留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

Spatial Characteristics and Influencing Factors of Urban Resilience from the Perspective of Daily Activity: A Case Study of Nanjing, China

Honghu SUN Feng ZHEN

SUN Honghu, ZHEN Feng, 2021. Spatial Characteristics and Influencing Factors of Urban Resilience from the Perspective of Daily Activity: A Case Study of Nanjing, China. Chinese Geographical Science, 31(3): 387−399 doi:  10.1007/s11769-021-1201-0
Citation: SUN Honghu, ZHEN Feng, 2021. Spatial Characteristics and Influencing Factors of Urban Resilience from the Perspective of Daily Activity: A Case Study of Nanjing, China. Chinese Geographical Science, 31(3): 387−399 doi:  10.1007/s11769-021-1201-0

doi: 10.1007/s11769-021-1201-0

Spatial Characteristics and Influencing Factors of Urban Resilience from the Perspective of Daily Activity: A Case Study of Nanjing, China

Funds: Under the auspices of National Social Science Fund of China (No. 20AZD040), the Program B for Outstanding PhD Candidate of Nanjing University (No. 202002B103)
More Information
图(5) / 表ll (5)
计量
  • 文章访问数:  119
  • HTML全文浏览量:  26
  • PDF下载量:  32
  • 被引次数: 0
出版历程
  • 收稿日期:  2020-09-12
  • 录用日期:  2021-01-05
  • 刊出日期:  2021-05-05

Spatial Characteristics and Influencing Factors of Urban Resilience from the Perspective of Daily Activity: A Case Study of Nanjing, China

doi: 10.1007/s11769-021-1201-0
    基金项目:  Under the auspices of National Social Science Fund of China (No. 20AZD040), the Program B for Outstanding PhD Candidate of Nanjing University (No. 202002B103)
    通讯作者: ZHEN Feng. E-mail: zhenfeng@nju.edu.cn

English Abstract

SUN Honghu, ZHEN Feng, 2021. Spatial Characteristics and Influencing Factors of Urban Resilience from the Perspective of Daily Activity: A Case Study of Nanjing, China. Chinese Geographical Science, 31(3): 387−399 doi:  10.1007/s11769-021-1201-0
Citation: SUN Honghu, ZHEN Feng, 2021. Spatial Characteristics and Influencing Factors of Urban Resilience from the Perspective of Daily Activity: A Case Study of Nanjing, China. Chinese Geographical Science, 31(3): 387−399 doi:  10.1007/s11769-021-1201-0
    • The concept of ‘resilience’ emphasizes the system’s ability to resist, absorb, transform and adapt to disturbances, shocks, or uncertainties without changing its basic conditions (Holling, 1973; 2001). The main source of its theoretical development originated from concern about the engineering resilience of a single steady state, and then developed into an explanation of ecological resilience with a multi-equilibrium state. At present, the theory has progressed to reveal the dynamic evolutionary resilience of the social-ecological composite system with the self-organization, self-learning, and self-adaptation characteristics (Gunderson, 2000; Folke, 2006). Because of its innovative analysis and solution to problems, such as dynamic adjustment, guidance according to circumstances, multi-dimensional interaction, and active response, the theory of resilience has been developed and applied in the fields of ecology, disaster science, urban geography, urban planning, and other disciplines (Gunderson and Holling, 2002; Pickett et al., 2004; Folke et al., 2005; Cimellaro et al., 2010). Currently, the combination of climate change, environmental risks, economic inequality, social polarization, and accelerated urbanization has seriously threatened the sustainable development of cities (Michelle, 2017). It can be said that most of the social, economic, and environmental contradictions are mainly concentrated in cities, and the resilience of cities needs to be strengthened in terms of multi-dimensionality and pertinence. Based on previous studies (Ahern, 2011; Cimellaro, 2016; Meerow et al., 2016; Zhang and Li, 2018), urban resilience can be defined as the ability of individuals, communities, institutions, enterprises, and other actors to survive, adapt and develop under various chronic pressures (such as traffic congestion and environmental pollution) and acute shocks (such as various man-made or natural disasters). Here, urban resilience can be regarded as a collection of capabilities of a city in the face of urban contradiction (various pressures, disturbances, or risks), its core connotation emphasizes that although urban contradiction is inevitable, it is also an opportunity and internal driving force for promoting urban development.

      At present, the theory and practice of urban resilience have made some progress. The research on urban resilience issues is cross-integrated, but the interpretation of urban resilience in the context of security still occupies the mainstream, such as climate change, disaster risk, ecological security, and economic crisis (Leichenko, 2011; O’Brien, 2013; León and March, 2014; Fang et al., 2018). However, in the context of macro security, urban residents are actually facing broader and longer-term life conflicts, and they need to constantly adapt to the fast-changing pace of urban life and pursue life improvement (transformation) and development (upgrade). The concrete micro daily life, especially the daily life in the urban transition period, has multi-level and diversified supply and demand contradictions, which directly touch the core rights and interests of urban residents (Xi, 2017). It can be seen that a single security perspective has hindered the cross-study and expansion of urban resilience to a certain extent, and there are few resilient interpretations and responses to various livelihood issues in daily life (Sun et al., 2019). As regards the study content of urban resilience, the conceptual framework and theoretical analysis of urban resilience are mostly carried out (Cote and Nightingale, 2012; Tyler and Moench, 2012; Wilkinson, 2012). Studies generally involve the dimensions of ecological environment, social and economic development, infrastructure construction, governance capabilities, and institutional guarantees (Ostadtaghizadeh et al., 2015), as well as various scales of communities, cities, and regions (Meerow et al., 2016). However, most evaluations of urban resilience are loose, qualitative, and single, which is difficult to apply in urban space (Carpenter et al., 2001; Cavallaro et al., 2014; Brown et al., 2018). Simultaneously, the existing researches on the impact mechanism of urban resilience mainly focus on the internal mechanism of the resilience system at the theoretical level, serving the construction of the theoretical framework and evaluation index system, and rarely analyzing the effect of the impact factors outside the resilience system on the independent resilience variables. Although some resilience cases involve simple impact mechanisms, they often have unclear definitions and overlapping connotations. The research on the impact mode, degree of action, and regulation direction of external factors on urban resilience is still insufficient (Guo and Zhang, 2015; Zhao et al., 2020). As far as its practical application is concerned, planning practice of urban resilience as a single theme is relatively rare. The regulatory concept, path, and vision have not been regularly integrated or decomposed into existing mature urban planning or governance issues. Therefore, the regulatory practice of urban resilience is not sustainable and long-term (Desouza and Flanery, 2013; Xu et al., 2018).

      Considering the above, this study looks at the contradiction between supply and demand in the daily life of China’s urban transformation period as the background, combines it with the connotation of urban resilience, and comprehensively reveals and characterizes the state of urban resilience in daily life through the perspective of dynamic interaction between time-space behavior and environment. In addition, based on geospatial big data (Batty, 2013; Hashem et al., 2016; Qin et al., 2016; Zhen et al., 2017; Wang et al., 2018), this paper examines Nanjing as an example to identify the spatial characteristics of urban elasticity and analyze its influencing factors. Finally, the optimization strategy is also put forward.

    • Nanjing is an important central city in the eastern China. In the near future, the vision of Nanjing’s urban development goal is to be an ‘innovative famous city, beautiful ancient capital’. However, the relatively high cost of living, the shortcomings in improving people’s livelihoods, and the unpopularity of the public services all have a negative impact on livability and the attraction of new talent. This research project takes the main urban area of Nanjing as a case study (Fig. 1), with the boundaries ranging from east to the ring road, south to the Qinhuai New River, and west and north to the Yangtze River, covering an area of 278 km2. This area has the densest population distribution and the most perfect comprehensive infrastructures in Nanjing, and it also has the highest cost of living and the most frequent urban disease outbreaks in Nanjing. Therefore, the shaping of urban resilience in this area is of great significance to the realization of Nanjing’s development vision.

      Figure 1.  Spatial location of the main urban area of Nanjing

    • The selected urban geographic big data has high accuracy and coverage, which can well represent the daily activities and activity environment of urban residents, and can be divided into five categories (Table 1). Digital elevation data and satellite remote sensing images are used to characterize the ecological environment; Wechat activity map data are used to characterize the concentration of residents’ space-time activities; Baidu Map POI (point of interest) and life service theme review website data are used to characterize infrastructure resources; online traffic maps are used to extract vector road network data; and real estate and recruitment information data are used to reflect livelihood support. The data acquisition period took place within one complete week in May 2017. In order to unify the accuracy of data analysis, improve the speed of calculation and eliminate the local multi-collinearity of explanatory variables in influencing factor analysis, the minimum analysis unit of spatial feature analysis is set to a 100 m × 100 m regular grid, and the minimum analysis unit of influencing factor analysis is set to a 500 m × 500 m regular grid. In addition, all data are pre-processed by data cleaning, data validation, standardization, and space projection coordinate calibration.

      Table 1.  Daily activities and activity environmental data of urban residents

      DimensionData contentData volume (10000)Data sources
      Ecological base Digital elevation data Geospatial Data Cloud of Chinese Academy of Sciences (http://www.gscloud.cn/)
      Landsat 8 Satellite Remote Sensing Images (Water Area Density, Vegetation Coverage) Geospatial Data Cloud of Chinese Academy of Sciences (http://www.gscloud.cn/)
      Infrastructure resources Shopping, leisure, education, medical treatment and transportation data of five kinds of point infrastructures, longitude and latitude 8.3 Baidu map open platform (http://lbsyun.baidu.com/)
      Road network Vector Data of Urban Road Network (Expressway, Main Road, Secondary Road, Branch Road, Metro Line) Baidu map open platform (http://lbsyun.baidu.com/)
      Space-time activity Value of activity agglomeration, time, longitude and
      latitude (24 hours a day, a complete week)
      1295 Tencent map open platform (https://lbs.qq.com/)
      Livelihood support Housing Price, Source, Longitude and Latitude of Residential Quarters 0.3 Internet real estate information service platform (http://nanjing.fang.com/)
      Recruitment, salary, educational background, length of service, longitude and latitude 7 Recruitment information platform (https://www.51job.com/)
    • At present, urban transformation is paying increased attention to improving the daily quality of life of urban residents. However, the main contradiction of the current urban development also exists in the daily life of the city, which is manifested in the difficulty of meeting the demands of urban life in a balanced and adequate way, that is, coordinating the contradiction between supply and demand in daily life. Obviously, the contradiction between supply and demand in daily life has brought the most widespread disturbance to the daily life of urban residents, and urban residents have to adapt to it. It can be seen that daily life also entails the core connotation of urban resilience, which is to coordinate the contradictions of daily life in order to promote the high-quality sustainable development of daily life. When integrating the theories of urban resilience and space-time behavior, in the situation of contradiction between supply and demand of urban residents’ daily activities and environment system, urban resilience evolves under the mutual disturbance and adaptation of activity behavior and activity environment. In order to highlight the people-oriented value, activity resilience can be characterized by opportunities and constraints in time and space of activity behavior. Meanwhile, as the space carrier of activity behavior, the activity environment’s space elements, organization, and form have a profound impact on activity resilience. Therefore, for the collaborative governance and dynamic regulation of urban resilience, the spatial characteristics of activity resilience and the local influence mechanism of the activity environment need to be further explored. This specific study framework is illustrated in Fig. 2.

      Figure 2.  Study framework

    • Study methods include the following: activity resilience model, bivariate spatial autocorrelation, comprehensive evaluation model, multiple centrality analysis, and geographically weighted regression. Elaboration on each method is as follows:

      (1) Activity resilience model

      As mentioned above, activity resilience can be specifically characterized by activity scale (spatial opportunities and constraints) and activity stability (temporal opportunities and constraints), and it is the ability to coordinate the scale of activity space agglomeration and the stability of activity time agglomeration. Thus, activity resilience can be calculated by the coupling coordination model (Tang, 2015), which reflects the nature of resilience and simplifies the difficulty of measurement. Simultaneously, the grading criteria of the coupling coordination model can also reveal the evolution stage and state of activity resilience. This model mainly includes the scale index of activity space agglomeration, the stability index of activity time agglomeration, the sustainability index of activity space-time agglomeration, and the resilience index of activity space-time agglomeration, abbreviated as activity scale index, activity stability index, activity sustainability index, and activity resilience index, respectively. The formulas are as follows.

      $$ SIASA = {{\left({AAS{A_{WD}} \times 5 + AAS{A_{WE}} \times 2} \right)} / 7} $$ (1)
      $$SIATA = {{\left({SDAT{A_{WD}} \times 5 + SDAT{A_{WE}} \times 2} \right)} / 7}$$ (2)
      $$SIASTA = {w_1} \times SIASA + {w_2} \times SIATA$$ (3)
      $$ RIASTA = \sqrt {2\sqrt {\frac{{SIASA \times SIATA}}{{{{\left({SIASA + SIATA} \right)}^2}}}} \times SIASTA} $$ (4)

      In Formula 1, SIASA (Scale Index of Activity Space Agglomeration) refers to the scale of activity space agglomeration in a complete week, AASAWD is the average of activity space agglomeration on weekday, and AASAWE is the average of activity space aggregation on weekend. In Formula 2, SIATA (Stability Index of Activity Time Agglomeration) represents the stability of activity time agglomeration in a complete week, SDATAWD is the standard deviation of activity time agglomeration on weekday, and SDATAWE is the standard deviation of activity time agglomeration on weekend. In Formula 3, SIASTA (Sustainability Index of Activity Space-time Agglomeration) is a comprehensive evaluation of activity space agglomeration scale and activity time agglomeration stability. For sustainability, it is equally important to maintain the scale and stability of activities, so the weights w1 and w2 are 0.5. In Formula 4, RIASTA (Resilience Index of Activity Space-time Agglomeration) expresses the essence of activity resilience, that is, the ability to sustain the coordination and sustainability of spatial and temporal agglomeration of activity. In addition, according to the previous research experience (Ding et al., 2015), the resilience grade standard is divided into 10 intervals between 0 and 1; each interval represents a resilience grade, and each grade corresponds to a kind of resilience state (Table 2), thus forming a continuous ladder that can more intuitively reflect the degree and the stage of resilience.

      Table 2.  Grading criteria for activity resilience

      Serial numberInterval of resilience indexGrade of resilience
      1[0.00,0.10]Extreme lack of resilience
      2(0.10,0.20]Serious lack of resilience
      3(0.20,0.30]Medium lack of resilience
      4(0.30,0.40]Mild lack of resilience
      5(0.40,0.50]Almost lack of resilience
      6(0.50,0.60]Reluctant resilience
      7(0.60,0.70]Primary resilience
      8(0.70,0.80]Medium resilience
      9(0.80,0.90]Good resilience
      10(0.90,1.00]Excellent resilience

      (1) Bivariate spatial autocorrelation model

      In order to investigate the spatial correlation between activity resilience and activity scale, activity stability, activity persistence, a bivariate spatial autocorrelation model was used to measure the global trend and local spatial heterogeneity of spatial autocorrelation between two variables (Dale and Fortin, 2009). In this paper, the adjacency criterion is adopted. If region i and region j have common boundaries, the spatial weight Wij is 1; otherwise, it is 0.

      (2) Comprehensive evaluation model

      The comprehensive evaluation model involves mainly the construction of the number index, category index, and quality index of activity infrastructures in influencing factors. Due to the different allocation criteria of different infrastructures, the number, type, and quality of infrastructures in the cell grid can not be directly and simply summed up without any difference. Instead, it needs to be revised by weighting according to the actual allocation ratio: the higher the number of infrastructures, the lower the service level of infrastructure sharing. The importance of such infrastructures in the whole infrastructure system is lower, so the weight given to them is also smaller. According to this basic principle, the ratio of the reciprocal of the total amount of each infrastructure to the sum of the reciprocal of the total amount of all infrastructures is taken as the weight of this kind of infrastructure. The formulas are as follows.

      $$W = \dfrac{{\dfrac{1}{{T{N_i}}}}}{{\displaystyle\sum\limits_{i = 1}^n {\dfrac{1}{{T{N_i}}}} }}$$ (5)

      In formula (5), W is the weight of infrastructures, and TNi (Total Number) is the total number of infrastructures in category i.

      $$ NII = \sum\limits_{i = 1}^n {{w_i}} \times N{I_i},\;CII = \sum\limits_{i = 1}^n {{w_i} \times } C{I_i},\;QII = \sum\limits_{i = 1}^n {{w_i}} \times Q{I_i} $$ (6)

      In formula (6), NII, CII, QII refers to the number index, category index, and quality index of infrastructures, respectively. NIi,CIi, QIi is the number, category and quality of infrastructure, respectively, in category i. wi is the weight of infrastructure i.

      (1) Multiple centrality assessment

      Based on the multiple centrality assessment (Porta et al., 2008), the centrality of network nodes in the integrated road network is measured by using the urban network analysis (UNA). The topological relationship and actual distance between road lines are taken into account to reflect the spatial organization characteristics of the traffic network in the resilience impact factor. Taking time as distance cost, according to the actual speed of the Nanjing main urban area (Amap, 2018), vehicle and subway speed are set to 30 km/h and 60 km/h, respectively. Closeness centrality, betweenness centrality and straightness centrality are selected to measure the accessibility, hub and directness of road network nodes.

      (2) Geographically weighted regression

      Geographically weighted regression (GWR) is a local regression model that is a spatial local extension based on the classical multiple linear regression model. For the spatial data with non-stationarity, the model can reflect the influence degree of the variables in different geographical locations in the region, and can thus explore the spatial differentiation characteristics of the factors affecting the explanatory variables (Brunsdon et al., 1996; Cavallaro et al., 2014; Wei et al., 2020). In this paper, the GWR model is used to reveal the spatial variation of the effect of activity environment factors on activity resilience.

    • As shown in Fig. 3, activity resilience presents a composite structure of ‘circles and clusters’ in the spatial distribution, and it ranges in eight grades ranging from extreme lack of resilience to moderate resilience, accounting for 0.1%, 1.0%, 4.0%, 6.0%, 8.0%, 14.0%, 36.0%, and 31.0% of the main urban area, basically showing an increasing law. More than 80% of the regions have activity resilience, and the average level of the activity resilience index in the main urban area is 0.61, which is in the primary resilience range. On the overall level, the daily activities in the main urban area are resilient, and the spatial distribution is relatively balanced, reflecting the limited daily life contradictions in the main urban area. Specifically, regions lacking resilience show an increasing trend of ‘circles’ of resilience from inside to outside, embedded in areas with activity resilience that is mainly distributed in low activity intensity areas such as mountains, green spaces, and water bodies in the outer suburbs. In addition, as the city’s commercial center and urban transportation hub, Xinjiekou, located in the old city area, has high activity intensity and close regional connections, but it also lacks activity resilience to a certain extent, which may be caused by violent activity fluctuations. As far as the regions with activity resilience are concerned, the spatial differentiation is more gradual and continuous, showing a decreasing trend of ‘circles’ of resilience from inside to outside. The moderately resilient areas are mainly distributed in the old city area and the adjacent northeastern part of the Hexi area and the central and western parts of the Tiebei area. The various functions of cities in these areas are relatively mature. There are several clusters in the surrounding areas of the main city that are generally self-contained and have a certain degree of independence. Areas with preliminary resilience are basically distributed around moderately resilient areas covering the Hexi and Chengnan city-center areas such areas are mainly dedicated to the development of modern service industries, so the daily activities in these areas have a certain degree of professionalism and exclusivity. Other areas with primary resilience are mostly remote residential areas. The areas that are barely resilient are mainly distributed in intermittent bands around the primary resilient areas, mostly in the direction of urban development and expansion; activity popularity is gradually forming in these areas.

      Figure 3.  Spatial distribution of activity resilience in the main urban area of Nanjing

    • At a confidence level of 99%, the global spatial autocorrelation coefficients between activity resilience and activity scale, and activity stability are 0.77 and –0.75. It can be seen that activity resilience is significantly positively and negatively correlated with activity scale and activity stability, respectively (P-value ≤ 0.05). As shown in Figs. 4a, 4b, the local spatial autocorrelation relationship between activity resilience and activity scale, and activity stability presents a significant and oppositely positive and negative trend of agglomeration characteristics, both showing ‘center-periphery’ separated spatial agglomeration mode; between the ‘center’ and the ‘periphery’ is a large area with insignificant autocorrelation agglomeration characteristics. Specifically, the local spatial autocorrelation between activity resilience and activity scale is mainly ‘high-high’ and ‘low-low’ positive agglomeration. The ‘high-high’ mode area mainly occupies the central area of the main city and is distributed in most areas of the old city area and the northeast of the adjacent Hexi area. In the Tiebei, Zijin-Xuanwu, and Chengnan areas, there are scattered clusters of ‘high-high’ mode. The ‘low-low’ mode is mainly distributed in the periphery of the main urban area, mainly in the ‘mountain, water, forest, and lake’ areas. The spatial autocorrelation between activity resilience and activity stability is dominated by negative agglomeration patterns of ‘low-high’ and ‘high-low’. Compared with the spatial autocorrelation agglomeration characteristics between activity resilience and activity stability, there is such a general rule: in the same distribution area, there is an opposite distribution trend. In short, it can be seen that the scale of activities in these central areas is appropriate, but the stability is insufficient, whereas the activities in the peripheral areas are more than stable, but the scale is not good. In addition, in the Xinjiekou area of the city center, there are very few areas with a ‘low-high’ pattern between activity resilience and activity scale, and a ‘low-low’ pattern between activity resilience and activity stability, which shows that the scale of activities in this area is excessive, but the stability is very poor.

      Figure 4.  Local spatial autocorrelation between activity resilience and activity scale (a), and activity stability (b) in the main urban area of Nanjing

    • The activity environment is the space carrier of activity behavior, which has the characteristics of materiality and immateriality, as well as place and flow. Therefore, according to the principles of difference, representativeness, and comprehensiveness, the impact of the activity environment on activity resilience can be analyzed from four main dimensions: ecological environment, infrastructure environment, location environment, and social environment (Zhao et al., 2020). Ecological environment shapes the base of activity space and also reflects the distribution of ecological resources; infrastructure environment includes shopping, leisure, education, medical and transportation resources, and is measured from quantity, quality, and type; location environment reflects the spatial organization ability of activity environment from the connection intensity, accessibility, and efficiency of the road network node; and social environment mainly covers livelihood capacity in the socio-economic sense, including residential opportunities and employability. The explanation and construction of explanatory variables can be viewed in Table 3.

      Table 3.  Description of explanatory variables

      Activity environmentExplanatory variableVariable description
      Ecological environment Water area density Reflect hydrophilic environment using the proportion of water area in statistical cell grid to characterize it
      Vegetation coverage Reflect the greening level using NDVI (Normalized vegetation index) to characterize it
      Elevation Reflect the whole terrain using the average elevation in the
      statistical cell grid to characterize it
      Infrastructure environment Infrastructure quantity Reflect the scale of infrastructure quantity and sum the quantity according to
      the weight given by infrastructure allocation criteria
      Infrastructure type Reflect the diversity of infrastructure types and sum the type according to the
      weight given by infrastructure allocation criteria
      Infrastructure quality Reflect the level of infrastructure quality and sum the quality according to the
      weight given by infrastructure allocation criteria
      Location environment Betweenness centrality Reflect the connection intensity of road network nodes
      Closeness centrality Reflect the connection accessibility of road network nodes
      Straightness centrality Reflect the connection efficiency of road network nodes
      (Straight line connection between nodes)
      Social environment Residence opportunity Reflect the housing pressure, considering the number of housing
      sources and housing prices
      Employment opportunity Reflect the employment pressure, considering the number of recruiters, salary level,
      educational requirements, and length of service

      Based on the principle of geographically weighted regression (GWR), the global regression model of ordinary least squares (OLS) was used to test 11 primary indicators such as the collinearity diagnosis and significance test. Finally, seven explanatory variables were selected: vegetation coverage, water density, infrastructure quantity, infrastructure type, straightness centrality, residence opportunity, and employment opportunity. It can be seen that the four dimensions of the activity environment all have an impact on activity resilience. By comparing the results of OLS and GWR regression, it is found that the fitting effect of the two models is high (Table 4). Because the GWR model takes into account the spatial impact effect and has higher fitting accuracy, it can be selected to analyze the spatial local impact differentiation.

      Table 4.  Comparison of fitting results between OLS and GWR model

      Regression modelResidual SquaresSigmaR2Adjusted R2
      OLS5.650.070.8410.84
      GWR4.850.060.8630.86
    • As shown in Table 5, the variation range, average value, and coefficient of variation (absolute value) of the regression coefficient of each influencing factor are heterogeneous, which indicates that there are significant differences in the effect of influencing factors on activity resilience, whether within or among the influencing factors. Specifically, water area density, vegetation coverage, straightness centrality, and employment opportunity have a gradually deepening overall negative impact on activity resilience, and the effects of vegetation coverage and straightness centrality fluctuate greatly. It may be because the poor accessibility of ecological resources such as water systems and green spaces restricts participation in daily activities, and the efficiency of straight-line arrival of urban road networks may limit the possibility of diverse activities in time and space. In areas that are simply rich in employment resources, the separation of job and residence will also cause strong temporal and spatial fluctuations in activities. Both infrastructure type and infrastructure quantity have an overall positive impact on activity resilience, although their impact fluctuation is small. The diversity and quantity of infrastructure still play a fundamental role in supporting activity resilience, and the quality-driven mechanism of activity resilience has not yet been formed. In addition, the residential opportunity has both positive and negative effects on activity resilience, with large fluctuations that result in weak positive effects on the overall level. Obviously, areas with many living opportunities are often not the urban areas with the highest cost of living and the most complete facilities. These areas often have both support and restrictions on daily activities.

      Table 5.  Statistical characteristics of explanatory variables

      Regression coefficientWater area densityVegetation coverageResidence opportunityEmployment opportunityInfrastructure typeInfrastructure quantityStraightness centrality
      Range of variation[–0.09, –0.03][–0.21,–0.04][–0.01,0.15][–0.52, –0.02][0.34,0.48][0.32,0.70][–0.35,–0.05]
      Average value–0.06–0.110.09–0.400.430.41–0.12
      Coefficient of variation (Absolute value)0.260.410.290.210.090.20.57

      Fig. 5 shows the difference in local spatial influence on activity resilience. Although the negative impact of water density on activity resilience is weak, it still shows an increasing trend from west to east (Fig. 5a) mainly due to the extremely uneven distribution of water systems, particularly in the central and eastern regions, and far from the central city, this further strengthens the constraints on activity resilience. Compared with water area density, the negative effect of vegetation coverage is higher and more complex, showing an increasing trend from southwest to northeast (Fig. 5b), mainly related to the distribution pattern of vegetation and population. Because the southwest urban area is a new urban area, the vegetation coverage is low, the population is relatively concentrated, and the activities here are concentrated and frequent. The northeast part of the central urban area is covered with large-scale mountain green space such as Zijin Mountain and Mufu Mountain, which leads to a scattered population distribution and poor activity scale and stability. Residential opportunities primarily play a mainly positive role, and their spatial differentiation basically shows a decreasing trend from southeast to other directions (Fig. 5c). The southeast is close to the center of the city but is mainly distributed with some universities and some undeveloped areas. The living cost here is relatively low and the housing supply is relatively tight. In this case, residential opportunities here have the greatest positive impact on activity resilience. Overall, employment opportunities have the greatest negative impact on activity resilience and tend to increase from southwest to northeast (Fig. 5d), primarily because the northeastern part of the main urban area is far away from the main activity areas of residents, and it is more likely to have a problem of a job-housing imbalance. The positive impact of infrastructure type is significant, tends to decrease from southwest to northeast (Fig. 5e). The reason for this is that most of the southwest areas are new urban areas and the population is relatively concentrated, but the supporting types of living infrastructures are not complete enough, so the impact on activity resilience is relatively significant. The positive impact of infrastructure quantity is as significant as that of infrastructure type and shows a trend of increasing from the middle to the periphery (Fig. 5f). The middle area covers the main urban areas, including the old and new urban areas. The scale of activities in the old city is high, but the stability is low. On the contrary, the scale of activities in the new city is relatively low, but the stability is high. Finally, although the amount of infrastructure in the region varies greatly, activity resilience is similar, which reflects that the effect of infrastructure quantity in this region is weaker than that in other regions. The spatial differentiation characteristics of the negative effect of straightness centrality are similar to that of infrastructure quantity (Fig. 5g). It is also because the straightness centrality of road network nodes in this area is complex and diverse, but the change range of activity resilience is small.

      Figure 5.  Spatial differentiation of regression coefficients of influencing factors in the main urban area of Nanjing

    • Macro security resilience is the cornerstone of urban development, while micro activity resilience can be the vision of urban development. Activity resilience reflects the most concentrated and frequent contradiction between supply and demand in the urban man-land system and it also reveals the longest and most extensive disturbance and adaptation that urban residents bear. Investigating the characteristics of activity resilience in the main urban area of Nanjing and the influencing factors of the activity environment on it, the study found that the distribution of activity resilience is relatively balanced, but the level is average. In general, the scale of activity space agglomeration and the stability of activity time agglomeration are mutually exclusive. Simultaneously, the urban environment is not well integrated into the scene of daily life activities—its comprehensive support, dynamic guarantee, and high-quality guidance for activity resilience have not yet formed. To a certain extent, this reflects that the high density and high rhythm of urban residents’ daily life activities are also highly separated, excessive, and inefficient in time and space, and the support of the activity environment is extensive, disordered, and unbalanced. This is mutually confirmed by the problems of urban sprawl, separation of job and residence, low space quality, separated space function, and excessive agglomeration in China’s big cities (Li and Fang, 2016; Song et al., 2017; Yue et al., 2020). Therefore, the optimization of activity resilience is a set of systematic means, and special attention should be paid to collaborative governance and dynamic regulation.

      This study on urban resilience from the perspective of daily activities is a supplement to the study on urban resilience from the perspective of traditional security, as well as a response to the main social and economic contradictions in the transition period of urban development in China. This perspective regards disturbance and adaptation as opportunities for urban development and improvement of people’s livelihoods. Although this study initially established the theoretical framework, comprehensive index model, and influencing factor index system of urban resilience, and carried out empirical analysis based on urban big data in the main urban area of Nanjing, there are some limitations. Firstly, the inner relationship between urban contradictions and urban resilience needs to be further improved to build a more universal theoretical framework of urban human settlement resilience. Secondly, due to the limitation of data sources, the spatial mobility and temporal persistence of activities are not included in the activity resilience index model. Thirdly, the revealing of the influence mechanism is relatively weak, so the multi-scale and non-linear analysis of influencing factors needs to be strengthened.

    • Based on the connotation of urban resilience and the main contradictions of social and economic development in China, supported by multi-source geographic big data, this paper puts urban resilience in the context of daily activities, identifies its spatial characteristics, and reveals the local impacts of the activity environment on activity resilience in multi-dimension, In addition, the corresponding optimization ideas and countermeasures are put forward. The main conclusions and understandings of the study are as follows:

      (1) The spatial distribution of activity resilience presents a composite structure of circles and clusters. Areas with no resilience are basically distributed in ‘mountains, waters, forests, lakes’ areas, showing an increasing trend of ‘circles’ from inside to outside, and forming a cluster structure. The spatial differentiation of areas with activity resilience is more gradual and continuous, occupying most of the built-up areas of the main city, and mainly showing a decreasing trend of ‘circles’ from inside to outside, and some clusters are also formed in some areas of the main city. Overall, the spatial distribution of activity resilience is relatively balanced. Most areas are resilient, but the average level is only in the primary resilience range. In addition, there are significant positive global autocorrelation and negative global autocorrelation between activity resilience and activity scale, and activity stability. Moreover, the local spatial autocorrelation characteristics with opposite trends of positive and negative effects appear at the same spatial location, indicating that there is an antagonism between the activity scale and activity stability for activity resilience.

      (2) As the carrier and support of activity behavior, the four dimensions—ecological environment, infrastructure environment, location environment, and social environment—all have an impact on activity resilience. The degree and direction of the influencing factors among different dimensions and regions are heterogeneous. In the dimension of the ecological environment, water bodies and green areas have not been well integrated into daily activities, which have a certain negative effect on activity resilience. The positive effect of the infrastructure dimension on activity resilience is determined by the quantity and type of infrastructures, and the quality of infrastructures has not yet played a significant role. In the dimension of the location environment, straightness efficiency of the road network nodes is a negative impact factor, which may limit the flexible selection of active paths. The influencing factors of the social environment did not all play a positive role. Due to huge regional impact differences, residence opportunity has only a weak positive impact on activity resilience. Employment opportunities may not match the spatial distribution of residence opportunity, which makes it difficult to balance the scale and stability of activities at the same time, thus having a strong negative impact on activity resilience.

      (3) Improving the resilience of urban daily activities should be rooted in the main contradictory scenes of urban daily life. For activity resilience itself, a high degree of coordination between activity scale and stability should be ensured in the dynamic, connection, and integration to improve the quality of activity resilience, reduce the cost of activity resilience, and promote the balance of activity resilience. For the activity environment, while protecting the quality of the ecological environment, it is necessary to improve the accessibility of daily activities to the ecological environment. In terms of activity infrastructures, after meeting the basic needs for the quantity and type of activity infrastructures, the balanced supply of high-quality activity infrastructures should be strengthened. With regard to location environment, a multi-center, distributed, and networked transportation organization network should be developed to reduce the disorderly, excessive, and inefficient flow and agglomeration of activities. Finally, as far as the social environment is concerned, while increasing employment and residence opportunities, more attention should be paid to the job-housing balance.

参考文献 (45)

目录

    /

    返回文章
    返回