2.1 Urban Anti-Pandemic Resilience Evaluation System
This evaluation system includes resistance, recuperability, and adaptability to describe the urban resistance disturbed from factors outside the system, to recover quickly after being damaged, self-learning in the face of risks and applying it to handle it next time. This is particularly important in the case of COVID-19 pandemic because the society was affected by microorganisms with certain transmission capacity and spread in wide range of people. This system is divided into three parts: risk factors that promote the spread of the pandemic, repression factors that inhibit the spread of the pandemic, and adaptability factors that examines the ability of self-learning and application of cities. To select optimal indexes, we have collected data in three sectors, (1) socioeconomics, (2) management and governance and (3) urban design and transportation.
The socioeconomic indicators are population aggregation, age structure, population quality, medical staff salary, social governance, and economy. The connection between population density or aggregation with infectious are ambiguous, because most studies have different relationship such as Carteni et. al. (2020) found positive whereas Boterman et al. (2020) found negative relationship. In terms of medical staff salary, which play an important role in their job satisfaction (Zhou H. et all., 2018). According to Sharifi (2020) and Cattivelli V. (2020), social governance such as voluntary activities and social works have huge influence on eliminating inequalities during pandemic. When it comes to economic condition, population in poor condition often hard to obtain medical interference comparing population with good condition, this is more common in slums (Biswas, P.P., 2020). In addition, cities low in diverse economic structure are vulnerable during pandemic, especially cities with mining municipalities, with large care centers, and shrinking cities are the most vulnerable ones (Krzysztofik R. et all., 2022). The governance is always coming front while dealing with index related management (Chu Z. et al. 2021)
Indexes related transportation and urban design are population migration, access to medical resource and health care, and composition of medical workers. Population migration across cities and intra-city are often considered as the risk factors of pandemic (Vignier N. et al., 2018). The health care and composition of medical workers are the crucial component during the pandemic. The involvement of medical institutions and staff is always challenging because the distribution is often inequal though the WHO guidelines suggest to distribute by patients’ priority (Mannelli C., 2020). In this dimension, increasing the number of medical institutions in a reasonable distribution encourage the development of private hospitals, and improve the training of medical human resources to promote the resilience (Núñez A, et all., 2021).
In terms of quality, balanced coverage of basic medical services should be guaranteed for population general medical demand, especially for patients with chronic noncommunicable diseases (Reshetnikov, A., et all., 2023). For another, the development of emergency medicine in cities are important to handle patients with acute complications caused by Covid-19 and other acute diseases. In addition, the development of remote medicine is beneficial for quantity and quality.
This study collected required information of 30 different cities in China (Fig. 1). All the collected data were divided into 3 parts: risk factors, repression factors, and adaptability factors. These primary factors were evaluated by four secondary indexes. Some of secondary evaluation indexes were further classified into tertiary evaluation indexes (Fig. 2).
The tertiary evaluation indexes were aggregated into secondary evaluation indexes applying Analytic Hierarchy Process (Saaty, 2005). Then, the Criteria Importance Through Intercriteria Correlation (Diakoulaki et al., 1995) and Entropy Weight Method (Zhu, et al., 2020) was used to weighting 12 secondary evaluation indexes. Finally, these weighted indexes were evaluated by Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) method (Hwang & Yoon. 1981). The data were processed using EXCEL 2020 for data storage and management, ArcGIS 10.8 and R program language 4.3.1 for data analysis and visualization, and SPSS 26 for Weight Analysis and TOPSIS.
2.2 Indicator System
The primary evaluation indexes consist of risk factors, repression factors, and adaptability factors. Each factor further divided into 4 secondary evaluation indexes where some secondary evaluation indexes further divided into tertiary evaluation indexes (Fig. 3).
2.2.1 Risk Factors
Risk factors refer to factors that promote the spread of SARS-CoV-2 virus as well as Covid-19 pandemic, which include migration of people, population density and age composition. Migration of people includes population moving across cities and intra cities as a percentage of the total population. Data were collected from Baidu Migration Big Data Platform (Baidu Migration Data 2022; Wang, 2021) and population density is replaced by aggregation index (Tan et al., 2018). To calculate this index, Night Light Data were obtained from the NPP-VIIRS and processed by median filtering method (Zhong et al., 2019) and logarithmic transformation. Age structure of population is the proportion of people who are older than 60 years old and younger than 14 years old used to describe the distribution of susceptible populations in cities (China Population Census Yearbook, 2020).
2.2.2 Repression Factors
Repression factors refer to those factors that inhibit the spread of Covid-19 pandemic. Access to medical resources refers to the convenience of residents to reach medical facilities such as hospitals and community health service centers. Hospital and health center’s coverage area were created by buffers considering three and one kilometers according to recommendation of China National Health Commission (2022) to calculated the population. Access to health care was estimated by the number of hospital beds per 1000 people. Composition of medical workers is the number and weighting of city’s doctors, nurses, and medical technicians. Data of number of medical stuff and its salary, as well as hospital beds are obtained from China National Statistical Yearbook (2021).
2.2.3 Adaptability Factors
Adaptability factors were used to quantify urban self-learning and practical capacity. Population quality consists of education situation of cities and government expenditure on education. Social Governance was used to evaluate the social service and social work of cities. The economy index includes economic aggregate, economic structure and economic efficiency that were collected from the China Integrated City Index 2020 (Zhou, et al., 2021). The governance is mainly referring to administrative efficiency of e-government in China under the background of urbanization, including democracy, political corruption, and public trust in government (National Academy of Governance, 2021).
2.3 Evaluation Model
The indicator system was constructed after selecting, collecting, and processing all required data. An AHP tool was used to weight the relative importance of doctors, nurses, and medical technicians in the pandemic. After that, CRITIC and EWM were used as average to weight twelve secondary indexes. Post-weighted evaluation index was used in TOPSIS (Table 1) to evaluate urban anti-pandemic resilience of 30 cities.
Table 1
TOPSIS methodological steps
Step | Overview | Formula |
1 Normalization | Sum of squares | \({x{\prime }}_{ij}=\frac{{x}_{ij}}{\sqrt{{\sum }_{i=1}^{n}{x}_{ij}^{2}}}\)① |
2 Forwarding | Reciprocal for indexes which are the smaller the better | \(Reciproca{l}_{index}=\frac{1}{index}\) |
3 Original matrix and weighted matrix | TOPSIS | \(x=\left\{\begin{array}{ccc}{x{\prime }}_{11}& \cdots & {x{\prime }}_{1j}\\ ⋮& \ddots & ⋮\\ {x{\prime }}_{i1}& \cdots & {x{\prime }}_{ij}\end{array}\right., {X}_{ij}={x{\prime }}_{ij}\times {W}_{i}\) \(X=\left\{\begin{array}{ccc}{X}_{11}& \cdots & {X}_{1j}\\ ⋮& \ddots & ⋮\\ {X}_{i1}& \cdots & {X}_{ij}\end{array}\right.\)② |
4 The best and worst vectors in the matrix | \({C}^{+}=\{\left(\underset{i}{\text{max}}{X}_{ij}|j\in R\right),\left(\underset{i}{\text{min}}{X}_{ij}|j\in P\right)\}\) \({C}^{-}=\{\left(\underset{i}{{min}}{X}_{ij}|j\in R\right),\left(\underset{i}{\text{max}}{X}_{ij}|j\in P\right)\}\)③ |
5 Ideal solution distance | \({d}_{j}^{+}=\sqrt{{\sum }_{j=1}^{n}{({X}_{ij}-{C}^{+})}^{2}}\) \({d}_{j}^{-}=\sqrt{{\sum }_{j=1}^{n}{({X}_{ij}-{C}^{-})}^{2}}\)④ |
6 Ranking | \({D}_{j}=\frac{{d}_{j}^{-}}{{{d}_{j}^{+}+d}_{j}^{-}}\)⑤ |
NOTES |
①\({x}_{ij}\) denotes the original matrix while \({x{\prime }}_{ij}\) denotes the normalized matrix |
②\({x{\prime }}_{ij}\) denotes index \(j\) of city \(i\) while \({X}_{ij}\) denotes the corresponding weighted index |
③\({C}^{+}\) and \({C}^{-}\) denotes the best and worst vectors, \(R\) and \(P\) denotes the risk and repression factors, respectively |
④\({d}_{j}^{+}\) and \({d}_{j}^{-}\) denotes the distance between evaluation object with positive ideal solution and negative ideal solution, or optimal matrix vector and worst matrix vector |
⑤\({D}_{j}\) denotes the fit degree between the evaluation object and the positive ideal matrix |
⑥\(i=\text{1,2},\dots ,30, j=\text{1,2},\dots ,12\) |
Step 1 Normalization of the original data matrix could exclude the impact of dimension and numerical size.
Step 2 Forward processing makes the larger the value of all the processed indicators, the better.
Step 4 The cosine method is used to find out the optimal matrix vector and worst matrix vector and the Euclidean Distance could be calculated.
Step 5 The Euclidean Distance the basis for evaluating projects’ pros and cons.