Measuring the Evolution of Urban Resilience Based on the Exposure–Connectedness–Potential (ECP) Approach: A Case Study of Shenyang City, China

Abstract

Resilience is a new path to express and enhance urban sustainability. Cities suffer from natural shocks and human-made disturbances due to rapid urbanization and global climate change. The construction of an urban resilient developmental environment is restricted by these factors. Strengthening the comprehensive evaluation of resilience is conducive to identifying high-risk areas in cities, guiding regional risk prevention, and providing a scientific basis for differentiated strategies for urban resilience governance. For this study, taking Shenyang city as a case study, the resilience index system was constructed as an ECP (“exposure”, “connectedness”, and “potential”) framework, and the adaptive cycle model was introduced into the resilience assessment framework. This model not only comprehensively considers the relationship between exposure and potential but also helps to focus on the temporal and spatial dynamics of urban resilience. The results show that the exposed indicators have experienced three spatial evolution stages, including single-center circle expansion, multicenter clustering, and multicenter expansion. The potential index increased radially from the downtown area to the outer suburbs, and the low-value area presented a multicenter pattern. The spatial agglomeration of connectivity indicators gradually weakened. The results reflect the fact that the resilience level of the downtown area has been improved and the resilience of the outer expansion area has declined due to urban construction. The multicenter cluster pattern is conducive to the balance of resilience levels. In terms of the adaptive cycle phases of urban resilience, the first ring has gone through three phases: exploitation (r), conservation (K), and release (Ω). The second and third rings have gradually shifted from the exploitation (r) phase to the conservation (K) phase. The fourth ring has entered the exploitation (r) phase from the reorganization (ɑ) phase. The fifth ring and its surrounding areas are in the reorganization (ɑ) phase. The results provide specific spatial guidance for implementing resilient urban planning and realizing sustainable urban development.

1. Introduction

As a complex social ecosystem, the cumulative pressure induced by the smog, flooding, traffic jams, and heat waves, caused by the extensive expansion of urban space or the acute impact caused by natural and human-made disasters such as earthquakes, typhoons, and terrorist attacks, have caused various degrees of impact on and disturbance to the safe development of the urban environment [1,2]. Since the beginning of the 21st century, cities began to enter an era of uncertainty while accumulating social wealth. The uncertainty factor facing the development of urban security has increased significantly, and the potential threats it poses have also continued to increase [3,4]. The ever-increasing urban volume and potential risks not only reduce a city’s development potential and anti-interference ability but also seriously restrict a city’s sustainable and high-quality development processes [5,6]. For this reason, building a safe, healthy, and sustainable urban development environment has become a popular and rapidly emerging global issue [7,8]. As a new paradigm of safety science research, resilience is regarded as a key way to cope with the high degree of uncertainty in urban challenges [9,10,11,12], and the topic has aroused widespread concern and discussion in academia, social organizations, and governments. A resilient city plan provides a comprehensive analysis framework for the assessment of global urban resilience proposed by the Rockefeller Foundation of the United States [13]. Cities such as Rotterdam in The Netherlands and Chicago and New York in the United States have proposed urban resilience development plans in response to climate change [14]. China has also included resilience as an important aspect in the construction of urban physical examinations and urban safety systems. In October 2016, the “Habitat Three New Cities Agenda” clarified the challenges and goals of resilient cities and defined the vision of future cities as cities that are sustainable and resilient [15]. In these regions, international and organizational resilience development plans have noted the importance of resilient urban construction. Although the uncertainties faced by various regions seem to be incomparable, they all regard resilience as a long-term driving force to achieve the sustainable development of urban systems [16].

The concept of resilience is used in various disciplines and different fields, including engineering, psychology, anthropology, and ecology [17]. It is mainly used to describe the ability of materials, individuals, organizations, and the entire social ecosystem to withstand external pressures and absorb shocks [18]. In the field of ecology, the concept of resilience was first introduced into ecosystem research by Holling. He used resilience to describe the magnitude of the disturbance that an ecosystem could absorb before changing its own structure and to measure the ability of the ecosystem to spontaneously reorganize after the disturbance [19,20]. This concept emphasizes the stability and adaptability of the ecosystem [21], which is considered to be the origin of modern resilience theory [22]. In the 1990s, with the increasing uncertainties facing cities, the concept of resilience was gradually introduced into the field of urban planning [23,24,25] and was mainly used to explore the sustainable development of urban systems after disasters such as climate change, floods, and epidemics [26,27,28,29,30]. In this process, the research object of resilience has changed from a single ecosystem to a complex social-ecological system, emphasizing that the state of the system has changed from single-state equilibrium and polymorphic equilibrium to dynamic nonequilibrium [31]. The connotation of resilience has been further enriched. It is believed that resilience should be regarded not only as a restoration to the initial state of the system but also as the change, adaptation, and change ability stimulated by the complex social ecosystem in response to pressure and constraints [32]. Although the theory of resilience provides a new way of understanding the complex social ecosystem and promoting the sustainable development of cities [33], there is still no consensus on the definition of the concept of urban resilience [34]. From the perspective of disaster management, some scholars regard urban resilience as the potential of a system to absorb damage to reduce the changes, impacts, or uncertainties caused by disturbances, shocks, or natural disasters and to successfully adapt to these adverse events [35,36,37,38,39,40]. In recent years, academic circles have continuously explored the relationship between resilience and cities. Scholars have generally realized that urban resilience construction is a nonlinear process, and multiple forces on temporal and spatial scales will affect the dynamics and unbalanced changes of the urban system [41,42,43]. For this reason, Meerow made a relatively complete and rigorous definition of urban resilience after systematically combing through relevant literature, and it has been widely used in many resilience studies [44,45,46,47]. She believes that urban resilience is the social ecology and social technology network of the urban system and all its components that span temporal and spatial scales to maintain or quickly restore required functions, adapt to changes, and quickly transform systems that limit current or future adaptive capabilities in the face of disturbances [22]. This definition not only emphasizes the importance of temporal and spatial scales but also regards resilient cities as an ideal state. Although academic circles have different definitions of the concept of urban resilience, they are generally the same in terms of urban resilience construction strategies. They believe that versatility, redundancy and modularity, biological and social diversity, multiscale networks and connectivity, and adaptive planning and design should be the principles and strategies generally followed in the construction of urban resilience [48,49,50,51,52]. The diversification of the urban resilience concept and strategy literature has laid a foundation for quantifying urban resilience research. In the past ten years, to better establish the connection between theory and practice and to scientifically guide the construction of resilient cities, scholars have developed various tools and models to measure urban resilience from different perspectives [53,54]. Index sets or methods such as hybrid socio-physical networks [55,56], spatiotemporal dynamic resilience systems [57,58], input–output analysis [33], landscape ecological models [59,60], and system simulation models [61,62] have been used to measure resilience in urban society, economy, ecology, and disasters. Overall, the research literature on urban resilience has mostly focused on the concept of urban resilience and on specific cases of urban resilience construction. In terms of the quantitative measurement of urban resilience, urban disaster risk is the research object, and the ability of cities to respond to or adapt to natural disaster events is mostly discussed from the perspective of short-term engineering resilience. Additionally, cities are often used as the basic unit in the process of quantitative measurement, and less attention is given to the heterogeneity of the internal space of the city or the evolutionary characteristics on the temporal scale.

It is well known that urban transformation is one of the key processes of China’s current socioeconomic development [63]. Building a “resilient city” has become an important goal of urban development and transformation. Scientifically quantifying urban resilience can help effectively guide the construction of resilient cities [45]. As an individual city, the quantitative measurement of its internal resilience is the key to urban management and planning [45,54,64], and the grid-based method provides a new approach for revealing the heterogeneity of its internal space [65,66]. In view of this, this paper is based on the theory of resilience evolution, starting from the heterogeneity within the city, taking the 1 km sample plot as the basic unit, combining it with landscape ecology theory and spatial analysis methods to conduct a comprehensive measurement of urban resilience, and identifying the location of the area. In the adaptability stage, the corresponding differentiation strategy is proposed. The specific research objectives of this paper mainly include: (1) the comprehensive use of landscape ecology theory and spatial analysis methods to construct an ECP analysis method and incorporate it into the adaptive cycle system to form a temporal and spatial dynamic urban resilience framework; (2) the use of remote sensing image data, socioeconomic statistics, and planning atlases, to provide a base for the analysis framework to carry out the dynamic evolution characteristics of urban resilience; and (3) based on the stage characteristics of adaptability, differentiated strategies are proposed for urban resilience; at the same time, the information can be used by planners and cities. Managers provide knowledge about the development of urban resilience and enrich the research cases of quantitative measurement of urban resilience.

This article is organized as follows. This section outlines the relevant research background and related literature and is followed by the theoretical framework. Section 3 shows the data, sources, indicator system and methods required in this paper in detail. Section 4 shows the evolution of urban resilience through the “ECP” analysis framework and analyzes the adaptability stage of each region within the city. Section 5 analyzes and explains the combination of the “ECP” framework and adaptive cycle theory and proposes a resilience-oriented urban zoning development strategy. Finally, Section 6 summarizes our research results and their implications for urban resilience planning.

2. Theoretical Framework

As a complex adaptive system involving humans and nature, cities are considered to be typical natural environmental systems. Developing resilience is an important way to ensure the long-term healthy and sustainable operation of this complex adaptive system. This article believes that the interaction of the three attributes of exposure, connectivity, and potential jointly drives the dynamic process of urban resilience based on the concept and characteristics of resilience. “Exposure” mainly refers to chronic pressure from the city itself and acute external shocks. The exposed elements have a disruptive effect on the urban development process and are one of the basic elements used to promote or slow the resilient development of cities. The state of development after experiencing risks is important feedback from cities regarding exposure. For example, some cities will experience a rapid recovery after experiencing risks and are more resilient than they were in the previous state of urban development, while some cities will experience collapse and the urban development environment will become worse. The urban system needs to consume a great amount of resources and energy in the development process, and these “potentials” provide certain support for the city. “Potential” mainly refers to the attributes of the evaluation object in social-ecosystem resilience. In the construction of a resilient urban environment, “potential” is relatively limited and should be the surplus value after the city’s own development is consumed. A good urban development model should be that the city consumes “potential” while continuously increasing “potential”, thereby providing strong support for sustainable urban development. The response of connectivity to urban exposure reflects its “dual” attribute characteristics. Exposure can spread through connectivity and can also be effectively mitigated through it. In this article, blue and green landscape systems such as woodlands, grasslands, and waters are effective sources for risk mitigation and use their spatial interaction to explore the mitigation speed after experiencing risks based on the “source–sink” theory. In general, the three-dimensional analysis of urban resilience from the perspective of “ECP” not only pays attention to the spatial heterogeneity of resilience but also considers the interaction relationship in the construction of the resilient urban development environment.

Cities have the characteristics of dynamic evolution. Resilience development is an important goal of the urban system environment, and its dynamics are also one of its basic attributes. Gunderson and Holling explained in detail the relationship between the resilience of a system and the adaptive cycle phase [67]. They noted that the adaptive cycle model is the core content of evolutionary resilience theory, which provides theoretical support for understanding and portraying the complexity and resilience development phase of the social ecosystem. The adaptive cycle system regards urban development as having multiple life cycles, and each cycle can be divided into four phases: exploitation (r), conservation (K), release (Ω), and reorganization (α) [45]. The time dimension of urban resilience development can also be fully reflected in the four phases of the adaptive cycle system. In the exploitation (r) phase, the expansion of urban construction lands increases the exposure of the city, and the connectivity and potential are gradually weakened. After a long period of material accumulation, urban exposure continues to increase, and the connectivity and potential are at low levels. The system becomes less flexible and more susceptible to internal and external disturbances while the city enters the conservation (K) phase. The system may suddenly collapse, and the original accumulation will be released quickly when the city’s exposure reaches a certain threshold. At the same time, the city’s potential and connectivity will gradually increase, and the city will enter the release (Ω) phase. In the restructuring phase, the city’s exposure and connectivity may be low, while the potential is at a high level in the reorganization (α) phase. The greatest uncertainty and opportunity for innovation exists in this phase, and a new adaptive cycle may proceed from this phase.

This paper embeds the analysis dimension of “exposure connectivity potential” (ECP) into the adaptive cycle model to form a spatiotemporal dynamic assessment framework for urban resilience (Figure 1). Among them, the spatial scale highlights the spatial heterogeneity of the resilience situation from the three dimensions of urban resilience, which belongs to the static dimension of the space–time scale. In contrast, the dynamic dimension of the space–time scale extends the analysis of resilience on the time scale in an adaptive cycle. Phase identification and development characteristics are combined to provide differentiated strategies for achieving resilient development in different regions based on the adaptive cycle model.

3. Materials and Methods

3.1. Area of Study and Data

Shenyang, which is located south of the Northeast Plain in the central area of Liaoning Province in China, is the provincial capital and deputy provincial city of Liaoning Province. In 2015, the average population in the urban area was 5.921 million, the built-up area was 465 square kilometers, the total urban production was 58.91 million yuan, and the urbanization rate was 80.55%. The city is near several rivers and mountains, including the Liaohe, Hunhe, Puhe, and Beisha Rivers and the Qipan, Shiren, Meteorite, and Xiang Mountains. The urban area mainly includes new city areas, such as Sujiatun, Zhangshi, Daoyi, Hu Shitai, and Xinchengzi (Figure 2a).

As the only megacity in Northeast China, its highly concentrated population and economic conditions create many uncertain factors in the city. Simultaneously, it is an important base for heavy industry in China. The industrial system still significantly affects urban development and increases environmental pollution. During this period of rapid urbanization in Shenyang, urban construction land is rapidly expanding, and uncertain disturbance factors also increased significantly. These interfering factors include not only chronic pressures such as air pollution and heat island effects from the city itself but also acute impacts such as sudden fires and rainstorms from the external environment. For example, a huge fire occurred in the commercial city in 1996, a huge flooding event occurred in the urban area in 2010, and a severe haze pollution event occurred in 2015. Ecological security and other issues, such as the reduction of biodiversity and the loss of ecological service functions caused by grassland degradation and forest fragmentation, are also involved in the process of urban expansion. These threats may seem to be simple engineering accidents or ecological or natural disasters, but their origins lie in a typical human–land relationship during the urbanization process, which is closely related to the urban human settlement environment.

To present the spatial characteristics of urban resilience more accurately and identify the resilience of different spatial units, the study area was divided into 3279 complete-site evaluation units (1 km × 1 km) and 404 incomplete-site evaluation units (Figure 2b) based on urban land-use types, according to the system sampling method of equal intervals. In this study, the data required to calculate urban resilience indicators included remote sensing images, flooding location data, population density data, land-use data, and socioeconomic statistics. The data sources are shown in

Table 1. Sources of study area data.

Data	Source	Accuracy
Landsat Remote Sensing Image Data	Geospatial Data Cloud (China) (http://www.gscloud.cn/, at 21 July 2018)	30 m
Haze Data	Atmospheric Composition Analysis Group (http://fizz.phys.dal.ca/~atmos/martin/, at 2 August 2020)	0.01°
Flooding Location Data	Shenyang City Master Plan Atlas (1980, 1996, 2011), Shenyang Planning and Design Institute (http://www.syup1960.com/, at 10 September 2020)	Vector
Population Density Data	Chinese Academy of Sciences Resource and Environmental Data Center (http://www.resdc.cn/, at 7 December 2020)	1 km
Urban Resources and Energy Utilization	China Economic and Social Development Statistics Database (http://tongji.cnki.net/kns55/navi/navidefault.aspx, at 12 February 2021)	-
Distribution of Shenyang City Ring	Shenyang Natural Resources Bureau (http://zrzyj.shenyang.gov.cn/, at 5 May 2021)	Vector

.

3.2. Indicator System

This study selects four indicators closely related to landscape patterns, including heat islands, haze, flooding, and habitat degradation, to characterize urban exposure based on the urban development status of Shenyang city. In addition, the urban system needs to consume a large amount of material resources and energy in the development process. The “potential” of the city provides support for urban development and mitigates urban risks. The ecological carrying capacity of a city can fully reflect the potential status of a city from the perspective of ecological footprint theory. Generally, the city’s adaptability and resilience to exposure are severely damaged, and urban development lacks resilience due to the excessive exceedance of the ecological carrying capacity of the city in an ecologically deficit city. Connectedness often reflects the speed of mitigation of risk sources by cities. From the perspective of source–sink theory, the negative effects of the “source” landscape in urban development can be absorbed and offset by the “sink” landscape. The distance between the “source” and “sink” landscapes directly affects the speed of diversion of urban risks. Overall, the effective combination of exposure, connectedness, and potential helps to understand the impact of risk sources on risk receptors and provides a more effective way to monitor the urban resilient development environment (

Table 2. Indicator system for urban resilience assessment based on the “ECP” framework.

Dimension (Weights)	Dimension Meaning	Indicators (Weights)	Indicator Meaning	Normalization Method
Exposure (0.5)	Natural disturbance	Urban heat island effect (0.15)	Urban thermal environment	UH > 28 °C was assigned as 1 and probability decreases by 0.1 for every 0.5 °C decrease
		Urban flooding (0.175)	Urban risk status under rainstorm	Negatively normalized from 0 to 1
	Human interference	Urban smog (0.125)	An important sign of air pollution	Negatively normalized from 0 to 1
		Urban habitat degradation (0.075)	The extent of the impact of human activities on biodiversity	Negatively normalized from 0 to 1
Connectedness (0.2)	Resolution speed	Average distance index of the “source–sink” landscape (0.200)	Ecosystem’s sharing distance of urban risk	Positively normalized from 0 to 1
Potential (0.3)	Support level	Urban ecological footprint (0.300)	The potential status of urban development after internal consumption	Positively normalized from 0 to 1

).

3.2.1. Indicators for “Exposure” Criterion

In the process of urban development, urban exposure factors include not only chronic pressure from the city’s own development, such as haze, heat waves, and habitat degradation, but also acute impacts from the external natural environment, such as acute rainstorms and sudden fires. These exposure factors directly affect the process of urban resilience development. In this article, four indicators closely related to landscape patterns were selected, including heat islands, haze, flooding, and habitat degradation, to characterize urban exposure according to the current status of urban development in Shenyang (Table 2).

Urban heat waves will not only reduce air quality but also endanger the physical and mental health of residents. When the temperature is 28 °C or higher, people will feel discomfort, which can easily lead to heatstroke and mental disorders. When the temperature continues to be higher than 34 °C, it can lead to an increase in the incidence of heart, cerebrovascular, and respiratory diseases. In addition, rising temperatures will accelerate the photochemical reaction speed, increase the ozone concentration in the atmosphere near the ground, and affect human health [68]. Landsat data are used to retrieve the land surface temperature based on the atmospheric correction method. According to the specific temperature situation in Shenyang, the average temperature in each grid was counted and the grid average temperature above 28 °C was set as the temperature level that threatens the health of urban residents, and its risk probability was 1. For every 0.5 °C decrease in temperature, the risk probability decreased by 0.1.

The continuous expansion of impervious surfaces has encroached on lakes, woodlands, and grasslands within the city, resulting in insufficient water storage space in the city. The disastrous phenomenon of “watching the sea in the city” often occurs when the relatively slow drainage capacity of a city is hit by heavy or continuous rainstorms. The urban operating system is basically in a state of paralysis, seriously affecting the urban economic development and the normal life of residents and even threatening the safety of residents’ lives and property [69]. In this paper, the area with severe water accumulation (the depth of water accumulation exceeds 0.5 m) was set as the area where flooding occurred, and the kernel density was estimated with 1 km as the search radius. The average density values in the grid were counted and normalized. The greater the density value was, the greater the risk probability was.

Haze caused by air pollution has a significant impact on human health, the climate environment, and sustainable urban development. Industrialization has created more material accumulation for people, and society has developed rapidly, but it has also brought environmental problems such as air pollution. In large urban areas with high-density populations, buildings, and frequent traffic congestion, the exposure of smog pollution to the urban population has increased significantly. In 2013, the International Agency for Research on Cancer (IARC) listed PM 2.5 as a human carcinogen [56]. As an important pollutant of air pollution in urban areas, PM 2.5 is one of the most important “culprits” of urban smog. The average value of PM 2.5 in the grid was counted and the range standardization for normalization was used, with the value range from 0 to 1. The greater the density value was, the stronger the exposure was.

Habitat diversity provides ecological well-being for humans by ensuring the stability and resilience of ecosystem functions. It also reflects the interaction of human activities on regional ecosystems. Habitat quality provides reliable models and indicators for measuring the level of biodiversity [70]. Driven by the rapid urbanization strategy, the enhancement of human activities has changed the distribution pattern and functional status of regional habitats by influencing the circulation of material and energy flows between habitat patches and has caused a large amount of habitat loss, habitat fragmentation, and habitat degradation. The urban habitat quality was analyzed based on the habitat analysis module in the InVEST model, and its value ranged from 0 to 1. The average habitat quality index in the grid was calculated. The higher the index was, the better the habitat quality was.

3.2.2. Indicators for “Connectedness” Criterion

In the process of urban development, ecological space is an important guarantee affecting the restoration of a city after encountering risks, and it is also one of the basic elements of constructing a resilient urban environment. The connectivity between ecology and construction land has an auxiliary role in mitigating or accelerating risk exposure (Table 2). In this article, the urban landscape is divided into two types of “source” and “sink” landscapes based on the “source–sink” landscape theory. The “source” landscape refers to the type of landscape that can promote the development of ecological processes. “Sink” landscapes are a type of landscape that prevents or delays the development of ecological processes. Urban landscapes can be divided into grey landscapes (such as buildings and roads), blue landscapes (water bodies), and green landscapes (green vegetation). In the process of urban exposure or stress generation, the “source” is mainly from the grey landscape, while the blue and green landscapes act as “sinks”. For a city, the larger the area of blue and green landscapes is, the better. When the areas of blue and green landscapes are constant, their spatial configuration is particularly important. For example, a balanced green landscape has a better effect on reducing urban heat islands and urban flooding. This paper starts from the “source–sink” landscape theory by calculating the average distance between the source and sink landscapes to characterize the connection between the source–sink landscape and describe the rate of mitigation of exposed elements by the ecological space. The shorter the average distance between the source and sink is, the better the connectivity, and the better the mitigation speed.

3.2.3. Indicators for “Potential” Criterion

The potential of urban development can provide support for alleviating exposure and is one of the important indicators that reflects the sustainable development of cities. In this article, urban potential should be the remaining development potential of the city and the ability or intensity to resolve urban risks after the city has experienced residents’ material energy consumption (Table 2). It can be used to characterize the development potential state of the region by comparing the ecological footprint demand of human activities in the region and the ecological carrying capacity provided by the natural ecosystem from the perspective of ecological footprint theory. The ecological footprint is also an important concept of sustainable development. It can be used to convert the human consumption of nature into the area of cultivated land, forests, water bodies, grasslands, and other land uses to support this consumption, and these areas can be used for comparison with the real land cover area. Conclusions such as ecological surplus and ecological deficit information can be obtained. Generally, due to the excessive exceedance of the ecological carrying capacity of the city, the urban adaptation and resilience capacity of the ecologically deficit city is severely impaired, and urban development lacks resilience, while the ecological surplus of the city has a higher potential for resilience development. In this paper, the ecological footprint index was used to characterize the urban development potential, and the urban development potential index was obtained by normalizing the ecological footprint index. The higher the index was, the better the urban development potential was.

Generally, the relevant indicators of urban resilience based on the grid scale were handled positively and normalized to values from 0 to 1, as detailed in Table 2. The weights of the indicators were scored using the AHP method [71]. Among them, exposure is considered to be the most important factor affecting the development of urban resilience, and related indicators of exposure have received greater weights. The “potential” of a city is the most important supporting element for alleviating exposure; it represents the strength of the city to resolve risks through its own capabilities, and the weight is relatively moderate. The connectivity between the “source” and “sink” landscapes represents the speed of mitigation of the external environment or elements to the city’s exposure, and the weight is relatively small.

3.3. Methods

3.3.1. Habitat Quality Analysis

The InVEST model developed by Stanford University is an evaluation model of ecosystem service functions and decision-making functions that has been used widely internationally, and the model provides a relatively accurate analysis method for habitat analysis [70]. Using this model can better measure the degradation of the regional ecosystem to assess the ecosystem service capabilities. The specific calculation method is as follows:

$$Q_{xj}=H_j\left[1-\left(\frac{{{\displaystyle D}}_{xj}^z}{{{\displaystyle D}}_{xj}^z+{{\displaystyle K}}^z}\right)\right]$$

(1)

where Q_xj is the HQ of grid x in habitat type j; H_j is the habitat suitability of grid x for habitat type j and its range is [0,1]; D_xj is the habitat degradation degree in grid x in the habitat type j; k is a half-saturation constant; and z is a normalized constant, generally 2.5. Among them, the calculation method of habitat degradation is as follows:

$$D_{xj}={\displaystyle \sum _{r=1}^R{\displaystyle \sum _{y=1}^{Y_r}{r}_y}}\left(\frac{{\omega }_r}{{\displaystyle \sum _{r=1}^R{\omega }_r}}\right)i_{rxy}{\beta }_x{S}_{jr}$$

(2)

$$\begin{array}{cc}{i}_{rxy}=1-\left(\frac{d_{xy}}{d_{r\max }}\right)(\mathrm{if} \mathrm{linear})& i_{rxy}=\exp \left(\frac{-2.99d_{xy}}{d_{r\max }}\right)(\mathrm{if} \mathrm{exponential})\end{array}$$

(3)

In Formulas (2) and (3), D_xj is the habitat degradation degree of grid x in habitat type j; R indicates the number of threat sources; ω_r is the weight of threat source r; Y_r is the number of threat sources; r_y represents the stress value of grid y; i_rxy is the stress value of grid y versus grid x. β_x is the accessibility of the threat source to grid x; S_jr is the sensitivity of the habitat type j to threat source r. d_xy is the Euclidean distance between the location x of the habitat and threat source y; d_rmax is the maximum interference radius of threat source r.

This paper defined woodland, grassland, arable land, and water as the habitat landscape. Urban construction land, rural residential areas, industrial and mining land, railways, expressways, national roads, provincial roads, and county roads were taken as threat sources. The threat radius of the threat source and the distance attenuation method of its impact on the habitat, the suitability of the habitat, and its relative sensitivity to various threat sources are all determined by the recommended values of the model and related literature [70,72].

3.3.2. Ecological Footprint Model

Based on the ecological footprint model, population and energy consumption statistics, a grid-scale urban ecological footprint index was constructed from the perspective of ecological supply and demand. In this article, this index is used to characterize the city’s development potential after its own consumption [59]. The formula is presented as follows:

$${UP}_i=\frac{{EC}_i\ast \left(1-12\%\right)}{{EF}_i}=\frac{\left(1-12\%\right)*{\displaystyle \sum r_k\times y_k\times a_{ik}}}{P_i\ast {\displaystyle \sum r_k\times c_j/w_j}}$$

(4)

where UP_i is the ecological footprint index of grid i, EC_i is the ecological carrying capacity of grid i, EF_i is the ecological footprint of grid i, r_k and y_k represent, respectively, the equilibrium factor and yield factor of productive land of type k, a_ik is the area of the k-type productive land in grid i, P_i is the population of grid i, C_j is the per capita consumption of j commodity in the city, and w_j represents the global average production of the consumer commodity j. According to the World Commission on Environment and Development (WCED), 12% of the ecological carrying capacity is used to protect areas of biodiversity.

3.3.3. “Source–Sink” Landscape Average Distance Index Model

The balance of the “source–sink” landscape spatial configuration is one of the important indicators of urban connectivity. The average distance index of the “source–sink” landscape can be used for measurement; this paper constructed a grid-scale-based “source–sink” landscape average distance index measurement model, which is presented as follows [59]:

$${UC}_i={\displaystyle \sum \frac{L_j\ast A_{ij}}{L_{ijk}}}={\displaystyle \sum L_j\ast A_{ij}/\left({\displaystyle \sum _{j=1}^n\min \left(d_{ijk}\right)}/m\right)}$$

(5)

where UC_i is the morphology resilience index of grid i, A_ij is the proportion of the category “source” landscape j in grid i, L_ijk is the average distance from the “source” patch j in grid i to the “sink” patch in the study area, d_ijk represents the distance from each cell of the “source” patch j in grid i to cell k in “sink” patch m, and n is the number of cells in the “source” patch in grid i and the number of cells in the “sink” patch in the study area. L_j is a constant that represents the average distance of the “source” landscape and the “sink” landscape in the study area. In this paper, the L_j values were set as the average distance index between construction land, unused land, and cultivated land and the “sink” landscape (woodland, grassland, and water), with values of 730.67, 731.52, and 340.99, respectively, in Shenyang city in 1985.

4. Results

4.1. Spatial Evolution of Urban Resilience Indicators

Figure 3 not only shows the evolution of the spatial pattern of the six indicators of urban resilience but also reflects the relationships among exposure, connectivity, potential, and urban resilience. Urban exposure factors such as the thermal environment, flooding, smog, habitat degradation, and potential factors all showed a continuous weakening from the city center to the suburbs with the expansion of the city, and their spatial influence showed a continuous expansion trend. The connected elements showed the opposite trend with the exposed elements and potential elements.

The high-risk grids of the urban thermal environment were mainly distributed in the urban built-up areas where the population is highly concentrated, the roads are relatively dense, and the building density is highly concentrated in each period. Large park areas, such as Beiling and Dongling and the Hunhe River, have become “cold islands” of the central city. The high-risk grid distribution of the urban thermal environment has a generalized trend driven by the expansion of urban construction land. The high-risk grid gradually expanded from the main urban area to the south of the Hunhe River, the sub-city area of Xinchengzi, and the ecological corridor area of Qipan Mountain. The distribution of high-risk grids shifted from agglomeration in the central urban area to expansion in the periphery.

Affected by the lagging drainage network construction, the continuous increase in the proportion of hard surface coverage, and the lack of internal water storage and stagnation space, the risk of urban flooding in Shenyang showed a significant expansion trend. Most of the flooding points in Shenyang were concentrated in the main urban area, and the high-risk areas for flooding have roughly gone through the evolutionary stages of “three clusters–two areas–peripheral expansion”. This evolutionary pattern was significantly correlated with the “internal filling–peripheral expansion” trend of Shenyang’s expansion of urban construction land.

The spatial distribution of the high-value grid of urban smog risk in Shenyang has gradually changed from a “central cluster” to a “south–north strip”. From 1985 to 2005, Shenyang formed a high-risk pattern of “one master and two associates”. The main urban area was a concentrated area with a high risk of smog, and two groups of regions, Sujiatun and Xinchengzi, successively became secondary peak areas with a high-risk distribution of smog. In 2015, the high-value grid of haze risk formed a “north–south axis” pattern. This pattern was not only closely related to the dominant direction of urban expansion but was also affected by the synergistic influences of terrain, wind direction, and ecological corridor layout.

The risk of habitat degradation in Shenyang was significantly correlated with the spatial pattern of land use. The overall level of habitat degradation risk in woodland, grassland, rivers, and other areas was relatively low, and the habitat degradation risk in construction areas was relatively high. Affected by the expansion of urban construction, the number of habitat degradation risk grids in Hunnan, Xinchengzi, Daoyi, Hushitai, and other cluster areas showed an increasing trend. The rapid expansion of the central urban area has gradually compressed ecological spaces such as Qipan Mountain and Shiren Mountain. The low-risk grid of ecological degradation in the eastern ecological corridor area gradually became fragmented. The ecological degradation low-risk grid in the eastern ecological corridor area was fragmented.

The average distance index of the “source–sink” landscape in Shenyang is increasing. This phenomenon indicates that the speed of the urban ecosystem’s response to risks has been significantly improved. The high-value grid of the average distance index of the “source–sink” landscape expanded significantly with urban expansion, and the low-value grid showed a continuous shrinking trend. From 1985 to 1995, the expansion of construction land in Shenyang was dominated by internal filling. The expansion of the “source” landscape has increased the segmentation of cultivated land, and the expansion of rural settlements has enhanced the spatial coupling of the “source” and “sink” landscapes. From 1995 to 2015, the average distance of the “source–sink” landscape steadily increased. Low-value grids continued to shrink and form agglomeration patterns in contiguous areas of cultivated land. It was mainly affected by the diversification of the expansion model of construction land (expansion of outer circles and enclave clusters) and ecological infrastructure construction.

Similar to the situation in many large cities in China, the urban ecological footprint of Shenyang continues to be higher than the ecological carrying capacity, and the ecological deficit is becoming severe. The high-value grid of the ecological footprint index has gradually changed from circular to semi-circular due to the expansion of urban construction and rapid population agglomeration. The urban capacity to cope with risk interference continues to weaken. During the study period, due to the high concentration of the population and the depletion of resources in the central urban area, its ecological ability to resolve urban risks was generally insufficient. The ecological footprint index of the ecological barrier area in the eastern part of Shenyang has continued to decline, such as Qipan Mountain and Meteor Mountain. The low-value grids in the eastern region have become gradually distributed from points to slabs and have formed a continuous development pattern with the central urban area.

4.2. Spatial Evolution and Transect Characteristics of Urban Resilience

4.2.1. Spatial Evolution of Urban Resilience

The low-level resilience grid of Shenyang was mainly located in the central urban area of various periods and showed significant expansion with the expansion of construction land. The low-resilience area and the expansion of construction land were closely related (Figure 4). In 1985, grids with a low level of resilience were mainly distributed in the main urban area, while the Sujiatun group area formed the second lowest value of resilience. Overall, the low-resilience grid of this period formed a spatial pattern of “one master and one pair” in Shenyang. In 1995, the Sujiatun group gradually matured and became another resilient low-value hot spot after the main urban area. During this period, the main urban area expanded to both the southwest and the south, and the highly resilient grid between the Sujiatun group and the main urban area gradually disappeared. In 2005, the low-resilience grid crossed the Hunhe River and expanded rapidly to the south of the Hunhe River. With the development and construction of Hunnan and the main urban area, the low-resilience grid gradually formed a cluster of development, and the resilience level of the Sujiatun group was further reduced. During this period, the resilience level of the Xinchengzi and Hushitai groups decreased, and new low-value centers of resilience were formed. The level of resilience in Qipan Mountain declined, which was affected by the construction of the Shenyang-Fushun New District. In 2015, the urban low-resilience grid of Shenyang formed a sprawling expansion. Hushitai and Xinchengzi have become new areas with low resilience due to factors such as rapid population, expansion of construction land, and industrial agglomeration. Under the influence of the spread and expansion of urban construction land and the construction of new districts, the eastern ecological barrier area’s ability to mitigate urban risks has continued to weaken, and the level of regional resilience has declined.

4.2.2. Transect Characteristics of Urban Resilience

In this study, the center point of the resilience level of Shenyang city in 1985 was used as the central grid, and four transects passing through the central grid were selected from the south–north direction, east–west direction, northeast–southwest direction, and southeast–northwest direction (Figure 2b). The grid resilience index value was calculated, and the resilience evolution curve diagram of the four transects was generated (Figure 5).

In the north–south direction, low-value areas of the resilience index appeared 16 and 27 km north of the central grid, which were the Hushitai and Xinchengzi groups, respectively. Driven by the expansion of the central urban area, the low-value areas of the resilience index have significantly expanded to the north and south. The resilience level of the downtown area has been slightly improved, and the resilience level of the Hunhe River is generally higher. In the east–west direction, the urban resilience index showed a downward trend year by year in the 8–25 km interval to the east and west of the central grid. The resilience index of the Zhangshi group and the Wangjia group both declined significantly under the influence of urban sprawl. In the northeast–southwest direction, at 5 km in the northeast of the central grid, the resilience level showed a downward trend and migrated to the northeast. Affected by the construction of Shenfu New District and its own development in Qipan Mountain, the ecosystem has been destroyed, and the level of resilience has weakened. In the direction of the southeast–northwest transect, the urban resilience level showed a fluctuating decline 5–18 km southeast of the Hunhe River in 2015, and there were three low values. The Daoyi group in the northwest showed a low resilience index, and the value of area had an expanding trend. The Daoyi group in the northwestern region has become a low-value area of the resilience index and has a trend of expansion. On the basis of the transect analysis, the following three conclusions were also found.

First, the urban resilience level of Shenyang has shown a downward trend overall, and the spatial distribution of resilience levels has significant regional characteristics. Blue and green landscape areas have a low intensity of human activities, high potential, and strong connectivity. They are the bases for urban risk mitigation, and the overall resilience of the region is at a relatively high level. The main landscape of the urban groups and downtown areas was construction land, which has a relatively low level of resilience; the transformation of agricultural land to non-agricultural land and the gradual increase in the intensity of social and economic activities have become important factors affecting the decline of the resilience level.

Second, the resilience level of the downtown area has been slightly improved, while the resilience level of the peripheral expansion area declined in various time periods. The downtown area is a key area for urban renewal and transformation. The internal population, industrial areas, and residential areas of relocation or decentralized development have resulted in a certain increase in urban potential and a downward trend in exposure. The emigration or decentralized development of the urban population, industries, and residences has resulted in an increase in potential and a downward exposure of the downtown area. Additionally, the spatial distance between the central city and the “source” landscape has been shortened due to urban expansion. The hysteresis of infrastructure construction, the concentration of the population, the relocation of industrial areas, and the transformation of land use contributed to the decline in the resilience level of the newly added expansion areas in each period.

Third, urban resilience was related to the scale of cities and towns, and the multicenter cluster pattern was conducive to achieving a balanced development of resilience levels. The resilience level of the urban groups and the downtown area showed a continuous decline and a gradual recovery trend in Shenyang city, respectively. Although the overall urban resilience level declined, its spatial equilibrium level gradually increased. The downtown area was developed earlier and is in a fully developed zone, and its population and industries will inevitably undergo relocation or transformation after reaching a certain scale. Moreover, the comprehensive management of the downtown area has further promoted the gradual recovery of the region’s resilience level. The cluster area plays the role of “population interception” and “risk sharing” in the downtown region. Although the rapid agglomeration of population and industries in the group area has promoted the economic development of the expanded area, it has also brought many potential interference factors to the region’s resilient development environment, such as habitat degradation and increased smog.

4.3. Adaptive Cycle Phase of Urban Resilience Evolution

4.3.1. Evaluation of the Three-Dimensional Elements of Urban Resilience

The urban resilience of Shenyang has the characteristics of ring expansion. To show the adaptive cycle stage of urban resilience in each ring, the research used the Shenyang city ring line to divide the area (Figure 2b) and calculate the exposure, connectedness, and potential of each ring area (Figure 6).

There are differences in the dynamic changes of the risk exposure of each ring in Shenyang. Among them, the risk exposure has increased most significantly in the third ring. The risk exposure within the first ring gradually decreased with time, while the second and third rings showed a significant growth trend, and the exposure outside the fifth ring was at a relatively low level during the study period. The risk in the fourth ring increased rapidly in 2015 and was affected by urban expansion. Urban connectivity showed an improvement during the study period, especially in the third ring. The potential of urban areas in the third ring was at a low level, but the potential in the first ring increased slightly. The dynamic evolution of the potential in the second ring was relatively stable, and the third ring had a rapid downward trend. The development potential of cities outside the fifth ring was relatively abundant. The potentials of the fourth ring and the fifth ring rapidly declined from 1995 to 2015. In general, the low-value grid of urban resilience was basically located within the third ring road and spread to the fourth ring road. The environment for resilient development within the third ring deteriorated, and disturbance factors in the fourth ring in Shenyang increased. The resilience of the fifth ring road and its surrounding areas was mostly due to cultivated land or forests. The level of urban resilience was higher in the fifth ring and its surrounding areas.

4.3.2. Adaptive Phase Characteristics of Urban Resilience

Figure 7 shows the evolution of the adaptive phase in each ring in Shenyang. The urban resilience in each year has the characteristics of coexisting multiple phases at the same time as being affected by geographic heterogeneity and development timing. With the continuous expansion of the central city, the complexity of its stages gradually deepens.

In detail, the urban development intensity showed a gradual weakening development trend from the first to the third rings in 1985. The intensity of human activities was the strongest in the first ring, and the third ring was at a relatively weak level. The exposure within the first ring increased, and the connectivity and potential were at low levels. The development impacts of the fourth and fifth ring areas were small, the urbanization development process was slow, and the urban ecological background conservation was generally better. Outside the fifth ring, the security development potential was high, and the exposure and connectivity were low. During this period, the first to third ring areas were in the exploitation (r) phase of the adaptive cycle. The fourth and fifth rings and their periphery were in the reorganization (α) phase. In 1995, the internal filling and external expansion of construction land reduced urban resilience. The exposure within the first ring reached a high point. The first ring zone will experience severe chaos, and eventually, the system will collapse. The increase in the intensity of urban development has also led to an increase in exposure within the second and third rings. The potentials of the fourth and fifth rings declined, but connectivity increased. During this period, the level of resilience in the periphery of the fifth ring was generally stable. The first ring area of the city entered the conservation (K) phase. The second and third rings were in the exploitation (r) phase. The fourth and fifth rings and their periphery were still in the reorganization (α) phase. In 2005, affected by factors such as industrial relocation, urban renewal, and functional adjustment, the first ring entered the release (Ω) phase. In the release phase, the exposure within the first ring began to decrease, the potential within the region improved, and the connectivity was stable. When encountering a strong disturbance that exceeds the recovery threshold, the accumulated social stability will be severely damaged in a short time. The second ring entered the conservation (K) phase. The exposure and connectivity in the region increased, but the potential continued to decline. The third ring is the key area of urban expansion, which was in the exploitation (r) phase. Regional exposure accumulated rapidly, potential decreased, and connectivity improved. The fourth ring, fifth ring, and their surrounding areas were still in the reorganization (α) phase. Although the exposure of these areas has increased, the potential and connectivity provided a good mitigation effect for urban resilience development. In 2015, the first ring was still in the release (Ω) phase, in which the exposure further decreased, and the potential and connectivity gradually increased. The second and third rings were in the conservation (K) phase. Regional exposure continued to grow, development potential declined, and connectivity improved. The potential exposure factors in the fourth ring increased due to the expansion of urban construction land. Additionally, population agglomeration and industrial development have continuously consumed the regional potential. The region has gradually entered the exploitation (r) phase. The fifth ring and its surrounding areas were still in the reorganization (α) phase.

In general, the urban resilience evolution of Shenyang has the characteristics of coexisting in multiple phases, which are affected by the factors of geographical heterogeneity and development timing. The complexity of its stages gradually deepens with the continuous expansion of construction land. The first ring has gone through three phases: exploitation (r), conservation (K), and release (Ω). The second and third rings have gradually shifted from the exploitation (r) phase to the conservation (K) phase due to the expansion of urban construction land. The fourth ring has entered the exploitation (r) phase from the reorganization (α) phase. The fifth ring road and its periphery have low exposure, sufficient potential, and improved connectivity. The areas were in the reorganization (α) phase during the study period.

5. Discussion

5.1. Advantages of “ECP” and Adaptive Cycle Theory in the Framework of Urban Resilience Analysis

The interaction of the three attributes of exposure, connectedness, and potential together drives the dynamic development of urban resilient environments. A three-dimensional evaluation index system of urban resilience of “exposure–connectedness–potential” is constructed from the perspectives of element movement and landscape pattern processes based on clarifying the spatial heterogeneity and interaction of resilience. This indicator system not only takes into account resilience thinking and connotation, but also enables resilience to be well quantified and breaks through the traditional post-analysis model of resilience [59]. In addition, resilience is an important part of the urban system environment, and dynamics are one of its basic attributes [22,33]. To this end, this paper further nested the resilience comprehensive evaluation system into the adaptive circle model based on adaptive circle theory, forming a spatiotemporal dynamic analysis framework and model for resilience evaluation. It provides a quantitative analysis method for the analysis of urban resilience conditions and stage characteristics in different regions and provides a basis for the implementation of differentiated optimization strategies for regional resilience environments.

The urban development process does not always have a complete adaptive cycle. To this end, the paradigm of “trading space for time” was used in this paper [73]. This paper focused on the spatial heterogeneity of urban resilience and discussed the adaptation phase of urban resilience over time and its characteristics based on this paradigm. The speed and intensity of the city’s response to exposure were different in the same period due to the effects of the natural and human environments. Therefore, the three-dimensional attributes of urban resilience, which contain exposure, connectedness, and potential, can be regarded as the basic characteristics of each adaptation phase on the time scale. Compared with previous resilience studies that use cities as the basic unit, this research pays more attention to revealing the temporal and spatial evolution of resilience in cities, which can enrich the spatial scale cases of urban resilience research [54,64]. Through the temporal and spatial differentiation characteristics of urban resilience at the grid scale, the transition speed and frequency of the urban development stages can be more accurately and comprehensively understood and provide a basis for the timely formulation of urban development strategies.

5.2. Urban Zoning Development Countermeasures Based on Resilience Orientation

Exposure factors are always changing with the dynamic development of cities. Urban managers should adhere to the principles of systems, adaptation, dynamics, and redundancy, and the urban management method should be changed from “command and control” to “learning and adaptation” [74]. In the process of urban resilient environment construction, it is suggested that while considering the urban resilient development phase on the time scale, the city should also take precautions against risks on the spatial scale [45]. In addition, exploring the interrelationships among exposure, potential, connectivity, and urban resilience can help distinguish the focus of risk control to adapt to the development of resilience by adjusting the development model or changing the regional biophysical conditions. This paper proposed spatial planning strategies for the different adaptation stages of each ring in Shenyang based on the spatial heterogeneity of urban resilience and urban functional zoning.

Both the third and the fourth ring roads are in the exploitation (r) phase of resilient development. In the process of urban expansion, managers should pay attention to strengthening the construction of ecological infrastructure, optimizing the spatial layout of ecological landscapes (blue, green) and gray landscapes, and reducing industrial energy consumption to promote the transformation and upgrading of traditional industries. The second ring is in the conservation (K) phase. Managers should actively adjust the functional pattern within the region, adding small-scale ecological patches and ecological corridors to promote the recovery of the function and connectivity of the internal ecosystem of the city. In addition, it should promote the gradual migration of the population to the third and fourth rings to reduce the ecological footprint in the area and increase the intensity of risk reduction. The first ring is in the release (Ω) phase of resilient development. Urban planning should focus on adapting to risk exposure, looking for opportunities for urban innovation and transformation and upgrades, such as actively advocating for low-carbon consumption and green travel activities for residents, appropriately allocating small green patches that adapt to resilient development, and improving the connectivity and integrity of the ecosystem. The fifth ring and its surrounding areas were in the reorganization (α) phase of resilient development during the study period. This area is an ecological barrier or an area with weak development intensity. The intensity of social and economic activities is lower, the ecological conservation status is better, and the level of resilience is relatively high. Urban planners should focus on controlling and reducing intensified disturbance risks from human activities to maintain the resilience and supporting ability of ecosystems.

5.3. Limitations and Prospects for Further Study

The main difficulty in the specific quantification and phase characteristics analysis of urban resilience is the determination of the specific indicators of urban resilience [60]. This paper selected representative exposure elements of Shenyang based on clarifying natural disturbances and considering disturbances such as heat islands, haze, flooding, and habitat degradation. The methodology of landscape ecology has been used to construct the potential and connectedness indicators of urban resilience. However, the exposure is different due to urban natural and cultural environments. It is necessary to clarify its representative element indicators when analyzing specific urban resilience conditions [25], such as water pollution and contamination risks in water management in urban areas. The parametric substitution of landscape ecology models and methods based on the persistence of resilience thinking will further improve the perspective of resilience city research and help enrich the urban resilience measurement method system. In the process of resilience city research, the threshold problem is an important cognitive basis for identifying the transition of urban adaptability. Using the system dynamics model to adjust the parameters and comprehensively simulate the urban resilience process is of great significance for guiding the health and sustainable development of the city. These issues will also be the focus of future research.

6. Conclusions

The city is a complex social-ecological system. Whether it is the cumulative pressure brought about by its own rapid development or the acute impact caused by natural and human-made disasters, it has caused varying degrees of disturbance to the urban development environment. Although there are various case studies on urban resilience, few are focused on static spatial patterns with dynamic temporal processes. On the basis of clarifying the representative elements of exposure in Shenyang, this paper used the theory and methods of landscape ecology to establish a three-dimensional analysis framework comprised of three key criteria, i.e., “exposure”, “connectedness”, and “potential”, and introduced an adaptive cycle model to divide the characteristics of its dynamic temporal processes into phases. The results showed that the exposure indicators roughly experienced three spatial evolution stages: single-center circle expansion, multicenter cluster development, and multicenter expansion. The high-value grid was spatially transformed from the dual-core development of the downtown Sujiatun area to the multicenter model of the downtown Sujiatun–Xinchengzi area. The spatial agglomeration trend of connectivity weakened, the urban potential declined, and a relatively stable low-value area formed in the downtown area. The resilience level of the downtown area improved slightly, and the resilience of the urban peripheral expansion areas declined at various stages. Urban resilience is related to the scale of towns, and the multicenter cluster pattern is conducive to promoting the balanced development of the resilience environment. These spatial patterns show that artificial adjustments based on different zoning conditions can reduce urban exposure and improve urban resilience to a certain extent, have a positive effect on urban risk prevention and management, and promote urban resilience development. In addition, the dynamic development of the city was divided into four stages based on adaptive cycle theory, and the spatial planning strategy for each stage was proposed from the perspective of resilient city planning.