The ‘scale effect’ is a major issue in geographic research and a significant worry for geographers (Petrović et al., 2022; Andersson et al., 2023). Numerous empirical studies have shown that research adhering to the fundamental paradigm of geographic research, such as spatial patterns, temporal characteristics, elemental coupling, and the like, are dependent on scale (Wang et al., 2013, 2021; Qu et al., 2018). The finding implies that in such processes, the research object has temporal or spatial scale differentiation characteristics (Yigitcanlar et al., 2015), making it difficult to reveal the scientific essence of the research problem if the scale is not selected properly (Wu, 2013; Shirazi et al., 2020). Issues such as a lack of urban spatial differentiation management, loose nested urban system relationships, and the challenges associated with implementing multiscale planning schemes must be resolved to fully improve the territorial spatial governance system and increase the capacity of modern governance. As regards geographical heterogeneity and multiscale spatial inter-feeding processes, the classification of urban sustainable development type zones from the standpoint of spatial coupling is crucial for fine-grained analysis of complex urban systems.

An urban district is a collection of streets. Districts are the management space for urban development and construction and, in comparison with streets and communities, form a relatively complete socioeconomic development system, have distinctive regional functional positioning (Zhou et al., 2016), and primarily conduct urban key project construction, public resource allocation, and other responsibilities (Tian et al., 2010). A street is typically the primary spatial carrier for implementing urban programs. Streets have to carry out relatively independent urban activities including housing, employment, transit, and recreation (Huang et al., 2023), i.e., the everyday life of urban residents is heavily dependent on the streets where they reside. The community is the basic cellular unit of urban development and has the highest sensitivity and the smallest spatial granularity. The people of a community might have comparable economic and social traits, but the community itself typically concentrates on residential functions rather than entire urban functions (Berardi, 2013; Shirazi et al., 2020), which cannot fully define the meaning of sustainable development.

This study combined the differences in focus and the interdependence between the spatial scales at district and street levels to meet the needs of cities in terms of both development scale and spatial accuracy (Figure 1). The focus was on the central issue of sustainable urban development and urban sustainability research was developed from two scales, i.e., management scale and utilization scale at district and street levels. The functional coordination of microspace is the support path of macrospace sustainability, whereas the economic, social, and environmental sustainability of macrospace is the main aim of microspace development. For instance, improving production efficiency is necessary for sustainable economic development at the macrolevel, which pushes the production space toward intensive development, whereas choosing localized business patterns in microspace supports the optimal arrangement of urban industries and leads to more sustainable industrial development (O’Connor et al., 2018). Improved living conditions, traffic accessibility, and practical and efficient transportation, i.e., all the requirements for improving the quality of life of residents, would increase the total livability and sustainability of a city (Zhu et al., 2020). The conditions for pedestrians in communities could be improved by rationalizing road design, enhancing the effectiveness of public transportation, and lowering private car trips, thereby also cutting back on carbon emissions (Mouratidis, 2021). Rationalization of the labor market would attract more people and resource concentrations, but could also lead to social unrest and an increase in crime, which would be detrimental to long-term social progress (Shirazi et al., 2020).

Our study area comprises 16 districts based on the administrative boundaries of the districts of Beijing, and 330 streets overall based on the administrative divisions at street level and data from the seventh census in 2020. Urban areas were assessed in terms of economic, social, and environmental development scales based on the selection of spatial scales for urban sustainability evaluation and analysis of the hierarchical structure of spatial units. Streets were measured in terms of the degree of suitability of the built environment for living conditions. As regards healthcare, e.g., urban scale focuses on the overall standard of care, leading to the indicator ‘number of healthcare beds per 10,000 people’ being chosen. However, street scale should concentrate on the accessibility of healthcare facilities in various spatial units to reflect the habitat experience of residents at microscopic scale. The distinction between district and street size sustainability evaluation could contribute to the fining of urban governance, deepening understanding of the complex urban system, and broadening understanding of what urban sustainability evaluation means.

As mentioned, Beijing was chosen as the case study for this research (Figure 2). Beijing has made outstanding advancements in economic and social construction after more than 40 years of swift growth in the reform and opening up. At the end of 2020, the regional GDP of the city was 3,610.26 billion RMB (Chinese yuan), or 3.56% of the total GDP of the country. The city had a population of 21.89 million, with 19.164 million living in urban areas, and the urbanization rate was 87.55%. The city has a total area of 16,410.54 km², and is divided into 16 administrative districts, namely Dongcheng, Xicheng, Chaoyang, Haidian, Shijingshan, Fengtai, Fangshan, Tongzhou, Shunyi, Changping, Daxing, Mentougou, Huairou, Pinggu, Miyun, and Yanqing.

Beijing, the capital of China, is considered and ideal area for research. The city has a developed economy and sizable population and contributes nearly 4% to the national GDP despite occupying only 0.17 percent of the total land area of the country. However, as the population of the capital slowly increases toward overcrowding, resource depletion and infrastructure overload will increase the cost of urban development, i.e., create agglomeration diseconomies. The city will deteriorate once these agglomeration diseconomies take hold. A number of industries with high pollution levels and high energy consumption as prominent characteristics have been removed gradually from the capital in accordance with the proposal “to clarify the strategic positioning of the capital city, adhere to and strengthen the core functions of the capital, and adjust and decentralize non-capital core functions.” Whereas the demands for the design of urban infrastructure, urban governance, habitat construction, and ecological environment have increased gradually, the pursuit of economic rewards in urban development has decreased (Li and Zhang, 2015). The districts of Beijing have gradually improved their functions over the past number of years, displaying diverse characteristics, complex functions, and high development levels. However, new challenges to sustainable development have emerged, such as significant regional development differences, a lack of spatially differentiated management, loose nested relationships of urban systems, and challenges for implementing multiscale planning schemes, which all urgently require in-depth research.

The database for this study was divided into three sections, namely, first, vector data for administrative districts, e.g., the municipal, county, and street boundaries of Beijing. These data derive from the Resource and Environment Science and Data Center of the Chinese Academy of Sciences. Second, statistical data covering the period 2011-2020, sourced from, e.g., statistical yearbooks published by the governments of Beijing and each district, statistical bulletins, and the seventh population census in 2020. Third, to create the POI database, the coordinates of various facilities, including bus stops, hospitals, parks, and green spaces in Beijing were obtained from the POI database of AutoNavi.

Figure 4 shows the spatial distribution characteristics of sustainable development type zones of Beijing based on district and street scales.

The Z-scores of each factor were standardized, clustered, and partitioned using the Ward system clustering method, and each municipal district in Beijing was classified into three types based on the clustering, namely synergistic development, ecologically biased, and low growth. These classifications were made at district level using the economic, social, and environmental subsystem scores and the overall scores of each municipal district in Beijing from 2011 to 2020 as classification factors. The average indices for the districts of synergistic development, including Dongcheng, Xicheng, Chaoyang, and Haidian, were 0.411, 0.308, 0.235, and 0.478. These four districts had the highest total scores because of their major economic advantages, as well as achieving good scores for social and environmental factors. For ecologically biased areas, including Mentougou, Huairou, Pinggu, Miyun, and Yanqing, the average indices of each categorization component were 0.107, 0.239, 0.489, and 0.309. The environmental subsystem scores for these five districts were the highest and differed significantly from the economic and social subsystem values. The average categorization factor indices for low-growth areas, including Shijingshan, Fengtai, Fangshan, Tongzhou, Shunyi, Changping, and Daxing, were 0.144, 0.126, 0.339, and 0.200. These seven districts lack clear benefits in all areas of the economy, society, and environment, and their degree of overall sustainability was constantly poor or even declining.

The evaluation indices and complete scores of the built environment were aggregated at street level using K-means to create five different habitat sustainability zones, namely highest, high, medium, low, and lowest. The highest habitat sustainability zone had the best built environment, with 111 streets distributed in a ring around the central regions of Xicheng and Dongcheng, and a mean value of 0.826. This type of location had high scores overall; however, the average non-main road area score was only 77.78%, placing it third among the five types of areas and showing a somewhat weak street network design for a region of this type. The two core districts of Xicheng and Dongcheng dominated the monocentric distribution of the 95 streets that made up the mean built environment score of 0.823 of this high habitat sustainability type area. The key element restricting the sustainability of the built environment in this area was the unusually high population density of this type of district, with a mean value of 21,621 persons/square kilometer. With 80 streets distributed in a ring pattern outside the sixth ring road of Beijing, the mean value of the built environment score of the medium habitat sustainability type area was 0.697. With 27 streets, mostly in the far-flung outskirts of the city, the mean value of the built environment score of the low habitat sustainability zone was 0.492. The mean value of the built environment score of the lowest habitat sustainability area, with 17 streets distributed in a dotted pattern in the urban perimeter, was 0.416. The medium, low, and lowest habitat sustainability zones eventually experienced a decline in population density. The provision of public service facilities, such as schools and hospitals, was insufficient owing to the progressive distance from the urban center, with comparatively few bus and subway stops, thereby detrimentally affecting the convenience of travel of the residents. The diversity and accessibility of facilities were, therefore, key elements limiting the sustainable growth of the built environment.

For the cross-tabulation study, district-scale sustainability type zones were allocated to each street, and the frequency of occurrence of street built environment type zones in each category was counted (Table 3). With a significance level of 0.000, the findings passed the chi-square test, indicating substantial differences between the built environment levels of various sustainable development type zones.

The value of each row, measured horizontally, represents the percentage of each sustainable development type zone in each built environment type zone. The type zone with the highest value corresponded to the primary sustainable development pattern of the built environment type zone. Districts with low-growth types overlapped with streets with highest habitat sustainability types, and districts with synergistic development overlapped more with streets with high habitat sustainability types. Streets with medium habitat sustainability had the highest overlap with ecologically biased areas, followed by those with low habitat sustainability and lowest habitat sustainability. Districts focused on synergistic development had the largest proportion of high habitat sustainability streets overall (66.02%). Medium and low habitat sustainable streets accounted for a higher percentage (50.57% and 26.44%, respectively) in ecologically biased districts. In low-growth type zones, there were more highest and medium habitat sustainable streets (50.71% and 25.71%, respectively). Overall, ecologically biased and low-growth type districts were mixed with various street built environment type areas, as shown by the greater number of all types of street built environment type areas in both types of regions.

Based on the findings of the cross-tabulation analysis and spatial location, the spatial distribution maps of the district sustainable development type areas and the street built environment type areas were abstracted into a conceptual map (Figure 5) to identify the spatial coupling relationship between them. The synergistic development type areas primarily overlapped with high habitat sustainability type areas, ascribed to the areas with coordinated development of the three subsystems of economy-society-environment having rich and diverse public service facilities and high accessibility, giving residents relatively convenient living conditions and a good overall level of built environment. Although human density was relatively low in places with high ecological quality, these areas lacked diversity and accessibility owing to their distance from urban facilities. Therefore, ecologically biased areas primarily overlapped with areas with medium and low built environments. The highest habitat sustainability streets, primarily located on the periphery of the city, had lower population densities than the city center but better infrastructure development and easier access to public services, overlapping with the low-growth zones.

Specific sustainable development type zones are created by the law of spatial differentiation of economic, social, and environmental factors. The functions of these zones in various locational areas affect how the physical environment of the street is constructed. Specific built environment type zones are created as a result of the spatial distribution of different infrastructure types, resulting in variances in built environment conditions. The economic, social, and environmental development of a region is influenced by the production and living activities of the residents living in various built environment conditions. The interplay between street-scale spatial units and spatial units at the district level creates sustainable development type zones and built environment type zones. Figure 6 shows the scores for the different zone types. A city is a massive, intricate system comprising nested combinations of various levels of space (Sharifi and Murayama, 2013; Zumelzu et al., 2019), and economic, social, policy, spatial, and other factors influence spatial coupling (Kong et al., 2022). The resultant scale-space effects have implications for urban planning and the transformation of sustainable development.

The supply of social public resources should be optimized in terms of both structure and quality in the highest habitat sustainability streets found in the low-growth districts. These streets did not excel in any economic, social, or environmental areas relative to the extent of development at district level, and their overall sustainability levels were constantly poor. Interestingly, after 2016, the social subsystem scores for each district changed, with the average score falling by 0.023 over the course of the five-year period. This finding was ascribed to the growing population of the category, which exacerbated the lack of social and public resources and detrimentally affected the quality of life. Such streets have a POI entropy of 3.823 and a wealth of infrastructure. Accessibility to basic services and public transportation was good (0.882 and 0.880, respectively), but the architecture of the road network was somewhat poor, which made commuting more difficult. In the future, this type of street should manage population density while implementing the decentralization of the primary activities of the capital city and ensuring a sufficient supply of varied resources for public services. Focusing even more on the microspace, pedestrian paths for the built-up area should be designed carefully to avoid cutting across any street precincts and to enhance the quality of the walking environment.

Priority should be given to population deconcentration for high habitat sustainability streets situated in synergistic development-type districts to increase the level of public service protection. These streets performed well in social and environmental aspects and had major economic advantages at the district-level development scale. The economic, social, and environmental subsystem scores of such areas recorded average annual growth rates of 9.33%, 6.22%, and 4.11% from 2011 to 2020, respectively. However, with time, a decline occurred in their improvement levels for social and environmental sustainability. The social security and employment expenditures as a percentage of GDP performed poorly among all the districts, indicating the shortcomings of sustainable development of this type of region. With a mean POI entropy score of 3.911 and an average travel time to diverse public facilities of approximately 30 minutes, these streets counted among the best in terms of public facility diversity and accessibility. However, the extraordinarily high population density was a significant barrier to sustainable development of the built environment. In the future, the emphasis of this type of street should be on social well-being, particularly to increase investment in employment and social security. Further emphasizing the microspace, the focus should be on population decongestion, encouraging residents of urban centers to move to nearby neighborhoods and cities, thereby lowering population density and enhancing livability.

The land use mix for medium sustainable streets in the low-growth region should be improved by adding more supporting facilities, which would result in a greater concentration of different production activities to boost overall economic strength. These streets scored poorly on the development scale in terms of economic, social, and environmental factors, as well as integrated sustainability. Over the five-year period from 2016 to 2020, the social and environmental subsystem scores of low-growth areas declined on average by 2.30% and 13.01%, respectively. This result was ascribed to the general economic strength of this type of street, leading to, on the one hand, an insufficient supply of various public resources within the streets, and its level of social well-being lagging behind that of the synergistic development type of district. On the other hand, its environmental protection and pollution prevention pressure was high, which restrained sustainable environmental development to some extent. The lack of diversity and accessibility of public service facilities, with a POI entropy index of 3.467 and an average travel time to various infrastructure facilities of nearly an hour, was the most significant barrier to the improvement of the built environment quality of these streets. By increasing the density of different amenities and enhancing their diversity and accessibility, this type of street could enhance the daily life for residents in the future. Such actions could also direct the clustering of business types in microspace through the logical arrangement of additional facilities, thereby driving economic growth and encouraging overall development strength.

Priority should be given to increasing the accessibility of all public services in the area while maintaining a green and low-carbon orientation in the construction process for medium, low, and lowest habitat sustainability streets located in ecologically biased zones. The average growth value of the environmental subsystem score for such locations from 2011 to 2020 was 0.489, whereas the average increases for the economic and social subsystem scores were only 0.107 and 0.239, respectively. These streets had excellent environmental advantages. In future, the emphasis should be on energy conservation and use, as well as increasing investment in environmental protection, along with per capita domestic electricity consumption rising and the proportion of GDP invested in environmental protection declining among environmental subsystem indicators. Low population density and good adaptability for road network design characterized such streets. The average journey time to various forms of infrastructure, however, was more than an hour, i.e., travel convenience was low owing to relatively few metro stations and bus stops on the periphery of the city. Whereas the POI entropy index of both low and lowest habitat sustainability type streets was lower than 3, with clear stratification from other regions, the provision of public service facilities such as education and medical care was insufficient owing to the distance from the city center. To further increase the accessibility of facilities in such regions, this type of street should concentrate on expanding the allocation of public transportation and services such as education and medical care. However, the issue of energy consumption must be considered when building infrastructure, and high technology could be employed to boost energy usage efficiency, thereby retaining the environmental benefits.

Sustainable urban research is currently more concerned with multiscale spatial governance than with the long-term viability of urban subsystems. This study adopted a multiscale perspective to condense the factors influencing the development of urban spatial units at various scales, built a theoretical framework for evaluating urban sustainability based on inter-scale differences and correlations, thoroughly evaluated the urban sustainability level, explored the patterns of urban sustainable development, and proposed targeted development strategies to assist decision-makers in taking appropriate development measures.

The study discovered that at district scale, the sustainability level of the economic, social, and environmental subsystems in each district of Beijing showed alternating trends and could be classified into three types of areas through cluster analysis, namely synergistic development, ecological biased, and low-growth types. At street scale, the spatial circle structure of the built environment sustainability level of Beijing was obvious. The areas with the highest scores were located mostly in the second circle, showing an ‘inverse core’ trend. These areas were divided into five types of zones, namely highest, high, medium, low, and lowest habitat sustainability. The built environment sustainability levels of streets and the economic, social, and environmental sustainability levels of districts were differentiated spatially at a large scale, i.e., there were notable differences in the built environment sustainability levels of various sustainable development type districts. Each sustainable development type zone featured built environment type zones with a high degree of overlap. The streets in the synergistic development type zones were all of the highest and high habitat sustainability types, ascribed to areas with coordinated economic, social, and environmental development having highly diverse and accessible public service facilities within their streets and a good overall level of built environment. Owing to a lack of diversity and accessibility in ecologically high-quality areas of the capital city, despite their relatively low population density, as a result of their distance from various urban facilities, ecologically biased districts primarily overlapped with medium and low habitability sustainable streets. Compared with the city center, outlying areas were mainly low-growth areas, with good infrastructure development, easy access to public services, and relatively moderate population densities, which accounted for the highest percentage of highest habitat sustainable streets being located in such low-growth areas.

As cities first started to form, the idea of sustainable development emerged. The objectives of sustainable urban development have changed gradually, and the scale of the research focus has evolved dramatically, as contemporary urban challenges have grown more complicated (Figure 7). Early on, the concept of sustainability was conveyed primarily through the wise use of resources. Although the concept of sustainability did not originate until the 20th Century, agrarian societies had long expressed concepts similar to it (Gong, 1996). Environmental problems, such as air, water, and noise pollution, and resource waste have evolved since the Industrial Revolution owing to accelerated urban expansion. The United Nations Conference on the Human Environment urged, “…an urgent need to regulate urbanization, pay attention to population growth, and design human settlements to avoid adverse environmental impacts” in recognition of the significant burden that population growth places on the environment. An important development in understanding urban population capacity is the carrying capacity of resources and the environment (Sun et al., 2020; Hsu et al., 2021; Wang, 2022). After World War II, capitalist nations experienced rapid economic growth, but the ecological limitations of dense populations, intensive production, and expanding transportation systems were primary barriers to urban development (Fu et al., 2020). There is broad consensus among academics that the interaction between urbanization and natural restrictions is a main factor contributing to urban problems (Seto et al., 2011). A reduced level of living comfort, lack of competitive investment environment, and slowing of economic growth are the main effects of the deteriorating ecological environment on urbanization (Fang et al., 2017). Maintaining sustainable economic, social, and ecological well-being and progressing toward a dynamic balance of the three factors have become the emphasis of sustainable urban construction because of the social and environmental effects of the rapid urbanization process are recognized (Mori and Yamashita, 2015; Liang et al., 2022). By the 1990s, the spatial imbalance of economic and social activities brought about by rapid urban growth had become so dire that sustainable urban development had to be extended down to the neighborhood level (Luederitz et al., 2013). “The Bristol Accord” stipulates that sustainable communities should, “meet the demands of existing and future people, and provide quality services, equitable opportunities, and a good quality of life for all” (UK Presidency, 2005). Over time, urban governance has been improved by increasing microspatial sustainability. The degree of connection between diverse areas has increased progressively in recent years as a result of the regular mobility of various elements within cities. It is critical to conclude the shift from elemental governance to systemic governance and to break the research bottleneck of isolated work in which each region is divided and various systems are independent to meet the practical needs of ‘one map’ of territorial spatial governance. The modern era has made accurate diagnosis and evaluation of complicated multiscale spatial relationships a critical road to increasing the efficacy of urban government policies and initiatives. The status quo issues, development objectives, and realization paths vary greatly between various urban locations in megacities like Beijing, Shanghai, and Shenzhen, and the spatial coupling interactions are intricate and varied. Each metropolitan area’s distinctiveness and continuity are taken into account by the multi-scale evaluation framework of sustainable cities, which also offers a thorough multi-scale control system for implementation.