Community Resilience Evaluation and Construction Strategies in the Perspective of Public Health Emergencies: A Case Study of Six Communities in Nanjing

Abstract

The theory of resilience has undergone three stages: engineering, ecological, and evolutionary. It has been developed in various professional fields, focusing on research scales such as urban resilience and community resilience. As the smallest unit of urban composition, the community serves as the principal carrier of numerous emergencies at the grassroots level. Its resilience construction level is somewhat connected to the city’s safe development. However, there is still a lack of a systematic evaluation framework for assessing community resilience, and studies from the perspective of public health safety also lack scientific quantitative results and dynamic analysis. In order to fully understand the connotation of resilient community in the combination of epidemic prevention and control, this study employs literature crawling and high-frequency vocabulary screening to construct a three-level resilience index. Taking into consideration both physical and social factors, a community resilience evaluation system with 4 core indicators, 14 secondary indicators, and 39 tertiary indicators is established by employing the resilience matrix (RM) framework and Analytic Hierarchy Process (AHP). It set up a collection quantification path based on the properties of multivariate data and weighted the indicators using the Delphi method. Taking the typical community in Xuanwu District, Nanjing, as the research sample, the differentiated performance during the COVID-19 pandemic is analyzed, and a systematic evaluation and scoring are conducted. The resilience composition and improvement directions of each sample are interpreted and analyzed to support the formulation of future sustainability strategies as much as possible. The study developed an evaluation approach combining three time periods and four response dimensions to demonstrate a relationship between complex factors and community resilience. The expandable resilience evaluation system offers a wide range of applications and serves as a scientific reference for strengthening community resilience, which is critical for urban sustainability.

1. Introduction

In recent years, cities around the world have been hit by natural and man-made disasters such as fires, earthquakes, and public health problems. The strength of these shocks is increasing at the regional, national, and global levels. These repercussions have posed numerous needs for a set of urban resilience strategies to face unforeseen issues in the future [1]. Urban space serves as the fundamental carrier of external elements, determining whether a city can accommodate positive external factors and withstand negative external factors [2]. Resilience is critical to the long-term growth of cities and communities. Community plays the most fundamental role in urban management, and research on community resilience is gradually gaining attention in China. Compared to complex adaptation theory, resilience theory is more comprehensive. It emphasizes the dynamic nature of how communities deal with public disasters, making it more applicable to addressing the diverse challenges faced by communities [3]. Given this context, research on community resilience during public health emergencies is of significant importance. In a way, this guarantees the health and safety of community members by helping to give more detailed guidelines for crisis management and community planning, improving the ability of communities to respond to public health emergencies [4]. In terms of spatial vitality, resilient infrastructure enhances a community’s ability to withstand external shocks, enabling quick recovery and adjustment in the face of challenges. This helps maintain the vibrancy and attractiveness of the community space [5]. Furthermore, resilient infrastructure enhances resource usage efficiency and lowers transportation costs in addition to producing compact and varied spatial forms [6]. In terms of spatial efficiency, resilient communities ensure the sustained functionality and efficient utilization of various spaces, enhancing residents’ quality of life and contributing to the overall sustainable development of the community [7]. Furthermore, improving community resilience through public health safety enables a more in-depth study of local resources within the context of sustainable development. This strategy enables communities to better plan for future unpredictability, increasing their ability to respond successfully. It has a considerable impact on localized development, which is essential for long-term growth [8].

2. Literature Review

2.1. Resilience Theory

The application of resilience theory has undergone three stages: engineering, ecological, and evolutionary. In the 19th century, resilience was extensively applied in the field of engineering, which is known as Engineering Resilience [9]. In 1973, Canadian ecologist Holling first introduced the concept of resilience in the context of ecological systems while researching how to address environmental disasters [10]. Early in the twenty-first century, Gunderson and Holling established the Adaptive Cycle Theory, which served as the foundation for Evolutionary Resilience. According to this idea, systems develop throughout the course of four phases: exploitation, conservation, release, and reorganization [11].

Community Resilience Theory represents an extension and expansion of urban resilience theory to grassroots communities [12]. Following Mileti’s introduction of the concept of Disaster-resilient Community in 1999 [13], numerous scholars and organizations have conducted research on resilient communities, each proposing their own perspectives on what constitutes a resilient community. In the 2000s, the United Nations’ UNISDR defined community resilience as the ability to achieve and sustain normal functioning of a community under potential disaster impacts [14]. Community resilience is a process and aptitude rather than an outcome that is characterized by stability as opposed to adaptability. It is the speed and capacity of a community to recover from the aftermath of an incident and the ability to both respond to and recover from disasters as well as to promote post-disaster reconstruction [3,15,16]. Twigg suggested that communities should have the ability to respond to or maintain specific functions and structures in the face of disaster pressures [17]. In the 2010s, scholars clarified that community resilience is the ability of community members to investigate and utilize community resources in unpredictable and changing environments. They also highlighted that community resilience is the process by which communities recover to their pre-disaster levels and is the ability to prevent, withstand, and mitigate health emergencies [18,19]. Communities go through a process of self-improvement and self-learning while fending off calamities, Lopez-Marrero observed, reaching a new equilibrium. This viewpoint stresses adaptability as well as resistance [20]. Bruneau described resilience as the ability to prevent shocks, absorb them, and swiftly return to a state of stability [21]. According to Berkes, social resources and community traits—such as initiative and self-organization—combine to produce resilience [22]. In the 2020s, Lucie emphasized that community resilience is not only the ability to resist disasters but also the capacity to recover and adapt to new environments after a crisis [23]. Shaikh argued that community resilience includes multiple aspects such as socio-economic factors, infrastructure durability, community social support networks, and government emergency response capabilities [24]. These studies discuss the components and meaning of community resilience from many viewpoints, providing a basic framework for identifying the variables that influence community resilience.

Focusing on the themes of urban renewal, community resilience, and sustainable development, 670 articles were retrieved from the core collection of the Web of Science. After importing the literature into CiteSpace for keyword clustering and examining the research hotspots in resilience theory since 1990, it is clear that the research predominantly focuses on the built environment and organizational management (Figure 1). Furthermore, the notion of community resilience has been continually developed in the context of managing public health events as a result of their frequent occurrence. This gives communities more thorough instruction and motivates everyone to work together to create a stronger, longer-lasting health security network. According to Jewett and Mah, social disparities are made worse by health emergencies, which highlight these inadequacies in social cohesiveness [25]. As a vital instrument for attaining social cohesiveness, the notion of community resilience makes it possible to adjust to shock occurrences and lessen their effects. Fransen and Peralta presented the dynamic viewpoint of community self-reliance in disaster response through a review of the literature and case studies. They emphasized the necessity of fostering social cooperation and setting up support systems while highlighting the common difficulties communities confront in defending against disasters [26]. Global research on resilience in urban development is becoming more in depth and comprehensive, providing a solid foundation for localizing community resilience research in China.

2.2. Resilience Assessment

Chinese scholars have mostly increased their research on community resilience evaluation based on international studies in the field of resilience assessment systems. Peng offers a typical perspective, positing that the evaluation process for resilient communities contains a set of competences, a growth process, and a developmental goal [27]. Scholars have localized and enhanced evaluation models to reflect the realities of communities. The RATA (Resilience Assessment Tool for Urban Areas) evaluation technique is commonly used in urban areas [28]. Scholars like Yun have used RATA as a foundation to select localized indicators for resilience communities, resulting in an evaluation approach tailored to existent communities in China [29]. The PSR model was adapted into the “DPSRC” mode and then used to investigate social resilience in Guangzhou’s international communities [30]. The vitality and resilience of historical and cultural areas have also been appreciated, and a number of explorations have been established based on the PSR concept [31,32]. In addition, new technologies have contributed to resilience research, including the development of a community emergency capacity assessment index system and a public safety risk assessment system [33]. The DEA (Data Envelopment Analysis) [34,35] and TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) are used to assess community resilience [36]. Guo et al. aimed to reduce catastrophe susceptibility and improve post-disaster recovery skills in older urban populations. They developed a catastrophe risk assessment approach for urban aging communities that includes the stages of “prevention, resistance, avoidance, and rescue” [37].

The RM (Resilience Matrix) framework, put out by Linkov, is regarded as one of the main frameworks for examining alterations in community resilience when it comes to resilience matrix construction. The phases of a disaster’s occurrence are broken down into four stages by the RM, which is a 4 × 4 two-dimensional framework: preparation, absorption, recovery, and adaptation. It divides the resilience system into four categories at the same time: organizational, social, physical, and informational, for a total of 16 distinct assessment units. This framework facilitates the measurement, statistical analysis, and comparison of resilience across time and space, providing a visual depiction of gaps and deficiencies among different systems [38]. Based on this method, Fox measured the resilience index of New York State communities during Hurricane Sandy and guided communities in enhancing resilience [39]. Chen proposed an urban resilience assessment model for rainfall catastrophes that takes into account characteristics including mechanisms for resistance, recovery, and adaptation. He discovered that a key element in how cities react to the threat of rainstorms is resilience [40]. Shi classified and evaluated general communities, elder communities, and urban village communities according to their features. She developed an evaluation method for community resilience that considers defensive capacity, recovery ability, and vulnerability [41].

However, existing research on resilience indicators in China focuses mostly on the material and spatial components of the “pre-disaster resistance” period with no systematic development of indicators at the social organizational level. This constraint makes it difficult for communities to function across the pre-disaster, disaster, and post-disaster stages. In this context, the goal of this research is to create a complete, scientific, and actionable indicator system for assessing community resilience. During this process, the RM can be enhanced and used to incorporate public health safety within the research framework. It seeks to improve systemic resilience and community interaction assessment by incorporating easily observable indicators. It should be noted that unlike traditional community resilience evaluation, resilience focus on public health emergencies has some unique characteristics in terms of the built environment, emergency space, critical facilities, and organizational behavior, and its resilience assessment must include targeted indicators.

3. Materials and Methods

The study employed three cycles and four dimensions to determine the degree of community resilience and comprehensively illustrated the resilience features through graphical representation. Firstly, the RM was employed in the study to structure and construct resilience dimensions across the built environment, emergency spaces, critical facilities, and organizational behavior. Secondly, indicators of community resilience were chosen using literature crawling and high-frequency vocabulary screening. Thirdly, indicators were categorized hierarchically, and weights for each level of indicators were determined using the Delphi method and the analytical hierarchy process (AHP). Based on weight data and scoring criteria, indicators of each community were measured using Fuzzy Comprehensive Evaluation (FCE). The measurement results encompass both quantitative and qualitative aspects. Qualitative results were derived from the concentrated feedback of local residents through survey questionnaires, while quantitative results were computed using software such as ArcGIS and Space Syntax [42]. Finally, the resilience level measurement results of communities underwent accuracy validation, visualization, and horizontal–vertical comparisons. Optimization suggestions were proposed to address resilience gaps in different communities (Figure 2). Details about the software version and the website are provided in the Appendix A.

3.1. Evaluation System

The study employed literature crawling and high-frequency vocabulary screening methods to examine the indicators of the resilience evaluation system at all levels, encompassing comprehensive Chinese and English literature to obtain scientifically grounded data. The study conducted a search on two large databases, CNKI (China National Knowledge Infrastructure) and Web of Science, using the keywords “community resilience” and “public health security”. A total of 3829 articles were retrieved and subjected to steps including duplicate screening and full-text screening. After removing comparable and low-quality literature, 925 research papers were identified and categorized based on publishing influence, author authority, citation volume, and so on. Then, the literature including community resilience evaluation indicators had been included in the indicator screening procedure. Taking into account the research on China’s national conditions, reference citation frequency, and authoritative degree of publications, 25 Chinese and 25 English publications were ultimately chosen and included in the analysis. It can be observed that foreign scholars have reached a certain level of agreement in the field of community resilience assessment, emphasizing research in the areas of community governance, social relationships, and organizational management as critical for assessing community resilience, with a focus on social elements. Domestic Chinese researchers, on the other hand, are especially concerned with infrastructure, geographical issues, organization, community capital, ecology, population, social capital, and technology. More than 40% of the domestic evaluation literature focuses on infrastructure, community space, and the environment, emphasizing the significance of material and physical factors.

Public health emergencies, such as infectious disease outbreaks as food safety concerns, differ from other urban disasters. They are frequently unexpected and unpredictable, and they can spread quickly, necessitating immediate measures to respond. This includes isolating patients, locating the source of infection, carrying out large-scale disinfection and prevention efforts, and so on. Furthermore, the response system for public health emergencies differs from that of other urban disasters, and there are unique policies and processes for responding to and disclosing information about such events. Therefore, in addition to taking into account the three fundamental dimensions of organization, environment, and facilities when setting up the resilience evaluation system for this study, it is also necessary to incorporate the unique context of public health catastrophes and include “emergency space” as one of the four dimensions (first-level indicators) for community resilience evaluation. Referring to the RM, further indicator selection will be based on the complete cycle of pre-disaster preparedness and prevention, disaster impact and response during the event, and post-disaster recovery and adaptation (

Table 1. Dimensions explanation.

Dimension	Preparedness and Prevention Phase (P1)	Impact and Response Phase (P2)	Recovery and Adaptation Phase (P3)
Built environment (A1)	(A1,P1) To maintain a good community environment and enhance the friendliness of public spaces, encouraging residents to participate in outdoor activities.	(A1,P2) Restricting external traffic flow at community entrances and equipping open public spaces with enhanced epidemic prevention functions to ensure residents’ physical and mental well-being.	(A1,P3) Establishing community parks, pocket green spaces, and other public recreational areas, and utilizing linear greenery as a natural barrier to reduce health risks.
Emergency spaces (A2)	(A2,P1) Planning adequate isolation spaces and layout of refuge areas, ensuring sufficient evacuation areas.	(A2,P2) Always ensure the security of emergency spaces and strive for unobstructed emergency routes.	(A2,P3) Expanding the number of emergency spaces, repair damaged areas, and meeting the dual requirements of emergency and daily use.
Critical facilities (A3)	(A3,P1) Increase the redundancy of community facilities and cultivate residents’ awareness of using safety facilities.	(A3,P2) Fully utilize community hospitals, sports facilities, leisure and health centers, and other health facilities for emergency interventions to minimize residents’ health injuries.	(A3,P3) Accelerate the restoration of postal, express delivery, and other transportation facilities to meet the dynamic needs of integrating community services during and after pandemics.
Organizational behavior (A4)	(A4,P1) Conduct early warning and prevention information campaigns; perform safety hazard inspections.	(A4,P2) Initiate emergency rescue and evacuation operations; formulate disaster response plans.	(A4,P3) Announce the disaster situation and ongoing efforts; promote community spirit of mutual assistance; enhance the level of health activities for residents.

).

The indicator selection is based on the literature screening mentioned above. Firstly, through the literature review and identification of key indicators, a high-frequency indicator library of 88 indicators was established to meet the four characteristics outlined in Table 1. Indicators related to the built environment (A1) primarily assess the sustainability of natural resources and environmental quality, such as the quality of the pedestrian environment and public green spaces. Indicators related to emergency spaces (A2) focus on the physical layout and structure of the community, such as emergency isolation spaces, epidemic evacuation routes, and dual-use spaces for both normal and emergency situations. Indicators for critical facilities (A3) involve the infrastructure and services provided by the community, including healthcare, education, and waste management facilities. Indicators for organizational behavior (A4) focus on the social structure and governance within the community, such as community participation, emergency response mechanisms, and organizational coordination capabilities. Secondly, the relevance ratio of all indicators in the high-frequency indicator library was calculated, and indicators with a relevance greater than 50% were selected to establish a preliminary three-level evaluation system. Thirdly, indicators not approved by more than 5 out of 40 experts were removed, completing the preliminary indicator processing. Finally, the remaining indicators were combined and separated into three stages: preparedness and prevention, impact and response, and recovery and adaptation. This method produced a community resilience evaluation system that included 4 core indicators, 14 secondary indicators, and 29 tertiary indicators. And setting specific evaluation criteria for each tertiary indicator helps to understand the community’s performance in dealing with public health emergencies and provides data support for formulating more effective response strategies (

Table 2. Community resilience evaluation index system and quantification methods.

First-Level Indicator A	Phase	Second-Level Indicator B	Third-Level Indicator C	Measurement Methods
Resilience of built environment A1	P1	Pedestrian and bicycle lane B1	C1 Street visual comfort	Semantic segmentation
			C2 Perception of street scale	Street height-to-width ratio
		Land use B2	C3 Land development intensity	Building density formula
			C4 Land use diversity	Land use formula
	P2	Transportation space B3	C5 Road integration	SpaceSyntax
			C6 Road connectivity	The ratio of intersections to sidewalks
	P3	Open space B4	C7 Spatial coverage	The ratio of open space area to the total community area
			C8 Morphological compactness	Compactness Index formula
Resilience of emergency space A2	P1	Emergency shelter signage system B5	C9 Signage utility	Questionnaire
			C10 Layout rationality	Field research
		Supply storage space B6	C11 Spatial coverage	Service coverage of supply points
			C12 Material supply level	Two-Step floating catchment area method
	P2	Emergency defense space B7	C13 Accessibility of places	The shortest distance from shelter to hospital
			C14 Coverage of places	Shelter service area
			C15 Safety of emergency access	Road congestion
	P3	Post-pandemic integration area B8	C16 Operability of post-pandemic transition	Percentage of operable space units
			C17 Scale of spatial planning for post-pandemic transition	Area of the epidemic prevention space
Resilience of critical facilities A3	P1	Public service facilities B9	C18 Facility equity	Location entropy index
			C19 Facility coverage	Public facility service coverage
	P2	Emergency support facilities B10	C20 Provision of healthcare facilities	Two-step floating catchment area method
			C21 Accessibility of healthcare facilities	Two-step floating catchment area method
	P3	Post-pandemic integration facilities B11	C22 Number of available existing facilities	Field research
			C23 Facility maintenance	Field research
Resilience of organizational behavior A4	P1	Preventive baseline conditions B12	C24 Residents’ disaster awareness	Questionnaire
			C25 Community disaster preparedness level	Questionnaire
	P2	Emergency preparedness level B13	C26 Level of resident activity	Standard deviational ellipse
			C27 Community organizational capacity	Python
	P3	Recovery adaptability B14	C28 Healthiness of activities	Questionnaire
			C29 Restoration participation	Python

).

3.2. Weights

A method of expert scoring was used to establish the weights. This study expanded the number of experts to 40 in order to give a wide variety of expert professional backgrounds. There are forty people involved in the selection process, including specialists in the fields of architecture, urban planning, and comprehensive disaster prevention as well as other pertinent persons including street workers, medical personnel, inhabitants of the community, and personnel involved in community epidemic prevention.

First, determine the relative importance of indications based on expert opinions. Second, create judgment matrices for all factors in the indicator layer to determine the weights of each indicator. Perform consistency checks on the weight results and compute the composite weights of indicators. The composite weights are calculated by multiplying the indicator weights by the weights of their parent indicators. Finally, weighted calculations are performed on the indicators of a community within the research area by combining the composite weights with resilience scoring standards. All indicator scores are added together to obtain the community’s comprehensive resilience score. As shown in the formula (1), W_i represents the weight assigned to the indicator, S_ij represents the comprehensive resilience score of a certain community, and S_i represents the score of the indicators involved in that community.

$${\mathrm{S}}_{\mathrm{ij}}={\displaystyle {\sum }_{\mathrm{i}=1}^{\mathrm{n}}{\mathrm{S}}_{\mathrm{i}}}\times {\mathrm{W}}_{\mathrm{i}}$$

(1)

Among the first-level indicators, A1 (built environment) has the highest weight value, making it crucial for enhancing community resilience. A2 (critical facilities) and A3 (emergency space) have similar weight values and are regarded as secondary significance, whilst A4 (organizational behavior) has the lowest weight value.

In the weight value of second-level indicators, B1 (slow road) has a weight of 0.4606 in A1 (built environment resilience), which is multiplied by the weight of A1 to obtain a combined weight of 0.1586, ranking first among all second-level indicators. B8 (epidemic prevention and control combined area) ranks first in A2 (emergency space resilience), which is followed by a combination weight in third. B10 (emergency support facilities) is the second most important factor in A3 (critical facility resilience), after B1 (slow traffic highways), with a combined weight ranking second. In A4 (organizational behavior resilience), B14 (resilience to adaptation) comes in first, which is followed by the combination weight (

Table 3. Weight of indicators.

First-Level Indicator	Weight	Second-Level Indicator	Second-Level Combined Weight	Third-Level Indicator	Third-Level Combined Weight	Ranking
Resilience of built environment A1	0.3444	B1	0.1586	C1	0.0529	9
				C2	0.1057	1
		B2	0.0257	C3	0.0086	26
				C4	0.0171	18
		B3	0.0470	C5	0.0117	23
				C6	0.0352	12
		B4	0.1131	C7	0.0754	3
				C8	0.0377	11
Resilience of emergency space A2	0.2111	B5	0.0205	C9	0.0068	28
				C10	0.0137	22
		B6	0.0220	C11	0.0055	29
				C12	0.0165	19
		B7	0.0505	C13	0.0202	15
				C14	0.0101	25
				C15	0.0202	15
		B8	0.1181	C16	0.0591	6
				C17	0.0591	6
Resilience of critical facilities A3	0.2472	B9	0.0489	C18	0.0326	13
				C19	0.0163	20
		B10	0.1212	C20	0.0606	4
				C21	0.0606	4
		B11	0.0771	C22	0.0578	8
				C23	0.0193	17
Resilience of organizational behavior A4	0.1972	B12	0.0415	C24	0.0103	24
				C25	0.0311	14
		B13	0.0475	C26	0.0079	27
				C27	0.0400	10
		B14	0.1082	C28	0.0927	2
				C29	0.0155	21

, Figure 3).

The combination weight value of C2 (scale perception of slow traffic streets) is first among the third-level indicators, indicating its central role in the community built environment. In contrast, the low weight assignment ranking of variables such as C11 (spatial coverage) and C9 (identification diversity) demonstrates that these characteristics are ignored in community resilience creation (Table 3).

4. Samples and Results

4.1. Study Area

The study area is located in Xuanwu District, Nanjing City, Jiangsu Province. This area is a core part of Nanjing’s central urban district, which is characterized by a blend of historical culture and modern facilities. It is an important residential area in Nanjing’s central urban district and a typical area with a concentration of old residential buildings. As shown in Figure 4, the six communities selected for the study are Suojin Village Community (I), Bancang Community (II) Zixincheng Community (III), Jiangwangmiao Community (IV), Huayuan Road Community (V), and Yingtie Village Community (VI). Among these, Communities V and VI were designated as high-risk areas during the COVID-19 pandemic and experienced community lockdown. Communities I and II are recognized as comprehensive disaster reduction model communities in Nanjing with certain disaster prevention and mitigation capabilities. Communities III and VI are high-density urban communities. These six communities differ in spatial layout, facility configuration, and other aspects, representing different types of communities. The evaluation results have significant implications for broader application.

• Suojin Village Community (I) is established in the 1980s, the community covers an area of 0.45 square kilometers and has a population of 12,000. It located in the central part of Xuanwu District, bordered by Xuanwu Lake Street to the east, Taipingmen and Houzaimen Street to the south, Xuanwu Lake and Xuanwumen Street to the west, and the Shanghai–Nanjing Railway and Hongshan Street to the north.
• Bancang Community (II) was established in the 1980s, covering an area of 0.28 square kilometers with a population of 6800. It is bordered by the Purple Mountain to the south, Jiangwangmiao Community to the east, Xuanwu Lake to the west, and is adjacent to Suojincun Street.
• Zixincheng Community (III) was established in the 1980s, covering an area of 0.35 square kilometers with a population of 6000. It is bordered by Purple Mountain to the east, Xuanwu Lake to the west, Baima Park and Bei’anmen Street to the south, and Ningxi Road to the north.
• Jiangwangmiao Community (IV) was established in the 1990s, covering an area of 0.32 square kilometers with a population of 6000. It is bordered by Ningxi Road to the south, National Highway 312 to the west, and Huaxin West Road to the north.
• Huayuan Road Community (V) was established in the 1990s, covering an area of 0.47 square kilometers with a population of 13,700. Garden Road runs through the community, which is bordered by Huaxin West Road to the east, Nanjing Forestry University to the west, Garden Road Neighborhoods 5 and 8 to the south, and Xuanwu Avenue to the north.
• Yingtie Village Community (VI) was established in the 1990s, covering an area of 0.57 square kilometers with a population of 13,000. It is bordered by the Jingwu Overpass to the east, Yingtuo Huayuan Road Community to the south, the East Long-Distance Bus Station to the west, and Xuanwu Avenue to the north.

4.2. Resilience Evaluation Results

Based on the measurement results of indicators, combined with indicator scoring criteria and weights, we determined the indicator scores for six communities (

Table 4. Comprehensive resilience evaluation of sample communities.

Sample Communities	I	II	III	IV	V	VI
Comprehensive resilience evaluation	3.1158	3.502	2.2415	2.5979	3.3753	2.5028
Ranking	3	1	6	4	2	5

). The entire status of community resilience offers a gradient of Community II > Community V > Community I > Community IV > Community VI > Community III.

Community I (3.1158) and Community II (3.502), identified as Comprehensive Disaster Reduction Demonstration Communities of China by the government, have shown great resilience and adaptation capacities during public health emergencies. Community IV (2.5979) and Community III (2.2415) both have intermediate overall scores. The community’s total resilience assessment is somewhat lower as a result of the relatively low scores for disaster prevention and reduction among the three level indicators of the two. Community V (3.3753) and Community VI (2.5028) are both high-density urban areas. However, when compared to the other four communities, their third-level indicator ratings differ significantly. Community VI, for example, has low facility supply capacity, limited open space coverage, and insufficient engagement in community repair, earning an overall resilience ranking of only 5. It is worth noting that Community V is not a full disaster reduction demonstration community, but the resilience level of indicators at all levels exceeds the standard line with the most comprehensive resilience level ranking second.

4.2.1. First-Level Indicator

As shown in Figure 5, there are certain variations in the scores of indicators in each dimension. The built environment (A1) has the highest overall score, followed by the critical facilities (A3), and emergency space (A2) and organizational behavior (A4) have lower overall scores.

Specifically for each community, Community I stands out with a high score of 0.827 in emergency spatial resilience, highlighting its strong spatial management and rapid response capabilities. Communities II and V demonstrate balanced resilience levels, showing strong overall capabilities across different resilience dimensions. Community III performs well in built environment resilience (0.7175), but its emergency spatial resilience (0.4554) and organizational behavioral resilience (0.4955) are relatively weaker. Community IV scores well in built environment resilience (1.0235), but it lags behind in emergency spatial resilience (0.4248) and organizational behavioral resilience (0.3552), indicating a need to address gaps in community emergency governance and spatial epidemic conversion. Community VI significantly leads in built environment resilience with a high score of 1.2629, demonstrating robust foundational support when facing public health emergencies.

4.2.2. Second-Level Indicator

Figure 6 depicts the scores of second-level indicators of the sample communities. Community I excels in public service facilities (B9 = 5) and recovery adaptability (B14 = 4.5) while scoring lower in open space (B9 = 2) and pedestrian and bicycle lane (B1 = 2). Community II performs in a balanced manner. It performs quite well in terms of land use (B2), public service facilities (B9), and recovery adaptability (B14); all three categories have scores over 4, while open space (B4 = 2) has a lower score. Community III scores reasonably well on public service facilities (B9 = 3.5) and recovery adaptability (B14 = 3.5). However, it receives lower scores in emergency support facilities (B5 = 1.5) and preventive baseline conditions (B12 = 1.5) among other areas. Community IV performs well in supply storage (B6 = 4) and emergency support facilities (B123 = 4.5) but falls behind in transportation space (B3 = 1.5) and post-pandemic integration area (B11 = 1.5). Community V excels in material storage space and emergency support facilities (both 4.5) but ranks lower in recovery adaptability (B10 = 2.5). Community VI excels in open space (B9 = 4) and transportation space (B3 = 4) but falls short in supply storage space (B6 = 1) and emergency support facilities (B10 = 1.5).

4.2.3. Third-Level Indicator

As indicated in Figure 7, it is evident that the six communities show significant differences in their third-level indicator resilience scores. Community I demonstrates a balanced resilience composition, particularly excelling in land development intensity (C3 = 4) and facility equity (C18 = 5). Community II scores high in multiple indicators but shows a slightly weaker performance in material supply level (C12 = 2). Community III performs extremely poorly in accessibility of places (C13 = 0) and spatial coverage (C7 = 1), highlighting critical areas needing attention and improvement. Community IV excels in specific indicators such as material supply level (C12 = 5) and the accessibility of healthcare facilities (C21 = 5), indicating strengths in critical resource allocation and accessibility. However, it shows significant deficiencies in land use diversity (C4 = 1) and scale of spatial planning for post-pandemic transition (C17 = 1). Community VI performs exceptionally well in level of resident activity (C26 = 5), demonstrating their potential in promoting community activities and participation.

4.3. Validity and Reliability of the Empirical Results

The empirical results obtained from the above process were tested for reliability using the Brown–Forsythe ANOVA to calculate differences in resilience scores among communities. As indicated in

Table 5. Analysis of variance (ANOVA) results.

Analysis Items	Name	Sample Size	Average Value	Standard Deviation	Brown F	p
Resilience evaluation	Community I	29	0.03	0.05	2.340	0.046
	Community II	29	0.03	0.04
	Community III	29	0.08	0.14
	Community IV	29	0.08	0.11
	Community V	29	0.04	0.06
	Community VI	29	0.05	0.07
	Total	174	0.05	0.09

, there is significant variability in resilience evaluation across different communities (Brown F = 2.340, p = 0.046 < 0.05).

The evaluation of community resilience from the standpoint of public health events begins with the features of the complete response cycle of public health events. “Emergency space” is added as one of the four dimensions of community resilience evaluation, forming a system with four primary indicators, fourteen secondary indicators, and twenty-nine tertiary indicators. This system is based on the traditional dimensions of built environment, critical facilities, and organizational behavior in community resilience evaluation combined with the unique background of public health events. Drawing on conventional resilience research, this method re-examines the intricate makeup and natural association of community resilience, highlighting the community’s capacity to adapt to public health disasters. In order to improve the quality and credibility of factor collection and prevent misleading conclusions caused by data quality issues, AHP and word frequency screening are combined during the evaluation system’s establishment. This process involves filtering and screening vocabulary data and high-frequency publications, deleting duplicate data, merging similar items, filling in missing values, and more. The community resilience assessment system’s application scope may be narrowed or broadened in further studies by changing and screening keywords, broadening the selection region, and using other techniques.

Taking advantage of the benefits of the digital age in data acquisition, it employed the ArcGIS platform to integrate subjective evaluation and objective data: field research, survey questionnaire distribution, Python, space syntax, and so on, collecting data and standardizing them to obtain resilience scores for each. The methodical examination of three-level indicators can reveal the resilience features of each community at the macro, meso, and micro levels, which correspond to the three-level “spatial–social” composition of urban life, community life, and resident life. In the process of establishing community resilience, we can begin with the secondary indicators at the meso level and then provide targeted supplements to low-scoring projects depending on the composition of the tertiary indicators in the reference secondary indicators (Figure 8).

• Suojin Village Community (I) established an efficient information communication mechanism to ensure that residents were kept up to date with the latest pandemic developments. Among the interviewed residents, there was widespread satisfaction with the community’s pandemic prevention and control performance. However, it performs poorly in pedestrian and bicycle lane (B1). Thus, it is advised that the walking and cycling systems be improved, the public transit system be made more accessible, the transportation network and transfer facilities be laid out sensibly, and the environment for slow traffic be improved in the ensuing resilience enhancement construction. Furthermore, the community slow traffic road system’s design and optimization must be combined in order to offer adaptable motor vehicle management plans, such as restricting parking during public health emergencies to lower safety risks.
• Bancang Community (II) performs well in public service facilities (B9) and recovery adaptability (B14). Residents expressed universal satisfaction with the community’s pandemic prevention and control performance: during the pandemic, the town plaza was rapidly converted into an emergency center, significantly slowing the spread of the virus. However, it performs poorly in open space (B4). Thus, it is advised that open green spaces be used for more purposes and that outdoor activity areas be planned with consideration for the local climate and the demands of the occupants throughout the ensuing resilience enhancement construction. In addition, it may guarantee comfortable slow traffic by improving the accessibility and connection of public areas like parks and block green areas. Simultaneously, adjust the spatial layout in accordance with the requirements of the inhabitants, such as by including rest spaces and lights in the park, to provide a better open area for everyday community activities.
• Zixincheng Community (III) performs poorly in emergency support facilities (B10). The majority of the citizens were dissatisfied with the community’s pandemic prevention and control performance. It is recommended to improve the configuration of medical equipment to ensure meeting various medical needs. Balanced layout and adding facilities to fill gaps and expand service coverage are essential. Additionally, establishing a 15-min disaster prevention and epidemic prevention zone and increasing facilities such as health stations can enhance epidemic prevention capabilities.
• Jiangwangmiao Community (IV) performs poorly in transportation space (B3). It is recommended to optimize the punctuality of public transportation and integrate non-motorized transportation, increase the density of bus stops, and reduce waiting times. Additionally, it is crucial to strategically allocate public transportation, medical facilities, and open spaces, establish a network of slow traffic and life services covering the community, and promote the development of a healthy community.
• Huayuan Road Community (V) performs poorly in emergency defense space (B7). It is recommended to optimize emergency shelters to respond to public health emergencies. It is suggested to establish construction standards that match the community, renovate public buildings to meet disaster response needs, and consider public and commercial facilities as potential shelters. Establishing and updating relevant databases for the rapid conversion of space use is also recommended.
• Yingtie Village Community (VI) performs poorly in supply storage space (B6). It is recommended to improve community emergency material reserves by establishing dedicated storage facilities. Implementing efficient material storage and rotation systems, and integrating community resources to optimize emergency provisioning, are crucial steps. Ensuring the seamless supply and utilization of materials in both emergency and normal situations, covering all residents and organizations, will enhance emergency response capabilities.

5. Discussion

5.1. Optimization Strategy from a Full Cycle Perspective

5.1.1. Optimization Strategy for Preparation and Prevention Phase

Table 6. Evaluation of community resilience in P1.

Dimension	I	II	III	IV	V	VI
Built environment	0.3501	0.5287	0.3172	0.6872	0.6344	0.5287
Emergency spaces	0.1522	0.1165	0.0892	0.1263	0.1618	0.063
Critical facilities	0.2445	0.2445	0.1793	0.1141	0.163	0.0815
Organizational behavior	0.0931	0.1345	0.0517	0.0517	0.1034	0.0828
Total	0.8399	1.0242	0.6374	0.9793	1.0626	0.756

and Figure 9 show that during the preparation and prevention phases (P1), Community V has the highest overall resilience level (1.0626), whereas Community III has the lowest (0.6374). Zixincheng Community (III) has the lowest value of the built environment. Yingtie Village Community (VI) has the lowest value of emergency space as well as the lowest value in the critical facilities. In Zixincheng Community (III) and Jiangwangmiao Community (IV), the organizational behavior dimension has the lowest value. Corresponding strategies are proposed for the dimensions where communities lack resilience during this stage.

In terms of the built environment, it is advised that Community III use a variety of space use methods to ensure the environmental safety of residential areas. Walking and cycling should be encouraged by establishing a fair ratio of motorized to non-motorized lanes in order to reduce pollution from motor cars and expand inhabitants’ alternatives for community activities. In terms of emergency space, Community VI is encouraged to build new shelters or enlarge existing ones in order to better satisfy the needs for emergency shelters by making sensible use of the resources already available to the community. In terms of vital infrastructure, Community VI should raise the standards of the road network, optimize its structures, and strengthen its road connectivity. To improve their accessibility to healthcare facilities, residents should be strongly encouraged to participate in activities close to public transportation lines.

5.1.2. Optimization Strategy for Impact and Response Phase

Table 7. Evaluation and summary of community resilience at various dimensions in P2.

Dimension	I	II	III	IV	V	VI
Built environment	0.1759	0.1524	0.1055	0.0586	0.1407	0.1876
Emergency spaces	0.202	0.1919	0.0707	0.1212	0.1515	0.1212
Critical facilities	0.4848	0.4242	0.1818	0.5454	0.5454	0.1818
Organizational behavior	0.1116	0.1437	0.1037	0.0716	0.1516	0.0795
Total	0.9743	0.9122	0.4617	0.7968	0.9892	0.5701

and Figure 10 show that during impact and response phases (p2), Community V has the highest overall resilience level (0.9892), while Community III has the lowest (0.4617). Jiangwangmiao Community (IV) has the lowest built environment value and the lowest organizational behavior value. Zixincheng Community (III) has the lowest emergency spaces value, while Zixincheng Community (III) and Yingtie Village Community (VI) have the lowest critical facilities values. Corresponding strategies are proposed for the dimensions where communities lack resilience during this stage.

To provide a safe living environment, Community IV needs to regulate land use that poses a pollution risk and establish buffer zones between polluting sites and residential areas. Emergency space resilience in Community III may be enhanced by strengthening emergency evacuation routes, constructing emergency shelters, and improving the efficiency of material reserves. In terms of critical facilities, Community VI is advised to re-integrate health facilities and land resources, rationally layout facilities such as hospitals and clinics, enhance residents’ accessibility to medical services, provide more medical options, and promote residents’ medical behaviors. It is advised that the organizational behavior dimension in Community IV focus on boosting citizens’ crisis awareness while also fostering information exchange and resource sharing within the community.

5.1.3. Optimization Strategy for Recovery and Adaptation Phase

Table 8. Evaluation and summary of community resilience at various dimensions in P3.

Dimension	I	II	III	IV	V	VI
Built environment	0.2262	0.2262	0.2262	0.2262	0.3393	0.4524
Emergency spaces	0.4728	0.4728	0.2955	0.1773	0.4137	0.2364
Critical facilities	0.1543	0.2313	0.212	0.1349	0.1928	0.1928
Organizational behavior	0.4483	0.541	0.3401	0.2319	0.3091	0.2009
Total	1.3016	1.4713	1.0738	0.7703	1.2549	1.0825

and Figure 11 show that during the recovery and adaptation periods, Community II has the highest comprehensive resilience level (1.4713), whereas Community IV has the lowest (0.7703). Four Communities (I, II, III and IV) have the lowest values for the built environment. Jiangwangmiao Community (IV) has the lowest value of emergency spaces. Suojin Village Community (I) has the lowest value for the critical facilities. In Yingtie Village Community (VI), the organizational behavior dimension has the lowest value. Corresponding strategies are proposed for the dimensions where communities lack resilience during this stage.

To enhance the resilience of the built environment in Community I, a multifunctional corridor centered on health should be created, connecting important community facilities such as hospitals, schools, parks, and shopping areas. These corridors should be designed as well-landscaped walking and cycling paths to encourage physical activity among residents while also serving as rapid evacuation routes in emergencies. To address the weakness of emergency spaces in Community IV, vehicles equipped with necessary medical equipment should be provided, offering mobile medical and mental health services. During the recovery phase, these mobile units can be swiftly deployed to the areas within the community that need them the most, offering a range of services from psychological counseling to basic medical check-ups. Regarding critical facilities, Community IV should ensure redundancy in critical infrastructures such as energy and communications. Through smart management and technological innovation, it is essential to ensure these facilities can quickly resume normal operation after a disaster. To improve the resilience of the organizational behavior dimension in Community VI, attention should be paid to the interaction and cooperation between the community management and residents, fostering the community’s self-recovery capability and mutual aid spirit. Establishing a disaster management knowledge-sharing platform and regularly organizing emergency drills can enhance the overall resilience of the community.

5.2. Practical Application in Real-World Circumstances

Firstly, field study on typical communities in Xuanwu District, Nanjing, was conducted to ensure the evaluation system’s correctness and applicability. The research also suggested particular optimization strategies for resilience gaps in different communities as well as scientific references for community resilience in similar communities in the same region. Secondly, technologies such as space syntax and ArcGIS are easy to use in practical operations and may effectively assist the assessment of community resilience. Community managers may identify and solve resilience weaknesses in real-world operations by using the comprehensive resilience evaluation outcomes that are produced by combining quantitative data and qualitative input. The “evaluation design” connection can give the potential resilience level after operation at any moment, offering scientific support for accurate operation. Finally, this study provides a universal resilience matrix customized for China’s specific circumstances. Targeted indications can be included for additional expansion in applications down the road. It is feasible to better understand and strengthen the community’s capacity to respond to unexpected public health events and support the community’s sustainable growth by utilizing this systematic assessment framework and improved data processing techniques.

Complex elements and community resilience were found to be related, opening the door to multivariate data interpretation and scientific recommendations for enhancing community resilience. Resilience analysis is made more scientific by this study’s mix of an objective and subjective evaluation method and data-processing path. The scalability of the open research framework allows it to be extended outside the sample area to a broader range of regions. Additionally, indicator supplementation might be used to improve targeting.

5.3. Limitations of the Study

As the most basic component of complex organisms, community resilience is both complicated and specialized. Conventional resilience assessment models concentrate on the urban environment and use intricate machine learning and algorithmic methodologies, which makes it challenging for users to comprehend how these models function internally. This lack of transparency limits users’ trust in the analysis results and may lead to misuse or misunderstanding, ultimately affecting the quality of decision making [43]. This study has streamlined the typically complicated model to be suitable to the study of resilience at the mesoscale and microscale. This simplification allows for a faster evaluation of resilience, but it will result in a drop in accuracy. As a result, in future study, the composition of each matrix may be adjusted based on the assessment job cycle and application circumstances.

Community resilience involves numerous social and spatial elements, and it is influenced by national conditions and regional factors [44]. It is also strongly related to the degree of management. As a result, the indicators used must strike a compromise between universality and specificity. The suggested three-level assessment method in this study necessitates the proper incorporation of targeted resilience indicators based on the individual situations of many communities throughout its development and expansion. Additionally, the effectiveness of resilience analysis is also influenced by the differences in goals between modelers and users. The background and objectives of the modelers may differ from those of the users, leading to communication issues during the model development and usage process. In this study, the expert panel was broadened to guarantee coverage across several areas. However, future studies can include more precise talks on the background domain makeup of experts.

It should be noted that this study combines resilience research into the full cycle of public health emergency response and develops an evaluation system. Nonetheless, the resilience assessment model is based on static assumptions and may have biases in dynamic environments [45]. This mismatch may cause the model to fail in the face of new or changing conditions, reducing its usefulness in actual applications. As a consequence, while carrying out community environmental construction based on the results of resilience assessment, it is vital to begin with the whole lifecycle governance of public health events, establish a long-term system, and improve management.

6. Conclusions

The study re-examined the space–social security role of complex features in communities during normal and epidemic periods and developed a three-level evaluation system using methods including literature crawling, high-frequency screening, and hierarchical analysis. The resilience composition was broadened to include four components: built environment, emergency spaces, critical facilities resilience, and organizational behavior. It investigates community resilience in the context of abrupt public health emergencies, broadens the perspective of community resilience evaluation, and supplements research on community resilience evaluation systems under Chinese national conditions. In terms of samples and data, six typical communities in Nanjing (including two national comprehensive disaster reduction demonstration communities) were screened, and data collection covered the entire COVID-19 period, resulting in a resilience assessment of typical communities over the previous five years and a reference for community resilience construction. On the one hand, the research uses both subjective and objective approaches to assess indicators. Various resilience indicators were examined using methods such as data crawling and quantitative calculation, and resident feedback was added to augment the limitations of quantitative analysis, resulting in a more thorough depiction of community resilience features. On the other hand, hierarchical data processing utilizing several methodologies can produce more accurate resilience evaluation findings; additionally, the individual benefits and disadvantages of each community in terms of resilience were clearly displayed through visualization results. Specific recommendations for strengthening community resilience were made by horizontally comparing resilience levels in different communities and vertically assessing changes in resilience in the same community over time.

It provides a full cycle resilience matrix to provide suggestions to construct a resilient community over the entire cycle. The resilience evaluation system presented in this article comprises four dimensions and three stages of public health incident development, resulting in a full cycle resilience evaluation that serves as a reference for the “evaluation–renovation” of resilient communities. In the renovation process of resilient communities, it is critical to specify the emphasis of different stages and types of public health catastrophes, stressing the complete dynamic governance of the whole process, including the pre-disaster, during-disaster, and post-disaster stages. This strategy seeks to improve the community’s readiness, response, recovery, and adaptive skills in the face of risk disruptions.

The combination weight index also highlights the significance of features at various times and stages, making it an effective tool for renovating resilient communities throughout time. In the pre-disaster phase, communities need to scientifically evaluate their capability to handle various public health events and establish dynamic monitoring and management mechanisms along with risk assessment systems. This allows for the timely identification of potential health risks within the community, such as public health security and natural disaster risks. During the disaster phase, communities should integrate all emergency resources and facilitate multi-party sharing among social organizations and individual residents. Based on the health risk warning system, communities should conduct a dynamic tracking of residents, disaster prediction, and systematic emergency rescue guidance. In the post-disaster phase, communities should encourage social participation and develop localized long-term mechanisms for comprehensive lifecycle governance of public health events, promoting refined management. Additionally, communities need to regularly update standards and implementation details for responding to public health events, considering their infrastructure capacity and residents’ actual needs. Furthermore, communities should conduct targeted technical and decision-making support based on different risk types.