Research on the Infrastructure Resilience System and Sustainable Development of Coastal Cities in the Bohai Sea, China: A Multi-Model and Spatiotemporal Heterogeneity Analysis Based on CAS

Abstract

In recent years, urban risk incidents have become more common. Enhancing infrastructure resilience is not only crucial for adapting to climate change and addressing natural disasters but also serves as a key cornerstone for sustaining urban sustainable development. The research objects for this study are 17 coastal cities in the Bohai Rim region of China. Based on the Complex Adaptive System (CAS) theory, from the multi-dimensional perspective of urban sustainable development, a resilience evaluation index system covering five subsystems, namely transportation, water supply and drainage, energy, environment, and communication, is constructed. Employing panel data from 2013 to 2022, this study develops the entropy weight–TOPSIS model to quantify resilience levels, and applies the obstacle degree model, geographical detector, and Geographically and Temporally Weighted Regression (GTWR) model to analyze influencing factors. The main research results are as follows: (1) The regional infrastructure resilience shows a slow upward trend, but the insufficient synergy among subsystems restricts urban sustainable development; (2) The primary barrier is the drainage and water supply system, and the environmental and communication systems’ notable spatial heterogeneity will result in uneven regional sustainable development; (3) The influence of driving factors such as economic level gradually weakens over time. Based on the above research results, the following paths for resilience improvement and urban sustainable development are proposed: Improve the regional coordination and long-term governance mechanism; Focus on key shortcomings and implement a resilience enhancement plan for water supply and drainage systems; Implement dynamic and precise policy adjustments to stimulate multiple drivers; Enhance smart empowerment and build a digital twin-based collaborative management platform.

1. Introduction

As a fundamental safeguard for residents’ livelihoods and the core support system for urbanization, urban infrastructure leverages its resilience as a critical defense against multifaceted risks [1]. Moreover, it serves as a foundational guarantee for advancing high-quality development in urban complex adaptive systems amid uncertain environments. In recent years, urban risk events have exhibited increasing frequency and complexity, exerting significant negative impacts on residents’ well-being. For instance, Hurricane Sandy, which affected 24 U.S. states [2], and the 2023 torrential rain-induced floods in Zhuozhou, Hebei, both inflicted severe damage on urban infrastructure [3]. Risk factors are elements that exert adverse impacts on urban infrastructure systems, including economic crises, natural disasters, and emergent incidents. Given the potential compound consequences of diverse risk factors, societal expectations and systemic demands for urban infrastructure resilience have been increasing. Urban infrastructure resilience refers to a system’s capacity to absorb, recover from, and adapt to risks or disasters, thus maintaining operational continuity [4,5,6].

Urban infrastructure systems must have strong resilience mechanisms to increase their capacity for resistance in order to successfully minimize and routinely defend against such threats. China’s Long-Term Goals and the 14th Five-Year Plan propose to align with current urban development philosophies and trends through 2035, start pilot projects for contemporary urban construction, and create livable, smart, green, culturally lively, resilient, and innovative cities.

2. Literature Review

The term “resilience” originates from the Latin word “resilio”, which originally meant “to return to a previous state”. It was first coined by ecologist Holling, who applied it to the field of ecology. The concept was first introduced by ecologist Holling and applied to the field of ecology [7]. Not until the 1990s was the term “resilience” gradually extended to other disciplines [8]. Resilience is widely applied in engineering, economic, and social fields [9,10,11], and has been institutionalized in national infrastructure protection strategies [12]. During this period, scholars also began to focus on urban resilience [13], and some explored it from the infrastructure perspective [14,15,16]. As China’s urbanization progresses, enhancing the resilience of urban infrastructure has become an inevitable trend in urban development, and scholars have systematically explored infrastructure resilience using various theoretical frameworks and methods.

In terms of theoretical frameworks, research has evolved from foundational concepts to more integrated systems. Early work by Rehak et al. established a tripartite framework for critical infrastructure resilience, centering on robustness, recovery, and adaptability [17]. Building on such concepts, Chen et al. adopted the Pressure-State-Response (PSR) model to dynamically measure urban infrastructure resilience [18]. Further refining these approaches, Chinese scholars have proposed a comprehensive evaluation system structured around five key subsystems—transportation, water supply and drainage, energy, environment, and communications—offering a multidimensional perspective on urban resilience [19,20]. In recent years, the role of ESG (Environmental, Social, and Governance) frameworks and institutional factors in sustainable urban governance has gradually gained attention [21,22], which provides a new perspective for understanding the multi-dimensional driving mechanism of urban infrastructure resilience. Methodologically, diverse techniques reflect the complexity of resilience assessment. Studies range from the application of Principal Component Analysis (PCA) and fuzzy comprehensive evaluation in Wuhan [23], to the use of Copulas analysis for watershed infrastructure in Florida [24], and the fuzzy VIKOR method for flood resilience evaluation [25]. Particularly prevalent is the integration of weighting methods—either subjective, objective, or combined—with the TOPSIS comprehensive evaluation, providing a robust toolkit for comparative urban resilience measurement [26,27,28]. Geographically, while significant attention has been devoted to major Chinese urban agglomerations such as Beijing-Tianjin-Hebei [25], Chengdu-Chongqing [29], the Yellow River Basin [30], the Yangtze River Delta [31], the Central Plains [32], and the Pearl River Delta [33], a conspicuous gap remains. The coastal cities of the Bohai Rim region, despite their economic significance and vulnerability to complex disasters, have not received commensurate scholarly attention, underscoring a critical area for future research.

Although existing literature offers multi-dimensional research perspectives and methods, a noticeable gap remains in empirical studies specifically focused on evaluating the resilience of urban infrastructure in coastal cities along the Bohai Rim. This paper aims to address the issue of urban infrastructure resilience by conducting a comprehensive evaluation of infrastructure system resilience and analyzing its influencing factors. Based on the results derived from the evaluation and factor analysis, targeted recommendations for improving resilience are proposed. In order to analyze the current state and shortcomings of urban infrastructure resilience in the region, this study uses 17 coastal cities in the Bohai Rim as empirical objects. It then constructs an evaluation index system based on the entropy weight method, or TOPSIS, and proposes specific resilience enhancement strategies. The goal is to provide theoretical underpinnings and policy references that will enable the cities to withstand future multiple risks, achieve high-quality development, and revitalize their urban areas.

3. Characterization of Infrastructure System Resilience from a CAS Perspective and Elemental Components

To systematically analyze the intrinsic mechanisms of urban infrastructure resilience, this section employs Complex Adaptive System (CAS) theory to characterize and deconstruct the infrastructure resilience system from three dimensions: diversity, adaptability, and synergy. The theoretical construction in this section aims to provide a conceptual foundation and methodological basis for subsequent empirical analysis, thereby enhancing the systematicity and explanatory power of the research.

3.1. Characterization of Infrastructure Resilience from a CAS Perspective

Complex Adaptive Systems (CAS) theory, a third-generation systems theory, was systematically proposed by Professor Holland of the University of Michigan in 1994 [34]. According to this theory, a system comprises multiple heterogeneous agents with learning, adaptive, and evolutionary capabilities. Each agent exhibits nonlinear active interaction behaviors and continuously adjusts itself to drive the evolution of the entire system via its adaptive capacity [35]. The urban infrastructure system is a typical complex adaptive system, and its resilience not only relies on the performance of individual subsystems but also on the synergy and adaptive capacity among subsystems. Therefore, this paper chooses CAS theory as the research framework to analyze the interaction mechanism of urban infrastructure resilience.

Based on CAS theory, urban infrastructure resilience is characterized as follows [35]:

(1)

Diversity: The urban infrastructure system is a complex system that encompasses many aspects of the city and requires a diversity of resources, technologies and approaches to be integrated in order to reduce the level of urban risk. Interactions between the subsystems of transportation, energy, water supply and drainage, environment, and communication, as well as between the subsystem parts themselves, enable the system’s diversity. As time goes on, the elements that are shown in the sub-systems vary dynamically. System diversity evolves through the incorporation of new elements and the phasing out of obsolete ones.

(2)

Adaptability: The infrastructure system’s subsystems are comparatively independent and have the ability to modify their operational parameters in accordance with internal logic, creating an adaptive response mechanism to changes in the environment. Urban infrastructure systems continue to enhance their non-material level of resilience capacity, or “soft resilience,” as a result of the experience and information gained during the risk adaptation process. The self-adaptation of every component of the system contributes to its overall stability, which is achieved by the self-adaptation of all components.

(3)

Synergy: The multifaceted network of urban material bases that makes up the urban infrastructure system keeps the entire urban complex system operating steadily by facilitating resource sharing and connectivity amongst its constituent parts. The synergistic qualities of the urban infrastructure system are formed by the coupling relationship and interaction mechanism between the subsystems and their constituent pieces. This synergism is a crucial trait for the implementation of the system’s overall function. The resilience of urban infrastructure is positively connected with the degree of system synergy; a high degree of synergy not only improves the functional correlation between components but also increases resource allocation efficiency and the system’s overall risk-coping capacity.

3.2. Infrastructure Resilience System Components and Response Mechanisms

The urban infrastructure resilience system is a complex system defined by synergy and adaptability. It comprises five key subsystems—energy, transportation, environment, water supply and drainage, and communication—each with unique functions, strong interconnections, and interdependencies, forming a complex network structure. own specific functions and is highly interconnected and interdependent, forming a complex network structure. The system enhances urban infrastructure resilience via an “attack-resistance-adaptation-recovery” mechanism [36], and the strength and efficiency of this responsiveness are reflected in the system’s resilience level, as shown in Figure 1.

4. Urban Infrastructure Resilience Evaluation System and Model Construction

To scientifically assess the level of infrastructure resilience in coastal cities of the Bohai Rim region, this chapter constructs a multidimensional evaluation system based on CAS theory and establishes corresponding quantitative models. This section aims to clarify the rationale for indicator selection, data sources, and model construction methods, thereby providing a theoretical foundation and methodological support for subsequent empirical analysis. Through a systematic combination of models, this chapter strives to comprehensively and objectively reveal the spatiotemporal evolution characteristics and driving mechanisms of infrastructure resilience in the Bohai Rim region.

4.1. Selection of Infrastructure Resilience Evaluation Indicators

The urban infrastructure resilience evaluation framework includes five dimensions: energy, transportation, environment, water supply and drainage, and communication. Drawing on the evaluation framework and relevant research literature, 21 indicators were selected for each subsystem to measure the resilience of urban infrastructure systems [19,37,38,39]. The indicator selection is as follows: For the transportation system, indicators include road mileage, transportation capacity, and car ownership. For the water supply and drainage system, indicators fully consider a city’s water supply and drainage capacity and level, focusing primarily on water supply, water discharge, and related facilities, and water discharge facilities; for the energy system, energy use and ownership are selected as the criteria for evaluating the resilience of the city’s energy; and for the environment, the indicators are used to measure the resilience of the city’s infrastructure system. system is a key subsystem for sustainable development of the city, and the selected indicators include greening and pollution level; communication system is the core support structure for urban information transmission, economic operation, public service, emergency management and scientific and technological innovation, and it has a decisive impact on the functioning and resilient development of the city, and the selected indicators include the Internet, cell phone and other aspects. The infrastructure system spans multiple urban dimensions, and the indicator framework highlights its inherent diversity. Through the above screening, we finally constructed the urban infrastructure resilience evaluation system, as shown in

Table 1. Urban infrastructure resilience evaluation system.

Target	Standard	Indicator	Attribute
Level of resilience of urban infrastructure	Energy system resilience	Per capita natural gas use (m3/person)	+
		LPG supply per capita (t/million people)	+
		Per capita heat supply (GJ/person)	+
	Transportation system resilience	Road area per capita (m2/person)	+
		Cabs per 10,000 people (units/10,000 people)	+
		Road passenger traffic per capita (persons)	+
		Bus ownership per 10,000 people (vehicles per 10,000 people)	+
		Road freight per capita (t/person)	+
	Environmental system resilience	Greening coverage of built-up areas (%)	+
		Sewage discharge (t)	−
		Green space per capita (m2/person)	+
		Area of green space in parks (hectares)	+
		Sewage treatment (t)	+
	Resilience of water supply and drainage systems	Per capita daily domestic water consumption (liters/person)	+
		Per capita water supply (m3/person)	+
		Length of drainage pipes per capita (km/ten thousand people)	+
		Length of water supply pipes per capita (km/ten thousand people)	+
	Communication system resilience	Internet users per 100 population (households/100 population)	+
		Cell phone subscribers per 100 population (households/100 population)	+
		Postal operations per capita (yuan/person)	+
		Telecommunications per capita (yuan/person)	+

Note: “+” Indicates positive support elements, “−” Indicate negative support factors.

.

4.2. Data Sources

The data studied in this paper come from the China Urban Statistical Yearbook and the China Urban Construction Statistical Yearbook from 2013 to 2022. Missing data are minimal and imputed using the mean difference method.

4.3. Infrastructure Resilience Evaluation Model

4.3.1. Entropy Weighting Method to Determine Weights

The entropy weight method is an approach based on the principle of information entropy for determining the objective weights of indicators. It is primarily used to address the issue of comprehensive evaluation of multiple indicators and can offer a relatively objective method for determining indicator weights. The entropy weight method does not need to rely on expert opinion or subjective preference of decision makers and assigns weights completely based on the characteristics of the data itself, which reduces the interference of human factors and improves the objectivity and scientificity of the evaluation results. This method can effectively handle the comprehensive evaluation issues of a multidimensional and multiscale nature, which highly aligns with the inherent diversity characteristics of infrastructure systems. By using an objective weighting method based on information entropy, it applies consistent weighting to the indicators of each subsystem, not only avoiding subjective bias but also indirectly revealing the relative contribution of different subsystems to the overall resilience composition, thus providing a quantitative basis for identifying the shortcomings of collaboration within the system. The specific steps are as follows:

The original matrix is constructed according to the evaluation index system, and the original matrix is constructed by selecting n indicators for m years:

$$X= \mid x_{ij}{\mid }_{m\times n}\left(i=\mathrm{1,2},\cdots ,m;j=\mathrm{1,2},\cdots ,n\right)$$

(1)

Standardization of raw data. In the evaluation system, there are various indicators, which represent different meanings, resulting in the existence of different quantitative outlines, and there is a huge difference between the indicators. To eliminate the influence of such scale differences, this study employs the extreme value method to standardize the data, confining values to the range of 0 to 1.

The normalization formula for positive indicators is as follows:

$$S_{ij}^{+}={\displaystyle \frac{x_{ij}-min\left\{x_{1j},x_{2j},\cdots ,x_{nj}\right\}}{max\left\{x_{1j},x_{2j},\cdots ,x_{nj}\right\}-min\left\{x_{1j},x_{2j},\cdots ,x_{nj}\right\}}}+0.0001$$

(2)

The normalization formula for negative indicators is as follows:

$$S_{ij}^{-}={\displaystyle \frac{max\left\{x_{1j},x_{2j},\cdots ,x_{nj}\right\}-x_{ij}}{max\left\{x_{1j},x_{2j},\cdots ,x_{nj}\right\}-min\left\{x_{1j},x_{2j},\cdots ,x_{nj}\right\}}}+0.0001$$

(3)

where $S_{ij}$ denotes the value of the $j$ indicator in year $i$ after normalization; $x_{ij}$ is the raw data of the j indicator in year i; $min\left\{x_{1j},x_{2j},\cdots ,x_{nj}\right\}$ is the minimum value of the $j$ indicator; $max\left\{x_{1j},x_{2j},\cdots ,x_{nj}\right\}$ is the maximum value of the $j$ indicator.

Specific gravity $P_{ij}$, information entropy $e_j$, and utility value $d_j$ are calculated for the normalized data.

① Calculation of the share of indicator $j$ in year $i$ $P_{ij}$

$$P_{ij}=S_{ij}÷\sum _{i=1}^m S_{ij}$$

(4)

② Calculate the information entropy of the j indicator $e_j$

$$e_j=-k\sum _{i=1}^n P_{ij}\ln \left(P_{ij}\right)$$

(5)

$$k={\displaystyle \frac{1}{\ln \left(a\right)}}, 0\le e_j<1$$

(6)

where k is a constant and a is the number of coastal cities in the Bohai Rim studied in this paper. The entropy value can be used to determine or assess the degree of change in the indicators in the indicator system. If the information entropy of the indicators is greater, the lower the variability of the indicator data, the smaller the weight of the indicators; conversely, the greater the weight of the indicators.

③ Calculate the information utility value of the $j$ indicator $d_j$

$$d_j=1-e_j$$

(7)

Calculation of the weight of indicator j $w_j$

$$w_j=d_j/\sum _{j=1}^n{d}_j$$

(8)

Style: $0\le w_j\le 1$.

4.3.2. TOPSIS Modeling to Determine Resilience Levels

TOPSIS was first proposed by YOON in 1995 [40]. The TOPSIS model, often referred to as the “ideal solution ranking method” or the “optimal worst solution method”, is a multi-criteria decision analysis method. The core concept of this method is to find the best solution in the decision matrix, i.e., to find the solution that is closest to the ideal solution and farthest from the worst solution. TOPSIS assumes that any decision choice should be as close as possible to the positive ideal solution and as far away as possible from the negative ideal solution. This study focuses on the infrastructure resilience of 17 cities within the region, aiming to enable both inter-city comparisons and longitudinal analyses of evolutionary trends over a ten-year period. The TOPSIS method was applied within a multi-dimensional resilience evaluation framework to measure the relative distance of each evaluated object from the ideal state. This approach ensures uniform and objective comparability of resilience levels across different cities and time points, thereby effectively supporting cross-sectional and temporal comparative analysis. The overall urban infrastructure resilience score and the resilience scores of the different dimensions are calculated by the TOPSIS model based on the weights of the calculated indicators. The specific steps are as follows:

Determine the weighted normalization matrix. Based on the weights and normalized values of the urban infrastructure resilience indicators obtained from the calculation, the weighted normalization matrix for urban infrastructure resilience is calculated with the following formula:

$$q_{ij}=S_{ij}\times w_j$$

(9)

where $q_{ij}$ is the weighted normalized value of the j indicator in year i.

Determine the positive ideal solution $Q^{+}$ and the negative ideal solution $Q^{-}$ of the matrix:

$$Q^{+}=\left(\mathit{max}\left\{q_{11},q_{21},\cdots ,q_{m1}\right\},\cdots ,\mathit{max}\left\{q_{1n},q_{2n},\cdots ,q_{mn}\right\}\right)=\left(q_1^{+},q_2^{+},\cdots ,q_n^{+}\right)$$

(10)

$$Q^{-}=\left(min\left\{q_{11},q_{21},\cdots ,q_{m1}\right\},\cdots ,min\left\{q_{1n},q_{2n},\cdots ,q_{mn}\right\}\right)=\left(q_1^{-},q_2^{-},\cdots ,q_n^{-}\right)$$

(11)

Calculate the Euclidean distance T between each indicator of urban infrastructure resilience and the optimal value $Q^{+}$ and the worst value $Q^{-}$.

(i)

Optimal Euclidean distance $T_i^{+}$:

$$T_i^{+}=\sqrt{\sum _{j=1}^n {\left(q_j^{+}-q_{ij}\right)}^2}$$

(12)

(ii)

the worst Euclidean distance $T_i^{-}$:

$$T_i^{-}=\sqrt{\sum _{j=1}^n {\left(q_j^{-}-q_{ij}\right)}^2}$$

(13)

Calculate the closeness of the metrics to the ideal solution $C_i$, and the composite resilience score.

$$C_i={\displaystyle \frac{T_i^{-}}{T_i^{+}+T_i^{-}}},0<C_i<1$$

(14)

$C_i$ The larger the value, the higher the city’s comprehensive infrastructure resilience score, i.e., the higher the city’s level of infrastructure resilience; conversely, the lower the city’s level of infrastructure resilience.

4.4. Obstacle Degree Model Exploring Infrastructure Resilience Barriers

The Obstacle Degree Model is a theoretical framework for assessing and managing the challenges and barriers that an individual or organization faces in achieving its goals [41]. The model focuses on identifying, quantifying and understanding the key factors that may inhibit or delay success. In order to better identify the paths and enhancement measures for resilience development of coastal cities in the Bohai Rim, this paper uses the Obstacle Degree Model to find out the main reasons that constrain the resilience development of cities in this region.

Calculation of indicator deviation I

$$I_{ij}=1-S_{ij}$$

(15)

where Indicator deviation $I_{ij}$ denotes the gap between the actual value of the indicator and the optimal value, and $S_{ij}$ denotes the standardized value of the $j$ indicator for the $i$ year.

Calculation of the indicator barrier $O$

$$O_{ij}={\displaystyle \frac{w_j{I}_{ij}}{\sum _{j=1}^n{w}_j{I}_{ij}}}$$

(16)

$O_{ij}$ The larger the value of the indicator, the more it hinders the development of urban resilience, i.e., factors that affect the improvement of urban resilience; conversely, it has little impact on urban resilience; and when $O_{ij}$ is 0, it means that the indicator has no impact on urban resilience.

4.5. Geographic Detector Probes Spatial Heterogeneity of Infrastructure Resilience

Geographic detectors are a statistical model for spatial analysis, detecting spatial dissimilarities as well as a set of statistical methods to reveal the driving forces behind them [42]. In this paper, geographic detectors are used to explore the factors influencing the resilience of infrastructure in coastal cities in the Bohai Rim region, as well as the interactions among the factors, which are expressed as:

$$q=1-{\displaystyle \frac{\sum _{h=1}^L N_h{\sigma }_h^2}{N{\sigma }^2}}$$

(17)

where L denotes the total number of indicators, $N_h$ and $N$ are the number of stratum samples and total samples of indicator $h$ ($h$ = 1, 2…L) in the study area, ${\sigma }_h^2$ and ${\sigma }^2$ are the variance of the $Y$ value of indicator h and the whole area, respectively. $q$ is the explanatory power of the influencing factors on the resilience of the urban infrastructure, whose value range is [0, 1]; the bigger the value, the more powerful the explanatory power of this factor on the resilience of the urban infrastructure, and the more obvious the spatial heterogeneity.

4.6. Time Geographically Weighted Regression Model to Explore Infrastructure Resilience Drivers

The spatio-temporal geographically weighted regression model, first proposed by Brunsdon, adds the temporal dimension to the traditional geographically weighted regression model to explore the non-stationarity of the data in both the temporal and spatial dimensions [43]. The driving factors of urban infrastructure resilience exhibit significant variations across spatial locations and over time, and the Geographically and Temporally Weighted Regression (GTWR) model effectively captures the complex and dynamic nature of resilience evolution in urban infrastructure systems. Therefore, in this paper, we explore the spatio-temporal relationship between the drivers and urban infrastructure resilience in both the temporal and spatial dimensions through a GTWR model.

$$Y_i={\beta }_0\left(u_i,v_i,t_i\right)+\sum _{k=1}^p {\beta }_k\left(u_i,v_i,t_i\right)X_{ik}+{\epsilon }_i$$

(18)

where $X$ and $Y$ are the explanatory variables and the explained variables, $\left(u_i,v_i,t_i\right)$ is the spatial and temporal coordinates of the i city, ${\beta }_0\left(u_i,v_i,t_i\right)$ is the intercept term, $p$ is the number of explanatory variables, ${\beta }_k\left(u_i,v_i,t_i\right)$ is the estimated coefficient of the k explanatory variable, and ${\epsilon }_i$ is the residual of the model.

5. Results

5.1. Overview of Coastal Cities in the Bohai Rim Region

With a total coastline length of over 5800 km and an area of 77,000 km², the Bohai Sea is a semi-enclosed shallow sea that is surrounded on three sides by the mainland and has a population of about 270 million [44]. Although the region’s average urbanization rate is approximately 65%, population density is highly unevenly distributed, with the Beijing-Tianjin-Hebei urban agglomeration presenting a “core-edge” structure and the Shandong Peninsula and Liaodong Peninsula having a relatively balanced distribution, comprising a land area of roughly 5.4% of the country. With the largest port group in northern China, which includes Tianjin Port, Dalian Port, Qingdao Port, and other top-tier ports, the Bohai Rim region enjoys a unique and advantageous geographic location. A strong basis for economic development is provided by the region’s abundance of natural resources, including coal, iron ore, oil, and other resource deposits. The 17 cities that make up the Bohai Coastal Cities are: Qinhuangdao, Tangshan, and Cangzhou in Hebei Province; Qingdao, Dongying, Yantai, Weifang, Weihai, Rizhao, and Binzhou in Shandong Province; Dandong, Jinzhou, Yingkou, Panjin, Dalian, and Huludao in Liaoning Province; and Tianjin, a municipality that is directly under the central government. Natural disasters, including excessive rainfall, flooding, and frequent earthquakes, occur in the area as a result of the city’s ongoing expansion. The Bohai Rim’s sustainable and superior development is aided by this paper’s assessment of the region’s urban infrastructure’s resilience (Figure 2).

5.2. Comprehensive Evaluation of Infrastructure Resilience

According to the selected 17 coastal cities in the Bohai Sea from 2013 to 2022, the data are substituted into Equations (1)–(8), from which the subsystem weights under the infrastructure resilience evaluation system, as well as the weights of each indicator, can be obtained, as shown in

Table 2. Infrastructure resilience indicator weights for coastal cities in the Bohai Rim.

Target	Standard	Weights	Indicator	Weights
Level of resilience of urban infrastructure	Energy system resilience	0.166	Per capita natural gas use (m3/person)	0.048
			LPG supply per capita (t/million people)	0.077
			Per capita heat supply (GJ/person)	0.041
	Transportation system resilience	0.224	Road area per capita (m2/person)	0.033
			Cabs per 10,000 people (units/10,000 people)	0.047
			Road passenger traffic per capita (persons)	0.037
			Bus ownership per 10,000 people (vehicles per 10,000 people)	0.052
			Road freight per capita (t/person)	0.055
	Environmental system resilience	0.279	Greening coverage of built-up areas (%)	0.028
			Sewage discharge (t)	0.010
			Green space per capita (m2/person)	0.043
			Area of green space in parks (hectares)	0.085
			Sewage treatment (t)	0.112
	Resilience of water supply and drainage systems	0.132	Per capita daily domestic water consumption (liters/person)	0.041
			Per capita water supply (m3/person)	0.018
			Length of drainage pipes per capita (km/ten thousand people)	0.042
			Length of water supply pipes per capita (km/ten thousand people)	0.031
	Communication system resilience	0.199	Internet users per 100 population (households/100 population)	0.050
			Cell phone subscribers per 100 population (households/100 population)	0.036
			Postal operations per capita (yuan/person)	0.065
			Telecommunications per capita (yuan/person)	0.049

.

Based on the weights of the obtained evaluation index system, the comprehensive infrastructure resilience scores of each city in different years were calculated through Equations (9)–(14), as shown in

Table 3. Results of Infrastructure Resilience Evaluation for Coastal Cities in the Bohai Sea, 2013–2022.

Cities	Year
	2013	2014	2015	2016	2017	2018	2019	2020	2021	2022
Dandong	0.14	0.20	0.25	0.19	0.23	0.21	0.21	0.10	0.16	0.15
Dalian	0.51	0.48	0.46	0.44	0.47	0.53	0.46	0.49	0.48	0.51
Yingkou	0.20	0.20	0.21	0.25	0.30	0.30	0.28	0.32	0.33	0.32
Panjin	0.32	0.33	0.31	0.32	0.32	0.34	0.35	0.32	0.29	0.36
Jinzhou	0.19	0.23	0.22	0.22	0.26	0.25	0.29	0.31	0.30	0.22
Huludao	0.21	0.17	0.23	0.26	0.31	0.22	0.26	0.34	0.27	0.28
Qinhuangdao	0.29	0.30	0.23	0.28	0.29	0.30	0.30	0.28	0.29	0.28
Tangshan	0.30	0.23	0.23	0.28	0.27	0.27	0.26	0.28	0.28	0.27
Tianjin	0.58	0.57	0.60	0.59	0.63	0.60	0.57	0.58	0.59	0.59
Cangzhou	0.28	0.26	0.28	0.27	0.26	0.30	0.30	0.27	0.29	0.31
Binzhou	0.23	0.19	0.20	0.21	0.29	0.28	0.26	0.28	0.29	0.25
Dongying	0.37	0.31	0.34	0.34	0.37	0.39	0.38	0.37	0.39	0.41
Weifang	0.27	0.22	0.21	0.21	0.25	0.22	0.25	0.24	0.24	0.26
Yantai	0.33	0.29	0.31	0.30	0.32	0.29	0.30	0.27	0.29	0.31
Weihai	0.36	0.31	0.34	0.36	0.39	0.38	0.38	0.35	0.35	0.38
Qingdao	0.45	0.51	0.45	0.47	0.43	0.52	0.48	0.49	0.51	0.54
Rizhao	0.21	0.20	0.22	0.20	0.23	0.21	0.22	0.18	0.22	0.24

.

Table 3 reveals that the infrastructure resilience of the 17 coastal cities along the Bohai Sea has not improved significantly over the past decade. The infrastructure resilience of these 17 cities is divided into three tiers: Tianjin, a municipality directly under the central government, leads the first tier; Qingdao and Dalian follow in the second tier; and the remaining 14 cities have roughly equivalent infrastructure resilience, forming the third tier. In the third tier, Weihai and Dongying have higher levels of infrastructure resilience, while Dandong and Rizhao lag behind. Dandong and Rizhao exhibit a clear downward trend in infrastructure resilience after 2019.

Based on the results in Table 3, the resilience level of urban infrastructure in the region over the years was averaged and plotted in Figure 3. An analysis of the average urban infrastructure resilience reveals an overall upward trend in the region’s resilience over the past decade. The resilience level realized a year-on-year growth trend in 2014–2017, which may be attributed to the proposal of the Beijing–Tianjin–Hebei Cooperative Development Strategy in 2014 and the Outline of Cooperative Development of the Bohai Rim Region approved by the State Council in 2015. This highlights the adaptive nature of infrastructure systems that evolve to improve resilience. Enhancing infrastructure resilience is a complex and challenging task that involves the entire infrastructure system, and the average infrastructure resilience score indicates a slow improvement rate. The slow improvement rate of resilience also reveals insufficient synergy within the system. The nonlinear interactions among various subsystems have failed to form an effective synergistic effect, thereby restricting the leap of overall resilience.

5.3. Evaluation of Resilience of Different Subsystems of Infrastructure

Based on the resilience scores of the cities in the region for each year, the average resilience scores of the subsystems in the region for each year were taken and plotted in

Table 4. Average level of resilience of infrastructure subsystems in coastal cities in the Bohai Rim, 2013–2022.

Year	Average Subsystem Resilience Level
	Communication System	Water Supply and Drainage	Environmental Systems	Transportation System	Energy System
2013	0.31	0.37	0.23	0.32	0.31
2014	0.28	0.35	0.24	0.32	0.28
2015	0.25	0.35	0.25	0.36	0.32
2016	0.28	0.37	0.25	0.35	0.33
2017	0.29	0.46	0.23	0.39	0.33
2018	0.36	0.38	0.25	0.40	0.27
2019	0.34	0.40	0.26	0.39	0.24
2020	0.33	0.41	0.27	0.37	0.26
2021	0.34	0.40	0.26	0.39	0.26
2022	0.39	0.41	0.26	0.39	0.27

.

Based on the five dimensions in the evaluation system, the strengths and weaknesses are explored in each dimension. Table 4 shows that the energy system resilience level in the region is in a decreasing trend, while the communication, water supply and drainage, environmental system, and transportation system resilience levels are in an increasing trend. Communication, water supply and drainage, and transportation systems have similar levels of resilience, while environmental systems are lagging behind in the region. This reflects a significant lack of synergies between the energy and environmental subsystems in the regional infrastructure system.

5.4. Analysis of Spatial and Temporal Differences in Infrastructure Resilience

Geostatistical trend analysis is an important component of GIS geostatistical analysis [45]. This paper clusters cities with different resilience indexes according to GIS software 10.8.2 using the Jenks natural breakpoint method [46] and integrates the resilience level evaluation results of 2013–2022 with GIS software to analyze the resilience development of urban infrastructure from spatial characteristics. In which the resilience scores of the four years of 2013, 2016, 2019, and 2022 are selected to be combined with GIS to generate a visualization map, as shown in Figure 4.

There are significant variations in the degree of urban infrastructure resilience with respect to its temporal and spatial evolution. Tianjin, Dalian, and Qingdao are the cities in the region with the highest resilience levels, according to the spatial and temporal distribution map from 2013; however, the resilience level of the urban infrastructure in the region has decreased from 2013 to 2016, with the exception of Tianjin, Dalian, and Qingdao; The overall level of infrastructure resilience in the region has increased when compared to 2016, according to the regional and temporal distribution map of 2019 and 2022. Examine the 2019 and 2022 spatial and temporal distribution maps. GIS-based spatiotemporal analysis reveals low synergy in infrastructure system resilience across the region. When comparing these three cities with the highest resilience levels in the region, Dalian exhibits a weak geographical spillover effect, while Tianjin and Qingdao’s development effectively enhances the infrastructure resilience of neighboring cities. It is impossible to divorce Tianjin, Qingdao, and Dalian’s infrastructure resilience scores from their respective city attributes, geographical locations, and contextual factors. Tianjin, Qingdao, and Dalian are strategic centers at the national level: Qingdao serves as the center of the Shandong Peninsula Blue Economic Zone, Tianjin is a key component of the Beijing-Tianjin-Hebei coordinated development strategy, and Dalian is an essential entry point for the revival of Northeast China. These cities have a high concentration of resources within their borders because they have long profited from governmental preferences and significant infrastructural investments. One of the four biggest cities in China, Tianjin boasts the biggest port in the country’s north, a sophisticated air, sea, and land transportation system, and a significant educational edge. Because of its strategic location, Qingdao, the second-biggest city in Shandong Province, has developed its coastline and drawn numerous businesses and skilled individuals. In the early years of New China, Dalian, a major industrial city with a considerable foundation in machinery and petrochemicals, was situated near the southernmost tip of the Liaodong Peninsula.

Because the Bohai Rim coastal area is made up of three provinces and one municipality, the lack of cross-provincial coordination mechanisms and unified resilience standards may explain the poor synergy in infrastructure resilience across the region. Regional development gaps have been exacerbated as a result of local cities’ independent planning and building, which has made it challenging to establish effective cooperation in early disaster warning, emergency response, and resource allocation.

5.5. Barrier Factor Measurement

Based on the weights of each indicator calculated by the entropy weighting method and the standardized data, the degree of deviation I and the degree of obstacle O are measured by Equations (15) and (16). The calculated degree of obstacle of each sub-system of each city in each year is averaged, and the degree of influence of different sub-systems of the 17 cities in the coastal region of the Bohai Rim on the region is observed for the last ten years, as shown in

Table 5. Results of Average Obstacle Degree of Subsystems in Bohai Rim Coastal Cities.

Year	Average Subsystem Handicap
	Communication System	Water Supply and Drainage	Environmental Systems	Transportation System	Energy System
2013	0.09	0.59	0.22	0.24	0.18
2014	0.12	0.69	0.16	0.20	0.19
2015	0.11	0.61	0.19	0.19	0.21
2016	0.17	0.53	0.20	0.18	0.21
2017	0.12	0.61	0.17	0.17	0.26
2018	0.09	0.69	0.20	0.18	0.20
2019	0.10	0.74	0.16	0.22	0.17
2020	0.11	0.70	0.21	0.20	0.16
2021	0.10	0.73	0.20	0.18	0.17
2022	0.17	0.32	0.30	0.22	0.19

.

An analysis of Table 5 shows that the water supply and drainage system in the region has an average obstacle degree score of 0.62, exerting the greatest impact on infrastructure resilience. The environmental, transportation, and energy systems have an average obstacle degree score of approximately 0.2, which has a relatively stable impact on urban infrastructure resilience but shows no obvious downward trend. The communication system has an average obstacle degree score of around 0.1, having the least impact on regional infrastructure resilience. By analyzing the factors affecting the resilience of infrastructure in the region through the obstacle degree model, it can be seen that the urban water supply and drainage system is the first problem to be solved, the weakness of this subsystem has severely undermined the adaptability of the entire infrastructure system, leaving it lacking sufficient response capabilities and alternative solutions when confronting risks unique to coastal cities, such as seawater intrusion and waterlogging; followed by the development of the environment, transportation, and energy systems have encountered bottlenecks, and new development paths need to be found in order to further reduce their impact on the resilience of urban infrastructure.

Geographically speaking, the region’s maritime characteristics greatly worsen the effects on the metropolitan water supply and drainage system. Long-term issues in the region include seawater intrusion brought on by climate change and tidal influences, as well as inadequate capacity and efficiency of the current treatment facilities, which reduces the stability of the water supply and drainage. Rotterdam, the Netherlands, has created a multi-level water management system to meet tidal and climatic influences, turning metropolitan areas into adaptable water control systems in contrast to other coastal cities of a similar nature [47]. This strategy provides a workable route for resilient development in low-lying cities across the globe. To address water scarcity and urban waterlogging, Singapore has proposed the “Water Duality Approach” [48]. Through technology, planning, and governance, the country has integrated water supply and drainage into a closed-loop, resilient water system, achieving a transformation from “dual opposition” to “dual unity”. In contrast, coastal cities around the Bohai Rim continue to rely on land-based designs for water systems, lacking an integrated land–sea operational structure, which makes them less able to adapt to marine circumstances. From an economic perspective, transportation and energy projects have historically dominated infrastructure investment in the Bohai Rim’s coastal cities. The percentage of investment devoted to water supply and drainage infrastructure is comparatively low when compared to comparable coastal towns such as Yokohama in Japan [49] and San Francisco in the United States. The two main problems caused by this disparity in investment are inadequate emergency facilities and postponed pipeline renovations, both of which seriously compromise the infrastructure’s resilience.

5.6. Geographic Detector Measurement

5.6.1. Divergence and Factor Detection

In this study, indicator data from 2013, 2016, 2019, and 2022 are selected. These data are discretized using the natural breakpoint method in ArcGIS 10.8, with urban infrastructure resilience as the dependent variable and each indicator as the independent variable. Geographic detectors are then used to measure the explanatory power of the indicator system on the spatial heterogeneity of urban infrastructure resilience in the region, as presented in

Table 6. Explanatory Power of Infrastructure Resilience Influencing Factors for Coastal Cities in the Bohai Rim.

Variant	2013	2016	2019	2022	Average
×1	0.42880	0.33289	0.40044	0.40822	0.39259
×2	0.57308	0.25886	0.32718	0.29813	0.36431
×3	0.18363	0.27807	0.23289	0.11414	0.20219
×4	0.08128	0.19201	0.17155	0.04610	0.12273
×5	0.15587	0.06412	0.47673	0.19088	0.22190
×6	0.18291	0.15690	0.37022	0.25002	0.24001
×7	0.35934	0.26725	0.25665	0.56203	0.36132
×8	0.22466	0.07854	0.22969	0.08580	0.15467
×9	0.42654	0.02455	0.13838	0.55355	0.28575
×10	0.30240	0.81327	0.79513	0.74140	0.66305
×11	0.47722	0.41763	0.27158	0.13872	0.32629
×12	0.63003	0.76116 **	0.70728	0.52472	0.65580
×13	0.74677 **	0.83392 *	0.79513	0.75818	0.78350
×14	0.22830	0.17162	0.33784	0.17596	0.22843
×15	0.48112	0.25459	0.24095	0.40561	0.34557
×16	0.16182	0.23323	0.14860	0.19337	0.18425
×17	0.15735	0.07255	0.02158	0.14156	0.09826
×18	0.62941 *	0.48545	0.54729 *	0.42194	0.52102
×19	0.83401 ***	0.75906 *	0.74932 ***	0.66801 *	0.75260
×20	0.51987	0.80915 *	0.74932 *	0.47183	0.63755
×21	0.75658 **	0.82665 *	0.85655 ***	0.86197 ***	0.82544

Note: * denotes p < 0.1, ** denotes p < 0.05, *** denotes p < 0.01.

.

Looking at the explanatory power values in Table 6, it can be seen that wastewater discharge ×10, green space in parks ×12, wastewater treatment ×13, Internet subscribers per 100 inhabitants ×18, cell phone subscribers per 100 inhabitants ×19, postal services per capita ×20, and telecommunication services per capita ×21 have a higher explanatory power for the resilience of the region’s urban infrastructure than, and that these indicators are distributed mainly in the environmental system and the communication system. These indicators are primarily distributed across environmental and communication systems.

5.6.2. Interaction Detection

Interaction detection is a method used to assess how the interaction of two or more factors impacts the dependent variable. It determines whether interactions exist between factors by comparing the influence of individual factors with that of multiple overlapping factors. Based on the results of divergence and factor detection, the above seven indicators are ranked e1–e7 in order from smallest to largest in the original ordinal number, and the interaction effect of the seven indicators on infrastructure resilience in the region is investigated. The assessment process is to first calculate the explanatory power q (e1) and q (e2) of two factors e1 and e2 on the dependent variable, and then calculate the explanatory power q (e1 ∩ e2) of the two factors interacting together, compare the magnitude relationship between the three q values, and derive the type of interaction between the two factors, as shown in

Table 7. Type of interaction of the two factors on the dependent variable.

Basis of Judgment	Interaction Type
q (e1 ∩ e2) < Min(q(e1), q(e2))	nonlinear weakening
Min(q(e1), q(e2)) < q (e1 ∩ e2) < Max(q(e1), q(e2))	Single-factor nonlinear attenuation
q (e1 ∩ e2) > Max(q(e1), q(e2))	two-factor enhancement
q (e1 ∩ e2) = q(e1) + q(e2)	stand alone
q (e1 ∩ e2) > q(e1) + q(e2)	nonlinear enhancement

.

As shown in Figure 5, interaction detection results for each factor, the type of interaction between every two factors is two-factor enhancement, indicating that the explanatory power of the spatial differentiation of infrastructure resilience when any two factors are combined is higher than that of a single factor independently. The interactive explanatory power is enhanced in the period of 2013–2016, but an overall decreasing trend is shown in the period of 2016–2022, in which the number of Internet users per 100 people e4 is significantly weakened in relation to the amount of wastewater discharged e1, park green space area e2 and postal services per capita e6 with sewage discharge e1, park green space area e2, and sewage treatment e3 weakened significantly. This suggests that the popularization of communication infrastructure may equalize Internet coverage and reduce regional differences, leading to a decrease in the explanatory power of the interaction; the implementation of environmental governance in the Bohai Rim region after 2015, such as the “Ten Water Principles” and the “Sponge City” policy, may have reduced the amount of wastewater discharges through standardized measures. Standardized measures reduce the spatial heterogeneity of sewage discharge, thus weakening its synergistic effect with other factors; The independent development of environment and communication system, such as the reliance of environmental governance on administrative means, while communication technology relies on market-driven, may lead to the weakening of the linkage between the two, which is reflected in the declining explanatory power of the interaction. The weakening of synergies between the factors can, at the same time, be mapped to the increasing differentiation of synergistic development in the region from year to year.

In the Coastal cities along the United States, the transportation and energy systems have stronger explanatory power for resilience [50], which stands in sharp contrast to the Bohai Rim region. This difference stems primarily from industrial structure: North American cities are predominantly service-oriented, with high reliance on transportation commutes and stable energy supply, whereas the Bohai Rim remains heavily dependent on heavy chemical industries, placing greater emphasis on environmental governance pressures and the need to address gaps in communication infrastructure. Meanwhile, in cities along the European North Sea coast, the environmental system also demonstrates high explanatory power [51], but the communication system generally has lower explanatory power. This is because Europe has achieved balanced coverage of communication infrastructure through the “Digital Single Market” policy. In contrast, the Bohai Rim region, affected by administrative barriers and economic disparities, suffers from uneven distribution of communication resources, making this a key driver of resilience divergence.

5.7. Analysis of Infrastructure Resilience Drivers in Bohai Rim Coastal Cities

5.7.1. Selection of Indicators

Based on the diversity of infrastructure systems and the characteristics of the development of the coastal area around the Bohai Sea, as well as the results of related research [52,53], five factors, namely, economic level, industrial structure, environmental health, human capital, and urban level, were selected from the economic, geographic, and social aspects to be analyzed. And the selected indicator variables were tested for multiple covariances by the variance expansion factor, as shown in

Table 8. Drivers of urban infrastructure resilience.

Driving Force	Norm	VIF
economic level	GDP per capita (person/million dollars)	2.044
industrial structure	Share of secondary sector in GDP (%)	1.857
environmental health	Road sweeping and cleaning area (10,000 square meters)	3.665
human capital	University students per 10,000 population (persons per 10,000)	1.979
urban level	Population density (persons/km2)	2.637

.

Increasing economic levels provide sufficient financial support for the construction, maintenance, and upgrading of urban infrastructure. Higher GDP per capita may imply that the government and social capital have the ability to invest more resources in disaster-resistant design of infrastructure, such as transportation networks and energy supply, as well as rapid post-disaster restoration, thus enhancing the resistance and resilience of the infrastructure system. Therefore, GDP per capita is chosen to characterize the economic level in this paper.

The characteristics of the industrial structure are closely related to the development of resilient urban infrastructure systems. A high proportion of secondary industries implies a more concentrated demand for energy, transportation and other infrastructure. Heavy industrial production relies on stable power supply and large capacity transportation network; if the industrial structure is too much in favor of high energy consumption and high load industry, it may increase the pressure of infrastructure operation and reduce its resilience; on the contrary, if the industrial structure is upgraded to technology-intensive, it may promote the infrastructure to the direction of intelligent and efficient development, and enhance the resilience. Therefore, this paper selects the proportion of the secondary industry in GDP to characterize the industrial structure.

Sanitation is particularly important for the maintenance of urban infrastructure. The area of road sweeping and cleaning reflects a city’s efforts to maintain the surface environment of its infrastructure. Regular cleaning can reduce road dust, garbage accumulation on the road structure erosion, prolonging the service life of the road; in coastal areas with a humid climate, it is easy to lead to corrosion of infrastructure, strengthening the cleaning can reduce the environmental factors on the damage to the facilities, and thus enhance its durability and resilience. Therefore, this paper selects the road cleaning area to characterize environmental health.

High human capital means sufficient support for skilled personnel. A high concentration of university students indicates strong innovation potential, which can advance infrastructure-related technological R&D and enhance the design level and operation and maintenance efficiency of facilities. In addition, the participation of high-quality talents in infrastructure management can optimize the emergency response mechanism and enhance the adaptability of the system in the face of emergencies. Therefore, this paper selects the number of college students per 10,000 people to characterize human capital.

There is a positive correlation between urban growth and population size. Population density directly reflects the degree of urban infrastructure load. The high population density in the Bohai Rim region places higher demands on transportation, water supply, power supply, and other facilities, and overloading may lead to accelerated aging of facilities, reduced service efficiency and lower resilience. However, on the other hand, population concentration also prompts cities to pay more attention to the intensive and multi-functional design of infrastructure in planning, and to enhance the system resilience by optimizing the allocation of resources. Therefore, this paper selects population density to characterize the level of town.

5.7.2. Model Testing

Based on ArcGIS software, urban infrastructure resilience drivers in the coastal region of Bohai Rim are analyzed by OLS and GTWR models, and see

Table 9. Comparison of model parameters [54].

Parameters	OLS	GTWR
AICc	−454.317	−540.261
R2	0.688	0.934
Adjusted R2	0.641	0.932

the results show that GTWR model is better than OLS model in AICc, R2 and adjusted R2, which proves that GTWR model is better in fitting effect. Therefore, the GTWR model was selected to analyze the drivers of the region in this paper.

5.7.3. Analysis of the Time Evolution of Drivers

The drivers of urban infrastructure resilience in the Bohai Rim coastal region are regressed through the GTWR model and plotted as a box plot with time evolution, as shown in Figure 6.

From the overall trend of economic level, the average regression coefficient shows a decreasing trend, indicating that the overall “positive driving effect” of economic level on infrastructure resilience is weakening from 2013 to 2022. This may be due to the evolution of the stage of urban development, the initial stage of economic inputs to enhance the resilience of the direct strong, later need to be more diverse factors synergistic, purely economic level of the marginal contribution of the reduction; or Bohai Rim region after the growth of infrastructure stock, the resilience to enhance the use of economic resources to improve the efficiency of the structural adequacy of the requirements of the crude economic inputs are diluted. 2013–2022 economic level of the impact of the infrastructure of the region resilience is largely positive, but the box line for 2021–2022 is negative, implying that the economic level of some cities in these two years shows a negative impact on resilience. It is possible that economic resources are tilted towards emergency response under the impact of the new crown epidemic, squeezing the investment in long-term construction of resilience; or cities blindly pursuing economic growth rate, resulting in the increase in economic level not being translated into enhanced resilience, or even having a negative effect.

The average regression coefficient of industrial structure in 2013–2022 is close to zero and has little fluctuation, indicating that the overall impact of industrial structure on resilience is positive and weak. The Bohai Rim has long promoted “industrial upgrading and resilience synergy”, such as the Beijing-Tianjin-Hebei Synergistic Development Plan, which emphasizes the transfer of industries and complementary infrastructure. However, in the early days, the industrial structure was dominated by traditional heavy industry, which had a low demand for disaster prevention infrastructure, resulting in a weak industry-resilience correlation. In 2018, although there are industrial transformation policies, the proportion of traditional industries is still high, and the driving force of new kinetic energies, such as high-end manufacturing and the digital economy on resilience, has not yet been fully manifested, so the overall impact is maintained as “weakly positive”. The regression coefficients for some cities in 2021–2022 are also negative. This may be due to the fact that during the epidemic, infrastructure resilience inputs such as old pipeline network renovation and smart platform construction were put on hold, leading to a “conflict between industrial activities and resilience building”, and the short-term negative effect is evident.

The overall regression coefficients of environmental sanitation are positive from 2013 to 2022, indicating that environmental sanitation is a positive driver of infrastructure resilience in general. This may be due to the fact that the Bohai Rim region has long practiced the synergistic policy of “ecological + infrastructure”, such as the Bohai Rim Cooperative Development Program. Environmental sanitation provides “ecological background support” for infrastructure resilience, but because traditional infrastructure emphasizes function over ecological synergy, the initial positive effect is slow to be released, and the overall strength of the impact is weak.

The average regression coefficient of human capital from 2013 to 2022 is close to 0 for a long time, and the fluctuation is very small, which indicates that human capital has a weak positive impact on infrastructure resilience, and the mode of action is relatively stable. Bohai Sea coastal cities human capital foundation exists, but the early urban infrastructure focus on “scale expansion”, the response to “resilience needs” lagging behind, the human capital advantage has not been fully converted into resilience to enhance the power; the later part of the human resources policy, but resilience of infrastructure to the Although there are talent policies in the later stage, the adaptability of the resilience infrastructure to “composite talents” is insufficient, resulting in the strength of human capital driven resilience has always been weak, and the overall trend remains stable.

The average regression coefficient of urban level in 2013–2022 is positive and decreasing in the early stage, and then converges to 0 and stabilizes in the late stage, which means that the positive driving effect of urban development level on resilience gradually decays, and the influence tends to be weaker and stable in the late stage. Bohai coastal cities in the early stage of rapid development, urban development directly drive the resilience of infrastructure investment, the level of urban development and resilience to enhance the positive correlation is strong; but with the advancement of development, the traditional “scale expansion” development model is difficult to adapt to the resilience of the needs of some of the city’s development resources tilted to the non-resilience of the field, so that the positive drive to attenuate. The positive driving force is attenuated. Later, along with the “Beijing-Tianjin-Hebei Cooperative Development Planning Outline” and other policies to promote the transformation of the urban development model, urban development and resilience of the construction of the suitability of adjustment, the impact tends to stabilize, but the overall weak correlation.

In the early stage of system development, single driving factors such as economic investment and urbanization can directly and effectively enhance the adaptability of the system. As the system evolves into a more complex stage, the improvement of overall resilience no longer relies on the linear input of a single factor, but rather depends more on the nonlinear synergetic effects of multiple elements.

6. Conclusions and Suggestions

This study develops an evaluation framework for infrastructure resilience covering five dimensions: energy, transportation, environment, water supply and drainage, and communications. Employing methodologies such as the entropy-weighted TOPSIS method, obstacle degree model, geographical detector, and GTWR model, this study conducts a comprehensive assessment and driving factor analysis of infrastructure resilience for 17 coastal cities in the Bohai Rim region from 2013 to 2022. The main findings are as follows: (1) Although resilience levels have improved steadily over the period, inter-system coordination remains weak and regional disparities are pronounced; (2) The water supply and drainage system is identified as the most critical obstacle, while environmental and communication systems are the main drivers of spatial differentiation; (3) The influence of traditional driving factors has weakened, underscoring the growing need for diversified and coordinated governance mechanisms.

China’s 14th Five-Year Plan advocates for the development of “resilient cities”. The goal of enhancing urban infrastructure resilience is to strengthen cities’ resistance and rapid recovery capabilities against natural disasters, climate change, and other uncertainties, while ensuring the continuity and safety of urban operations. Continuity and safety of urban operation. Based on the infrastructure resilience of Bohai Sea coastal cities, the following recommendations are proposed:

(1)

Improve the regional coordination and long-term governance mechanism. First, establish differentiated resilience standards. For high-resilience cities like Tianjin and Qingdao, explore higher resilience construction standards that align with international benchmarks. For cities with low resilience, develop basic improvement plans to address gaps. Second, establish a dynamic resilience assessment and feedback system, updating data annually and using tools like geographic detectors to continuously monitor changes in obstacle factors and drivers, enabling dynamic policy optimization and precise resource allocation. In the future, experiences from cross-sectoral collaborative governance can be drawn upon [55], and mechanisms for information disclosure and transparency should be strengthened [56], so as to enhance the overall effectiveness and social recognition of urban resilience building.

(2)

Focus on key shortcomings and implement a resilience enhancement plan for water supply and drainage systems. First, address high-obstacle issues by prioritizing the systematic renovation and upgrading of aging water supply and drainage networks in resilience-vulnerable cities such as Dandong, Rizhao, and Binzhou. Use corrosion-resistant materials and embed IoT sensors to establish real-time monitoring and early warning systems. Second, respond to coastal characteristics by piloting the integration of “blue-green-gray” infrastructure in cities such as Tianjin, Dalian, and Qingdao, including the construction of ecological seawalls, rain gardens, and underground storage tanks to effectively prevent and control seawater intrusion and waterlogging risks. Finally, establish a regional coordination mechanism by developing a water supply scheduling and drainage emergency linkage platform for the Bohai Rim urban agglomeration, enabling cross-city sharing and optimal allocation of water resources and emergency resources to break administrative barriers.

(3)

Implement dynamic and precise policy adjustments to stimulate multiple drivers. First, optimize the investment structure by establishing a resilience special fund, shifting investment from previously prioritized areas such as transportation and energy to areas with shortcomings such as environmental governance, water supply and drainage, and digitalization. Second, promote green industrial transformation by incorporating resilience indicators, such as rainwater recycling and low-carbon transformation, into the assessment of enterprises in industrial cities such as Tangshan and Dongying, ensuring coordinated development between industrial growth and resilience building. Finally, strengthen talent development and innovation by leveraging regional universities and research institutions to cultivate interdisciplinary talent proficient in infrastructure operation and digital technology, and establish “resilience labs” for policy simulation and impact assessment.

(4)

Enhance smart empowerment and build a digital twin-based collaborative management platform. First, promote multi-system data integration by developing a “city information model” that covers transportation, energy, environment, water supply and drainage, and communication systems. Use AI algorithms to simulate coupling risks under extreme scenarios, enabling intelligent scheduling and preventive maintenance of infrastructure. Second, emphasize communication resilience by enhancing redundant backups for communication infrastructure in resilience-vulnerable areas, such as underground optical cables and satellite communication links, to ensure unimpeded information channels during disasters and bridge the digital divide.

This study adopts a holistic-subsystem analysis framework to systematically and comprehensively evaluate the infrastructure resilience of coastal cities in the Bohai Rim, effectively revealing the spatial and temporal evolutionary characteristics and current status of the region. Based on the differentiated analysis of the development level of each subsystem’s resilience, the study can formulate targeted policy guidance and development strategies for city managers to realize the systematic improvement of urban infrastructure resilience. This study still has limitations in terms of the theoretical framework and methodology. First, the study mainly focuses on the independent assessment of the resilience level of each subsystem and fails to fully consider the coupling relationship and synergistic effect between subsystems; second, the study of the resilience level of each subsystem of the region’s infrastructure and the factors affecting the resilience development of the region’s infrastructure adopts the form of taking the average, and the results of the study will have a certain degree of bias.