Resilience assessment and optimization of mountainous county settlements based on society-space network: a case study of the Qinba Mountain area

Abstract

Under the influence of rapidly changing traffic conditions, the spatial structure of mountainous county settlements is often difficult to adapt to the evolution of social relations, which leads to issues such as disorganized spatial organization, imbalanced resource allocation, and mismatches between public service supply and demand. To this end, it is urgent to develop a resilient development model through the coordinated regulation of society–space systems. This study takes Lueyang County in the Qinba Mountain area as a case study, analyzes the county settlement system based on the society-space network theory, and establishes a three-dimensional research framework of “space simulation-resilience assessment-spatial optimization”. The study finds that the transportation network in the peripheral areas of Lueyang County has significantly improved over the past decade. The total social demand has increased by 44.32%, and a larger-scale industrial development belt was formed along National Highway G345. The rapid evolution of the transportation network has enhanced the collaboration, stability, and restorability of county settlements by 15.8%, 42.4%, and 22.2%, respectively. Additionally, this evolution has led to the emergence of 2 agglomeration development zones and 3 settlement clusters with strong buffering capabilities. Based on these findings, it is recommended to plan 8 coordinated development zones within the county, establish 19 passenger hubs, focus on developing 28 core communities, and outline 6 types of rural development models with corresponding planning guidelines. This study can provide theoretical support for the coordinated spatial development planning of mountain settlements as well as practical guidance for promoting urban-rural integration, village system optimization, resources allocation, and construction standards in the county.

Introduction

Mountainous terrain makes up about one-third of China’s land area. Due to its economic backwardness and limited social development, it has become a key region for resilience research. The Qinba Mountain area, which is the dividing line between northern and southern China, features steep slopes and deep ravines, which severely limits transportation options and causes prolonged economic underdevelopment in counties (Han, Wang, & Yang, 2023). As early as 2000 years ago, local residents and soldiers set up strongholds at strategic passes, gradually forming defensive settlements based on geo-social relations. With the rise of commercial activities around the mountainous areas, settlement spaces similar to post stations appeared along the sparse and narrow mountain roads based on commercial relationships (Chen, Mehmood, Liu, & Gao, 2022). Since then, the law of diminishing returns on cultivated land will compel the community to split. Kinship networks accompanied roads and slowly spread in mountainous areas, causing settlements to scatter along the roads (Chen, Liu, Feng, Li, & Li, 2022). Consequently, the relationship between local settlement space and transportation network has always been in a dynamic cycle of “terrain restricting settlement development - transportation breaking through terrain restrictions - settlement expansion again”. With the promotion of the national “village-to-village access” policy, the local road density has increased significantly by 80%-90% over the past decade. On the one hand, the rapid development of transportation has fostered export-oriented social relations and production methods. This has weakened traditional production cooperation and clan ties, reduced the endogenous development force of settlements, and disordered the original spatial organization (Li, Miao, Wang, Jiang, & Wang, 2022). On the other hand, convenient transportation has prompted a large number of young and middle-aged villagers to seek employment in county towns or adopt dual urban-rural livelihoods. Meanwhile, the activities of the elderly villagers left behind are limited, profoundly altering social relations such as community interaction, services, and consumption among settlements. This further exacerbates problems such as inefficient resource utilization and imbalance between supply and demand of public services (Zhou, Jiang, Niu, Li, Li, Zhang, & Liu, 2023). The rapidly changing transportation network in the Qinba Mountain area has given rise to a new social production and lifestyle. However, its positive impact is not compatible with the lagging spatial evolution of settlements, creating the root problems. Therefore, there is a need to coordinate the society-space relationships of county settlements based on the evolution characteristics of the transportation network and build a resilient development model. In the study of settlement resilience, the county-level units are usually considered relatively complete geographical units in mountainous areas. It also serves as a historically stable administrative unit and a spatial vehicle for implementing policies and planning. Consequently, it is commonly regarded as a settlement system for investigating the evolutionary patterns and resilience. (Chen, Wang, Qiang, & Lin, 2024). Unlike plain regions, most urban and rural settlements in mountainous areas are distributed along roads on narrow river terraces. Their spatial organization is closely related to the transportation network, with roads usually running through settlements and forming a symbiotic developmental relationship (Qin, Luo, Li, Tan, Zheng, Yu, Liao, & Li, 2022). Road hubs often serve as the core nodes for production, services, and social interaction within settlements. Although the branched road system brings dense human and material flows, it forms networked and flattened social connections. Some settlements located on the edge of space also play an important role because of their socioeconomic advantages or their hub position in organizational hubs (Yiannakou, Eppas, & Zeka, 2017). Therefore, in order to reconstruct the settlement space system and optimize resource allocation, it is necessary to comprehensively consider the social connections and transportation networks of county settlements. Spatial resilience is the contribution of spatial attributes to the feedbacks that generate resilience in systems. To operationalize and quantify it, consideration of asymmetries, connectivity, and information processing is warranted (Zhou & Hou, 2021a). It is vital to identify the impact of key settlements and their connections on county spatial resilience. Additionally, proposing dynamic, resilient spatial optimization schemes requires integrating the evolution characteristics of social networks and spatial networks.

Lueyang County is situated in the Qinba Mountain area, at the intersection of the Qinling Mountains and the Bashan Mountains. How to scientifically guide the reasonable layout of county settlement space, enable urban and rural settlements to better adapt to the opportunities and challenges brought by the rapid changes is an urgent problem to be solved. This involves coordinating the adaptive evolution of the society-space relationship to create resilient development models. The research objectives o are as follows: (1) To analyze the characteristics of transportation connection and functional connection of county settlements in mountainous areas based on society-space networks; (2) To evaluate the space resilience of urban and rural settlements using complex network indicators; (3) To propose the spatial organization models, settlement systems, and development guidelines for county units. The main contributions include: (1) In the field of spatial resilience research, a network analysis model that can coordinate physical space and social functions is constructed. (2) Through empirical research, a network research framework and quantitative evaluation indicators are provided for traditional society-space theory.

Literature Review

Settlement space resilience and resilient planning are cutting-edge topics in the field of sustainable development. This research trajectory traces back to the concept of adaptive management proposed by Holling. Its core lies in analyzing the stability and resilience of the system in response to uncertain disturbances, and formulating adaptive management strategies accordingly (Zhou & Hou, 2021b). Since 2009, studies on assessing space resilience based on spatial factors such as transportation, ecology, and infrastructure have become mainstream (Gao, Gong, Li, Yang, & Liang, 2024; Liu, Zhou, Li, Zheng, & Liu, 2024). Masnavi et al. (2019) argued that the spatial organization, layout, and attributes of the street network could reveal spatial resilience. It could be quantitatively studied by such indicators as diversity, redundancy, coherence, and efficiency. Janssen believed that a system had spatial network features, and thus its resilience could be revealed by the spatial network resilience (Zhou & Hou, 2021a). Especially in river valleys with scattered construction land and sparse road networks, the resilience level of transportation networks, which is the core components of county-level urban and rural spaces, directly affects the overall stability and sustainability of these systems (Liu, Zhou, Li, Zheng, & Liu, 2024; Yuan, Fan, Luo, Ren, & Li, 2025). Existing research primarily focuses on the buffering capacity and restorative capacity of the transportation network in response to congestion, disasters, and extreme events (Markolf, Hoehne, Fraser, Chester, & Underwood, 2019; Lang, Wan, Zhang, Zhang, Zhu, & Liu, 2024; Zhang, Mahadevan, & Goebel, 2019;). The research method generally adopts network modeling technology, and evaluates key nodes, important paths and system recovery time through indicators such as structural holes, betweenness centrality, and network efficiency (Markolf, Hoehne, Fraser, Chester, & Underwood, 2019). However, the urban-rural network formed by both physical and non-physical connections in the county settlement system is relatively independent (Wang & Zhang, 2021). Its spatial resilience not only includes the physical resilience of the transportation network, but also integrates multidimensional characteristics such as social resilience and economic resilience (Gonçalves & Ribeiro, 2020; Zhou, Wang, &Yang, 2019). For instance, Folke pointed out that the resilience of urban and rural settlements can be characterized by the self-organizing capabilities of teams and actor groups within social networks (Folke, Hahn, Olsson, & Norberg, 2005). Related research areas include rural settlement safety and disaster prevention systems (Cassidy & Barnes, 2012), resilience development paths in socio-ecologically vulnerable areas (Scott, 2013), and settlement system optimization based on social networks (Zhou, Jiang, Niu, Li, Li, Zhang, & Liu, 2023). It is worth noting that in special areas such as the Qinba Mountain area, transportation networks and social relations exhibit co-evolutionary characteristics, contributing equally to the resilience of settlement space. However, there is still a lack of spatial resilience assessment models and indicator systems that can coordinate the dynamic changes of social functions and spatial elements.

The theoretical origin of the society-space network model can be traced back to Cavallaro’s study on the measurement of urban system resilience using the complex network model. The model innovatively integrates urban and rural social attributes and other physical element information on the basis of traditional infrastructure networks, including spatial entities such as population distribution, residential layouts, industrial facilities, and service nodes, and quantifies spatial resilience by analyzing the topological structure and connectivity efficiency of complex networks (Cavallaro, Asprone, Latora, Manfredi, & Nicosia, 2014). Because its network structure embeds both social attributes and spatial elements, the model was originally defined as “Hybrid Social-Physical Complex Network”. The advantage of this modeling framework is its ability to simultaneously characterize the co-evolution of social functions and spatial elements in urban and rural settlements, thereby revealing the resilience formation mechanism of settlement systems in response to external disturbances. At the methodological level, the model analyzes the settlement system through the following two dimensions. Firstly, the analysis of network topology can reveal the spatial distribution characteristics of settlements. Secondly, network diffusion dynamics simulation can restore the spatial evolution process of settlements (Li, Yuan, Zhang, Li, Wang, & Hu, 2022). Typical applications include the use of the society-space network model by Varol’s research team to achieve quantitative representation of settlement space under multicultural and multi-governance conditions, and successfully applied it to administrative boundary demarcation, multi-scale space simulation (Varol & Soylemez, 2019). Related research results have penetrated into many fields such as urban-rural spatial organization, emergency evacuation planning, and agricultural landscape configuration. In terms of resilience assessment, independence, collaboration, connectivity, interdependence, and stability of settlements could be evaluated through indicators such as degree, structural holes, betweenness, clustering coefficient, and k-core of the participants’ network. (Fei, Wang, Guo, Ma, & Wei, 2022; Zhou & Hou, 2021b; Zhou, Jiang, Niu, Li, Li, Zhang, & Liu, 2023). Its methodological advantages are mainly reflected in three aspects: (1) The flexibility in network construction supports multi-data integration. (2) The fine space-time granularity allows dynamic monitoring. (3) The real-time simulation capabilities are suitable for evaluating complex scenarios such as self-organization efficiency, post-disaster recovery potential, and social vulnerability responses (Yang, Metcalf, & Mao, 2021; Yu, Yang, & Li, 2018). From the perspective of theoretical development, the society-space theory emphasizes that the coordinated adaptation of social functions and spatial elements is the crucial pathway to enhancing urban and rural settlement resilience (Cumming, 2011). This theoretical framework has successfully guided the resolution of practical challenges such as traditional village preservation, immigrant community planning, and urban-rural spatial differentiation (Liao, Gil, Yeh, Pereira, & Alessandretti, 2024; Shi, Wang, Fang, Xiao, & Huang, 2023). On this basis, the study selected Lueyang County in the Qinba Mountain area as an empirical case, analyzed the county settlement system by constructing a society-space network model, and used complex network indicators to evaluate spatial resilience characteristics, and finally proposed a resilience planning scheme that coordinates society-space co-evolution.

Data and Methods

Research scope and data source

The study area, Lueyang County, is situated in the western segment of the Qinba Mountain area, at the tri-junction of Shaanxi, Gansu, and Sichuan provinces (Fig. 1). The county covers a total area of 2831 km², with complex and diverse topography, significant altitude differences ranging from 587 to 2425 meters, and an average slope of 36 degrees. The valleys have narrow landforms, and agricultural lands are scarce. Rural buildings are dense, encroaching heavily on the river banks. Most external connections rely on a single, narrow road, whose limited capacity directly impacts the spatial resilience of settlements (Fig. 2). In terms of administrative divisions, Lueyang County currently includes 17 towns and 151 administrative villages. As of the latest statistical data, the county’s registered population is 176,900, with a total road network length of 1904 km. Notably, both the population density (6.25 people/km²) and road density (0.67 km/km²) are significantly lower than the national average. In recent years, against the backdrop of rapid development in transportation infrastructure, Lueyang County has faced severe population outflow and social network collapse. In 2023, the GDP decreased by 13.4% year-on-year. This further led to a decline in the spatial resilience of the county settlements.

Fig. 1: Location of the study area.

The figure shows the location of the Qinba Mountain Area in China and the detailed Digital Elevation Model (DEM) of Lueyang County.

Fig. 2: Distribution characteristics of settlements in Qinba Mountain area.

The figure shows that the confined valley terrain forces settlements to cluster along a single road and river, creating most external connections rely on a single, narrow road, whose limited capacity directly impacts the spatial resilience of settlements.

The study data encompasses various sources including administrative boundary data, road network data, remote sensing imagery data, statistical yearbook data, government report data, and field survey data specific to Lueyang County. Among them, the administrative boundaries and road network data are derived from the third national land survey conducted by the natural resources and planning department. Remote sensing imagery data were downloaded from the Geospatial Data Cloud¹. Statistical yearbooks and government report data were provided by the Lueyang County Library. Field survey data mainly include information on passenger routes and volumes, logistics routes and freight volumes, population mobility changes, and the distribution of enterprises and logistics points. These data were mainly obtained through interviews with officials from the transportation department and by distributing questionnaires to various villages in a sampling manner to verify and supplement existing data.

Research methodology

Based on the society-space network model, this study constructs a research framework of “spatial simulation-resilience assessment-spatial optimization”. First, by topologically processing the transportation facility network and the social demand network, a society-space network model is established to simulate the spatial evolution process of the county unit settlement. Secondly, the spatial resilience of the county settlement is evaluated based on the complex network theory, focusing on identifying the impact mechanism of key settlements and their correlation on the overall spatial resilience. Finally, based on the results of the resilience assessment, a dynamic adaptive spatial optimization strategy is proposed.

Space simulation

This study selected two time nodes, 2010 and 2020, to construct a county-level settlement society-space network model. The spatial subnetwork is mainly composed of various types of roads, while the social subnetwork consists of core social demands within the county, specifically focusing on passenger flow and logistics.

• (1) Construction of Traffic Subnetwork

  In view of the complexity of road conditions in mountainous areas, the simple road length cannot accurately reflect the actual connection strength. Therefore, this study used the travel time as the key indicator to measure the convenience of connection between villages and towns based on road traffic rates for different road categories. As shown in Equation (1), T_ij represents the highway connection weight between villages (or towns) i and j, where lg_ij, ls_ij, lx_ij, ly_ij, and lc_ij respectively represent the length of the expressway, national highway, provincial highway, county road, and township road between the two villages (or towns). On this basis, the spatial position and attribute characteristics of the topological points of the highway intersection are matched to the corresponding administrative village. If there are multiple road intersections in an administrative village, its geometric center and attribute weight value are taken as the representative.

  $${T}_{{ij}}=1/\left(\frac{l{g}_{ij}}{80}+\frac{{{ls}}_{{ij}}}{60}+\frac{{{lx}}_{{ij}}}{40}+\frac{{{ly}}_{{ij}}}{30}+\frac{{{lc}}_{{ij}}}{30}\right)$$

  (1)

• (2) Construction of Social Subnetwork

  Based on the flow characteristics of rural settlement elements obtained from field research, the study selected passenger flow and logistics data between settlements to build and standardize an indicator system (Table 1) (Huang, Song, & Song, 2020; Huang, Zheng, Tan, & Xie, 2022; Shcherbina & Gorbenkova, 2019). Referring to relevant research results, two indicators, rural population density and rural per capita annual income, were selected to measure the level of socioeconomic development. The frequency of passenger buses and the number of passenger stations were used to measure the level of passenger transport development. The number of rural enterprises and the number of logistics distribution stations were used to measure the level of rural logistics development. After weighted superposition calculation of each indicator, the comprehensive quality scores of development for each rural settlement in Lueyang County were calculated for 2010 and 2020.

  Table 1 Comprehensive development evaluation indicators and weights for rural settlements in Lueyang County.

  First, the positive differential method is applied to normalize the six core indicators, including population density, per capita annual income, passenger frequency, and number of passenger stations. Next, the Entropy Weight Method is used to calculate the information entropy (E_j) and corresponding weights (W_j) for each indicator (Eqs. 2 and 3). P_ij represents the proportion of standardized indicator data within the overall dataset.

  $${E}_{j}=-{\rm{Ln}}{n}^{-1}\times {\sum }_{i=1}^{n}{P}_{{ij}}\times {Ln}{{P}_{i}}_{j}$$

  (2)
  $${W}_{j}=(1-{E}_{j})/{\sum }_{i=1}^{n}(1-{E}_{j})$$

  (3)

• (3) Construction of society-space network simulation model

Identification of traffic network nodes and edges: The geometric centers of the village committee patches of 151 administrative villages are extracted as the traffic network nodes. Each node is assigned the comprehensive development quality of its respective rural settlement as an attribute feature. The actual road network undergoes topological processing to form connecting edges, where the highway connection weight T_ij is used as the quantitative attribute indicator for these edges.

Identification of social network nodes and edges: With the same spatial baseline, the geometric centers of the 151 administrative villages are also defined as social network nodes. Node attributes are also characterized by the comprehensive development quality of rural settlements. The social demand connections between settlements, including passenger flow and logistics, are topologically represented as connecting edges in the social network. Further, the K-means clustering algorithm is employed to eliminate edges with insufficient influence in the network (Gao, Yang, Dai, & Wang, 2024). Passenger flow and logistics are used as quantitative attribute indicators for these edges.

Given that the nodes of the social subnetwork and traffic subnetwork spatially correspond to the same settlement locations, this study achieves effective integration of the two subnetworks by sharing settlement nodes, thereby constructing a complete settlement society-space network model.

Resilience assessment

• (1) Collaboration

  Collaboration refers to the ability of settlement nodes to coordinate surrounding functional entities (Durugbo, Hutabarat, Tiwari, & Alcock, 2011). Nodes with good collaboration usually serve as bridges, which can facilitate stakeholder participation and efficiently allocate resources. In the society-space network, the structural holes (EC_i) of a node e_i indicate its dependency on other nodes in the network. The less the node is constrained by the surrounding nodes, the more significant its competitive advantage is, and the more likely it is to occupy the structural hole position. Therefore, structural holes can effectively evaluate the contribution of a single settlement node to the regional collaboration within the settlement system. The calculation formula is as follows:

  $${{EC}}_{{ij}}={{(P}_{{ij}}+{\sum }_{q}{P}_{{iq}}{P}_{{qj}})}^{2}$$

  (4)
  $${{EC}}_{i}={\sum }_{j}{{EC}}_{{ij}}$$

  (5)

• (2) Connectivity

  Connectivity reflects the ability of functional entities within the settlement system to connect through complex networks (Turnbull, Hütt, Inannides, Kininmonth, Poeppl, Tockner, Bracken, Keesstra, Liu, Masselink, & Parsons, 2018). Network hub nodes usually have stronger connectivity and handle the most frequent interactions. Betweenness can effectively measure its connectivity. Eb_i is defined as the proportion of times node e_i appears on the shortest path between all nodes. The calculation formula is as follows:

  $${{Eb}}_{i}={\sum }_{s,t\in E,s\ne t}\frac{{n}_{{st}}({e}_{i})}{{n}_{{st}}}$$

  (6)

• (3) Buffering

  Eigenvector centrality is used to evaluate the importance of a settlement node as a bridge for information transmission (Liu, Niu, Yu, Yang, Ma, Qiu, Wang, Liu, & Li, 2022). When external disturbances occur, the stronger the ability of the adjacent nodes, the more prominent the buffering ability of the node. Therefore, eigenvector centrality can effectively evaluate the contribution of settlement nodes in a certain area to the overall system’s buffering. Let A be the adjacency matrix (A_ij = 1 if nodes are connected, otherwise 0). The eigenvector centrality x satisfies:

  $$A{\bf{x}}=\lambda {\bf{x}}$$

  (7)

  where λ represents the maximum eigenvalue of the adjacency matrix A. Each component x_i of the eigenvector centrality x quantitatively represents the degree of centrality of node i in the network. The computation process involves ① constructing the adjacency matrix A, ② solving the eigenvalue and eigenvector of the matrix A, ③ identifying the eigenvector x corresponding to the maximum eigenvalue λ, and ④ normalizing x to ensure that the sum of all components is 1. Each component x_i of the eigenvector centrality x can accurately reflect the importance level of node i. A larger x_i indicates greater importance of node i within the network.

• (4) Stability

  Stability indicates the ability of settlement units to maintain interconnections during crises. k -core analysis shows that when every node in a subgraph is connected to at least k other nodes, the larger the k value and the higher the proportion of such subgraphs, the stronger the network resilience. Otherwise, the network is more fragile (Zhou & Hou, 2021b).

• (5) Restorability

Restorability is reflected in the ability to improve the connectivity of the transportation network or optimize the operating efficiency of the social network in crisis response (Yang & Pan, 2021). Network efficiency can quantify system restorability and reflect the ability to maintain connectivity after node/edge failure (Latora & Marchiori, 2001). The calculation formula is as follows:

$$E\left(G\right)=\frac{1}{N(N-1)}{\sum }_{{ij}\in E,i\ne j}\left(\frac{1}{{d}_{{ij}}}\right)$$

(8)

The connection efficiency between any two nodes e_i and e_j is determined by the inverse of their shortest path distance d_ij. When there is no connection path between the nodes, d_ij = +∞, and the corresponding connection efficiency is 0. Assuming that there are N nodes in the network, the overall network efficiency is defined as the arithmetic mean of the connection efficiency between all node pairs.

Spatial optimization

• (1) Spatial structure planning

  As modular units with similar functions within a social network, communities can be preliminarily defined as coordinated development units through network community division methods (Cai, Wei, Dai, Haung, & Li, 2024). The specific implementation steps are as follows. First, based on the regional social subnetwork, the “Communities” function in Pajek software is used to perform preliminary unit division, and the regional management plan and transportation development strategy are formulated in combination with the social connections and functional characteristics between actual settlements. Secondly, based on the collaboration and connectivity assessment results of the society-space network, the advantageous settlement nodes and their social demand connections within communities are identified. Thirdly, based on the comprehensive development evaluation of settlements, passenger and freight stations could be sited around settlements that demonstrate advantages in passenger transport and logistics indicators. Correspondingly, industrial facilities could be sited around settlements with favorable resource conditions and superior socioeconomic development levels. Finally, the core communities for priority development are established, the configuration of public service facilities is optimized, and the vulnerable links of the transportation network are identified and connecting lines are added to improve the road network structure.

• (2) Rural classification planning

In order to improve the resilience and stability of county unit development, administrative villages are classified into the following 6 development types based on the resilience assessment results and statutory planning requirements (Li, Miao, Wang, Jiang, & Wang, 2022): Centralized promotion village continues to exhibit outstanding collaboration, connectivity, and buffering capacity within the society-space network. The reserved development village has good spatial resilience. However, its improvement tends to be flat with the optimization of the road network. Resilient shrinkage village displays poor regional collaboration, connectivity, and buffering, as well as a slow downward trend. A characteristic protection village possesses unique attributes such as cultural, tourism, or conservation areas, as well as a strong collaborative capacity. Urban-rural integration village maintains close connectivity with urban nodes and has significant connections with county towns in terms of stability and buffering indicators. The relocation village has weak spatial resilience, with transportation network optimization showing no significant improvement in its indicators.

Research Results

Space simulation results

From 2010 to 2020, the number of rural nodes in Lueyang County remained relatively stable, while the transportation infrastructure network and social demand network showed significant changes. Specifically, the main east-west corridor was upgraded to National Highway G345. The Xibai Highway was newly built in Baishuijiang Town and Matiwan Town in the north of the county, and several village roads were added around Village Node 143 and Village Node 148, which greatly improved the traffic connectivity in the northern region. In addition, a denser network of village roads emerged around Lesuhe Town, Baiquesi, Jieguanting, Heihe, and Xiakouyi in the southern part of the county. Throughout the evolution of the transportation network, changes in social demand were mainly reflected in three dimensions. First, the total amount of social demand between settlements increased significantly, with the main social connections rising from 185 in 2010 to 276 in 2020. The newly added connections are mainly concentrated in the county center, Guo Town, Jieguanting Town, and Xiakouyi Town. Second, there was a notable increase in social demand intensity, evidenced by a near doubling of passenger routes. Logistics points increased from 80 in 2010 to 435 in 2020, and the number of rural enterprises rose from 800 to 2399. Third, driven by National Highway G345, the radiation range of Lueyang County was further expanded, and a larger-scale industrial development belt has been formed along the highway (Fig. 3).

Fig. 3: Settlement society-space network in Lueyang County.

The figure shows the structural expansion of the society-space network between 2010 and 2020. Key developments include the upgrade of the main east-west corridor and the densification of village road networks.

Results of resilience assessment

Collaboration

In the transportation network of Lueyang County in 2010, except for the areas covered by National Highway G345, Shitian Expressway and Luening Highway, the rest of the areas were identified as structural hole regions (Fig. 4). The settlement connections in these structural hole areas are greatly restricted by the terrain, mainly relying on single road connections, and the collaborative ability is weak. Administrative villages without structural holes are usually located along the main roads and play a role as regional centers. With 0.45 as the threshold for structural hole determination, these areas accounted for 77.48% of the region. By 2020, with the completion and opening of the Xibai Highway in the north, the structural hole issue for over ten rural nodes around Baishuijiang Town was eliminated, and two new regional centers at Village Node 150 and Village Node 134 were formed. Additionally, the completion of the Shitian Expressway in the southwest further enhanced the collaborative ability of the western and southwestern areas (such as around Village Node 65, Village Node 94, and Village Node 101). During this period, cross-terrain and cross-administrative transportation links emerged. If 0.45 was still used as the structural hole threshold, the proportion of regional structural holes decreased to 64.90%.

Fig. 4: Structural holes of society-space network in Lueyang County.

The figure shows the comparison of structure holes in the society-space network between 2010 and 2020. The smaller the value, the more significant the role of the node as a bridge connecting different groups or regions in the network.

As shown in Fig. 4, in the 2010 social network of Lueyang County, Village Nodes 45, 56, 65, 94, 95, and others occupied structural holes due to their strategic locations at transportation hubs. At the same time, rural settlements distant from the county seat also became structural hole areas due to inconvenient transportation. Taking 0.45 as the structural hole judgment threshold, the proportion of structural holes in this area reached 76.15%. By 2020, with the rapid growth in passenger flow and logistics, the overall structural hole value in the county dropped by 15.8%, and the transportation radiation range expanded further. Currently, the structural holes are mainly distributed in the northeastern edge of the county, and the overall structural hole proportion dropped to 64.24%. In addition, although the eastern part of Lueyang County did not see significant traffic enhancement over the past decade, the social demand connections showed notable improvements. Specifically, a tight regional social connection formed along the settlement line from Village Node 13 to Village Node 36 to Village Node 138.

Connectivity

In 2010, the primary transportation connections within Lueyang County were concentrated along the route starting at Village Node 10, extending through Village Node 3 to Village Nodes 38 and 41. The route is divided into two directions. One branch extends westward through Village Node 67 to Village Node 69, and another branch extends southward through Village Node 44 and Village Node 86 in turn and finally reaches Village Node 95 (Fig. 5). The western section of the route is mainly distributed along National Highway G345 and Shitian Expressway, while the eastern section extends along G345. Settlements with high transportation links are mostly distributed on both sides of these routes. By 2020, with the improvement of transportation conditions around Baishuijiang Town in the north and Heihe and Xiakouyi towns in the east, the traffic pressure of National Highway G345 can be effectively shared by surrounding provincial and rural roads. As a result, the betweenness index showed a more balanced distribution trend, and the accessibility of rural settlements significantly improved. The restrictive effect of terrain factors on regional rural development was weakened, and the impact of national highway changes on the region was further diminished. Additionally, the transportation hub function of the county town gradually weakened, with the traffic center shifting to the southwest around the Jieguanting Community.

Fig. 5: Betweenness of society-space network in Lueyang County.

The figure shows the comparison of the betweenness in the society-space networks between 2010 and 2020. The larger the value, the stronger its connectivity in the network.

As shown in Fig. 5, in 2010, the main social connections in Lueyang County were mostly concentrated in the area around the county town. Most of the other nodes with higher betweenness occupied the structural hole position and were scattered. This indicated that traffic conditions had a decisive influence on regional collaboration, but the cooperation radius at that time was relatively limited. By 2020, with the improvement in the county’s transportation environment, the radiating capability of the county seat over surrounding areas significantly increased. At this point, two important social connection paths formed within the county: both started from Village Node 81 in the southeast, passing through Village Node 86 to the county seat. One path extended northward through Village Node 25 to Village Node 138, while the other extended westward through Village Node 10 to Village Node 13. In real space, the former path is distributed along National Highway G345 and County Road X222, while the latter extends along G345. Settlements with strong connection intensity are mainly distributed along G345, X222, the Xibai Highway, and Lingyan Road. Compared with high-grade highways, provincial roads and county roads had a more significant impact on the connectivity between rural settlements.

Buffering

As shown in Fig. 6, in 2010, the transportation network of Lueyang County had buffering capacity primarily only in the county seat and its surrounding areas. Due to terrain constraints, other regions in the county had weak transportation connections with the central county seat, resulting in a high overall network vulnerability. By 2020, optimization of the transportation network led to the formation of four zones with strong risk resistance, namely the Fengjiaba-Matiwan area, Chengguan-Xingzhou area, the Baiquesi-Lesuhe area, and the Heiheba-Gaojiakan area. With the reduced influence of terrain constraints, the crisis-buffering performance of the transportation network improved significantly.

Fig. 6: : Eigenvector centrality of society-space network in Lueyang County.

The figure shows the comparison of the eigenvector centrality in the society-space networks between 2010 and 2020. The larger the value, the more connections a node has with important nodes, and the higher its own importance in the network.

As depicted in Fig. 6, in 2010, the risk resilience of Lueyang County’s social network was only reflected in the county seat, Village Node 25, Village Node 108, Village Node 86, Village Node 45, and their neighboring areas. Due to insufficient transportation connections, other regions find it difficult to establish effective connections with these core nodes, making the social network generally vulnerable. In 2020, the social network developed into three areas with strong buffering capacity, and its spatial distribution basically overlapped with the first three zones of the transportation network. However, the Fengjiaba–Matiwan area had yet to form stable social connections.

Stability

As shown in Fig. 7, the “k-core” analysis of Lueyang County’s transportation network in 2010 revealed a maximum core level of 4-core. High-core areas were limited and evenly spread across five regions within the county. Among them, the proportion of vulnerable areas with 2-core and below accounted for 41.72%, indicating that the overall network structure was significantly fragile. By 2020, the maximum core level of the transportation network increased to 5-core, with the proportion of vulnerable areas dropping significantly to 16.56%, thereby significantly enhancing network stability. At this time, the county’s transportation network formed a tiered structure comprising 1 primary core area and 5 secondary core areas.

Fig. 7: k-Core of society-space network in Lueyang County.

The figure shows the transition of the k-core from a fragile structure in 2010 (max 4-core) to a more stable tiered system in 2020 (max 5-core), reflects significantly enhanced network stability through the formation of primary and secondary core areas.

In Fig. 7, the social network’s “k-core” analysis in 2010 had a maximum core value of 4-core. The high-core areas were concentrated along National Highway G345 and Provincial Road Lueguan Road, though their number was relatively limited. The proportion of vulnerable areas (2-core and below) reached 64.90%, reflecting the fragile characteristics of the social network structure. In 2020, the maximum core level of the social network rose to 5-core, and the proportion of vulnerable areas decreased to 22.51%, indicating a notable improvement in network stability. The stable social network area formed two cores: one around the county seat and its surroundings, and the other in the southeastern area centered on Village Node 86, Village Node 94, and Village Node 101. The latter has a wider coverage and stronger stability, especially the settlements of Jieguanting, Heihe, and Xiakouyi towns are tightly interconnected, showcasing strong urbanization potential.

Restorability

In 2010, the network efficiency index of the transportation network was 0.71, and the nodes that affected the network efficiency were mainly concentrated around the county. By 2020, network efficiency improved by 3.14%. The nodes that affect the network efficiency are still concentrated around the county, and extend along the National Highway G315 and the Shitian Expressway in the east-west direction. In particular, the rural settlements located at the intersection of these major roads (such as Village Nodes 94, 38, and 150) emerged as key nodes crucial for regional resilience recovery. In comparison to the transportation network, the social network’s efficiency significantly increased by 22.2% over the decade, and its resilience to social crises was significantly enhanced. The nodes impacting social network efficiency were also concentrated around the county seat and along provincial and county roads.

Spatial optimization results

Spatial structure planning

As shown in Fig. 8, in 2020, the social network of Lueyang County formed a total of 8 coordinated development units. Under the guidance of transportation, the social connections between settlements within the county gradually break through the administrative boundaries of towns and villages, with more robust cross-regional passenger flow and logistics. Generally, the boundaries of these coordinated development units align closely with township borders, with each unit comprising one to four townships. Based on the results of the resilience assessment, the following management strategies are proposed: First, it should give priority to building 28 village nodes with outstanding society-space resilience, such as 13, 20, 19, and 38, as central communities. Second, it should improve the layout of passenger network stations by adding a Level 4 passenger station in the county seat, expand passenger stations in Xihuai Town, Matiwan Town, Guanyinsi Town, Hejiayan Town, and Xujiaping Town, and establishing 18 Level 5 passenger stations in central communities such as those at Village Nodes 12, 29, 50, and 140. Third, differentiated characteristic centers are set up according to the unit development type. For example, a commercial service center is built in the XZ-HXH-XJP unit (Xingzhou-Hengxianhe-Xujiaping unit), an industrial development center is built in the JGT-XKY unit (Jieguanting-Xiakouyi unit), and a leisure and tourism center is built in the WLD unit (Wulongdong unit), so as to improve the travel convenience of residents in the unit and promote industrial collaboration and sharing of public service facilities. Finally, cross-unit traffic connections are strengthened through local road optimization, such as adding connecting roads between the BSJ-MTW unit (Baishuijiang-Matiwan unit) and the WLD unit and the XZ-HXH-XJP unit to enhance the spatial linkage of the northern unit of the county. (In the text, XZ, HXH, XJP, etc., are the initial letter combinations of the township names.)

Fig. 8: Spatial structure planning of Lueyang County.

The figure shows eight coordinated development units represented by colored regions that transcend administrative boundaries. It identifies 28 high-resilience village nodes as central communities and optimizes the passenger station hierarchy. Differentiated functional centers are designated by unit type, and facilitating cross-unit spatial linkage through optimized road connections.

Rural classification planning

As depicted in Fig. 9, the villages of the centralized promotion type are mostly distributed at the entrances and intersections of national roads, provincial roads, and county roads. The villages of the reserved development type are mainly concentrated in the central Xumu Road, the southern Luezhong Road, the southeastern National Road G345, the Shitian Expressway, and the Luening Highway. They are showcasing significant development potential due to their robust industrial connections with surrounding areas. The villages of the characteristic protection type are concentrated on both sides of the Baocheng Railway; this area is also the early settlement focus area. Meanwhile, the villages of the resilient shrinkage type and the relocation type are mostly located in the far suburbs. The specific development guidelines of the six types of rural settlements are detailed in Table 2.

Fig. 9: Classification results of rural settlement in Lueyang County.

The figure shows six development types categorized by spatial resilience and its variations: centralized promotion (red) at road intersections; reserved development (orange) along major highways; characteristic protection (purple) along the Baocheng Railway; and resilient shrinkage (green) and relocation (yellow) in far suburbs. These types are integrated within synergistic development units (black boundaries) to optimize regional development potential.

Table 2 Development guides for rural settlements in Lueyang County.

Discussion and Conclusion

This chapter discusses the advantages of this study in urban and rural planning, analyzes the contribution of the society-space network model to settlement spatial layout, offers recommendations for the spatial arrangement of mountain county settlements, and puts forward the conclusions.

Discussion

Advantages in urban and rural planning

In mountainous and similarly complex topographical areas, the spatial form of county settlements typically aligns along transportation networks, presenting a branched pattern in physical space. However, the organizational patterns driven by social demand tend to form a networked structure. The findings of this study align with those of Hui (Hui, He, Ma, & Hui, 2014), indicating that the influence range of mountainous settlements is primarily determined by traffic conditions, resulting in a heterogeneous settlement system. Since this study takes into account both transportation networks and social needs, the planning results are more economical and targeted. In traditional urban and rural planning, performance evaluation and spatial optimization are mostly based on the administrative units of villages and towns, and statistical analysis is used as a research method (Tang, He, Zhou, Zeng, & Xiao, 2018). Due to the neglect of cross-administrative horizontal society-space relationships, the flow and interaction of elements within administrative regions may lead to compositional fallacies. Typically, inter-relationship models guide spatial management through spatial element matching. Such inter-relationships are static associations derived from long-term summaries (Haynes & Fotheringham, 1984). However, rapidly changing social relationships can lead to spatial mismatches. To solve the above problems, this study regards all settlements in the county as a closely connected urban-rural settlement system at the spatial level. By coordinating transportation networks and social needs, potential settlements and connections are identified, and their impact on spatial resilience is revealed. At the temporal level, the dynamic operation laws of the settlement system are better revealed through network changes, and dynamic and resilient spatial optimization solutions are proposed. Simultaneously, based on the intensity of network structure, this study proposes the concept of synergistic development units to coordinate the development of closely linked townships. In the planning and control of mountain settlements, this approach not only better achieves resource complementarity and public service sharing but also promotes industrial agglomeration and distinctive development. In the territorial spatial planning of Lueyang County, we drew upon the results of spatial zoning to guide industrial layout. Nevertheless, compared with plain areas, the population changes in mountainous areas are more significant, with seasonal migration characteristics (Wang, Wang, Du, Yi, Liu, & Tu, 2021). The data in this study mainly uses an annual statistical period. To develop more accurate planning schemes, it is also necessary to improve data accuracy, such as obtaining mobile phone signaling data and GPS trajectory data, and establishing a planning information platform to track the movement of people, homesteads, and cultivated land.

Contribution of the society-space network model to settlement spatial layout

The “society-space” concept is crucial for understanding the essence of settlements (Wang & Chiou, 2019). Settlements reflect social relationships and manifest in spatial distribution, form, appearance, and order. At the same time, settlement space also participates in the construction of social relations, reflecting the relationship of distribution, collaboration and control (Zhou & Xiong, 2019). However, current research is still insufficient in describing the sensitivity of short-term changes in settlement space and spatial accuracy, making it difficult to clearly depict the society-space dynamic changes of settlements in disturbances. By analyzing the changes in the society-space network, this study reveals the dynamic matching relationship between social demands and transportation networks. Based on a multi-stage resilience assessment, this study achieves a transition from static urban and rural planning to dynamic resilience planning. Additionally, by connecting social networks with physical components, the study can accurately identify real spatial elements, thus avoiding misleading analyses caused by hollow settlements in mountainous areas. Further comparison of this study with actual rural classification planning shows that most of them are in line with the actual situation. However, unlike the actual planning, county spatial layouts should not strive for balance under the guidance of transportation. Instead, specific developmental priorities for Lueyang County are proposed. The county seat should collaborate with national and express highways to prioritize development in southeastern towns. Moreover, the five types of space resilience assessment indicators proposed in this study demonstrate considerable validity and could be extended to spatial resilience assessments in mountainous regions. Notably, the eigenvector centrality indicator can comprehensively consider the advantages of the coordinated development of the evaluated settlement and the surrounding settlements, and the error is relatively smaller.

Recommendations for the spatial layout of county settlements in the Qinba Mountain area

In the county unit of Qinba Mountain area, the transportation network is a key factor affecting the social demands of settlements and even spatial evolution. Highways and national roads determine the collaboration of regional settlements and profoundly affect regional industrial collaboration and logistics layout. Provincial and county roads have a significant impact on the connectivity of rural settlements and can increase the service range of commercial and public facilities. Thus, the optimization of settlement space in mountainous counties should consider the following aspects: ① Based on the major external transportation links, logistics and passenger flow centers should be built at nodes with high connectivity and collaboration to promote the gathering of settlements on both sides of the road and improve the connectivity efficiency of regional settlements. ② In Qinba Mountain county units, suitable development areas, controlled construction areas, and relocation/merge zones account for approximately 10%, 60%, and 30%, respectively (Wu, Jing, Guo, Li, & Wei, 2016). County towns should be guided to develop preferentially in one direction, with 1-2 key developmental zones identified at major external contact hubs. ③ For non-key construction areas, the scale of construction land and cultivated land should be controlled, and residents should be gradually guided to move out. Rural idle land should be reclaimed as arable land to generate surplus indicators of construction land, which can be replaced to development zones through the network platform to feed back the countryside with land premium (Guo, Zhao, Li, Mao, He, & He, 2024). After the villagers are compensated, 60% of the funds will be reserved for road construction and industrial support. ④ Within the collaborative development zones proposed in this study, the logistics network and regional characteristic industries should be combined to focus on the development of a leading industry. The findings presented in this study has informed the territorial spatial plan for Lueyang County. The relevant recommendations could be adapted to guide spatial planning in analogous mountainous regions across Shaanxi and Gansu provinces.

Conclusion

In the study of the spatial resilience of county settlements, this study innovatively uses the society-space network to coordinate changes in social functions and spatial elements. It achieves a dual-focus examination of space resilience under the influences of both physical and social connections. This approach enhances experimental precision at both the temporal and spatial levels, which is helpful to propose effective spatial optimization strategies. Using Lueyang County in the Qinba Mountain area as an example, the study first utilizes a two-mode network to simultaneously reveal the changing characteristics of transportation and settlements, enriching the paradigm of county-level space resilience research. Secondly, it proposes five types of space resilience assessment indicators for mountainous counties. These indicators can identify key areas, settlements, and road connections, and enrich the resilience assessment indicator system as well as the model’s application scenarios. Lastly, by matching elements like communities, key nodes, and edges from network research with spatial structure planning in urban-rural planning, the practical application of research findings can be facilitated. The research shows that over the past decade, Lueyang County has improved its social needs by 44.32% by improving the transportation network in the marginal areas, and formed a distinct industrial development belt along National Highway G345. With the rapid evolution of the transportation network, the collaboration, stability and resilience of the county settlements increased by 15.8%, 42.4%, and 22.2%, respectively, resulting in 2 clustered development belts and 3 well-buffered settlement groups. Ultimately, the study proposes the formation of 8 collaborative development zones within the county, the addition of 19 passenger stations, the focused development of 28 center communities, and 6 types of rural development categories with guidance. This achievement can provide guidance for the coordinated development of mountain settlements, and serves as a basis for county-level urban-rural integration, village layout, resource allocation, and construction guidelines. The study has two prominent advantages. First, it considers both transportation networks and social demands, making planning outcomes more economically viable and targeted. Second, in the research of spatial resilience, it describes the dynamic socio-spatial changes during disturbances, enhancing sensitivity to temporal changes and accuracy in spatial element identification. Due to the complex social relations, significant population changes, and periodic migration characteristics in mountainous areas, future research will focus on increasing data precision, involving segmenting social relationship types based on the travel behaviors of citizens and villagers. The different types of society-space resilience and adaptability will be a direction for future research.