, we mainly choose rail transit. Considering the quickness of rail transit station transfer, this paper takes the straight-line distance between each spatial unit and its nearest rail station as the A₂ index and uses the ArcGIS nearest neighbor analysis tool for spatial analysis, and the expression is shown in the table, where D_ip is the spatial linear distance between spatial unit i and rail transit station, and then the minimum value Min (D_ip) is taken among all the linear distances. D₀ is the minimum distance constant, due to the different spatial scales of the whole study area, the range of values is different, this paper takes the value of 1500. If the distance between the spatial unit and the rail station is smaller, the larger A₂ is.

Lastly, we incorporate in the analysis distance from border vacuums (Bou), taking into account single-use buildings (5000 m² and above), large transportation infrastructures, single-use extensive service or administrative buildings, and also large parks (5000 m² and above).

3.3. Data Processing and Calculations

For the study area, we selected the conservation planning area (2020–2035) of Qingdao historic city given by Qingdao Natural Resources and Planning Bureau, and firstly corrected the projection coordinate system for the main road layers in the road network that were extracted by OSM, and then merged and physically interrupted them to generate a total of 5949 street links and checked all the nodes. In the second step, the spatial unit division of the traffic analysis zone (TAZ) is used to convert the road network into polygon elements and to delete, trim, and merge the irregular road network in order to partition the study area into irregular polygons (Figure 3a). In addition, the fragmented parcels with very small areas were merged with neighboring parcels for the convenience of statistics and calculation, and the study area was divided into 2281 study units after final processing. In the third step, this paper selects and reclassifies a total of 11 functional types including residential land, land for public administration and public service facilities, land for commercial service facilities, land for roads and transportation facilities, and land for green spaces and squares according to the Urban Land Classification and Planning and Construction Land Standard (GB50137-2011) that was issued by the Ministry of Housing and Urban-Rural Development of China, including 10,301 for shopping, 7451 for living services, 7384 for restaurants, 3100 for residential areas and services, 2633 for transportation facilities, 1891 for culture and education, 1860 for medical facilities, 1252 for government offices, 848 for financial services, 834 for recreational facilities, and 215 for natural landscapes, totaling 37,769 POI data for spatial analysis (Figure 3b).

The number of each type of POI in each study unit and within 10 m of the boundary was counted in diversity analysis, and the Shannon index was calculated for each unit. Subsequently, we standardized the data for extreme differences in view of the diversity of each variable unit, divided all the results of index quantification into 8 categories in ArcGIS using the natural discontinuity method [55], and finally summed up to obtain the graded results of diversity evaluation of each study unit (Figure 4).

In the calculation of the street scale and accessibility, the road data were vectorized and preprocessed through processes such as coordinate correction on the ArcGIS platform. By breaking all of the “link” intersections and checking all of the intersections, we obtained 5949 road network segments. To remove the errors in the network, we ran the prepare network tool on the transport vector layer in the sDNA toolbox and obtained a road network spatial database of 4863 road network segments. Finally, we performed “integrated analysis” on the output model, choose “Angle” weighting as the calculation type in the parameter settings, and set the analysis radius to 1500 m to calculate the local spatial scale, where the radius type is set to “continuous space” to obtain more accurate calculation results. The output of the integrated analysis is the result of the spatial syntax analysis of each index (Figure 5).

In the building age diversity analysis, we used urban real estate statistics and connected them to TAZ units to obtain Figure 6. For the concentration analysis, this paper uses 100 m resolution open space population density statistics that were provided by WorldPop opensource dataset. There were four public service facilities of transportation, culture, education, and healthcare that were selected, including 2659 POIs of transportation service facilities, 215 cultural service facilities, 1919 educational service facilities, and 1866 medical service facilities, totaling 6659 POIs of basic service facilities. Then, we imported the obtained data into ArcGIS and carried out the kernel density calculation of point elements separately. Among the parameter settings, the image element size was set to 5.6 m, and the search radius was set to 600 m based on the 10-min walking radius. Finally, we used the entropy weight method to assign weights to the above four POI types, and assigned 0.0956, 0.4427, 0.1488, and 0.3128 weight ratios in turn, and overlapped with TAZ units to analyze the grading results of the comprehensive evaluation of concentration (Figure 7).

In order to combine the 16 constituent indexes with the corresponding urban vitality criteria, we weighted them by applying the raw z-values. Among the six dimension indicators, Jacobs regards the first four as the primary conditions for the generation of urban vitality. She believes that these four conditions must work together to generate urban vitality, and that the absence of any one of them will prevent the generation of vitality, so we use equal weights for the first 4 indicators. The last two conditions are considered as secondary conditions that are more independent of each other, but they also play an important role. In this paper, we adopt a more direct idea of weight assignment, i.e., 2:2:2:2:1:1, which is a simpler and more direct expression of Jacobs’ criterion. The JANE index of urban vitality after overlay analysis is constructed as shown in (16).

$$JANE Index=\frac{1}{5}\left(Div\right)+\frac{1}{5}\left(Di{v}_A\right)+\frac{1}{5}\left(Sca\right)+\frac{1}{5}\left(Con\right)+\frac{1}{10}\left(Acc\right)+\frac{1}{10}\left(Bou\right)$$

(16)

4. Results

The result of mapping Jane Jacobs’ urban vitality criteria in Qingdao historic city is presented in different maps in Figure 4, Figure 5, Figure 6, Figure 7, Figure 8, Figure 9 and Figure 10. Comments on the side are made for each of the urban vitality conditions as well as for the final result in order to support their spatial interpretation.

4.1. Building Function Diversity

The spatial distribution of function diversity corresponding scores in Qingdao historic city is presented in Figure 4. Figure 4a–c show the spatial distribution of land use mixture (D1), residential and non-residential ratio (D2), and commercial and facility ratio (D3), respectively. The high values of D1 are mainly located in the central–eastern part of the country, with relatively low diversity in coastal land use, while D2 shows the opposite distribution pattern, with a relatively high ratio of residential to employment in the coastal zone. Figure 4d shows the weighted superposition of the three indicators using the entropy weighting method to obtain the composite index of Div.

Diversity was tested for spatial autocorrelation, given that the Global Moran’ s I was 0.29 and the z-score was 25.72, thus implying significant spatial correlation and clustering. The pattern of building function diversity in Qingdao is more complex: (1) The spatial pattern of the high-value area (Div ≥ 0.70) follows a banding pattern along Qingdao Jiaozhou road, and the further away from the main road, the more obvious the diversity decay, with a mean diversity index (Mdiv) of 0.77, including mean land use mix (MD₁) 0.75, mean residential-nonresidential ratio (MD₂) 0.31, and mean commercial-public facility ratio (MD₃) 0.82. Due to the limitation of the conservation plan of Qingdao’s historic city, the diversity of Zhongshan Road and Taidong streets are low and unevenly distributed, compared to the higher diversity of land use in the surrounding urban areas. (2) Anselin Local Moran’ s I test shows that some areas show significant differences in land use mix within a relatively short distance, such as the intersection of Laiyang Road and Qingyu Road and Huiquan Commercial Square, where there are many high-low outlier areas, which are characterized by high values of local diversity in single-use building functional areas. This contrast is further revealed by the disaggregated nature of the raw data, and this micro-contrast is likely to be weakened if larger spatial scales (e.g., municipal scales) are used. (3) The area with low diversity value (Div ≤ 0.30) is mainly located in the eastern mountainous region with a mean diversity index (Mdiv) 0.09, including mean land use mix (MD₁) 0.15, mean residential-nonresidential ratio (MD₂) 0.02, and mean commercial-public facilities ratio (MD₃) 0.05, for instance, Taiping Mountain, Baguan Mountain, and Qingdao Mountain generally have lower land use diversity. Within the core area of the old city, the southern coastal neighborhoods such as Sifang Road, Guanhai Mountain, and Baguan Mountain reflect a certain degree of homogeneity in land use and a low diversity index.

In urban conservation planning of Qingdao that was formulated in 2020, Qingdao puts forward three categories of control requirements for construction height, the first of which requires the existing height to be maintained in the core protection area and no change is allowed. The second category requires that the construction height in the historic district of the main city shall not exceed 18 m, and in the towns of Taixi and Taidong shall not exceed 24 m, and the new buildings shall not exceed the adjacent historic buildings. The third category requires that the building heights in other areas shall not affect the overall mountain overlook view area and protect the overall view corridor. Combined with the results of the diversity distribution, the construction height restriction has a significant impact on the diversity of building functions, and the diversity of the main urban area is generally lower due to the 18-m building height limit, however, there are also individual phenomena, such as Parkson Shopping Center, Guangde Li 1898, and Yue Hik Lai Shopping Plaza in the Zhongshan Road neighborhood along the southern coast, which are much higher than the control height due to their early construction, and these buildings are affected by the conservation planning and development These buildings are affected by the conservation planning, development is reduced, and the customer flow is also reduced accordingly. The height limit in Taidong Town is higher than that of the main urban area, and thus has higher diversity characteristics, and it gathers more people traffic, however, because the central area of Taidong Street is stricter than the peripheral height limit, so the diversity distribution shows the characteristics of low center and high surrounding.

4.2. Road Scale

The spatial distribution of the road scale score is presented in Figure. 5. Figure 5a to Figure 5d show the spatial distribution of block size (S1), road centerline density (S2), road centerline connectivity (S3), and road centerline betweenness (S4), respectively. The S1 high value area is mainly located in the port area along Jiaozhou Bay and the mountainous area in the southeast, and the street blocks are large in scale; the S1 low value area has a clear correlation with the S2 high value area, and the small scale street blocks increase the street density and have high centrality; S3 and S4 show the connectivity and passability of the local network. Figure 4d shows the Sca composite index that was obtained by weighted superposition of the four indicators using the entropy weighting method.

Road scale was tested for spatial autocorrelation, given that the Global Moran’ s I was 0.85 and the z-score was 71.70, thus implying significant spatial correlation and clustering. Since walking is more suitable for stimulating urban vitality than vehicular traffic, we focus on the spatial scale of people in the walking mode and set the search radius in the spatial syntax calculation to 1500 m, which is close to the 10 min walking circle range, covering central, daily, and active streets. According to Jacobs, smaller block sizes and shorter streets provide more favorable conditions for contact between people. The results of this index show two more significant spatial patterns: (1) High-value areas (Sca ≥ 0.52) identify areas that provide more public contact, which were planned during the German occupation period (1897–1914) in accordance with the European “garden city” concept of colonial planning, with a high-density road network, mainly in the central area of the old urban area, historic area of Taixi and Taidong, with a cluster layout. They are mainly located in the central area of the old city and in the historic areas of Taisi and Taitung and are laid out in clusters. These areas have a mean street-scale index (MSca) of 0.62, with a mean block area index (MS₁) of 0.22, a mean road density index (MS₂) of 0.73, a mean connectivity value (MS₃) of 0.58, and a mean betweenness value (MS₄) of 0.20. The newer urban development areas are considered to be areas with lower contact potential due to larger, more evenly spaced blocks planned and wider streets. (2) Low-value areas (Sca ≤ 0.30) that are located in the northern part of old urban center, which is bordered by Taidong historic area, and in southeast along Huiquan Bay and Taiping Bay, with a banding layout and mean street-scale index (MS₁) of 0.20, including a mean block area index (MS₂) of 0.67, a mean road density index (MS₃) of 0.25, a mean connection value of 0.38, and a mean penetration (MS₄) of 0.06. Various historic cities in the northern part of the city were planned during the massive land expansion program of the German occupation (1910–1914), with a low-density road network and lower contact opportunity values in these expansion areas compared to more compact road structures in old urban centers.

4.3. Building Age Diversity

The next criteria for urban vitality, building age diversity, allows us to distinguish the areas in which a larger presence of older buildings coexist with newer ones, from those that are newer and more uniform in terms of age. Building age diversity was tested for spatial autocorrelation, given that the Global Moran’ s I was 0.63 and the z-score was 52.97, thus implying significant spatial correlation and clustering. The results for this index reveal higher scores within the city’s historic center, where a larger number of old buildings are intermingled with more recent structures that were built as part of the densification that has been occurring over the course of the past 123 years. The results show that: (1) High-value areas (Div_A ≥ 0.45) that are distributed in addition to the historical centers, form important peripheral sub-centers in the southeast coastal area with a mean building age diversity index (MDiv_A) of 0.49, including a mean building age index (MDA₁) of 0.81, and a mean building age diversity (MDA₂) of 0.55, such as Yu Hill, Signal Hill historic conservation area, Huiquan Commercial Square, Badaguan, Taiping cape, due to its favorable climate and its historical use as a European retreat area, have more intact architectural types, mainly single-family houses, which are now mostly utilized as tourist resources such as resorts and sea baths. (2) Low-value areas (Div_A ≤ 0.18), which are mainly located in Badaxia Street west of the railway station, and Dengzhou Road Street north of East–-West Rapid Road, etc., show greater homogeneity in terms of building age, a feature that constitutes a negative factor in urban vitality. The mean building age diversity index (MDiv_A) 0.04, where the mean building age index (MDA₁) 0.07 and the mean building age diversity (MDA₂) 0.22 are identified in the Taixi and Taidong historic areas and Liaoning Road. These areas have lower housing and rental prices than higher value areas due to the government’s Affordable Housing, Housing Provident Fund (HPF) and Low Rent Housing programs since 1994, which subsidize land allocation and reduce taxes on the one hand, and provide financial assistance to individuals on the other, allowing lower income families to enjoy lower rents.

4.4. Concentricity

Figure 7a–d show the spatial distribution of building density (C1), The floor area density (C2), people density (C3), and public facility density (C4), respectively, which all show more obvious spatial differences. There is an obvious spatial mismatch between building density and population density, and Taitung Town, which has a high concentration of population, but fails to provide housing density that meets the demand. Public facilities show a multi-core distribution, and the same problem of uneven distribution exists. Figure 7e shows the composite index after the weighted overlay of the four indicators.

Concentricity was tested for spatial autocorrelation, given that the Global Moran’ s I was 0.65 and the z-score was 55.23, thus implying significant spatial correlation and clustering. Similar to the road scale evaluations, smaller street-scales, greater road density, connectivity, and betweenness also explained to some extent the distribution of concentration indexes. (1) High-value areas (Con ≥ 0.31) correspond first to high population and housing density neighborhoods in the central area of the old urban area and north of it, with a mean concentration index (MCon) of 0.28, including a mean building density index (MC₁) of 0.41, a mean floor area ratio index (MC₂) of 0.13, a population density index (MC₃) of 0.11, and a public facilities density index (MC₄) of 0.53. The local Moran’s index test (Anselin Local Moran’ s I test) shows that this area is mostly a high-high cluster area. (2) Although low-value areas (Con ≤ 0.14) show no clear trend in their spatial distribution, these areas are generally located in the urban fringe zone, corresponding to industrial sites or areas that are constrained by construction due to topography or the presence of large nature conservation parks, with a mean concentration index (MCon) of 0.05, where the mean building density index (MC₁) is 0.09, the mean floor area ratio index (MC₂) is 0.02, the mean population density index (MC₃) is 0.03, and the mean public facilities density index (MC₄) is 0.16. However, the presence of localized high-value areas can also be seen in the surrounding low-value areas. One explanation for this spatial pattern is that these areas are laid out with a high density of specific public service facilities, such as the historic district of Guanxiang Hill and its surrounding neighborhoods, which are dominated by medical service facilities.

In urban conservation planning of Qingdao, the requirements for building volume, layout, color, material, etc., are consistent with the historical features, that is, the environmental characteristics of small volumes of white walls, red tiles, and green trees, which attract Although the renovation of historical buildings for small entertainment facilities such as coffee shops and bars has limited the attraction of mixed functions to a certain extent, it has improved the attraction of tourists, and thus improved the characteristics of concentration. Such an agglomeration is most obvious in Zhongshan Road, Yushan, and Signal Hill neighborhoods.

4.5. Accessibility and Boundary Vacuum

The spatial distribution of the two complementary conditions for urban vitality are presented in Figure. 8. Accessibility is mainly explained by the “center-periphery” structural logic of spatial syntax, i.e., the geometric law of decay from the center to the periphery, and its spatial distribution was tested for spatial autocorrelation, given that the Global Moran’ s I was 0.92 and the z-score was 77.71, thus implying significant spatial correlation and clustering. Accessibility of Qingdao historic city shows a multi-polar layout, basically reflecting the public transportation network, with the shorter average distance from the geometric center of the city indicating more convenient transportation. The results show that (1) high-value areas (Acc ≥ 0.70) are concentrated in central areas of the old urban area with Zhongshan Road as the central axis, areas around the railway station and Taidong historic area, with a mean accessibility (MAcc) of 0.80, including a mean closeness (MA₁) of 0.69 and a mean traffic facility proximity (MA₂) of 0.05. The central area is characterized by a lower slope compared to the mountainous area, and its road network is planned to follow the topography and there are traffic calming areas, so the accessibility becomes higher. Another reason is the location of the railroad station at the end of the Jiaoji Railway, which connects the urban space. Under the influence of the east-west fast road network, closeness and public transportation accessibility appear to be multi-polar, and the Taitung urban area becomes the trend of urban structure in the 21st century. (2) The low-value area (Acc ≤ 0.30) is concentrated in port area along Jiaozhou Bay, mountainous area such as Taiping Mountain, and the southeast coastal area, with a mean accessibility (MAcc) of 0.19, including a mean proximity (MA₁) of 0.16 and a mean transportation facilities proximity (MA₂) of 0.12. This area locates at the edge of southeast urban area, where the road network is less segmented and more interrupted; reducing the distribution of interrupted street network will help create more “reach” opportunities for people.

In terms of distance from border vacuums, areas with high values are those that are distant to large infrastructures or large, single-used buildings that can potentially discourage street life. These boundary vacuum elements are mainly the Haibo River, Changle River, and Hangan Viaduct at the northern city boundary; Jiao Ning Viaduct extending eastward from the central old urban center; and the railway lines around the railway station. In the surrounding zone, the inner part of the old urban area is also covered with major parking lots in points and large parks such as nature reserves in polygon form. The Bou index measures the distance of each spatial unit from boundary vacuum elements, and its distribution pattern is tested for spatial autocorrelation, given that the Global Moran’s I 0.33 and the z-score is 27.81, so there are significant spatial correlations and clustering. The results show that: (1) High-value areas (Bou ≥ 0.63) are mainly located along the southeast coastline, including Signal Hill, Yu Hill, and Baguan Hill, due to the superior natural environment and distance from facilities such as railway lines and large parking lots. Besides, there are points of high value areas that are distributed along Liaoning Road and Changle Road in the north. (2) Although the boundary vacuum elements are mainly located at urban boundaries, it is worth noticing that within the central area of Qingdao’s historic city, there are also numerous streets that are close to the boundary vacuum and present low values. Since the planning of the German system at the end of the 19th century was motivated by the construction of Qingdao as a military base and industrial port, large infrastructures such as railroad stations, Jiaoji railroad, and ports were planned, and these infrastructural elements had a negative impact on human contact.

4.6. JANE Index in Qingdao Historic City

We then synthesized the six conditions of urban vitality in the JANE Index. Higher values of the JANE Index correspond to areas with a higher potential for urban vitality, while lower values indicate the lack of such conditions. The distribution of Qingdao historic city JANE index is shown in Figure 9.

The potential of urban vitality in Qingdao historic city presents a polycentric pattern, as we find high values of the JANE Index distributed in different sub-centers of potential vitality. In turn, different intensities of urban vitality potential are identified. As Jacobs’ advanced, not to be picked up at the city scale, but is at the district and neighborhood levels that this is properly understood. In this sense, one center and four sub-centers of urban vitality were identified in Qingdao historic city. (1) The west end of Qingdao East–West Rapid Road, the south end of Jiaoji Railway, Taiping Road, and Jiangsu Road enclose the central area with the highest urban vitality, and the north end has part of the junction between the south and north districts of Qingdao, such as Zhongshan Road Street and Sifang Street, with a mean JANE index of 0.49 and a standard deviation 0.11 and more uniform spatial distribution. The area has a long planning history, significant street morphological differences, and more historic buildings, and thus has more intensive crowd activity (Con), more neighborhood contact opportunities (Sca), higher building age diversity (Div_A), and street accessibility (Acc). The general pattern of spatial patterns of urban vitality is again validated by the center–periphery logic: closer to the center means that higher urban vitality is achieved, while those areas with lower or no potential vitality are incorporated into peripheral contours of city. (2) The streets of Taidong, Liaoning Road, Badaxia and Jiangsu Road constitute the four sub-centers of urban vitality, with mean JANE scores of 0.44, 0.45, 0.43, and 0.39 and standard deviations of 0.08, 0.07, 0.07, and 0.08, respectively, and a more uniform spatial distribution. Compared with the traditional 100 m×100 m grid in the historic center, these areas have a narrower street pattern, smaller block scale, and high housing and population density, and higher JANE indexes are mainly explained by two indicators of concentration (Con) and street-scale (Sca). (3) Similarly, it is evident that some edge areas have high JANE scores, such as Badaguan Street near Wushengguan Road and Dagang Street near Dagang Weisi Road, which are mainly explained by the street-scale (Sca) index. These areas with high-low cluster values confirm the multicenter trend of urban vitality. 4) The area with a low JANE index (JANE ≤ 0.30) is mainly located in northern Dagang Street, Yunnan Road Street, and the southern end of Badaxia Street, with a mean JANE index 0.24, 0.29, and 0.31 and a standard deviation 0.03, 0.05, and 0.07, respectively, and more uniform spatial distribution. The lower spatial vitality of this area is related to its location in the urban edge area, relative isolation, large street block area, low density, and single land use.

Figure 10 shows the grading results of the Qingdao historic city JANE index, and

Table 2. JANE index group statistics results.

Group	Count	Mean	Std. Deviation	Min.	Max.	Proportion
High Vitality	196	0.568	0.048	0.507	0.699	6.73%
Moderate Vitality	572	0.449	0.032	0.392	0.505	26.98%
Low Vitality	779	0.338	0.028	0.293	0.391	29.21%
Minimum Vitality	732	0.248	0.033	0.110	0.292	37.08%
Total	2279	0.356	0.104	0.110	0.699	100.00%

Source: compiled by the authors.

shows the statistics of TAZ units at each level. There were four urban vitality classes of Qingdao’s historic city that were identified by JANE cluster analysis, labeled as high vitality areas, moderate vitality areas, low vitality areas, and non-vitality areas. The analysis results show that high vitality areas account for about 6.73% of the total area of the study area, with a mean JANE index of 0.568, standard deviation of 0.048, and extreme values 0.507 and 0.699, mostly located in historic central areas, and a few areas in the above four JANE index sub-centers. These areas are characterized by a small-scale street network, a mixed distribution of old and new buildings, and a mix of commercial and public facilities, producing a high accessibility and diversity value, a feature that is further enhanced by the increased distance from the boundary vacuum. The main axis of urban vitality in this area is from Zhongshan Road southward to the Trestle Bridge, where traditional Qingdao neighborhoods, shopping malls, and Trestle Park converge, reflecting a high level of diversity and neighborhood access. On the eastern side of the old city, Guanhai Hill and the area to its south are highly accessible due to the radial road network, creating high vitality points. Finally, in addition to the historic center, the neighborhoods of Weihai Road Pedestrian Street in Taidong Street are characterized by a higher concentration of urban vitality due to higher population density and commercial facility density.

Secondary areas of moderate vitality zones account for approximately 26.98% of the study area, with a mean JANE index of 0.449, a standard deviation of 0.032, and extreme values of 0.392 and 0.505. In terms of the planned street patterns, the area is expected to exhibit a high potential level of street life, but the combination with other indexes reflects that central areas of the district are either categorized as moderate or even low vitality zones and tends to be a transitional buffer zone between high and low vitality zones. Compared with the neighboring high vitality zones, the approximate street scale and population density show that the mixed degree of building functions and density become the main reason for the formation of transition zones. For example, among Taidong streets, the layout of Weihai Road pedestrian street tends to have a positive impact on local vitality, while neighborhoods that are farther away from pedestrian streets have relatively lower vitality.

Finally, areas that are classified as low and no vitality account for 29.21% and 37.08% of Qingdao’s historic city, respectively, with low vitality areas having a mean JANE index of 0.338, a standard deviation of 0.028, and extreme values of 0.293 and 0.391, and no vitality areas having a mean JANE index of 0.248, a standard deviation of 0.033, and extreme values of 0.110 and 0.292, which are often are located in urban edge zones with low population and building densities and in close proximity to border vacuums. These two vibrant environments correspond to being located on the periphery of a moderately vibrant zone, presenting very low accessibility values. The analysis reveals that places that are in contact with agricultural, natural, or industrial areas tend to present significantly low values of urban vitality; on the one hand they usually are near extensive mountainous areas with significant slopes, such as Qingdao Mountain, Taiping Mountain, and Zhushui Mountain, and on the other hand they are located in functionally single areas with larger block scales, such as Dagang and Xiaogang in the north, and Taiping Cape and Huizhuan Cape in the southeast coast.

5. Discussion and Conclusions

This study aims to recontextualize Jacobs’ urban vitality principle through urban spatial analysis and analyzes the livability and urban vitality of Chinese historic cities in the context of rapid urbanization. The newly established JANE index consists of six criteria with a total of 16 indexes, and since urban vitality is considered a key measure of urban residents’ happiness and quality of life, this study contributes to a discussion of the nature of the drivers of urban vitality and the way that they are distributed in space. To this end, the study creates a framework of indexes based on the Delclòs-Alió establishment and integrates the theoretical and methodological proposals that were made by recent literature on the subject, and performs a spatial analysis based on GIS-sDNA, the results of which are applicable to Qingdao historic city with its different muscular morphology. Jacobs has a rather important role in the history of the field of modern urbanism. From a practical point of view, this paper will help planners to combine morphological conditions that are conducive to vibrant streets, urban environments, and neighborhoods, thus promoting community, local spatial practices, and neighborhoods.

The results of this study show that conditions of urban vitality in today’s cities are not necessarily related to centrality, a specific urban texture, or a certain income level, but may be a result of different combinations of certain urban characteristics. In this sense, this study validates the applicability of Jacobs’ theory to cities in developing Asia, where more street activity and vitality is not singularly concentrated in central areas, although these areas are often characterized by higher population densities and mixed uses. Qingdao’s historic urban vitality follows a multi-centered spatial pattern, with a more fragmented distribution of public services and coexists with a dense network of local businesses. Different combinations of indexes of urban vitality are distributed throughout the area, showing a multi-centered pattern covering a diverse range of urban texture with two characteristics: (1) The main areas with high potential values of urban vitality in Qingdao’s historic city correspond to the oldest historic city center, which has the original dynamics of all vitality indexes. The square grid street pattern that was planned along the topography creates the conditions for urban vitality. As Jacobs argues, the mix of land uses and the distribution of more intersections produce a more compact and complex form that provides constant access to citizens and a diversity of facilities and services. (2) Medium and high potential vitality values are found equally in parts of the city that are away from the Old City Center, such as the Taidong and Taixi historic areas. It has long been marginalized due to topographical factors and the racial segregation motive in historical planning. In recent years, the area has gradually formed a sub-center of urban vitality in Qingdao with the combination of different vitality conditions such as increasing population density and diverse land use development, which attracts and directs not only short-distance mobility but also people that are traveling long distances.

After China established the historic and cultural city protection system in 1982, protection concepts gradually developed from single buildings to overall historic environmental protection in historic districts. Qingdao historic city was incorporated into the national historic and cultural city protection list in 1994, as Figure 8 delineates 14 historic conservation areas and 2 historic areas. Qingdao government strictly protects the physical environment of the historic city but implements relatively few policies and practices to improve the social life of residents and promote the renewal and maintenance of old public service facilities. Qingdao’s historic conservation areas originated from the segregated division of European and Chinese districts in the urban planning during the German rule. Historic conservation areas, mainly in the former European district, are mainly villa residences, with large street block scale and high green space rate, but there is a problem of insufficient vitality, so in the renovation and reuse of such streets, commercial and public service facilities should be increased to improve the diversity of architectural functions. The former Chinese district is the main historical district that was originally for residential-commercial mixed function, with smaller street block scale, low green space rate, and poor livability. In the renovation and reuse of these kinds of streets, green space and public activity space should be increased. Based on the relationship between the Jacobs index system and the quantitative results, we established the following planning strategies to enhance the vitality of Qingdao’s historic city:

• Based on the results of diversity analysis, for single-function blocks, such as Badaguan and other blocks on the southeast coast, the mix of land use, residential-non-residential, and commercial-public facilities should be improved, and the interaction of various types of facilities should be exerted.
• Based on the results of diversity and old building analysis, we will take measures to preserve the old buildings in Zhongshan Road and Sifang Road and revitalize them to meet functional needs.
• Based on the results of small blocks analysis, some of the blocks along Jiaozhou Bay in the northwest maintain the historical block texture of short streets and small plots from the 1920s Japanese planning period, and the spatial structure of small blocks should be continued for new development sites.
• Based on the results of concentration and old building analysis, measures are taken to maintain a reasonable high density of buildings, roads, and population in the neighborhoods of Yushan and Signal Hill, and to consider the use of old buildings as subsidized housing, so as to avoid excessive tourism and gentrification of the historic districts.
• Based on the results of concentration, small blocks, and accessibility analysis, the density of bus route coverage and shared transportation facilities is reasonably increased for Zhongshan Road and other neighborhoods.
• Based on the results of boundary vacuum analysis, large scale infrastructure construction should be avoided for the central city, and the crossing of boundary vacuum elements such as viaducts and railroads should be avoided, because these elements cut the integrity of the historical urban form.

There are certain limitations in this study. Firstly, urban vitality is a broad and complex concept involving social, economic, cultural, and other multi-dimensional contents. Although this study characterizes urban vitality by various parameters of LBS data with objectivity, it cannot accurately express the subjective perception of the public. For example, the satisfaction and perception of an active population on the space of historical neighborhoods mean the research results may be biased due to the data characteristics. Follow-up studies should calculate more accurate variable data that are based on streetscape images and explore human perception characteristics in detail. Second, POI data cannot present time information, a limitation that prevents the unique seasonal characteristics of historic urban vibrancy from being effectively measured. Follow-up studies can combine multi-source spatio-temporal data such as social media punch card data and cell phone signaling data to establish a more accurate urban vitality index system.