Urban resilience evaluation based on the DRIVING FORCE-PRESSURE-STATE-IMPACT-RESPONSE (DPSIR) framework and BP NEURAL NETWORK: A case study of Hubei Province

doi:10.21203/rs.3.rs-4978441/v1

Download PDF

Article

Urban resilience evaluation based on the DRIVING FORCE-PRESSURE-STATE-IMPACT-RESPONSE (DPSIR) framework and BP NEURAL NETWORK: A case study of Hubei Province

https://doi.org/10.21203/rs.3.rs-4978441/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Building resilient cities has become an emerging risk management strategy, thus it is necessary to make a scientific evaluation on urban resilience. In this study, both the Driving Force-Pressure-State-Impact-Response (DPSIR) framework and the BP neural network were innovatively adopted to construct a comprehensive urban resilience evaluation model. Prefecture-level cities in Hubei Province were examined for empirical analysis. The results showed that: (1) Urban resilience is a dynamic process of change. The resilience level of cities in Hubei Province was influenced primarily by two major factors: driving force and response. (2) The urban resilience of cities in Hubei Province had been improving steadily from 2015 to 2021, but there was a spatial negative correlation among them. Owing to uneven development within Hubei Province, it can be apparently seen that Wuhan, the provincial capital, holds a dominant position. (3) Resource and environmental pressure has become the main obstacle to the construction of resilient cities in Wuhan. The primary limiting factors for other cities are the degree of socioeconomic growth and the capacity of the government to handle affairs. This study not only enriched the theory and methods of urban resilience evaluation, but also had important reference value for the government to formulate effective urban sustainable development strategies.

Earth and environmental sciences/Environmental social sciences/Sustainability

Earth and environmental sciences/Environmental social sciences

Urban resilience

Driving Force-Pressure-State-Impact-Response (DPSIR)

BP neural network

Obstacle diagnosis model

Sustainable development

The global urbanization trend is irreversible and ongoing, with an increasing number of people opting to live in cities. According to The World Urbanization Prospects 2022 report released by UN-Habitat, the urban population accounted for 56% of the global total in 2021, and is expected to grow to 68% by 2050, leading to increased urbanization across all regions(UN-Habitat, 2022). Natural disasters, abrupt public health incidents, climate change, and financial crises are just a few of the dangers and stresses that cities face as the most intricate system of human-environment interaction. These are the specific manifestations of the opposition and conflict between people and the environmental system in the rapid development of urbanization, and restrict the sustainable development of cities. To solve these problems, building economic, social, and environmental resilience, including appropriate governance and institutional structures, must become the core of cities in the future(Elmqvist et al., 2019; Ayyoob Sharifi, 2019; Ziervogel et al., 2017). Resilience emphasizes the ability to resist destructive challenges and recover from them, gradually garnering the interest of scholars worldwide (Amirzadeh, Sobhaninia, & Sharifi, 2022; Chelleri, Waters, Olazabal, & Minucci, 2015).

In 1973, Holling(Holling, 1973) first introduced the concept of resilience into the field of ecology. He believed that resilience is the ability of a system to absorb disturbances and reorganize while undergoing changes, to maintain its function, structure, characteristics, and feedback, with a focus on the coexistence of the system and challenges. Later, the theory of resilience was applied to various fields. Later, The application of resilience theory has been extended across diverse disciplines such as ecology(Alberti et al., 2003; Shamsipour et al., 2024), agriculture and biosciences(P. Lu & Stead, 2013), social science(Meerow, Newell, & Stults, 2016), environmental science(McPhearson, Andersson, Elmqvist, & Frantzeskaki, 2015), engineering(Liao, 2012), business management and accounting(Romero-Lankao & Gnatz, 2013). Cities, as an essential research subject in the academic field, naturally apply resilience thinking to urban studies(A. Sharifi & Yamagata, 2016). The ability of a city and its urban system—social, economic, natural, human, technological, and physical—to withstand initial harm to lessen interference—impacts, natural disasters, weather fluctuations, crises, or disruptive events—to adjust to changes and restrict the system's capacity to adapt in the present or the future(Ribeiro & Gonçalves, 2019). As a new approach to urban risk governance, resilience has garnered significant attention from countries worldwide. In 2013, the Rockefeller Foundation launched the "Global 100 Resilient Cities" project, aiming to enhance the resilience of cities(Spaans & Waterhout, 2017). That same year, the UK Department for International Development collaborated with the Rockefeller Foundation, the Asian Development Bank, and the Swiss Secretariat for Economic Affairs to establish the Urban Climate Change Resilience Trust Fund (UCCRTF). China's 14th Five-Year Plan (2021–2025) proposed the concept of "building resilient cities, improving urban governance, and strengthening risk prevention and control in megacities" incorporating the construction of resilient cities into the national strategic planning framework (Mu, Fang, & Yang, 2022).

Against such a backdrop, it is particularly important to scientifically quantify urban resilience. Constructing a reasonable indicator system and selecting scientific evaluation methods are crucial for quantitative research on urban resilience. Most scholars choose evaluation indicators in the corresponding fields based on the economy, society, system, ecology, and infrastructure involved in the complex concept of urban resilience(Qasim et al., 2016). Hudec et al. (Hudec, Reggiani, & Siserová, 2018)evaluated the urban resilience and vulnerability of various regions in Slovakia in the context of the economic crisis from 12 indicators in 3 dimensions of economy, society, and community connectivity. Zhang et al. (Zhang, Yang, Li, & van Dijk, 2020) developed a comprehensive urban climate change resilience index system encompassing six key dimensions: societal resilience, economic resilience, community resilience, infrastructure resilience, ecological resilience, and system resilience. Other scholars constructed evaluation index systems based on the core characteristics of resilience, which were the ability to resist and recover from external shocks(Jiao et al., 2023; X. Wang, Wang, & Shi, 2023). The most popular approach for evaluating urban resilience is to give assessment indicators weights, and then use those weights to do both qualitative and quantitative research. Weight assignment methods include analytic hierarchy Process (AHP)(Z. Z. Liu, Ma, & Wang, 2022; Rezvani, de Almeida, Falcao, & Duarte, 2022), expert scoring method, entropy method(D. C. Tang, Li, Zhao, Boamah, & Lansana, 2023), etc. In contrast, the System Dynamics (SD) model possesses noteworthy strengths in modeling the time-dependent evolution of complex systems and enabling the comparison of different development scenarios. It is a useful tool to evaluate and simulate the development and change of urban resilience(Datola, Bottero, De Angelis, & Romagnoli, 2022; Feofilovs & Romagnoli, 2021), which can not only reveal the synergies and trade-offs among indicators, but also accomplish the dynamic assessment and forecasting of urban resilience in various contexts. (X. L. Tang & Chen, 2023). In order to better study the distribution of urban resilience in terms of space and time and analyze the spatial dependence and correlation of resilience in different regions, many scholars have introduced the classical exploratory spatial data analysis method (ESDA) of geography into urban resilience assessment research(Chacon-Hurtado, Kumar, Gkritza, Fricker, & Beaulieu, 2020; H. Wang, Liu, & Zhou, 2023). However, when dealing with high-dimensional data in intricate urban systems, these methods frequently fall short of addressing non-normal and nonlinear processing demands.

In summary, there is no unified standard for constructing an urban resilience evaluation indicator system, and simply utilizing a small number of indicators cannot achieve the goal of evaluating urban resilience. It is necessary to establish a complete evaluation indicator system. The Driving force-Pressure-State-Impact-Response (DPSIR) model was first suggested by the Organization for Economic Co-operation and Development (OECD) in 1993(OECD, 2003). It is an optimization and development of the PSR model. The biggest advantage of using the DPSIR model is that it can clarify the logical relationships among many interconnected indicators and view the interaction between humans, nature, society, economy, resources, and the environment from a systematic analysis perspective. It has comprehensive, scientific, and regional characteristics. Given this, this study will adopt the DPSIR model to examine and establish a comprehensive framework for evaluating urban resilience. In addition, traditional statistical analysis methods heavily rely on assumptions of distributions and models that may not hold in complex scenarios, while machine learning provides greater flexibility and robustness. Therefore, in recent years, artificial intelligence technologies represented by machine learning have been tried and applied to urban resilience evaluation research(X. L. Chen et al., 2023; Kutty, Wakjira, Kucukvar, Abdella, & Onat, 2022; Xia & Zhai, 2022), which helps to have a more detailed understanding of urban resilience. BP neural network is one of the most used neural network models nowadays. It is a fairly traditional machine learning approach(Holyoak, 1987), exhibiting novel potential in urban resilience evaluation research(H. Lu, Zhang, Jiao, Wei, & Zhang, 2022).

This study was based on the theory of urban resilience, combining DPSIR with BP neural networks to construct an urban resilience evaluation model, and using Hubei Province, a core province in central China, as the research object for practical analysis. The main contributions were as follows: (1) Based on DPSIR, a theoretical framework and assessment index system for urban resilience were developed,which facilitated the analysis of causal and constraint relationships between urban resilience and natural, social, economic, resource, and environmental factors, as well as the dissection of complex interactive processes. (2) Based on the DPSIR analysis framework, multi-source data such as statistical data and remote sensing images were integrated, and the BP neural network method was used to quantitatively evaluate urban resilience from multiple dimensions. (3) Based on Moran's I index, Further investigation was conducted into the temporal evolution trends and spatial variations in urban resilience levels among the various cities in Hubei Province. (4) Through a horizontal comparison of cities of different sizes in Hubei Province, the limiting patterns of urban resilience construction and key factors affecting urban resilience levels were identified from the perspective of obstacles, giving the government a resource to support urban development that is sustainable.

2.1 The DPSIR framework

The DPSIR theoretical framework has been widely applied in the evaluation of complex environmental systems due to its advantages such as comprehensiveness, integrity, systematicness, and flexibility(Ruan, Li, Zhang, & Liu, 2019; Tscherning, Helming, Krippner, Sieber, & Paloma, 2012). The driving force, pressure, status, impact, and response are the five aspects that make up the assessment system of a complex system. Every dimension is made up of several indications. Originating from sociological research, this framework is applied as an organizational indicator system architecture in the field of environmental and sustainable development research.

DPSIR is not a static or isolated description of research objects, but a framework of organizational information and related indexes based on causal relationships(Levrel, Kerbiriou, Couvet, & Weber, 2009). There is a causality chain of driving force → pressure → state → influence → response(Malmir, Javadi, Moridi, Neshat, & Razdar, 2021; Manservisi et al., 2023). It should be noted that response will eventually have an impact on other parts. "Driving force" refers to the underlying cause of environmental change. "Stress" refers to the influence exerted by human activities on their immediate surroundings as well as the natural environment, and is a direct effect of environmental stress factors. "State" refers to the condition of the environment under the above stress. "Impact" refers to the impact of the state of the system on human health and socio-economic structure in turn. "Response" is the countermeasures and active policies adopted by human beings in the process of promoting sustainable development. This theoretical framework, which can reveal the causal relationships between the environment and human activities, has been widely applied in various fields such as water resources(Ntona, Busico, Mastrocicco, & Kazakis, 2023; Qin, Hu, & Xia, 2023), soil and its environment(Qu et al., 2020), agriculture(Moss, Evans, & Atkins, 2021), fisheries(Mangi, Roberts, & Rodwell, 2007), biology(Delgado et al., 2021; Zhao, Fang, Liu, & Liu, 2021), coastal zones, and marine resources for sustainable development evaluation(Borongan & NaRanong, 2022; Federigi et al., 2022) and ecological security assessment. It is also used in environmental management science decision-making and implementation, including geographic information systems and multi-criteria decision support systems.

2.2 Urban Resilience Mechanism Based on DPSIR Framework

Based on the DPSIR framework, conducting a resilience evaluation of the complex urban system can provide a more comprehensive and systematic analysis of the interactions between humans and nature, society, economy, resources, and the environment. Figure 1 illustrates the urban resilience mechanism based on the DPSIR framework.

According to the theoretical framework, the "driving force" in urban resilience encompasses sustainable urban development and the evolution of the ecological environment, primarily composed of social and economic driving forces. This driving force exerts a dual influence on urban resilience. On the one hand, economic growth and urbanization provide a fundamental guarantee for urban development and resilience against risks. On the other hand, they also lead to resource consumption and environmental degradation, thereby limiting urban resilience. "Pressure" is imposed on the urban system by the driving force, leading to the depletion of resources, environmental degradation, and affecting the inhabitants of the city. As cities attract a significant influx of people, housing and employment become the primary concerns. To alleviate the pressure on the living conditions of more people, it is necessary to construct new infrastructure and factories. However, this further aggravates land occupation, leading to increased waste emissions, energy consumption, and environmental problems. This "pressure" not only hinders the improvement of the city's resilience level but also burdens its "state". The current stage of the city's development as a result of the combined influence of "pressure" and "driving force" is referred to as the "state". The availability of resources and the environment in which people live both reflect this situation. In addition to reflecting the "driving force" and acting as a feedback mechanism for the "pressure," the "state" also points the way for the development of urban resilience. The "impact" refers to how the social and economic structure evolves and changes, and whether the ecological environment remains intact, under the combined influence of these three elements. The "response" represents the process by which the government promotes sustainable urban development through the implementation of environmental protection measures and the safeguarding of people's welfare, adjusting the driving forces and mitigating the pressures to achieve a harmonious urban environment.

2.3 Indicators System for Urban Resilience Evaluation

Based on the DPSIR model, this study comprehensively and systematically selects 34 suitable indicators around the five aspects of "driving force-pressure-state-impact-response" to finally construct an urban resilience evaluation indicator system, as shown in Table 1. The indicators can be divided into positive indicators and negative indicators.

Table 1

Indicators System for Urban Resilience Evaluation
Target level	Criterion level	Indication level	Description	Unit	Attribute	Reference	Serial number
Urban Resilience	Driving force	Gross Domestic Product	Regional economic aggregate	10,000 yuan	+	(Xia & Zhai, 2022; Zhao et al., 2021)	D1
		Gross Domestic Product per capita	Economic development level	yuan	+	(Xia & Zhai, 2022; Zhao et al., 2021)	D2
		Growth Rate of Regional Gross Domestic Product	The expansion of the financial sphere	%	+	(Shi et al., 2021)	D3
		Local General Public Budget Revenue	Financial guarantee of government	10000yuan	+	(Shi et al., 2021)	D4
		Annual Per Capita Disposable Income of Rural Residents	Living standards of rural residents	yuan	+	(Cao, Xu, Zhang, & Kong, 2023)	D5
		Annual per capita disposable income of urban residents	Living standards of urban residents	yuan	+	(Cao et al., 2023)	D6
		Urbanization Rate	Urbanization level	%	+	(Zhao et al., 2021)	D7
	Pressure	Volume of Sulphur Dioxide Emission	Pressure from industrial pollutant emissions	ton	-	(Jiao et al., 2023; Xia & Zhai, 2022)	P1
		Annual Mean Concentration of PM2.5	Pressure from air pollutant emissions	ug/m³	-	(Jiao et al., 2023; Xia & Zhai, 2022)	P2
		Average Wage of Employed Staff and Workers in Urban Non-Private Units	Employee's wage income	yuan	+	(Shi et al., 2021; H. Wang et al., 2023; Zhao et al., 2021)	P3
		Engel's Coefficient for Urban Residents	The quality of life for people in cities	%	-	(H. Lu et al., 2022; H. Wang et al., 2023)	P4
		Engel's Coefficient for Rural Residents	The quality of life for people in rural areas	%	-	(H. Lu et al., 2022; H. Wang et al., 2023)	P5
		Population Density	The population's pressure	person/km²	-	(X. Wang et al., 2023)	P6
		Annual Electricity Consumption	Resource consumption pressure	100 MWh	-	(Xia & Zhai, 2022)	P7
		Per Capita Daily Domestic Water Consumption	Pressure from water consumption	L	-	(H. Lu et al., 2022; H. Wang et al., 2023; Xia & Zhai, 2022)	P8
	State	Built-up Area	Area of urban construction land	km²	+	(Xia & Zhai, 2022)	S1
		Total Water Resources	The current status of resource storage	108cu.m	+	(Jiao et al., 2023)	S2
		Normalized Difference Vegetation Index	The amount and state of vegetation		+	(W. Liu, Zhou, Li, Zheng, & Liu, 2024)	S3
		Area of Green Land	The current status of green space area	hectare	+	(Cao et al., 2023)	S4
		Green Covered Area as % of Completed Area	Urban ecological level	%	+	(Jiao et al., 2023; H. Lu et al., 2022; Xia & Zhai, 2022)	S5
		Public Recreational Green Space Per Capita	The quality of living environment	m²	+	(Jiao et al., 2023; H. Lu et al., 2022; Xia & Zhai, 2022)	S6
	Impact	Primary Industry as Percentage of GRP	The extent to which the production level is lagging behind	%	-	(H. Wang et al., 2023)	I1
		Secondary Industry as Percentage of GRP	Industry level	%	+	(H. Wang et al., 2023)	I2
		Tertiary Industry as Percentage of GRP	Advanced level	%	+	(Xia & Zhai, 2022)	I3
		Energy Consumption per Unit of GDP	Energy utilization rate	%	-	(Zhao et al., 2021)	I4
		Number of Days with Good Air Quality	Urban air quality	day	+	(Hou et al., 2023)	I5
	Response	Ratio of Waste Water Centralized Treated of Sewage Work	Urban sewage treatment capacity	%	+	(H. Wang et al., 2023; Xia & Zhai, 2022)	R1
		Rate of Domestic Garbage Harmless Treatment	Waste management capacity	%	+	(H. Wang et al., 2023; Xia & Zhai, 2022)	R2
		Fixed Asset Investment in Municipal Utilities Construction	How responsive the government is to the realization of urban functions	10000yuan	+	(Hou et al., 2023; Ruan et al., 2019)	R3
		Road Sweeping and Cleaning Area	The overall hygiene level of a city	10000m²	+	(H. Lu et al., 2022)	R4
		Number of Hospitals	Urban medical rescue capacity	unit	+	(Jiao et al., 2023; Xia & Zhai, 2022)	R5
		Number of Employees Joining Urban Basic Pension Insurance	Urban pension welfare	person	+	(Z. Z. Liu et al., 2022)	R6
		Number of Employees Joining Basic Medical Care System	Urban medical welfare	person	+	(Z. Z. Liu et al., 2022; Shi et al., 2021)	R7
		Persons Covered by Unemployment Insurance	Urban unemployment relief capacity	person	+	(Shi et al., 2021)	R8

Based on the urban resilience theory and artificial intelligence technology, this study established an urban resilience evaluation research framework, as shown in Fig. 2.

First, the resilience of the five subsystems of urban driving force, pressure, state, impact, and response was determined using the BP neural network and the DPSIR evaluation index system. Second, urban resilience was obtained by combining the entropy method and the BP neural network calculation results. Then, based on Moran's I index, the spatial differences and temporal evolution trends of urban resilience levels among different cities in Hubei Province were further explored. Finally, the obstacle degree model was utilized to explore the barriers to improving urban resilience, based on the thorough elaboration of each city's resilience level in the study area.

3.1 Entropy method

The entropy method originates from the concept of physics(Maisseu & Voss, 1995), The system's degree of disorder can be gauged using the entropy value. Formulas (1) through (6) illustrate the precise technique of calculation.

Normalizing the raw data is the first step.

$$\:\begin{array}{c}{Positive\:indicator:x}_{ij}=({X}_{ij}-min{X}_{ij})/(maxX\_ij-minX\_ij)\#\left(1\right)\end{array}$$

$$\:\begin{array}{c}Negative\:indicator:{x}_{ij}=(max{X}_{ij}-{X}_{ij)}/(max{X}_{ij}-min{X}_{ij})\#(2)\end{array}$$

The standardized data of indicator j for city i is denoted by x_ij (i = 1,2,3,…,m; j = 1,2,3,…,n), and X_ij represents the original data. maxX_ij and minX_ij are the maximum and minimum values of the original data for indicator j, respectively.

The second step is to calculate the proportion of indicator j for city i, denoted as p_ij.

$$\:\begin{array}{c}{p}_{ij}={x}_{ij}/\sum\:_{i=1}^{n}{x}_{ij}\#\left(3\right)\end{array}$$

The third step involves calculating the entropy of indicator j, which is denoted as e_j.

$$\:\begin{array}{c}{e}_{j}=-1/\text{ln}n\sum\:_{i=1}^{n}{p}_{ij}\text{ln}{p}_{ij}\#\left(4\right)\end{array}$$

The fourth step is to calculate the coefficient of variation for indicator j, denoted as dj.

$$\:\begin{array}{c}{d}_{j}=1-{e}_{j}\#\left(5\right)\end{array}$$

The fifth step is to calculate the weight of indicator j, denoted as w_j.

$$\:\begin{array}{c}{w}_{j}={d}_{j}/\sum\:_{j=1}^{m}{d}_{j}\#\left(6\right)\end{array}$$

3.2 BP neural network

The BP (Back Propagation) neural network is a classic artificial intelligence learning algorithm that consists of an input layer, a hidden layer(Holyoak, 1987), and an output layer in a multi-layer feedforward network. Its main characteristic is the backward propagation of errors. Currently, there are various evaluation methods for urban resilience, and research on the application of artificial intelligence in this field is still limited. However, there is no unified standard for the optimal range of each indicator. Therefore, based on related studies(Jiao et al., 2023; H. Lu et al., 2022), this research adopts the method of direct interpolation to obtain training samples. The specific operation steps are as follows:

Step 1 Initialization of BP neural network.

In this study, a toughness evaluation system was constructed, consisting of five subsystems: driving force, pressure, state, influence, and response. First, each layer's node count has to be ascertained. The output layer has a single node that represents the subsystem's final toughness evaluation result; the number of nodes in the input layer corresponds to the number of indications for each subsystem. The basic learning rate for this study is set at 0.1, with an error tolerance of 0.001.

Step 2 Generation of Learning Samples

The standardized indicators are evenly divided into four levels between their minimum and maximum values. Assuming that after standardization, the maximum value of an indicator is 1 and the minimum value is 0, the levels are as follows: Level I for values between 0 and 0.25, Level II for values between 0.25 and 0.5, Level III for values between 0.5 and 0.75, and Level IV for values between 0.75 and 1. A total of 599 sets of data were evenly distributed within each level. Out of them, 20% were identified as test samples and the remaining 80% were chosen as training samples. There were ten repetitions of the training procedure.

Step 3 Calculate the actual output and the output of hidden layer neurons based on the input samples.

Step 4 Calculate the difference between the actual output and the expected output, and determine the errors in the output layer and the hidden layer.

Step 5 Based on the errors obtained from Step 4, the gradient descent approach is used to update the connection weights between the nodes in the input layer and the hidden layer as well as between the hidden layer and the output layer.

Step 6 Calculate the error function E and determine whether E has converged within the given learning accuracy (E ≤ proposed error). If it satisfies the condition, the learning is complete, and the resilience evaluation results for the five subsystems are obtained. Otherwise, proceed to Step 2 for further learning.

3.3 Moran's I

Spatial autocorrelation analysis is a method that studies whether there is a significant correlation between the attribute values of a certain element and the attribute values of its neighboring spatial elements in a geospatial context(Y. Chen, 2021; Hu, 2020). The most commonly used measure for evaluating association is Moran's I. The following is the calculating formula:

$$\:\begin{array}{c}I=\left(n\sum\:_{i=1}^{n}\sum\:_{t=1}^{n}{q}_{it}\left({r}_{i}-\stackrel{-}{r}\right)\left({r}_{t}-\stackrel{-}{r}\right)\right)/\left(\sum\:_{i=1}^{n}\sum\:_{t=1}^{n}{q}_{it}\sum\:_{i=1}^{n}{\left({r}_{i}-\stackrel{-}{r}\right)}^{2}\right)\#\left(7\right)\end{array}$$

r_i and r_t represent the resilience evaluation results for cities i and t, respectively, while $\:\stackrel{-}{r}$ denotes the average of all resilience evaluation results for the cities. q_it serves as a spatial weight matrix, with a value of 1 indicating spatial adjacency and 0 indicating non-adjacency.

3.4 Obstacle diagnosis model

To identify the obstacles that contribute to urban resilience, this study presents an obstacle degree model. To determine the obstruction degree of indicators, it usually uses the contribution degree and deviation degree.

$$\:\begin{array}{c}{M}_{ij}=({w}_{j}\times\:（1-{x}_{ij}）)/\left(\sum\:_{j=1}^{n}{w}_{j}\times\:（1-{x}_{ij}）\right)\#\left(8\right)\end{array}$$

M_ij refers to the obstacle degree of indicator j in city j, and the weight w_j of indicator j characterizes the contribution degree of this indicator.

4.1 Study area

Hubei Province is located in the central region of China and serves as an important hub linking the country's western inland parts and eastern coastline regions. It is also the most important province in the Yangtze River Basin, boasting geographical advantages, economic strength, cultural heritage, and abundant resources. In 2023, the gross domestic product of Hubei Province reached 5.580363 trillion yuan, maintaining its position as the 7th largest in the country. Hubei Province mainly administers 12 prefecture-level cities, as well as 1 autonomous prefecture and 4 directly-administered county-level administrative regions under the province.

4.2 Data sources

This study collected data from 12 prefecture-level cities in Hubei Province in 2015, 2018, and 2021. Earth Science Cloud Platform provided the remote sensing data. Monthly Normalized Difference Vegetation Index (NDVI) products were downloaded from this website and then the annual average NDVI was calculated to represent the growth status of regional vegetation. The China Urban Construction Statistical Yearbook, the China City Statistical Yearbook, local statistical yearbooks, and the statistical bulletins on the national economic and social development of each city were the sources of the socio-economic data. For individual missing data, we interpolated and supplemented them based on adjacent years.

4.3 Results of Urban Resilience Evaluation

Using the entropy method, the objective weights of the urban resilience evaluation indicators for Hubei Province were calculated. By summing the weights of indicators under each subsystem, the weight value for that subsystem was obtained (Table 2). After standardizing the sample data for Hubei Province in 2015, 2018, and 2021, the impact levels were set using direct interpolation to construct training data for the BP neural network. Finally, the resilience evaluation values for the five subsystems of each city in Hubei Province in 2015, 2018, and 2021 were obtained (Table 3, Fig. 3). The urban resilience evaluation values for each city were then calculated using the following formula:

$$\:\begin{array}{c}{R}_{i}=\sum\:_{i=1}^{5}{r}_{ij}\times\:{w}_{ij}\#\left(9\right)\end{array}$$

R_i represents the comprehensive resilience values of the city i, r_ij represents the urban resilience evaluation values of subsystem j in city i (j = 1,2,3,4,5), and w_ij represents the weight of subsystem j in city i.

Table 2

Weights of Evaluation Indicators for Urban Resilience in Hubei Province
Subsystem	Indicator Serial number	Weight
Driving force (0.213)	D1	0.058
	D2	0.022
	D3	0.013
	D4	0.078
	D5	0.010
	D6	0.020
	D7	0.012
Pressure (0.104)	P1	0.007
	P2	0.014
	P3	0.029
	P4	0.013
	P5	0.015
	P6	0.011
	P7	0.004
	P8	0.010
State (0.181)	S1	0.065
	S2	0.020
	S3	0.014
	S4	0.062
	S5	0.006
	S6	0.015
Impact (0.056)	I1	0.010
	I2	0.011
	I3	0.020
	I4	0.005
	I5	0.011
Response (0.446)	R1	0.004
	R2	0.003
	R3	0.129
	R4	0.094
	R5	0.037
	R6	0.046
	R7	0.068
	R8	0.064

Table 3

Evaluation Results of Urban Resilience in Hubei Province
Resilience Dimension	year	Wu han	Huang shi	Shi yan	Yi chang	Xiang yang	E zhou	Jing men	Xiao gan	Jing zhou	Huang gang	Xian ning	Sui zhou
Driving force	2021	3.81	1.58	1.19	2.22	1.97	1.75	1.50	1.45	1.32	1.04	1.29	1.20
	2018	3.25	1.06	0.67	1.40	1.31	1.25	1.09	0.84	0.82	0.50	0.80	0.69
	2015	2.48	0.63	0.45	1.12	0.99	0.86	0.74	0.56	0.53	0.28	0.45	0.43
Pressure	2021	2.24	2.45	2.73	2.64	2.82	2.44	2.72	2.18	2.31	1.99	2.57	2.88
	2018	1.99	2.85	2.39	2.37	2.40	1.88	2.40	2.02	2.08	2.11	2.55	2.56
	2015	1.35	2.40	2.05	1.99	1.76	1.45	2.04	1.37	1.35	1.65	2.18	2.29
State	2021	1.42	0.63	1.13	1.20	1.08	0.52	0.47	0.36	0.67	0.92	1.04	0.49
	2018	0.68	0.48	0.68	0.78	0.57	0.36	0.42	0.01	0.26	0.32	0.76	0.12
	2015	0.64	0.01	0.21	0.92	0.22	0.01	0.24	0.01	0.21	0.30	0.58	0.10
Impact	2021	2.79	2.51	2.51	2.31	2.41	2.36	1.53	1.92	1.70	1.49	2.28	2.02
	2018	2.84	2.47	2.55	2.28	2.15	2.31	2.05	1.90	1.50	1.61	2.03	1.45
	2015	2.62	2.52	2.36	2.26	1.99	2.04	1.90	1.55	1.21	1.21	1.63	1.78
Response	2021	3.82	1.29	1.30	1.52	1.54	1.19	1.30	1.36	1.44	1.09	1.19	1.17
	2018	3.53	1.16	1.23	1.46	1.39	1.04	1.22	1.23	1.35	0.90	1.01	1.12
	2015	3.05	1.17	1.04	1.36	1.14	1.44	0.88	1.13	0.78	0.57	1.12	1.03
Urban Resilience	2021	3.16	1.42	1.46	1.77	1.73	1.39	1.35	1.31	1.38	1.17	1.39	1.28
	2018	2.76	1.27	1.20	1.47	1.37	1.12	1.21	1.05	1.13	0.88	1.14	1.01
	2015	2.29	1.05	0.94	1.34	1.05	1.09	0.91	0.85	0.71	0.61	1.02	0.91

4.3.1 Driving Force Analysis

Tables 3 and Fig. 3 clearly present the results of the driving force evaluation for the 12 prefecture-level cities in Hubei Province. From 2015 to 2021, it is evident that all cities have maintained a sustained growth in driving force, exhibiting similar rates of increase. However, Wuhan, as the capital city of Hubei Province, significantly surpasses other cities in terms of driving force evaluation results, creating a noticeable gap. Wuhan's economic development level is high and leading, with its regional gross domestic product being several times higher than other cities, and its urbanization rate reaching 84.56% in 2021. This provides a solid economic and social driving force for the development of urban resilience in Wuhan. Ranked second and third are Yichang and Xiangyang, which, as two sub-central cities in Hubei Province, have also achieved significant improvements in economic development, urban functions, and regional influence. However, we also note that cities such as Suizhou and Huanggang have consistently ranked lower in the driving force rankings for several consecutive years, with Huanggang consistently at the bottom. These cities are experiencing challenges related to a lack of impetus for the development of urban resilience since they have relatively low levels of economic growth and need to enhance their rates of urbanization.

4.3.2 Pressure Analysis

After a detailed analysis of stress evaluation results, we observe that from 2015 to 2018, the 12 cities made significant progress in their ability to withstand various economic pressures, resource consumption, ecological and environmental degradation, and other issues. However, from 2018 to 2021, the rate of improvement in stress resistance slowed down in most cities, and there was even a regression in Huangshi and Huanggang. In terms of driving force, Wuhan has demonstrated significant advantages. However, in the stress evaluation, the performance of this city has been poor. Although the evaluation results have been steadily improving, Wuhan has consistently been among the bottom three cities in the rankings. This disadvantage is not unique to Wuhan, as its neighboring city of Xiaogan also exhibits a similar situation. Rapid urbanization has led to resource consumption and environmental degradation, and a large influx of population to Wuhan, coupled with the high cost of living in the capital city, has increased urban stress and become an obstacle to improving resilience.

On the other hand, cities such as Suizhou, Xianning, and Shiyan have relatively weak driving forces for urban resilience development, but their stress resistance capabilities have consistently remained at a high level. This is closely related to the local ecological environment protection and low population density. The lower population density also reduces the complexity and stress of cities when responding to emergencies, making these cities more flexible and efficient in dealing with pressures from various sources.

4.3.3 State Analysis

In the state evaluation results of the 12 prefecture-level cities in Hubei Province, it is clearly observed that the state evaluation values of Huangshi, Shiyan, Ezhou, Xiangyang, and other cities increased significantly from 2015 to 2018, while most other regions either remained unchanged or worsened. From 2018 to 2021, the states of almost all cities have seen significant improvements, with Wuhan achieving the highest state rating in the province in 2021. Yichang, as an important sub-central city, has also performed exceptionally well, consistently ranking among the top two. However, the cities of Xiaogan and Suizhou have relatively lagged behind economically, consistently ranking among the lowest in the province. Both cities are geographically adjacent to Wuhan, but this geographical advantage has not been translated into development momentum. Despite being close to the capital city, they have not fully benefited from Wuhan's radiation effect and resource advantages, which has limited their development potential to some extent.

4.3.4 Impact Analysis

In the evaluation results of the impact dimension, there was no significant "provincial capital city dominating" discontinuity observed. The evaluation scores of the 12 cities were evenly distributed across various ranges, and most regions remained relatively stable from 2015 to 2021, with exceptions for a few cities. Wuhan has consistently maintained its first-place position, benefiting from its sound economic structure and developed tertiary industry. Shiyan and Huangshi followed closely behind due to their excellent performance in the quantity of days with good air quality. There have been no appreciable shifts in the provincial rankings for Huanggang and Jingzhou, which have continuously been ranked lowest. In contrast, Suizhou exhibited significant instability, showing clear fluctuations. Jingmen also experienced a significant decline in 2021, falling from its original intermediate position to the second-to-last place.

4.3.5 Response Analysis

The evaluation results of the response dimension exhibit the same characteristics as those of the driving force, with Wuhan far outperforming other cities in terms of evaluation scores, resulting in a significant gap. By contrast, the differences among other cities are relatively minor. The government's management level plays a major role in determining how quickly cities in Hubei can respond and recover from risks and pressures including natural catastrophes, abrupt public health events, climate change, and financial crises. Factors such as government decision-making efficiency, emergency response mechanisms, and post-disaster recovery strategies directly impact the stability and resilience of cities during crises. On the other hand, during normal, stable times, if the social security system for residents, including pension security, unemployment assistance, and medical insurance, is properly maintained, and if the city's hygiene and cleanliness are well-kept, along with sufficient hospital capacity and service quality, these measures will contribute to enhancing the city's response level. As the capital city, Wuhan enjoys unique political advantages, resulting in the continuous concentration of provincial resources and welfare. This advantage is reflected not only in efficient governance, policy favoritism, and financial support, but also in the accumulation of medical talents, infrastructure construction, and the improvement of public services. During the COVID-19 pandemic, the Wuhan Municipal Government demonstrated excellent response capabilities and efficient recovery, benefiting from its past investments and construction in social security and public services. This experience fully demonstrates Wuhan's strong strength in responding to risks and challenges. It also further emphasizes the important role of the government in enhancing the city's response level.

As secondary central cities in Hubei Province, Yichang and Xiangyang are continuously improving their urban response capabilities, but compared to other regions, their advantages are not yet apparent, and they have not fully leveraged their role as supporting and leading cities. Notably, Huanggang City has consistently ranked at the bottom in the province's response dimension, exhibiting a certain gap compared to other cities. This raises concerns about Huanggang's ability to respond promptly and take effective measures when facing risks.

4.3.6 Urban Resilience Analysis

Based on the evaluation of the five subsystems, including driving force, pressure, state, impact, and response, this study has obtained urban resilience evaluation scores for the 12 prefecture-level cities in Hubei Province. Overall, from 2015 to 2021, the resilience level of cities has been steadily improving, exhibiting a positive development trend. However, the gap in resilience among cities has gradually expanded, and the issue of imbalance with the capital city dominating is prominent. Hubei Province is still dominated by cities with medium and low resilience.

From 2015 to 2021, the gap between Wuhan and Huanggang, which is at the bottom of the resilience rankings, has increased from 1.68 to 1.99, showing a clear upward trend. Although the two secondary central cities of Yichang and Xiangyang are continuously improving their resilience levels, their advantages are not yet apparent, and the gap has not significantly widened. The uneven development of urban resilience among cities in Hubei Province is closely related to the dominance of the driving force and response subsystems. Various factors such as economic development level, urbanization degree, resident welfare, and municipal management level have jointly influenced the progress of resilience construction among cities in Hubei Province, thus contributing to the gap in resilience levels among cities to some extent.

On the other hand, it is noteworthy that the evaluation scores for the impact dimension are very low, with a weight value of only 0.056. This indicates the existence of shortcomings in economic structure and living environment, and when cities are impacted, the coordinated development pattern among the primary, secondary, and tertiary industries can easily be disrupted.

5.1 Discussion on the characteristics of temporal and spatial evolution

To further explore the temporal evolution trends and geographical variations in urban resilience levels across the various cities in Hubei Province, and their temporal evolution trends, this study calculated the Moran's I index of the urban resilience results of each city in 2015, 2018, and 2021, and produced scatter plots (Fig. 4). At the same time, ArcGIS software was used to map the urban resilience of major cities in Hubei Province (Fig. 5). Based on the natural grouping intervals of urban resilience values in terms of attributes or space, the resilience levels of 12 cities were subdivided into four levels, thus visually demonstrating the differences in resilience levels among cities. This series of analyses not only helps us understand the geographical distribution characteristics of urban resilience but also provides targeted guidance for policymakers to promote balanced urban resilience improvement and sustainable development in Hubei Province.

The urban resilience Moran's I indices of the 12 prefecture-level cities in Hubei Province in 2015, 2018, and 2021 were − 0.204, -0.246, and − 0.238, respectively, all of which are less than 0. This indicates that there is a spatial negative correlation in the urban resilience among cities in Hubei Province. This means that cities with high resilience and cities with low resilience tend to cluster spatially, forming distinct spatial heterogeneity. Further observation shows that from 2015 to 2018, the absolute value of Moran's I index showed an upward trend, indicating that the spatial negative correlation among cities was gradually increasing during this period. Although the index value decreased in 2021, the absolute value remained high, demonstrating that there was still a sizable resilience difference between neighboring cities. These findings are significant for understanding the spatial distribution and dynamic changes of urban resilience in Hubei Province. This unevenness can also be seen on the urban resilience map of major cities in Hubei Province. From 2015 to 2021, Wuhan, as the capital of Hubei Province, located in the east, has maintained a high level of urban resilience, belonging to the fourth level, demonstrating its strong ability to respond to external shocks and pressures. However, this dominant position is not universally reflected in other cities. This shows that Wuhan has an obvious siphon effect, forming a strong spatial difference from other cities. To promote balanced regional development, it is necessary to further strengthen the cooperation and exchanges between Wuhan and its neighboring cities to realize resource sharing and complementary advantages.

In 2015, Yichang, Xiangyang, Ezhou, Xianning, and Huangshi reached Level II resilience, demonstrating relatively high urban resilience. However, Hubei Province has always emphasized Wuhan's core status as a "leading city" and is committed to strengthening its leading role. At the same time, to promote balanced economic development across the province, policies have also focused on promoting economic growth, urban functional improvement, and regional influence expansion in Xiangyang and Yichang, to leverage their "two-wing drive" role and jointly build a regional innovation center. Over time, especially after the Hubei Provincial Government increased its investment and attention to the secondary central cities of Yichang and Xiangyang, the economic and social driving forces of these two cities have continued to improve, and the gap in resilience levels between them and other cities has gradually widened. By 2018 and 2021, only Yichang and Xiangyang maintained their Level II resilience, while most other cities were mainly concentrated on Level III resilience. There has not been a large agglomeration center created in Hubei Province as a result of the two secondary central cities' location in the western portion of the province, distant from Wuhan, the provincial capital. The geographically dispersed leading cities cannot exert spillover effects to lead other cities to make progress together.

The uneven resilience phenomenon among cities in Hubei Province not only reveals the different abilities of cities in responding to challenges and crises but also reflects significant differences in resource allocation, policy support, and urban development strategies among cities. To promote balanced urban development and enhance overall resilience in the province, policymakers need to pay more attention to cities with lower resilience levels. By optimizing resource allocation, strengthening regional cooperation, and promoting innovation-driven development, the gap in resilience among cities can be narrowed.

5.2 Analysis of Urban Resilience Obstacles

To further explore the most significant determinants limiting the resilience levels of Hubei Province's twelve prefecture-level cities, this study introduced the obstacle degree model to identify the limiting indicators and their obstacle degrees of urban resilience levels. The total obstacle degree of the indications that each of the five subsystems contains determines its obstacle degree, and the outcomes are ordered.

From the perspective of the five subsystems, from 2015 to 2021, except for Wuhan, the ranking of the degree of constraint on the resilience levels of all other cities by the five subsystems remained unchanged, with response > driving force > status > pressure > impact. In Wuhan, the response and status subsystems were primarily responsible for the city's low level of resilience in 2015 and 2018, and in 2021, pressure and response primarily limited it. This change provides us with an opportunity for further research. The "S" curve theory proposed by American urban geographer Northam in his book "Urban Geography" in 1979 divides the process of urbanization into three stages: the initial stage (when the urbanization rate is less than 30%), the accelerated stage (when the urbanization rate is higher than 30% but lower than 70%), and the stable development stage (when the urbanization rate is higher than 70%)(M.Northam, 1979). During the second stage, while the city expands outward, it also begins to develop internally(Yang, Liu, & Zhang, 2017). At the same time, "megalopolis" and "megacity" emerged in some regions. Among the 12 prefecture-level cities in Hubei Province, only Wuhan has an urbanization rate exceeding 70%, which means that other cities are still in the second stage of urbanization development, characterized by "accelerated development." A crucial area of study throughout the urbanization process is urban resilience. The resilience levels of other cities are limited by the response and driving force dimensions. This is due to the significant impact of Chinese government policies on urban development, as well as the current focus on economic development in these regions. As a "megacity" in Hubei Province, Wuhan has entered the next stage of urbanization development, and economic and social driving forces are no longer limiting factors for urban resilience in the region. Instead, various pressures derived from the rapid economic and social development in the previous stage have become the main limiting factors. To address these issues, the role of the government remains crucial, which is reflected in the response subsystem.

From the perspective of individual evaluation indicators, this study calculates the cumulative obstacle scores for each of the 34 indicators for each city each year, ranks them, and obtains the top five indicators for each city in terms of obstacle degree (Table 4). The limiting factors for Wuhan's urban resilience are, in order: total water resources, fixed asset investment in municipal public facilities construction, road sweeping and cleaning area, normalized vegetation index, and green area. As the only "megacity" in Hubei Province that has entered the third stage of development, Wuhan often relied on resource consumption and environmental degradation as the cost of economic and social driving forces in the initial phases of enhancing urban resilience. However, with the deepening of urbanization, these sacrifices brought about by early development have now become the main factors limiting the further development of the city. The Wuhan Municipal Government must prioritize the sustainable use of resources and the preservation of the environment in the current stage of development. It must also increase fixed asset investment in the building of municipal public facilities and urban sanitation systems to support the development of urban resilience.

Table 4

Top five obstacle factors for prefecture-level cities in Hubei Province
	Wu han	Huang shi	Shi yan	Yi chang	Xiang yang	E zhou	Jing men	Xiao gan	Jing zhou	Huang gang	Xian ning	Sui zhou
Local General Public Budget Revenue		3	3	3	3	3	3	3	3	3	3	3
Built-up Area		5	5	5	6	5	5	5	4	5	5	6
Total Water Resources	1
Normalized Difference Vegetation Index	4
Area of Green Land	5
Fixed Asset Investment in Municipal Utilities Construction	2	1	1	1	1	1	1	1	1	1	1	1
Number of Employees Joining Basic Medical Care System		4	4	4	4	4	4	4	5	4	4	4
Road Sweeping and Cleaning Area	3	2	2	2	2	2	2	2	2	2	2	2
Persons Covered by Unemployment Insurance					5							5

The resilience limiting factors of other cities are, in turn, fixed asset investment in the construction of municipal public facilities, area of road cleaning, and local general public budget income. The number of employees covered by basic medical insurance is the primary concentration of the final two limiting criteria, built-up area, and unemployment insurance. For the other 11 cities, the economic impetus for urban resilience development is insufficient, limiting their ability to invest more resources in municipal construction and residents' welfare. The insufficient number of basic medical insurance participants leads to increased pressure on medical expenditure, forcing individuals to shoulder a heavier burden of medical expenses, which may bring a greater burden to people with financial difficulties and hinder the sustainable development of the medical security system. In the face of various crises, such as natural disasters, financial crises, and public health emergencies, the small number of people insured by unemployment insurance aggravates social and economic instability factors, hindering the city's stability and sustainable development. It is also worth noting that the insufficient built-up area has a limiting restricting effect on the growth of urban resilience. With the advancement of urbanization, the population is increasingly concentrated in urban areas, and economic activities are becoming more frequent. In order to accommodate the demands of population expansion and economic development, more built-up areas must be added to cities to increase their capacity for production and living. This will increase the degree of urban resilience. The expansion of built-up areas can also increase the construction of urban infrastructure such as roads, transportation, water supply, power supply, and communication, improve the comprehensive carrying capacity of the city, and provide strong support for the city's sustainable development.

Urban resilience research is not static but dynamic, emphasizing how urban systems demonstrate their strong adaptability and resilience in constantly changing and uncertain environments. Therefore, this study constructs a comprehensive evaluation system for urban resilience, grounded in the "Driving force-Pressure-State-Impact-Response (DPSIR)" framework. Quantitative assessments are conducted using BP neural networks, while spatial and temporal evolution trends as well as the obstacles to urban resilience in Hubei Province are also discussed. This research has made significant contributions to both academic and practical fields.

The empirical analysis results show that: (1) Urban resilience research involves the dynamic changes of driving forces, pressures, states, impacts, and responses. The "State" represents a relative equilibrium achieved through the combined influence of "driving forces" and "pressures". This equilibrium has a significant "impact" on society and the environment, motivating government managers to formulate "responses" aimed at enhancing the city's resilience to risks. The resilience level of cities in Hubei Province is mainly dominated by the two major factors of "driving forces" and "responses", which are significantly impacted by the degree of social and economic growth as well as the managerial skills of the administration. (2)From 2015 to 2021, the resilience level of cities has been steadily improving, exhibiting a positive development trend. However, the gap in resilience among cities has gradually expanded, and the issue of imbalance with the capital city dominating is prominent. Hubei Province is still dominated by cities with medium and low resilience. There is a spatial negative correlation between the urban resilience of cities, with cities tending to cluster spatially based on their high or low resilience. Wuhan exhibits an obvious siphon effect, forming strong spatial differences with other cities. (3) Wuhan has successfully entered the third phase of urbanization, in which economic and social drivers are no longer the main constraints on the resilience of the region's cities. On the contrary, various pressures brought about by rapid economic and social development in the previous stage, such as pressure on resources and the environment, have turned into major limiting factors. In contrast, other cities in Hubei Province are still in the second stage of urbanization. The ability of cities to tolerate different dangers is now mostly limited by the socio-economic development level and the management capabilities of the government. (4) The uneven urban resilience in Hubei Province not only highlights the differences in the ability of cities to cope with various challenges and crises, but also reflects their obvious differences in resource allocation, policy support, and urban development strategies. To promote balanced development and overall resilience in Hubei Province, policymakers need to focus more on cities with relatively low levels of resilience. By optimizing resource allocation, strengthening inter-regional coordination and cooperation, and providing policy support, it is expected to narrow the resilience gap between cities, and thereby promote the comprehensive, coordinated, and sustainable development of cities in the province.

Certainly, there are some constraints to this study. Firstly, it primarily centers on Hubei Province as the target of investigation, thus lacking a horizontal comparison with other provinces in China. Furthermore, the resilience evaluation conducted in this study is confined to 12 prefecture-level cities within Hubei Province. Counties-level cities should be included in future research projects to provide a more thorough assessment and comparison of resilience at different urban sizes. Second, considering the unique characteristics of China's social and economic environments, it is essential to modify the index selection in light of the unique data characteristics of other countries when implementing the urban resilience evaluation model developed in this work. This will still be a major area of interest for this study team's next investigations.

Funding

This work was supported by the National Key R&D Program of China (2022YFC3005704)

Author Contribution

Conceptualization, Y.L. ; methodology,Y.S.C. and Y.L.; validation, Y.S.C. and Y.L.; writing—original draft preparation, Y.L.; writing—review and editing, F.R. and Q.Y.D. ; All authors have read and agreed to the published version of the manuscript.

Data Availability

The raw data was sourced from the China Urban Construction Statistical Yearbook, the China City Statistical Yearbook, local statistical yearbooks, statistical bulletins, and the Earth Science Cloud Platform. The data are available from the corresponding author if there is a reasonable request.

Alberti, M., Marzluff, J. M., Shulenberger, E., Bradley, G., Ryan, C., & Zumbrunnen, C. (2003). Integrating humans into ecology: Opportunities and challenges for studying urban ecosystems. Bioscience, 53(12), 1169-1179. doi:10.1641/0006-3568(2003)053[1169:Ihieoa]2.0.Co;2
Amirzadeh, M., Sobhaninia, S., & Sharifi, A. (2022). Urban resilience: A vague or an evolutionary concept? Sustainable Cities and Society, 81. doi:10.1016/j.scs.2022.103853
Borongan, G., & NaRanong, A. (2022). Factors in enhancing environmental governance for marine plastic litter abatement in Manila, the Philippines: A combined structural equation modeling and DPSIR framework. Marine Pollution Bulletin, 181, 113920. doi:https://doi.org/10.1016/j.marpolbul.2022.113920
Cao, F., Xu, X., Zhang, C., & Kong, W. (2023). Evaluation of urban flood resilience and its Space-Time Evolution: A case study of Zhejiang Province, China. Ecological Indicators, 154, 110643. doi:https://doi.org/10.1016/j.ecolind.2023.110643
Chacon-Hurtado, D., Kumar, I., Gkritza, K., Fricker, J. D., & Beaulieu, L. J. (2020). The role of transportation accessibility in regional economic resilience. Journal of Transport Geography, 84. doi:10.1016/j.jtrangeo.2020.102695
Chelleri, L., Waters, J. J., Olazabal, M., & Minucci, G. (2015). Resilience trade-offs: addressing multiple scales and temporal aspects of urban resilience. Environment and Urbanization, 27(1), 181-198. doi:10.1177/0956247814550780
Chen, X. L., Yu, L. X., Lin, W. D., Yang, F. Q., Li, Y. P., Tao, J., & Cheng, S. (2023). Urban resilience assessment from the multidimensional perspective using dynamic Bayesian network: A case study of Fujian Province, China. Reliability Engineering & System Safety, 238. doi:10.1016/j.ress.2023.109469
Chen, Y. (2021). An analytical process of spatial autocorrelation functions based on Moran's index. PLOS ONE, 16(4). doi:10.1371/journal.pone.0249589
Datola, G., Bottero, M., De Angelis, E., & Romagnoli, F. (2022). Operationalising resilience: A methodological framework for assessing urban resilience through System Dynamics Model. Ecological Modelling, 465. doi:10.1016/j.ecolmodel.2021.109851
Delgado, L. E., Zúñiga, C. C., Asún, R. A., Castro-Díaz, R., Natenzon, C. E., Paredes, L. D., . . . Marín, V. H. (2021). Toward social-ecological coastal zone governance of Chiloé Island (Chile) based on the DPSIR framework. Science of the Total Environment, 758, 143999. doi:https://doi.org/10.1016/j.scitotenv.2020.143999
Elmqvist, T., Andersson, E., Frantzeskaki, N., McPhearson, T., Olsson, P., Gaffney, O., . . . Folke, C. (2019). Sustainability and resilience for transformation in the urban century. NATURE SUSTAINABILITY, 2(4), 267-273. doi:10.1038/s41893-019-0250-1
Federigi, I., Balestri, E., Castelli, A., De Battisti, D., Maltagliati, F., Menicagli, V., . . . Carducci, A. (2022). Beach pollution from marine litter: Analysis with the DPSIR framework (driver, pressure, state, impact, response) in Tuscany, Italy. Ecological Indicators, 143, 109395. doi:https://doi.org/10.1016/j.ecolind.2022.109395
Feofilovs, M., & Romagnoli, F. (2021). Dynamic assessment of urban resilience to natural hazards. International Journal of Disaster Risk Reduction, 62, 102328. doi:https://doi.org/10.1016/j.ijdrr.2021.102328
Holling, C. S. (1973). Resilience and stability of ecological systems. Annu. Rev. Ecol. Systemat, 4(4), 1-23.
Holyoak, K. J. (1987). PARALLEL DISTRIBUTED-PROCESSING - EXPLORATIONS IN THE MICROSTRUCTURE OF COGNITION - RUMELHART,DE, MCCLELLAND,JL. Science, 236(4804), 992-996. doi:10.1126/science.236.4804.992
Hou, X., Lv, T., Xu, J., Deng, X., Liu, F., Lam, J. S. L., . . . Han, X. (2023). Evaluation of urban public transport sustainability in China based on the Driving Force-Pressure-State-Impact-Response (DPSIR) framework——A case study of 36 major cities. Environmental Impact Assessment Review, 103, 107263. doi:https://doi.org/10.1016/j.eiar.2023.107263
Hu, L. Z. (2020). Negative Spatial Autocorrelation and Its Impacts on Georeferenced Data Analyses: With Case Studies of Cancer Incidences.
Hudec, O., Reggiani, A., & Siserová, M. (2018). Resilience capacity and vulnerability: A joint analysis with reference to Slovak urban districts. Cities, 73, 24-35. doi:10.1016/j.cities.2017.10.004
Jiao, L. D., Wang, L., Lu, H., Fan, Y. W., Zhang, Y., & Wu, Y. (2023). An assessment model for urban resilience based on the pressure-state-response framework and BP-GA neural network. Urban Climate, 49. doi:10.1016/j.uclim.2023.101543
Kutty, A. A., Wakjira, T. G., Kucukvar, M., Abdella, G. M., & Onat, N. C. (2022). Urban resilience and livability performance of European smart cities: A novel machine learning approach. Journal of Cleaner Production, 378. doi:10.1016/j.jclepro.2022.134203
Levrel, H., Kerbiriou, C., Couvet, D., & Weber, J. (2009). OECD pressure-state-response indicators for managing biodiversity: a realistic perspective for a French biosphere reserve. Biodiversity and Conservation, 18(7), 1719-1732. doi:10.1007/s10531-008-9507-0
Liao, K.-H. (2012). A Theory on Urban Resilience to Floods—A Basis for Alternative Planning Practices. Ecology and Society, 17(4). doi:10.5751/ES-05231-170448
Liu, W., Zhou, J., Li, X., Zheng, H., & Liu, Y. (2024). Urban resilience assessment and its spatial correlation from the multidimensional perspective: A case study of four provinces in North-South Seismic Belt, China. Sustainable Cities and Society, 101, 105109. doi:https://doi.org/10.1016/j.scs.2023.105109
Liu, Z. Z., Ma, R., & Wang, H. J. (2022). Assessing urban resilience to public health disaster using the rough analytic hierarchy process method: A regional study in China. Journal of Safety Science and Resilience, 3(2), 93-104. doi:10.1016/j.jnlssr.2021.12.003
Lu, H., Zhang, C., Jiao, L. D., Wei, Y., & Zhang, Y. (2022). Analysis on the spatial-temporal evolution of urban agglomeration resilience: A case study in Chengdu-Chongqing Urban Agglomeration, China. International Journal of Disaster Risk Reduction, 79. doi:10.1016/j.ijdrr.2022.103167
Lu, P., & Stead, D. (2013). Understanding the notion of resilience in spatial planning: A case study of Rotterdam, The Netherlands. Cities, 35, 200-212. doi:https://doi.org/10.1016/j.cities.2013.06.001
M.Northam, R. (1979). URBAN GEOGRAPHY. New York: JOHN WILEY & SONS.
Maisseu, A., & Voss, A. (1995). Energy, entropy and sustainable development. INTERNATIONAL JOURNAL OF GLOBAL ENERGY.
Malmir, M., Javadi, S., Moridi, A., Neshat, A., & Razdar, B. (2021). A new combined framework for sustainable development using the DPSIR approach and numerical modeling. Geoscience Frontiers, 12(4), 101169. doi:https://doi.org/10.1016/j.gsf.2021.101169
Mangi, S. C., Roberts, C. M., & Rodwell, L. D. (2007). Reef fisheries management in Kenya: Preliminary approach using the driver–pressure–state–impacts–response (DPSIR) scheme of indicators. Ocean & Coastal Management, 50(5), 463-480. doi:https://doi.org/10.1016/j.ocecoaman.2006.10.003
Manservisi, F., Banzi, M., Tonelli, T., Veronesi, P., Ricci, S., Distante, D., . . . Bortone, G. (2023). Environmental complaint insights through text mining based on the driver, pressure, state, impact, and response (DPSIR) framework: Evidence from an Italian environmental agency. Regional Sustainability, 4(3), 261-281. doi:https://doi.org/10.1016/j.regsus.2023.08.002
McPhearson, T., Andersson, E., Elmqvist, T., & Frantzeskaki, N. (2015). Resilience of and through urban ecosystem services. Ecosystem Services, 12, 152-156. doi:https://doi.org/10.1016/j.ecoser.2014.07.012
Meerow, S., Newell, J. P., & Stults, M. (2016). Defining urban resilience: A review. Landscape and Urban Planning, 147, 38-49. doi:https://doi.org/10.1016/j.landurbplan.2015.11.011
Moss, E. D., Evans, D. M., & Atkins, J. P. (2021). Investigating the impacts of climate change on ecosystem services in UK agro-ecosystems: An application of the DPSIR framework. LAND USE POLICY, 105, 105394. doi:https://doi.org/10.1016/j.landusepol.2021.105394
Mu, X. F., Fang, C. L., & Yang, Z. Q. (2022). Spatio-temporal evolution and dynamic simulation of the urban resilience of Beijing-Tianjin-Hebei urban agglomeration. Journal of Geographical Sciences, 32(9), 1766-1790. doi:10.1007/s11442-022-2022-5
Ntona, M. M., Busico, G., Mastrocicco, M., & Kazakis, N. (2023). Coupling SWAT and DPSIR models for groundwater management in Mediterranean catchments. Journal of Environmental Management, 344, 118543. doi:https://doi.org/10.1016/j.jenvman.2023.118543
OECD. (2003). Organisation for Economic Co-operation and Development (OECD) core set of indicators for environmental performance reviews. Retrieved from http://www.oecd.org/dataoecd/32/20/31558547.pdf
Qasim, S., Qasim, M., Shrestha, R. P., Khan, A. N., Tune, K., & Ashraf, M. (2016). Community resilience to flood hazards in Khyber Pukhthunkhwa province of Pakistan. International Journal of Disaster Risk Reduction, 18, 100-106. doi:10.1016/j.ijdrr.2016.03.009
Qin, X., Hu, X., & Xia, W. (2023). Investigating the dynamic decoupling relationship between regional social economy and lake water environment: The application of DPSIR-Extended Tapio decoupling model. Journal of Environmental Management, 345, 118926. doi:https://doi.org/10.1016/j.jenvman.2023.118926
Qu, S., Hu, S., Li, W., Wang, H., Zhang, C., & Li, Q. (2020). Interaction between urban land expansion and land use policy: An analysis using the DPSIR framework. LAND USE POLICY, 99, 104856. doi:https://doi.org/10.1016/j.landusepol.2020.104856
Rezvani, S., de Almeida, N. M., Falcao, M. J., & Duarte, M. (2022). Enhancing urban resilience evaluation systems through automated rational and consistent decision-making simulations. Sustainable Cities and Society, 78. doi:10.1016/j.scs.2021.103612
Ribeiro, P. J. G., & Gonçalves, L. (2019). Urban resilience: A conceptual framework. Sustainable Cities and Society, 50. doi:10.1016/j.scs.2019.101625
Romero-Lankao, P., & Gnatz, D. M. (2013). Exploring urban transformations in Latin America. Current Opinion in Environmental Sustainability, 5(3), 358-367. doi:https://doi.org/10.1016/j.cosust.2013.07.008
Ruan, W., Li, Y., Zhang, S., & Liu, C.-H. (2019). Evaluation and drive mechanism of tourism ecological security based on the DPSIR-DEA model. Tourism Management, 75, 609-625. doi:https://doi.org/10.1016/j.tourman.2019.06.021
Shamsipour, A., Jahanshahi, S., Mousavi, S. S., Shoja, F., Ansari Golenji, R., Tayebi, S., . . . Sharifi, A. (2024). Assessing and mapping urban ecological resilience using the loss-gain approach: A case study of Tehran, Iran. Sustainable Cities and Society, 103, 105252. doi:https://doi.org/10.1016/j.scs.2024.105252
Sharifi, A. (2019). Resilient urban forms: A macro-scale analysis. Cities, 85, 1-14. doi:10.1016/j.cities.2018.11.023
Sharifi, A., & Yamagata, Y. (2016). Principles and criteria for assessing urban energy resilience: A literature review. Renewable & Sustainable Energy Reviews, 60, 1654-1677. doi:10.1016/j.rser.2016.03.028
Shi, Y., Zhai, G., Xu, L., Zhou, S., Lu, Y., Liu, H., & Huang, W. (2021). Assessment methods of urban system resilience: From the perspective of complex adaptive system theory. Cities, 112, 103141. doi:https://doi.org/10.1016/j.cities.2021.103141
Spaans, M., & Waterhout, B. (2017). Building up resilience in cities worldwide – Rotterdam as participant in the 100 Resilient Cities Programme. Cities, 61, 109-116. doi:10.1016/j.cities.2016.05.011
Tang, D. C., Li, J. N., Zhao, Z. Q., Boamah, V., & Lansana, D. D. (2023). The influence of industrial structure transformation on urban resilience based on 110 prefecture-level cities in the Yangtze River. Sustainable Cities and Society, 96. doi:10.1016/j.scs.2023.104621
Tang, X. L., & Chen, N. (2023). A multi-scenario urban resilience simulation study based on the system dynamics model. Physics and Chemistry of the Earth, 132. doi:10.1016/j.pce.2023.103467
Tscherning, K., Helming, K., Krippner, B., Sieber, S., & Paloma, S. G. Y. (2012). Does research applying the DPSIR framework support decision making? LAND USE POLICY, 29(1), 102-110. doi:10.1016/j.landusepol.2011.05.009
UN-Habitat. (2022). World cities report 2022: envisaging the future of cities. Retrieved from https://unhabitat.org/world-cities-report-2022-envisaging-the-future-of-cities
Wang, H., Liu, Z., & Zhou, Y. (2023). Assessing urban resilience in China from the perspective of socioeconomic and ecological sustainability. Environmental Impact Assessment Review, 102. doi:10.1016/j.eiar.2023.107163
Wang, X., Wang, C. X., & Shi, J. L. (2023). Evaluation of urban resilience based on Service-Connectivity- Environment (SCE) model: A case study of Jinan city, China. International Journal of Disaster Risk Reduction, 95. doi:10.1016/j.ijdrr.2023.103828
Xia, C. H., & Zhai, G. F. (2022). The spatiotemporal evolution pattern of urban resilience in the Yangtze River Delta urban agglomeration based on TOPSIS-PSO-ELM. Sustainable Cities and Society, 87. doi:10.1016/j.scs.2022.104223
Yang, Y., Liu, J., & Zhang, Y. (2017). An analysis of the implications of China’s urbanization policy for economic growth and energy consumption. Journal of Cleaner Production, 161, 1251-1262. doi:https://doi.org/10.1016/j.jclepro.2017.03.207
Zhang, M. S., Yang, Y. G., Li, H. H., & van Dijk, M. P. (2020). Measuring Urban Resilience to Climate Change in Three Chinese Cities. Sustainability, 12(22). doi:10.3390/su12229735
Zhao, R., Fang, C., Liu, H., & Liu, X. (2021). Evaluating urban ecosystem resilience using the DPSIR framework and the ENA model: A case study of 35 cities in China. Sustainable Cities and Society, 72, 102997. doi:https://doi.org/10.1016/j.scs.2021.102997
Ziervogel, G., Pelling, M., Cartwright, A., Chu, E., Deshpande, T., Harris, L., . . . Zweig, P. (2017). Inserting rights and justice into urban resilience: a focus on everyday risk. Environment and Urbanization, 29(1), 123-138. doi:10.1177/0956247816686905

No competing interests reported.

Download PDF

Editorial decision: Revision requested
14 Nov, 2024
Reviews received at journal
13 Nov, 2024
Reviewers agreed at journal
02 Nov, 2024
Reviews received at journal
05 Oct, 2024
Reviewers agreed at journal
26 Sep, 2024
Reviewers invited by journal
25 Sep, 2024
Editor assigned by journal
25 Sep, 2024
Editor invited by journal
09 Sep, 2024
Submission checks completed at journal
06 Sep, 2024
First submitted to journal
26 Aug, 2024

You are reading this latest preprint version

Urban resilience evaluation based on the DRIVING FORCE-PRESSURE-STATE-IMPACT-RESPONSE (DPSIR) framework and BP NEURAL NETWORK: A case study of Hubei Province

Status:

Version 1

Abstract

Figures

1. Introduction

2. Construction of Urban Resilience Evaluation System

2.1 The DPSIR framework

2.2 Urban Resilience Mechanism Based on DPSIR Framework

2.3 Indicators System for Urban Resilience Evaluation

3. Methodology

3.1 Entropy method

3.2 BP neural network

3.3 Moran's I

3.4 Obstacle diagnosis model

4. Case study

4.1 Study area

4.2 Data sources

4.3 Results of Urban Resilience Evaluation

4.3.1 Driving Force Analysis

4.3.2 Pressure Analysis

4.3.3 State Analysis

4.3.4 Impact Analysis

4.3.5 Response Analysis

4.3.6 Urban Resilience Analysis

5. Discussion

5.1 Discussion on the characteristics of temporal and spatial evolution

5.2 Analysis of Urban Resilience Obstacles

6. Conclusion

Declarations

Funding

Author Contribution

Data Availability

References

Additional Declarations

Status:

Version 1