Classification and Spatial Differentiation of Subdistrict Units for Sustainable Urban Renewal in Megacities: A Case Study of Chengdu

Abstract

Sustainable urban renewal is an important approach to achieving high-quality urban development. The elements of megacities are diverse, and their structures are complex. It is critical to carry out the scientific classification of grassroots governance units based on the concept and needs of urban renewal to promote targeted sustainability evaluation and achieve the precise application of renewal design and planning. This study takes the jurisdiction of Chengdu City as an example and constructs a hierarchical dimension composite classification. For this classification, 128 grassroots governance units are divided into nine types, according to their obvious spatial differences. Based on the properties of these types, suggestions for evaluating and implementing urban renewal are proposed: (1) high-density central areas generally face the dilemma of complex and rigid needs and administrative weaknesses, so the development of public participatory governance is an urgent issue; (2) in transitional suburban zones, areas on and between the development axes are significantly different, indicating that extra attention should be paid to the fairness of the renewal of semi-urbanized areas; (3) outer areas are generally marginalized in urban renewal processes and destructive redevelopment behaviors should be avoided.

1. Introduction

Sustainable development has become a universal paradigm for economic growth, social equity, and environmental protection. Article 11 of the United Nations Sustainable Development Goals (SDG 11) clearly states that cities and communities are facing urgent social and environmental issues, which echoes the vision of harmonious and sustainable urban development in modern urban planning disciplines [1,2]. In China, with the population urbanization rate exceeding 65% and the urbanization process entering the “stock era”, the notion of urban renewal has shifted from the large-scale demolition and reconstruction of physical spaces to the comprehensive maintenance and improvement of urban economies, societies, and environments [3]. As sustainable development corresponds to urban renewal in terms of social, economic, and environmental sustainability, it has been recognized that urban renewal and sustainability should be combined [4,5], and the composite concept of “sustainable urban renewal” has been formed. Sustainable urban renewal was initially proposed as a problem-oriented concept as a criticism of large-scale demolition and redevelopment. It specifically emphasizes that the actions of urban renewal (such as planning, construction, investment, etc.) should also consider long-term socioeconomic benefits rather than just short-term economic stimuli or improvements to building environments. With the requirements of high-quality development deeply embedded in China’s socioeconomic development plans, the notion of sustainable urban renewal has gradually become goal oriented. Ye et al. [6] summarized research progress on sustainable urban renewal in China and pointed out that sustainable urban renewal is a modern governance model based on the potential value of built spaces and the coordinated development of resources, environments, economies, societies, and residents’ lives. Cao and Deng [5] further developed this definition, interpreting it as all urban renewal actions guided by sustainable development goals, such as economic growth, social inclusion, and environmental protection; however, they placed more emphasis on scientific decision making than previous authors. This article adopts the same definition as Cao, aiming to describe the different urban renewal attributes in megacities using a subdistrict unit classification to respond to the needs of societies, economies, and the environment.

Megacities are key areas for sustainable urban renewal. China has more megacities than any other country in the world. More than 200 million people reside in 19 megacities in China. On the one hand, Chinese megacities are invaluable regional centers that serve as the core of coordinated regional development; on the other hand, they are plagued by urban diseases due to their huge scale [7]. Their internal renewal needs are significant, highly complex, and unevenly distributed in space, forming a sharp contradiction with limited renewal driving forces [8]. To achieve precise, efficient, and fair sustainable urban renewal in such environments, it is necessary to categorize the internal spaces of megacities in terms of macro-requirements, external forces, and endogenous needs and formulate differentiated constraints and guidance based on their particular characteristics. Subdistrict units are important intermediate components that carry the properties of space within megacities. As local grassroots governance units, subdistrict offices are the middle ground between administrative instructions and social feedback [9]. In the process of descending the focus of social governance in China, these subunits have been endowed with more power and greater responsibility [9]. They are the responsible entities and main government forces for promoting the implementation and promotion of urban renewal projects [10]. Therefore, exploring the characteristics and types of grassroots governance units for sustainable urban renewal within megacities has important theoretical and practical significance.

The discussion of grassroots governance unit types occasionally appears as a prerequisite for sustainability evaluations based on the multiple-criteria decision analysis (MCDA) method, which has been widely applied in the field of sustainable cities because of its ability to structurally characterize conflicting multidimensional properties in urban renewal processes [11,12]. Representative works have used the analytic hierarchy process (AHP) method to screen indicators and confirm weights [13,14,15] and the fuzzy analysis method [16,17], the technique for order preference by similarity to an ideal solution (TOPSIS) method [18], and the multi-attribute value theory (MAVT) [19,20] to evaluate the sustainability levels of particular objects. An important consensus within this evaluation research is that methods should have both universality and adaptability. The former ensures full coverage, while the latter effectively responds to the properties of similar research objects. Therefore, it is necessary to define the characteristics of objects before evaluation.

At a macro-regional study scale, some rough empirical classifications have produced better effects, such as using developed European countries [18] or central Chinese cities as research objects [21]. However, in more detailed discussions at the district or even block level, empirical fuzzy prerequisites can easily lead to confusion. For example, in the sustainability evaluation of urban brownfields, objects in downtown residential areas and suburban agricultural belts are ranked and discussed according to different indicators and weights [22]. Yet, as a fuzzy concept, the definition, standards, and boundaries of “downtown residential areas” are not clear enough, making it difficult to accurately define each evaluation object. A common measure for addressing this problem is to skip the prerequisite types and expand the scope of indicators directly to comprehensively describe various environmental characteristics in a single set of standards. This not only increases the difficulty of weight matching but also confuses the different types of objects, reducing the interpretability of evaluation results.

Additionally, since current urban renewal practices are primarily carried out at the block and below scale, few studies specifically classify grassroots governance units based on sustainable urban renewal requirements [23]. In micro-scale case studies in China, empirical fuzzy classification is often used to describe the regional situation of the research object, such as ‘old town’ [24], ‘old community’ [25], ‘old neighborhood’ [26], etc. However, there is no unified standard among classification types, and fuzzy types cannot be accurate objects for local government classification and fine governance. Overall, the current specialized discussion regarding the types of grassroots governance units in sustainable cities needs to be deepened. It is necessary to combine qualitative and quantitative methods to realize scientific classification and to promote the efficient renewal and effective governance of megacities.

Therefore, this article combines the economic, social, and environmental dimensions of the sustainability concept with megacities’ three administrative levels. We constructed a spatial unit classification method with three steps: regional features characterization, internal attributes clustering, and set mapping analysis. Based on administrative hierarchy, the classification results correspond to the level of government authority. The internal and external commonalities of each research unit are effectively revealed through a three-step classification, highlighting high-value, unique resources. The study’s innovation lies in the multi-scale attribute constraints on grassroots governance units. The classification results aim to show the characteristics and spatial differentiation of urban renewal background attributes at the subdistrict scale. Through a classification method combining qualitative and quantitative analysis, this study verifies empirical expert perspectives on regional urban renewal planning based on object characteristics and mathematical analysis. Thus, we provide explicit strategic suggestions regarding the obstacles, key issues to be solved, corresponding workflow, and frameworks for urban renewal work in different types of subdistrict units. The findings of this study are expected to provide crucial empirical support and a decision-making basis for implementing urban renewal actions, designing renewal units, and the full-cycle operation and maintenance management of renewal projects.

2. Materials and Methods

The study framework is shown in Figure 1. This figure also presents the organization of Section 2. First, the case study area and data sources are introduced in Section 2.1. The general classification methods are discussed in Section 2.2, which includes a discussion on the classification indicator system formation. Section 2.3 describes the process of characterizing indicators and the approach to identifying subdistrict unit types.

2.1. Study Area and Data Source

Chengdu City (102°54′~104°53′ E, 30°05~31°26′ N) was selected as the case study, as shown in Figure 2a. Chengdu is the capital of Sichuan Province, which is located northwest of the Sichuan Basin. As one of the two poles of the Chengdu–Chongqing economic circle, Chengdu has a strategic position as both the core of its metropolitan area and the Chengdu–Chongqing urban agglomeration [27,28]. In the past two decades, Chengdu has experienced rapid growth, which has led to an extensive expansion of built-up areas, transforming it into a multi-core megacity. During this period, multiple satellite towns took shape while a complex land use transformation occurred. This represents a typical pattern of urban expansion and megacity formation in China [29]. Due to the diversity in the primary environment, construction timing, and resource inclination in the past decades, the characteristics of built areas show significant differences [30]. In 2021, Chengdu was selected as one of the first batches of urban regeneration pilot cities issued by the PRC’s Ministry of Housing and Urban–Rural Development.

The municipal districts of Chengdu referred to in this paper represent the area stipulated in Article 3 of the “Measures for the Implementation of Urban Organic Regeneration of Chengdu” issued in 2020. As Figure 2b shows, there are 128 subdistrict-level units, including 108 subdistrict offices and 20 towns. A town is an administrative unit at the same level as a subdistrict office but typically in remote areas. However, since towns share some features with the subdistrict offices in the study area, they were included in the study scope.

The required data were divided into four parts: (1) The land cover data were elicited from the GlobeLand30 dataset of the national geographic information resource directory service system for the year 2021. We used the Artificial Surfaces field to represent the urban built areas. (2) The administrative boundaries of subdistrict-level units were adjusted based on the 2018 data from the Chengdu Urban Planning and Design Institute, which was combined with the announced results in 2020. (3) The POI data were derived from a Baidu map POI extraction service. It was checked and cleaned according to the analysis requirements. (4) Other social and economic data were derived from the 2021 Yearbook of the Chengdu Statistics Bureau and the public data of the Chengdu Housing Leasing Service Platform.

2.2. Classification Concept and Indicator System

China has 19 megacities with a population of over 5 million comprising their main urban areas. These megacities are currently plagued by urban diseases such as spatial aging, insufficient public service supply, traffic congestion, and environmental pollution. The complexity of their urban system structure and elements has given rise to differentiated urban renewal needs scattered throughout the region. Due to increasing urban renewal, the elements involved are becoming diverse and multi-layered, emphasizing the government, market, and society’s joint participation [3]. At the municipal and district levels, policies and standards should impose constraints and provide guarantees and protection. In subdistricts, communities, and projects, the regenerating actions must be rooted in actual conditions while effectively addressing urban problems like traffic overload, lack of economic vitality, lack of social inclusion and fairness, resource shortage, and environmental pollution [31]. This indicates that the key information regarding urban regeneration varies. In practice, the uneven resolution of local statistical data results from this ‘differentiated attention.’ Generally, it is reasonable to select a widely recognized hierarchical system, such as current administrative levels, to concatenate attributes of different granularity.

Urban renewal and sustainable urban development are closely connected regarding their implication and objectives [32]. Urban researchers have basically reached a consensus to construct urban renewal perspectives from the three dimensions of the economy, society, and the environment [6]. Among them, the economic dimension primarily describes the sustainability of the local economy through its degree of local industry mixing, the ability to provide different types of jobs, and the employment situation of residents. The social dimension is broad, and its discussion elements include public safety, public welfare, community organizations, cultural heritage, historical heritage, etc., with strong local characteristics. The environmental dimension typically involves directly discussing urban stock material space, such as area, density, plot ratio, form type, resource status, etc. Due to different interpretations of these three dimensions, economic conditions are sometimes included in discussing the social environment [17,21]. In other cases, elements such as water, land, and air are classified as resources and separated from the environmental dimension for separate measurement [33]. Due to differences in research scale and objectives, expressions such as “space” and “building” are sometimes used to replace or specialize the environmental dimension [34]. However, the above changes are subtle adjustments within the three-dimensional model framework regarding the economy, society, and the environment, which should remain stable and unchanged as the basic framework [35]. The conceptual framework of classification is shown in Figure 3.

After establishing a classification concept model, this study conducted a secondary screening of the reference sources used in previous discussions. We primarily included research on the urban renewal mechanism and sustainability evaluation of Asian megacities and high-density cities since 2020. We excluded research that only discussed the provincial and municipal levels. Subsequently, the various indicators involved were organized according to this study’s hierarchy-dimension compounded concept. The results are shown in

Table 1. Preliminary list of classification indicators under the guidance of urban renewal.

Hierarchy	Dimension	Keyword	Measurement	Source	Corresponding Indicators
Municipal	Economy Society	Policy	National and provincial preferential policies	CAO, K.X.; DENG, Y. 2021 [5] KARA, C.; IRANMANESH, A. 2023 [15] LEE, J.H.; LIM, S. 2018 [13] AWAD, J.; JUNG, C. 2022 [14]	Regional Economic Policies
		GDP	GDP	CAO, K.X.; DENG, Y. 2021 [5]
		Superior Planning	Priority Zoning	CAO, K.X.; DENG, Y. 2021 [5] KARA, C.; IRANMANESH, A. 2023 [15]	Social Development Planning
District	Economy Society	Landscape	Guidelines for construction style	WANG, Y.; TANG, H. et al., 2020 [34] YANG, C.; XIN, L. et al., 2020 [23]	Historical and Cultural Heritage Reservation
		Regional Transport	Rail transit station density and line length	WANG, Y.; TANG, H. et al., 2020 [34] AWAD, J.; JUNG, C. 2022 [14]	Average Residential Rent
		Fiscal Capacity	General public budget	CAO, K.X.; DENG, Y. 2021 [5]	General Public Budget Income
		Industry	Industrial income Job provision	PENG, Y.; LAI, Y.N. et al., 2015 [12] CAO, K.X.; DENG, Y. 2021 [5]
Subdistrict District	Environment	Building Condition	Age, Ownership, Usage	PENG, Y.; LAI, Y.N. et al., 2015 [12] WANG, Y.; TANG, H. et al., 2020 [34]	Average Residential Rent Population Density
		Surrounding	LCZ type, Greening rate	PENG, Y.; LAI, Y.N. et al., 2015 [12] WANG, Y.; TANG, H. et al., 2020 [34] AWAD, J.; JUNG, C. 2022 [14]
		Quantity of Existence	Built-up Area Floor space	YANG, C.; XIN, L. et al., 2020 [23] CAO, K.X.; DENG, Y. 2021 [5]
	Society	Affordability of Housing	Average rent and price	PENG, Y.; LAI, Y.N. et al., 2015 [12] LEE, J.H.; LIM, S. 2018 [13] WANG, Y.; TANG, H. et al., 2020 [34]
		Culture	Heritage protection Local customs	PENG, Y.; LAI, Y.N. et al., 2015 [12] YANG, C.; XIN, L. et al., 2020 [23] KARA, C.; IRANMANESH, A. 2023 [15]	Historical and Cultural Heritage Reservation
		Public Service	Education coverage Health service coverage Safety	PENG, Y.; LAI, Y.N. et al., 2015 [12] LEE, J.H.; LIM, S. 2018 [13] YANG, C.; XIN, L. et al., 2020 [23] KARA, C.; IRANMANESH, A. 2023 [15]	Average Residential Rent The Ratio of Permanent Population to Registered Residence Population
	Economy	Local Commerce	Employment density, Convenient business service	PENG, Y.; LAI, Y.N. et al., 2015 [12] LEE, J.H.; LIM, S. 2018 [13] CAO, K.X.; DENG, Y. 2021 [5] AWAD, J.; JUNG, C. 2022 [14]
		Land Use	Development intensity Functional blending	PENG, Y.; LAI, Y.N. et al., 2015 [12] WANG, Y.; TANG, H. et al., 2020 [34] AWAD, J.; JUNG, C. 2022 [14]

.

In this Chengdu case study, the indicators obtained through discussion were integrated into the conceptual model to form an indicator system, as shown in Figure 4. At the municipal level, the development priority zoning from the local overall planning was adopted to describe social conditions, and we collected economic preferential policies to represent macroeconomic conditions. District units are the carrying space for the general public budget revenue [36] and historical and cultural protection policies [37]. After combining the indicators in pairs according to their dimensional focus, each subdistrict unit’s external conditions were described, revealing the priorities in the overall coordination process. At the subdistrict level, three composite indicators were selected to represent the socioeconomic conditions based on spatial environment. Population density was used to characterize the intensity of the built environment [38]. The ratio of the permanent population to the registered residence population was used to characterize each unit’s population mobility and social structure [39,40]. The average residential rent was used to characterize the accessibility of public services and the affordability of residents’ housing [41,42,43].

2.3. Quantification and Characterization of Indicators

First, we characterized the economic and social conditions at the municipal and district levels, which together constitute the external regeneration circumstances of the research unit. This method focuses on differentiation rather than precise evaluation; the importance of indicators was elaborated on during the screening and merging process. Thus, indicator weights were no longer considered, while a superposition method was adopted to intensify the differences between the classes.

Qualitative indicators should be assigned based on mutual exclusion and completeness in typological methods. The qualitative indicators involved in this study were divided into two categories for discussion. First is the mutually exclusive attributes. This pertains to attributes that do not overlap and can fully cover the research area after being merged, such as the development priority zoning explicitly determined in the overall plan. The second is inclusive attributes. This is where highly effective features can cover low-level features, such as economic preferential policies and historical protection efforts. For quantitative indicators such as the general public’s budget income, this study considered the results of natural breakpoints to divide subdistrict units into different levels.

Table 2. Assignment of external conditions of units and their corresponding meanings.

Assignment Value	Meanings of Value
	Social Development Planning ${\mathit{F}}_{\mathit{i}}$	Regional Economic Policies ${\mathit{E}}_{\mathit{i}}$	Historical and Cultural Reservation ${\mathit{C}}_{\mathit{i}}$	General Public Budget Income ${\mathit{B}}_{\mathit{i}}$
0		Ng.	Ng.
1	South and east expansion area Growth-oriented	With provincial development zone policies No national policies	With provincial reservation policies No national policies	Low
2	West and north Renovation area Growth-controlled	With national development and urbanization zone policies	With national reservation zones or other relics protection policies	Average
3	Central Comprehensive improvement and transformation			High

shows the assignment and corresponding meanings of the external feature indicators for the four units.

We added the assigned values of the same dimension to obtain the external condition value of each research unit:

$$E{\prime }_i=E_i+B_i, E{\prime }_i\in \left\{0,1,2,3,4,5\right\}$$

(1)

$$S{\prime }_i=F_i+C_i, S{\prime }_i\in \left\{0,1,2,3,4,5\right\}$$

(2)

Due to the conscious control of indicator size and number of breakpoints during the characterization process, the added results can directly demonstrate the level type of external social and economic conditions of each subdistrict unit. Here, a larger value of $E{\prime }_i$ indicates a more abundant external support, and a larger value of $S{\prime }_i$ indicates a more restricted urban renewal request.

After characterizing external conditions, we discussed the inner attributes at the subdistrict level. The selected model’s three indicators of population density, resident registered residence population ratio, and average housing rent were all continuous quantitative indicators. Therefore, the basic data were first normalized and mapped according to the dimensionless expression result of 0 to 1. Then, we conducted a cluster analysis to obtain results to comprehensively represent the three internal attributes. The method is as follows.

For the research subjects (subdistrict units) denoted as $T_1,T_2,T_3,\dots ,T_n$ measured by indicators $I_1,I_2,I_3,\dots ,I_p$, let $a_{ij}$ represent the actual performance value of subject $T_i$ concerning indicator $I_j$. The leaner Min–Max method was used to obtain normalized indicator values, such as the following:

$$r_{ij}=\frac{a_{ij}-a_{min\left(j\right)}}{a_{max\left(j\right)}-a_{min\left(j\right)}}$$

(3)

where $r_{ij}\in \left[0,1\right]$ is the normalized indicator value, $a_{max\left(j\right)}=max\left\{a_{1j},a_{2j},\dots ,a_{nj},\right\}$ is the maximum performance of the alternatives on indicator $I_j$, and $a_{min\left(j\right)}=min\left\{a_{1j},a_{2j},\dots ,a_{nj},\right\}$ is the minimum performance.

We assumed that the weights of each indicator were the same in clustering. We conducted the cluster analysis using the K-means tool in IBM SPSS statistics 19. K values were determined by combining the elbow method with the target number from management requirements. In this study, K = 4 was determined, and the number of iterations was 20, resulting in types donated as I, II, III, and IV.

Finally, the external feature assignment results and internal attribute clustering results were integrated through parallel set graphs. Based on these results, we visualized the combination relationships, discussed the characteristics of these relationships, and determined the final types.

3. Results

3.1. Mismatch of Regional Features

Figure 5 shows that the development priority zone from overall plans laid the foundation of the social attributes (upper, from left, 1). In the central part, subdistrict units are primarily filled with built-up areas, which are given development priorities closely associated with urban regeneration. The western and northern regions are adjacent to the ecological source area with lower suitability for construction due to rugged terrain. Their development priority is to control incremental construction and reduce the built area, leading to lower compatibility with regeneration goals. The eastern and southern regions are the most recently developed compared to the others and align with Chengdu’s next stage of urbanization. Thus, they have the weakest relationship with regenerating actions. Based on development priority zoning, the historical protection level identifies and distinguishes units with unique cultural heritage resources (upper, from left, 2).

The economic endowment spatial characteristics are also illustrated in Figure 5. Units in the eastern and southern regions benefit from the national development strategy, generally receiving more transfer payments and possessing stronger independent construction capabilities (upper, from left, 3 and 4). Meanwhile, units in the central urban area have a longer urbanization history and more complex built environments. Still, their local budgets are constrained compared to recently emerging urban areas. Moreover, due to limited land for construction use, they often lack economic preferential policies tied to local industrial parks. As such, these units are more likely to encounter industry recession and challenges regenerating the built area.

Figure 6 further reveals the above mismatch. The 4-2, 5-2, and 5-3 mapping relationships show that subdistrict units under high regenerating requirements lack corresponding resource endowments. They are prone to challenges in regeneration practices and may find it difficult to solve problems effectively. In contrast, the 1-3, 1-4, and 1-5 combination types indicate that many subdistrict units that are undervalued in the urban regeneration process have abundant fiscal capability, which exacerbates the risk of destructive construction.

3.2. Circular-Axial Spatial Pattern of Internal Characteristics

We obtained four internal attribute types of subdistrict units in Chengdu using K-means clustering. There were 41, 15, 36, and 36 units in the four types of I, II, III, and IV. Figure 7 illustrates the original attributes of each unit used for clustering. The scatter plot in Figure 8 presents the distribution and corresponding attribute characteristics of the four types. Although the spatial relation between research units was not explicitly considered during clustering, the distribution of the four cluster categories is highly consistent with Chengdu’s spatial development strategy. It exhibits a ring-axis composite feature, as shown in Figure 9. For ease of understanding, the four categories were named after their spatial location.

The I—outer-ring units contain the maximum total area, accounting for over 50% of the research area. Units of this type typically comprise towns with larger unit areas but less development. Thus, the number of cases in this type does not significantly exceed the other three types. Although Class I units are under the direct jurisdiction of the municipal government, they made limited progress in urbanization low population aggregation, mobility, and significantly inferior public service compared to other units in the inner ring. Social conflicts in these subdistrict units are relatively moderate. Generally, the I—outer-ring units are not only located in peripheral areas but also in the border zone concerning various urban regenerating actions.

The III—on-axis-type units coincide with the development axes connecting the central urban area with the surrounding satellite cities. In contrast, the II—inter-axis-type units are located between these development axes. The difference in population mobility between the two types indicates that although units of Class III typically have adequate public transportation resources and better jobs–housing balance, a large population of semi-urbanized groups still gather in Class II units; this suggests that they seek an optimal solution of lower settlement costs and easier access to public services. Based on China’s household registration system, Class II units act as a transit station for rural to urban populations. This means that Class II streets are more likely to face challenges in achieving equity and justice. Moreover, the tidal traffic, facility fairness, and service equalization between Class II units and other subdistrict units may be critical affairs during the urban regeneration process.

The IV—core-type units reflect the public perspective of the Old Town of Chengdu, generally possessing characteristics such as qualified infrastructure status, high regenerating requirements, and a complex internal environment. Based on the discussion in Section 3.1, we have reason to suspect that units of Class IV may be given high expectations without additional support and therefore face more obstacles in urban regeneration.

3.3. The Circular and Axial Structures Shape the Composite-Type Results

After combining the results of Section 3.1 and Section 3.2, we merged the two units of the type IV—core-Ag adjacent to axes into the inter-axis augmentative type according to the actual situation. One unit of the type II—inter-axis-C between the central area and a satellite city was integrated into the on-axis common type. Nine types of subdistrict units were ultimately defined. Their names, codes, abbreviations, counts, and proportions among all objects are presented in

Table 3. The 128 subdistrict unit type results.

Type Code	Abbreviation	Count	Percentage
IV—core-R	CR	34	26.6%
III—axis-R	AR	2	1.6%
III—axis-Ag	AA	12	9.4%
III—axis-C	AC	20	15.6%
III—axis-Lf	AL	5	3.9%
II—inter-axis-Lf	IA	14	10.9%
I—outer ring-Wb	OW	4	3.1%
I—outer ring-Ag	OA	12	9.4%
I—outer ring-Lf	OL	25	19.5%

. Their spatial distribution and the alluvial plots of typical types are shown in Figure 10.

The CR type had the maximum quantity with 34 cases, accounting for 26.6% of the total number of research units. Therefore, the CR type has a concentrated distribution in the central built areas, roughly the same as the IV—core-Ag type from Section 3.2, with high population density, relatively low population mobility, and complete public services. The alluvial plot indicates that this type of unit generally has a high level of goal orientation and policy constraints. Still, the financial capacity of their offices is not prominent nor even weak, making the external supply condition for this type mediocre.

Regarding units on the developmental axis, the AC type shows the highest proportion on Chengdu’s developmental axis with a total of 20 cases, accounting for 15.6% of the research units and 51.3% of the on-axis units. This type shares similar external supply conditions with the CR type, but due to its lower development intensity and alleviated demand, it will not face urgent supply–demand mismatch problems. There are 12 AA-type units, all located in the National Special Economic Strategy Zone, with top-level incremental potential. Several social problems that have emerged due to the slowdown in economic growth are still inconspicuous in the AA type. Furthermore, there are five laissez-faire units located at the end of the development axes, constituting the final AL type. There was another IA type, including units between axes. However, these were minimal and occurred under very different circumstances, which will be further discussed in Section 4.2.

Outer-ring units are located at the outermost edge of the circle axial structure and can be primarily divided into two types: the OL type and the OA type. The former includes 25 cases, accounting for 19.5% of the total research units; these were mainly distributed on the west and north sides of the study area. The latter includes 12 cases, accounting for 9.4%, which were distributed primarily in the east and south. The reason for the differences is the location relationship between these units and the development priority zoning as well as the motivation disparity elicited through multi-level economic policies.

4. Discussion

4.1. An Emphasis on Participatory Governance in High-Density Central Areas

The central area of Chengdu comprises CR-type units, which is one of the most typical high-density urban grassroots governance units. CR-type subdistricts had highly concentrated urban elements, and their various problems in urban renewal are also uncertain, fuzzy, and complex. In examining the evolution of urban vitality, Zeng et al. [44] demonstrated that the urban renewal process has elicited a trend of problem complexity in the central areas of megacities. Therefore, it is even more necessary for grassroots governance institutions to coordinate information and integrate resources to support problem-solving efficiency. However, the classification results in Section 3.3 reflect the reality of the mismatch between expectations, tasks, and capabilities in grassroots governance. In other words, urban grassroots should have the initiative to solve problems due to their information advantages and governance experience. However, they lack sufficient decision-making and execution capabilities due to inadequate formal authority allocation [45]. This “spatiotemporal mismatch” between resources and tasks has been a persistent obstacle to achieving governance efficiency improvement in megacities; it is also an urgent issue that urban renewal in the central area of megacities must face.

A critical survey in Nanjing [46] indicates that public participation is the buffer that mediates and ‘repairs’ the inconsistencies between different levels of governmental agencies. In fact, in the complex urban renewal process, relying solely on government power to fully control environmental remediation and conduct socioeconomic regeneration is extremely difficult; it is necessary to explore and guide social forces effectively. At the same time, the relatively stable resident structure and higher-level living needs of CR-type units provide a good foundation for public participation [47,48]. Therefore, promoting the contribution of renewal entities to greater social forces through participatory grassroots governance is an essential means of realizing urban renewal in high-density central areas.

Since Arnstein’s ladder of citizen participation [49], several scholars have conducted graded and classified discussions on public participation in urban governance. Among them, Hurlbert et al. [50] found that the possible effectiveness and value of public participation highly depend on the environmental conditions in which it operates. He and Hou [51] examined the localization of the citizen participation ladder in the context of China’s “big government” while suggesting that the focus should shift from power acquisition to examining the relationships between entities. Based on the classification of subdistrict scale units in the present study, two recommendations regarding the construction of a public participatory governance framework in the central urban areas of megacities are evident.

In the urban renewal process, there are distinguished public participation paradigms in different types of subdistricts, and there will be varying solutions to the contradiction between outcome-oriented governance and process–demand public participation. Classifying and studying these paradigms based on urban renewal characteristics and the needs of subdistrict scale units seems to be a feasible approach to addressing the above issues.

In central areas requiring social intervention in the renewal process, constructing fine-grained public participation scenarios using blocks or communities as spatial carriers will become a critical task while conforming to the overall characteristics described by subdistrict types. These can more precisely address on-site issues, thus facilitating in-depth discussions on specific evaluation criteria while achieving scenario-based optimization of governance frameworks through further studies.

4.2. Consideration Should Be Given to Equity in Peri-Urban Areas

Large cities in China have generally experienced transit-oriented expansion and construction, which is an effective means of coordinated development between cities and transportation to control urban sprawl [52]. On the one hand, transit-oriented development causes an uneven diffusion of the public service supply system that relies on the public transportation system from the city center to the periphery, forming a relatively low-supply inter-axis area. On the other hand, it gathers production factors along the axis, forming industrial depressions between axes. Generally, the low rental level of residential properties in the inter-axis area reflects poor accessibility to public services, infrastructure, and job positions, all of which pose a potential threat to spatial equity [53,54]. However, the analysis of population mobility in Section 3.2 indicates that compared to the edge areas, the inter-axis subdistricts are close in spatial distance to the city center. This makes them transit areas undergoing the process of population urbanization. This undoubtedly further exacerbates the risk of social barriers between semi-urbanized groups and urban residents.

Resident characteristics determine that the non-governmental social forces of the inter-axis units are far weaker than those in the central area. The most urgent needs of residents in these subdistricts are more concentrated between the safety and belonging levels, which indicates a stronger demand for external guarantees from the authority. In other words, promoting urban renewal in such peri-urban subdistricts ameliorates their equity through spatial planning, regulations, policies, etc.

Generally, targeted accessibility assessment concerning travel distance, time cost, and monetary cost [55] is a breakthrough point for precisely evaluating the sustainability of peri-urban areas. On this basis, enhancing the relatively scarce public services and infrastructure could effectively address the issue of equity in peri-urban areas. Notably, however, the ability of urban residents to access public services is significantly tied to the registered domicile in China, which transfers the equity problem to a housing problem in most cases. Under the strict control of incremental urban construction, the solution to housing problems relies more on integrating existing housing resources by official forces as well as circulating and implanting additional infrastructure space in the process of renovating existing buildings, thereby enhancing the capacity of public services.

4.3. Research Significance and Prospects

By quantifying and grading the regional features and clustering the attributes of the 128 subdistrict-level units in the municipal area of Chengdu, the type characteristics and their spatial distribution were comprehensively obtained. This can guide the implementation and evaluation of sustainable urban renewal in megacities. The present study’s findings indicate that when targeting the urban renewal process, various attributes within the jurisdiction of Chinese megacities exhibit significant spatial differences. These differences undoubtedly dissimilate the short-term goals and methods of urban renewal. Only when the spatial carriers for implementing urban renewal actions are clearly defined and their characteristics are clearly distinguished can the supervision and constraints on the urban renewal process better meet the actual needs of the local area.

Through dividing subdistrict unit types, the development of differentiated assessment standards and management norms for sustainable urban renewal can become more evidence-based. This can lead researchers to more accurately describe the unique problems within different types of spatial units. Taking public participatory governance in Chengdu City as an example, the key in the core subdistricts lies in the legal foundation, policy framework construction, and the effective coordination and integration of strong public will. At the same time, cultivating the willingness and ability to participate is more urgent in inter-axis areas. Evidently, the problem expression of public participatory governance becomes more dedicated and targeted through the subdistrict type descriptions. This provides clearer guidance for formulating incentive policies for public participation.

Additionally, classification work based on actual administrative boundaries can better respond to the power hierarchy in complex urban environments. Notably, the classification and evaluation results based on subdistrict units can directly correspond to the hierarchical authority of the subdistrict offices and their superior management institutions, exerting direct constraints and guidance on exercising their power. Consequently, the classification guides local governments to apply targeted, focused attention in implementing urban renewal actions while formulating feasible renewal plans based on different subdistrict attributes.

The classification results can be applied to optimizing the indicators in MCDA sustainability evaluation frameworks such as MAVT and TOPSIS. This makes them more applicable to evaluating small and medium-sized units, including subdistricts and communities. This study’s findings also elicit avenues for future work, such as verifying the optimization effect. For example, future researchers can use the indicator framework before and after classification optimization to evaluate the same group of objects separately and compare the evaluation results. Research should also attempt to apply this classification method to more megacities in China and Asia, examine the similarities and differences in their classification results, verify the effectiveness of this classification method, and explore the possibility of optimizing this method with big data and deep learning.

5. Conclusions

Taking the jurisdiction of Chengdu City as an example, nine types of units were obtained through a combination of qualitative and quantitative analysis. The analysis of their attribute mapping relationships and spatial distribution indicates a significant spatial mismatch between the social and economic external conditions. The findings also indicate a strong presence of the circle axial structure in the urban renewal process. Based on the classification results, three typical characteristics and targeted recommendations for sustainable renewal in megacities are provided below.

(1) High-density central urban areas should include public participation, especially participatory governance, as an essential component to bridge the disconnect between the power and responsibility of subdistrict offices during the transformation of urban governance. Further research of the general framework and application scenario of public participation in the urban renewal process should be conducted based on subdistrict unit classification. The present study provides a foundation for researching this topic regarding megacities.

(2) The issue of social equity is a significant obstacle to achieving sustainable development in semi-urbanized areas represented by inter-axis regions. This must be emphasized in the process of urban renewal. Considering China’s registered residence management system, taking the government as the main body to solve the housing problem with limited incremental construction will become key to sustainable renewal in the peri-urban areas of megacities.

(3) The outer edge area is a de facto preparatory urbanization area, and it is necessary to be vigilant against destructive development or redevelopment. Restricted by the recognition of urban renewal in the past, the risk of destructive construction is easily overlooked and has more potential hazards that need to be considered.