Spatiotemporal Characteristics and Determinants of Tourism Cooperation Network in Beijing–Tianjin–Hebei Region

Abstract

The optimization of the cooperation network is a key link to accelerate the high-quality development of regional tourism. Taking the Beijing–Tianjin–Hebei region as an example, this paper measures the tourism cooperation intensity with the modified gravity model, on which the original, binary, and Top networks are generated to identify the spatiotemporal evolution characteristics from the multi-dimensional difference–association–agglomeration model, and provide insight into the determinants by the GeoDetector model. The results show that (1) the cooperation network reveals a diffusion trend with Beijing–Tianjin as the main axis chain, and southward expansion, and the overall differences tend to moderate at a slow pace, among which the north is the weak area. (2) The robustness of the cooperation network association structure is enhanced, showing that the outgoing equilibrium is improved, while the cohesion is strengthened and accessibility changes little. Furthermore, the cities show a core–edge distribution pattern in terms of power roles. (3) The cooperation network shows the phenomenon of hierarchical agglomeration gradually with the expansion of network scale, and eventually evolves into two camps: the Beijing–Tianjin cooperation circle and the Shijiazhuang–Xingtai cooperation circle. (4) Tourism cooperation belongs to the locational traffic constraint type, and making up for the shortcomings of rural development is another key to further enhancing regional tourism cooperation. The future optimization of regional tourism cooperation needs to seek multifactorial promotion paths.

1. Introduction

Cooperation is an important model for the healthy and efficient development of regional tourism [1], which can effectively avoid vicious competition, promote rational allocation and overall optimization of elements within the tourism system, and accelerate regional tourism’s high-quality development. The Beijing–Tianjin–Hebei region is geographically close, with deep historical sources, firm industrial ties, and complementary resources [2,3]. It has superior conditions for tourism spatial cooperation and natural advantages for collaborative development [4], as well as being the first region in China to propose regional tourism cooperation. The earliest can be traced back to the 1980s and 1990s when the Beijing–Tianjin–Hebei region took the lead in establishing the Beijing-East Tourism Zone. Since 1987, the three areas have held regular “Beijing–Tianjin–Hebei Regional Tourism Cooperation Seminars” and reached a preliminary “Beijing–Tianjin–Hebei Consensus on Barrier-Free Tourism” in 2004, then officially signed the “Beijing–Tianjin–Hebei Tourism Cooperation Agreement” in 2007. In 2014, the coordinated development of the Beijing–Tianjin–Hebei region was elevated to a major national strategy, and tourism cooperation was regarded as a critical path. The following year, the 515 strategy proposed to break regional barriers and promote regional tourism integration, and “The 13th Five-Year Plan for Tourism Development” also included the Beijing–Tianjin–Hebei as one of the key points of strengthening trans-regional tourism city clusters. “The Beijing–Tianjin–Hebei Tourism Cooperative Development Action Plan (2016~2018)” responded positively, emphasizing the construction of a zero-distance interchange system for tourists, working together to build five tourism cooperative development demonstration areas, with basic completion in 2019. To further enhance the cultural driving force of Beijing–Tianjin–Hebei synergistic development, the three areas collaborated to legislate. Since 2023, the “Decision on Beijing–Tianjin–Hebei Synergistic Promotion of the Grand Canal Cultural Protection and Heritage Utilization” has been implemented, to provide the protection of the rule of law for the synergistic promotion of the Grand Canal cultural protection and heritage utilization. This shows that the Beijing–Tianjin–Hebei region continues to pay attention to tourism economic agglomeration, regional connectivity, and policy synergy efficiency.

At the same time, driven by the continuous innovation of internet and transportation technology, the three areas of Beijing, Tianjin, and Hebei are also developing towards a connected and interactive community, gradually evolving into a comprehensive system with multiple elements closely linked [5]. However, due to the differences in tourism resources, economic layout, political status, geographical location, etc., the gap in tourism benefits among the areas is increasing year by year, and the interests of all parties are difficult to coordinate, leading to independent development, insufficient cooperation, and a resulting lack of cohesion [6,7]. The antagonistic joint relationship of competition and cooperation has led to a disjointed and interconnected structure of association within the region. In the face of optimal adjustment of epidemic prevention and control measures, the recovery of the tourism market is gaining momentum, and the pursuit of a single administrative region tourism boost is likely to lead to wasteful competition and a blowout of tourists [8]. Driven by the maximization of symbiotic interests and the inherent needs of industrial development, it has become a consensus to move from isolated construction and territorial management to the construction of a connected and complete tourism cooperation network across administrative boundaries to seek common development of regional tourism. Therefore, exploring the basic laws of regional tourism cooperation, systematically cognizing the spatiotemporal changes of regional tourism cooperation networks, and gaining insight into the determinants of cooperation are important supports for optimizing the tourism development pattern of the Beijing–Tianjin–Hebei region, which are also key links for achieving high-quality regional tourism development.

2. Literature Review

As an important factor affecting regional tourism interrelationships and economic behavior, cooperation has been a hot issue in the tourism field. Most of the studies are based on symbiotic cooperation theory, coordinated development theory, and policy management theory, mainly focusing on cooperation agent behavior [9,10,11,12], cooperation model construction [13,14,15,16], cooperation countermeasure optimization [17,18], cooperation governance policies [19,20], and cooperation progress evaluation [21], with qualitative research methods. Particularly, the research focus on tourism cooperation driven by external forces or internal factors, which are common in events [22,23], traffic infrastructure [24,25,26,27], and related policies [28]. The research scope involves cross-border [29,30,31], interregional [32], and even among scenic spots [33,34] or hotels [35].

In recent years, scholars gradually realize that tourism cooperation has gone beyond the proximal relationship in the purely geographical sense and instead presents a more complex spatial network structure form [36]. The research begins to go deeper into the direction of reconstructing and optimizing the spatial network pattern of regional tourism cooperation. Thereinto, the construction of network structure is mainly based on agreements (formal and informal, global and regional) [37,38], web-based text data [39], interviews, and questionnaires [40]. As well as relying on statistics as the data source, a few scholars also attempt to measure the extent of tourism cooperation by the harmonized cooperation index [41] and gravity model [42,43]. On this basis, ArcGIS is used for visual display to form a spatial hierarchical network based on geographical location, focusing on the connection between network nodes and geographical location, as well as identifying the spatial allocation relationship between nodes and tourism elements, and then summarizing the spatial differentiation law of tourism cooperation network structure [44]. Further, ArcGIS spatial analysis tools, such as buffer zone analysis and overlay analysis, are used to reveal the influencing factors [45]. With the introduction of Social Network Analysis (SNA), this provided a new perspective for the study of tourism cooperation network, and quantitative spatial research to explore the characteristics of spatial network structure began to emerge. The mainstream research idea is to conduct threshold processing on the basic coefficient matrix, convert it into a binary matrix, and then measure network density, network type, node centrality, and cohesive subgroups with numerical values, to explore the node status and overall pattern of the cooperative network from both the overall and individual aspects [46,47,48]. It is an important contribution to the analysis of the tourism level of each node, the tourism development links between nodes, and the future direction of regional cooperation. In addition, researchers are also committed to establishing a correlation between changes in the structure of tourism cooperation networks and tourism flows, investigating the influence on bilateral tourism flows [49,50].

By reviewing the existing studies, we found that the existing tourism cooperation studies still have the following shortcomings: (1) The quantitative construction of regional tourism cooperation is insufficient, i.e., there is a lack of analytical basis for the spatial network structure. (2) The identification of cooperation network characteristics mostly focuses on a single geographical network or social network, i.e., the identification of spatial network structure features is not comprehensive and accurate. (3) The influencing factors of cooperation network distribution and evolution are still mainly discussed qualitatively, supplemented by simple statistical analysis or spatial analysis tools, and lack the interaction detection of influencing factors. In short, quantitative analysis and quantitative attribution is still the weak link in the existing research on tourism cooperation, while an increasingly improved cooperation network is bound to become an important cornerstone to guiding Beijing–Tianjin–Hebei regional tourism towards high-quality development. Given this, this paper takes the Beijing–Tianjin–Hebei region as an example, focuses on the scientific problem of identification of characteristics and detection of driving factors of regional tourism cooperation pattern evolution, carries out tourism cooperation intensity measurement and ephemeral analysis, summarizes the characteristics and laws of the cooperation network at multi-dimensional levels and examines several factors affecting tourism cooperation, to provide scientific support for the synergistic development, misplace development, and win-win development of tourism in the Beijing–Tianjin–Hebei region.

3. Materials and Methods

3.1. Research Framework

Regional tourism cooperation network is the projection of regional tourism cooperation relationship in geographic space. Due to the spatial heterogeneity of regional development, the spatial interaction is intricate and complex, but it is undeniably this complex spatial connection and linked spatial network structure that drives the flow and integration of tourism elements within the region. This paper first adopts the modified gravity model to quantitatively measure the intensity of inter-regional tourism cooperation as the side weights of the cooperation network. On this basis, we construct the feature extraction framework model of difference–association–agglomeration, and accordingly generate the original, binary and Top network to comprehensively analyze the spatiotemporal evolution characteristics of the cooperation network. Specifically: (1) Spatial difference perception. Aiming at the weighted directed original network with mutual relationships between any two nodes, GIS is used to describe the spatial differential evolution characteristics. (2) Spatial association identification. The original network is binarized and transformed into an unweighted directed binary network with selected thresholds, focusing on observing the connectivity between cities after the cooperation intensity reaches a certain standard. This part elucidates the overall association characteristics of the network and the functional roles of cities in the network with the help of common indicators of the SNA model. (3) Spatial clustering determination. Based on the cooperation intensity ranking results, the Top 1, Top 3, and Top 5 networks are generated orderly, and the spatial clustering characteristics of the networks are portrayed by using indicators from the complex network, and the closely connected city clusters are identified. Finally, Geo Detector is used to detect the dominant factors and interactions in the spatiotemporal evolution of the cooperation network patterns, to provide a scientific index for the precise enhancement of regional tourism synergy.

3.2. Methods

3.2.1. Improved Gravity Model

The measurement of tourism cooperation intensity is mainly based on the gravity model, while the classical gravity model to a certain extent does not fully reflect the cooperation links, mainly considering two limitations [51,52,53,54]: The first is the inability to distinguish the directionality and difference of tourism cooperation between cities, ignoring the key point that the status of tourism cooperation is not equal, which is spatially expressed as an undirected network. The second is the neglect of the spatiotemporal compression effect of modern transportation networks on the correction of the spatial distance between cities, which can easily cause bias in the research results. Therefore, this paper tries to improve the conventional model factors of the classical gravity model and further selects the abundance of tourism resources as variables to construct correction coefficients, to reflect the directional variability of tourism linkages. The formulation is as follows:

$$TC{I}_{ij}=k_{ij}\frac{\sqrt{P_i{V}_i}\sqrt{P_j{V}_j}}{O{D}_{ij}^2}$$

(1)

$$k_{ij}=\frac{r_i}{r_i+r_j}$$

(2)

$$r=5n_5+2.5n_4+1.75n_3+0.5n_2+0.25n_1$$

(3)

where TCI_ij is the tourism cooperation intensity between city i and city j; P and V represent the total number of tourist arrivals and total tourism revenue, respectively; OD_ij is the shortest travel time, which is obtained from GIS cost-weighted distance; k_ij is the correction coefficient; r represents the abundance of tourism resources; and n₅~n₁ represents the number of 5A~1A scenic spots, respectively. On this basis, based on the intensity of tourism cooperation in year t, the directed weighting matrix of regional tourism cooperation is established as follows:

$$TC{I}_t=\left[\begin{array}{cccc}0& TC{I}_{12,t}& \cdots & TC{I}_{1n,t}\\ TC{I}_{21,t}& 0& \cdots & TC{I}_{2n,t}\\ ⋮& & \ddots & ⋮\\ TC{I}_{n1,t}& TC{I}_{n2,t}& \cdots & 0\end{array}\right]$$

3.2.2. Geographical Network Analysis

Geographic Information System (GIS) is a computer system for collecting, processing, analyzing, expressing, and applying geographic information, which combines spatial information of geographic objects with relevant attributes and then represents them in the visual form [55,56]. In this paper, the tourism cooperation intensity TCI_ij is first performed by the raster calculator, and then the spatial superposition analysis method is used to superimpose larger values to generate the original cooperation network, with cities as network nodes and tourism cooperation intensity between cities as edge weights. The shades of color of linking line segments between nodes reflect the high or low gaps in tourism cooperation intensity.

3.2.3. Social Network Analysis

B Social Network Analysis (SNA) is a relatively mature method to study binary value networks in the social science field, based on relational data to visualize the relational structure of the research object [57]. Its basic principle is to retain the edges with greater strength of action in the network to identify the nodes in key positions and the groups with strong ties [58]. In this paper, we first determine the cut-off value, following the principle of retaining valid information and comparability, to pre-process the tourism cooperation intensity TCI_ij, and then transform the weighted directed original network into the unweighted directed binary network. Secondly, the overall characteristics of the cooperative network and the role of individuals are dissected by relevant indicators. The specific indicators and calculation methods are described in the literature [59,60].

3.2.4. Complex Network Analysis

Based on the ranking results of inter-city tourism cooperation intensity TCI_ij, the Top network is constructed to capture the agglomeration characteristics of the cooperation network at different scales. The Top 1 network reflects the proximity connections between neighboring cities and the formation of spatial agglomeration units. The Top 3 network shows the intertwined connections between nodes within agglomeration units and the proximity connections between agglomeration units. The Top 5 network further reveals the meritocratic connection characteristics between agglomerations on a wider range. In this paper, the degree distribution (k), average path length (L), and average clustering coefficient (C) in complex networks are selected as the parameters to describe the agglomeration of the Top network, and the specific evaluation indexes and calculation methods are described in the literature [61,62,63].

3.2.5. GeoDetector Model

The formation and evolution of the cooperation network are influenced by various factors, and most of the existing studies have been analyzed from the government dimension [17,31,32], tourism infrastructure [32], consumption level [31,36], security [31,36,64], etc., but less from a structuralist-based relational perspective to explore the influence of factor differences on cooperation networks. Therefore, synthesizing the existing relevant research results and fully considering the data availability and typicality, this paper takes tourism cooperation intensity TCI_ij as the explanatory variable and selects explanatory variables from three dimensions of destination attractiveness, origin travel capability, and spatial distance, then the constructed index system is shown in

Table 1. The indicator system of the tourism cooperation network.

Dimension	First-Class Index	Second-Class Index
Destination attractiveness	O1 Infrastructure	O11 Percentage of greenery coverage
		O12 decontamination rate of domestic waste
		O13 Public toilets per 10,000 people
		O14 Road area per capita
	O2 Reception facilities	O21 Number of scenic spots 4A and 5A per unit area
		O22 Number of travel agencies per unit area
		O23 Number of star hotels per unit area
	O3 Service level	O31 Percentage of tourism employees
		O32 Percentage of education spending
	O4 Environmental assurance	O41 Excellent rate of air quality
		O42 Vacation climate comfort index
Origin travel capability	D1 Population status	D11 Density of resident population
		D12 Percentage of urban population
		D13 Urban registered employment rate
	D2 Living standards	D21 Per capita urban disposable income
		D22 Per capita rural disposable income
		D23 per capita savings balance
	D3 Consumption level	D31 Per capita urban consumption expenditure
		D32 Per capita rural consumption expenditure
		D33 Urban Engel coefficient
		D34 Rural Engel coefficient
Spatial distance		ODt Shortest travel time

. Factor detection and interaction detection are used to identify the key factors and their interaction relationships, which affect regional tourism cooperation. The specific calculation methods are shown in the literature [65,66].

3.3. Data Collection

The research data mainly involve two types: (1) Economic and social statistics. The data are mainly from The China City Statistical Yearbook, China Tourism Statistical Yearbook, China Tertiary Industry Statistical Yearbook, The Statistical Yearbook, and The Statistical Bulletin of National Economic and Social Development of the corresponding year. (2) Geographic data. The administrative zoning map data is obtained from the National Earth System Scientific Data Sharing Platform (http://geodata.cn/, accessed on 1 January 2020), and the modern transportation network base map data are obtained by manual digitization based on the opening of traffic in previous years.

4. Results

4.1. Characteristic Analysis of the Cooperation Network

4.1.1. Spatial Difference Characteristics Perception Based on the Original Network

Based on the cooperation matrix $TC{I}_t=\left\{TC{I}_{ij,t}\right\}\left(i,j=1,\cdots ,13;2014\le \mathrm{t}\le 2020\right)$, the spatial topology of the cooperation network is visualized by ArcGIS, while the cooperation lines are hierarchically divided, making the spatial differences more prominent (Figure 1).

From the value of the cooperation lines, in the 78 directional cooperation relations shown, the upper limit of the highest-level cooperation line continues to rise from 11.56 in 2014 to 26.08 in 2020, the intensity of tourism cooperation gradually increases, and the inter-city ties become more solid. Meanwhile, the city combination of medium and high value increases significantly, indicating that the low cooperation intensity city combination has a catch-up effect on the higher combination. The top three city combinations are Beijing ← Tianjin, Beijing ← Langfang, and Tianjin ← Langfang, which gather 64.12%, 64.07%, 61.31%, and 56.66% of the total cooperation intensity in turn by only 3.85% of the path in the region, which are the core of the whole cooperation network, holding the absolute network power. However, it is noted that the proportion is continuously decreasing, which indicates that the centralization trend of the cooperation network is continuously weakening, the differences tend to moderate, and the inter-city tourism cooperation links are gradually developing in the direction of intensification and multidistrict. The reason for this is that, in addition to the own tourism advantages, the practicing of the “Four Vertical and Four Horizontal” high-speed railroad network system, while optimizing the tourism travel system and reducing the perceived distance of tourists, also relatively weakens the polarizing effect of the core cities brought by their location advantages and reduces the overall regional differences.

From the distribution pattern of cooperation lines, the network shows the spatial non-homogeneous characteristics of overall loosening and local aggregation, and tends to be highly dependent on the axis–spoke structure of the core nodes, specifically manifesting the diffusion trend of Beijing–Tianjin as the main axis chain and southward expansion. The overall pulse and pattern of the cooperation network highlight the intuitive gap between the Beijing–Tianjin advantageous area and other cities, and its peripheral driving effect from near to far. Although the connection is getting closer and more solid, the diffusion speed is slow and the long-range radiation capacity is still slightly insufficient. According to the spatial separation state of the cooperation line, the network can be roughly divided into three large areas:

• The central compact area, with the densest cooperation line, containing Beijing, Tianjin, Langfang, Tangshan, Cangzhou, and other cities, has a pivotal role in the key link from central to southwest. In particular, Beijing and Tianjin pull the development and evolution of the cooperation network, but the spillover around Beijing is low. The reason lies in the fact that Beijing, as a more mature tourist destination, is rich in resources, carrying a large tourism consumer crowd, which produces a tourism shading effect on Tianjin and Hebei. Moreover, Beijing tourists tend to be more cultured, more open-minded, and thus extremely picky about tourism products, while the tourism resources endowment of the surrounding cities is hardly excellent, and there is a lack of tourism products. Given the lack of coastal tourism resources in Beijing, in the future Tianjin can be considered as the core, driving Cangzhou, Qinhuangdao, and Tangshan to create a seaside leisure and resort development belt, attracting Beijing tourists.
• The southwest development area, which initially presents a simple linear structure, gradually forms an interlocking network structure as the construction of destinations accelerates and the network system of transportation is completed. It includes Baoding, Shijiazhuang, Xingtai, Handan, and Hengshui, among which Shijiazhuang is an important bridge node for the network to develop to the south of Hebei. Shijiazhuang not only has the “Red, Ancient and Green” tourism resources, which is an important bearing space for the tourism flow, but also actively connects with Xingtai, Handan, and other cities in south Hebei, and its status as a transit hub linking Beijing, Tianjin, and south Hebei is becoming increasingly obvious. Note that as the cities in the area have been connected by high-speed rail, attention should be paid to the phenomenon of the high-speed rail “double-edged sword”, focusing on new challenges such as increased competition in tourism along the route cities to continue to strengthen the division of labor in order to achieve a mutually beneficial regional win-win cooperation.
• The weak Northern area shows almost no line connection point structure; the network structure has not been formed. For a long time, there has only been a weak cooperative relationship between Qinhuangdao to Beijing–Tianjin–Tang, which restricts the development of regional tourism synergy. In the context of the Winter Olympics, the Beijing-Zhangzhou high-speed railway, which is a key supporting traffic project, opened to traffic at the end of December 2019. Zhangjiakou benefits from this transport corridor, making it significantly more accessible to other cities, solving its inherent geographical location disadvantage, easing the long-standing marginalized situation, and improving the amount and magnitude of changes in tourism cooperation, while, until the May Day holiday in 2020, Chengde was isolated due to the lack of high-speed rail access and the absence of high-value cooperation lines connecting it to major cooperation networks. In the future, development should be further accelerated to achieve high-speed rail connections between Chengde and other cities through policy guidelines, and at the same time to break the tourism seasonal bottleneck, expanding tourism. Compared with other areas with full high-speed rail lines, the future expansion of tourism cooperation in the northern region should be more extensive, which will help to further reduce the gap between cities in the Beijing–Tianjin–Hebei region in the future.

4.1.2. Spatial Association Characteristics Identification Based on the Binary Network

Through repeated experiments, this paper finally selects 0.4 as the breakpoint value to binarize the original network, mainly based on the following reasons: first, the value of tourism cooperation intensity accounts for 46.11% by 2020, retaining the main cooperation paths of different time sections while excluding the influence of chance on the authenticity of the data; and second, by initially highlighting the core nodes of the whole network, and then achieving the existence of linkage relationships of all urban nodes by 2020, the leakage network characteristics is lessened, making them more conducive to the exploration of network evolution laws.

The Overall Structural Characteristics of the Network and Its Evolution

In this paper, we mainly select the indicators of network density, intermediary central potential, and degree central potential to characterize the overall spatial association characteristics of the cooperation network (

Table 2. Overall structure index of the regional tourism cooperation network.

		2014	2016	2018	2020
Network density		0.1859	0.2436	0.3077	0.3333
Intermediary center potential		0.3355	0.2304	0.2906	0.3043
Degree central potential	Inward	0.4305	0.4583	0.6597	0.7222
	outward	0.4306	0.3681	0.3889	0.3611

).

The network density increased from 0.1859 in 2014 to 0.3333 in 2020, with an average annual growth rate of 10.22%, which shows that with the deepening of Beijing–Tianjin–Hebei synergistic development and tourism integration, interactions become more frequent and the robustness of the association structure gradually increases. However, the network density still does not reach 0.5, indicating that the cooperation network is relatively loose, many tourism cooperation linkage paths have yet to be scaled up, and that there is still much room for improvement.

Further, the centralization trend of the network is measured quantitatively. As for the degree central potential, the results show that the inward from 2014 to 2020 is always higher than the outward, and the gap continues to expand, which laterally verifies the research hypothesis that tourism cooperation between cities has significant directionality. Specifically, compared with the beginning of the period, the outward decreased by 16.13%, indicating that the outward equilibrium improved and the distribution of tourism sources tended to be decentralized. However, the inward increased by 67.74%, revealing that the cohesiveness strengthened, and tourism destinations may have gradually concentrated in a few cities. However, it is noted that the growth rate decreased after 2018, which, combined with the results presented in Figure 1, may be due to the increasing position of cities such as Shijiazhuang and Tangshan in the network, and the polycentric pattern slowing down its centralization trend. The intermediary center potential fluctuates between the interval [0.2304, 0.3355], proving that the degree of inter-city cooperation access is reasonable, and the role of intermediaries in the cooperation network does not change significantly.

Network Individual Location Relationships and Their Evolution

Further measurements of degree centrality (C_D), closeness centrality (C_C), and betweenness centrality (C_B)reveal that cities have different roles in the cooperative network, and their relative positions have not shifted significantly (

Table 3. Individual characteristic index of the regional tourism cooperation networks.

City	2014			2016			2018			2020
	CD	CC	CB	CD	CC	CB	CD	CC	CB	CD	CC	CB
Beijing	66.67	31.58	51.26	66.67	31.58	23.99	91.67	50.00	38.89	100.0	100.0	47.48
Tianjin	41.67	27.91	2.78	58.33	30.77	10.35	66.67	44.44	8.59	75.00	80.00	12.63
Shijiazhuang	25.00	27.91	24.24	41.67	29.27	24.24	58.33	42.86	7.07	58.33	70.59	7.58
Tangshan	33.33	27.27	0.51	33.33	27.27	0.51	33.33	38.71	0.51	41.67	63.16	1.01
Qinhuangdao	25.00	26.67	0.00	25.00	26.67	0.00	25.00	37.50	0.00	25.00	57.14	0.00
Handan	16.67	23.53	0.00	16.67	24.49	0.00	25.00	37.50	0.00	25.00	57.14	0.00
Xingtai	16.67	23.53	0.00	16.67	24.49	0.00	25.00	37.50	0.00	25.00	57.14	0.00
Baoding	8.33	25.53	0.00	25.00	26.67	0.00	33.33	38.71	0.00	25.00	57.14	0.00
Zhangjiakou	8.33	25.53	0.00	8.33	25.53	0.00	8.33	35.29	0.00	25.00	57.14	0.00
Chengde	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.00	8.33	52.17	0.00
Cangzhou	16.67	26.09	0.00	25.00	26.67	6.06	25.00	37.50	0.00	33.33	60.00	0.00
Langfang	25.00	26.67	0.00	50.00	30.00	0.00	50.00	41.38	2.53	50.00	66.67	2.53
Hengshui	0.00	0.00	0.00	0.00	0.00	0.00	25.00	37.50	0.00	25.00	57.14	0.00
Average	21.80	26.56	6.06	28.21	27.58	5.01	35.90	39.91	4.43	39.74	64.27	5.48
Standard	17.46	2.13	14.53	20.29	2.35	8.69	23.88	3.94	10.33	24.27	12.53	12.67

). Next, the evolutionary characteristics of the cities’ power roles are explored in depth.

First of all, comparing the degree centrality, we found that the overall trend was for growth and the average value of the four years was 21.80, 28.21, 35.90, and 39.74, indicating that tourism cooperation is growing closer and deeper. The degree centrality of Beijing and Tianjin is much higher than that of the cities in Hebei, playing the role of leader in the cooperation network. Beijing and Tianjin are rich in high-grade scenic spots, with a strong industrial base, especially Beijing, which also benefits from its position as the national political, economic, and cultural center and has the most connections with other cities. Tianjin comes second, and also shows strong core control and dominance in the cooperative network. In contrast, due to their economic base and resource abundance, cities in Hebei have a relatively weak influence and are in a subordinate position in the network, with Chengde, Zhangjiakou, and Hengshui being even more marginal and only associated with a few other cities. In addition, the growth rate of urban nodes is unbalanced, with cities in the central region such as Langfang and Shijiazhuang growing rapidly, while Chengde, Qinhuangdao, Tangshan, and Handan in the peripheral region grow slowly. As of 2020, the degree centrality of all cities is greater than 0, indicating that all cities are involved in tourism cooperation, while the degree centrality of Hengshui is still low and there is much room for cooperation.

Second, the ease of cooperation between cities is judged by the closeness centrality. The measurement results show that the average value was only 26.56 in 2014, and the connection between cities was relatively loose. After that, the average value increased steadily with an average annual growth rate of 15.87%, and the tightness between cities is getting stronger. By 2020, the average value had reached 64.27. However, at the same time, the difference between cities also showed a trend of expansion, and the standard increased from 2.13 in 2014 to 12.53 in 2020. During this period, the closeness centrality of Beijing, Tianjin, Shijiazhuang and Langfang is always higher than the average value, and this plays the role of the central actor in the cooperation network, making it easier to carry out tourism cooperation with other cities. Hengshui and Chengde have a low closeness centrality level in the early stage due to their remote geographical location, weak industrial development base, etc., cooperation is not obvious, although this has been improved in recent years.

Furthermore, the betweenness centrality is used to explore the ability to control resources in the tourism cooperation process. The results show that the average value reveals a U-shaped dynamic evolution trend, with relatively small changes compared to the beginning of the period. Only Beijing, Tianjin, Shijiazhuang, and Langfang have betweenness centrality greater than 0, indicating that most tourism cooperation relationships in the cooperation network are mediated and bridged through these three cities, which play the role of intermediaries. The betweenness centrality of Beijing is much higher than other node cities, and it occupies an absolute control position in the cooperation network, which indicates that Beijing has a strong influence and competitiveness in the tourism development of the Beijing–Tianjin–Hebei region. However, the betweenness centrality is significantly decreasing, and its role as the intermediary is weakening. Conversely, Tianjin and Langfang are actively assuming this role and gradually playing the role of conducting and pivoting in the cooperation network. In addition, more than half of the cities, such as Qinhuangdao, Handan, Xingtai, Baoding, Zhangjiakou, Chengde, and Cangzhou, whose betweenness centrality is always 0, have a strong dependency on the cooperation network, with a more passive role, receiving control and domination from other cities.

Finally, the degree centrality, closeness centrality, and betweenness centrality are summed with equal weights to obtain the power role coefficients, and then the cities are classified into four classes of core leaders, sub-core leaders, general collaborators, and marginal collaborators by the K-value clustering method. Among them, the core leader is Beijing, with all centrality indexes much higher than other cities, which has the strongest control over other cities and is the most important tourist destination and distribution center for tourists. The sub-core leaders are Tianjin, Shijiazhuang, and Tangshan, which are the secondary gathering centers in the middle, northeast, and southwest directions, respectively, and together drive the synergistic development of tourism in the Beijing–Tianjin–Hebei region. The general collaborators are Qinhuangdao, Baoding and other six cities, which are not prominent in the cooperative network. Usually, they lack high-level tourism resources, leading to a very limited attraction to tourists, and are neither important tourist destinations nor important distribution centers. However, they are geographically adjacent to the leaders, having certain locational advantages. Therefore, these cities need to strengthen the development of their tourism resources, make good use of their geographical advantage, maintain the depth and breadth of cooperation with the leader cities, and then achieve synergistic development with the leader cities. The marginal cooperators are Chengde, Hengshui, and Zhangjiakou, which are three cities with very low indicators of centrality, and are more isolated and in a disadvantageous position for cooperation. They need to strengthen the exchange of tourism activities with other cities in the future after perfecting infrastructure construction, improving tourism services level and enhancing tourism resources quality.

4.1.3. Spatial Clustering Characteristics Determination Based on the Top Network

Taking 2020 as an example, the Top 1, Top 3, and Top 5 networks are generated by Gephi, respectively (Figure 2). In the Top 1 network, there exist seven cities in the cooperation network with degree distribution k > 0, and the first connection is mainly concentrated in Beijing, Tianjin, Langfang, Tangshan, Handan, Shijiazhuang, and Xingtai. The average path length of the cooperation network is L = 1.741, and the average clustering coefficient is C = 0, proving that the network clustering characteristics have not yet been highlighted. Next, in the Top 3 network, the newly added city nodes with k > 0 include Cangzhou and Qinhuangdao. At this point, the average path length is L = 1.817, and the average clustering coefficient is C = 0.637. It can be seen that the level of spatial agglomeration is improved, and the network cohesion is enhanced. Compared with the Top 1 network, it is no longer simply a radiating connection between the core nodes and the peripheral nodes but shows a connection between cities and spatial agglomeration units. In the Top 5 network, except for Chengde and Zhangjiakou, the city nodes are connected to the network, and the internal nodes are intertwined with each other, making the network connections more intensive. The average path length decreases from 1.817 to 1.688, indicating that the accessibility of the cooperative network is improved and the accessibility between cities is enhanced. The average clustering coefficient is C = 0.584. The linkage, not only between neighboring agglomeration units, but also between core agglomeration units and core nodes in non-neighboring agglomeration units by virtue of the high-speed rail axis, reflects the character of meritocratic connection. For example, Shijiazhuang, which has a higher level of socio-economic development and better tourism resources, is preferentially connected to Beijing compared to the surrounding nodes.

The association division of the cooperative network is carried out through modularization to identify well-connected city clusters (Figure 3). The top 1 network is mainly divided into three associations, mainly depending on the geographical location. The Beijing–Tianjin association consists of the seven cities of Beijing, Tianjin, Langfang, Baoding, Cangzhou, Zhangjiakou, and Chengde; the Tangshan–Qinhuangdao association only has Tangshan and Qinhuangdao; and the Shijiazhuang-Xingtai association includes the four cities of Xingtai, Handan, Hengshui, and Shijiazhuang. Among them, the Beijing–Tianjin association takes Beijing, Tianjin, and Langfang as the axis chain, which has not yet formed a better tourism cooperation link internally, and the linkage development needs to be strengthened. It has a better cooperation relationship with the Tangshan-Qinhuangdao association, while the tourism cooperation with the Shijiazhuang-Xingtai association is slightly lacking. There is no change to the Top 3 network, which is also divided into three societies, except that Chengde is removed from the Beijing–Tianjin association and integrated into the Tangshan–Qinhuangdao association instead. In the Top 5 network, the cooperative network associations are integrated from three into two, and the Beijing–Tianjin and Tangshan–Qinhuangdao associations are merged into one. So far, the Beijing–Tianjin–Hebei regional city tourism cooperation is divided into two camps, and the cooperation links between the two camps need to be strengthened in the future.

4.2. Driving Factors for the Evolution of the Cooperation Network

On the basis of clarifying the evolution features of the cooperation network, quantitative attribution from 2014 to 2020 is carried out with the help of GeoDetector.

4.2.1. Core Driving Factors Identification

From the results of influence factor detection throughout the study period, 18 of the 22 independent variables passed the significance test at the 0.01 level, with a passing rate of 81.82%. As can be seen from

Table 4. Geographical detection results of influencing factors over the whole period.

	Factors	q-Value		Factors	q-Value
1	ODt Shortest travel time	0.2857	6	D21 Per capita urban disposable income	0.1001
2	O21 Number of scenic spots 4A and 5A per unit area	0.1862	7	D13 Urban registered employment rate	0.0994
3	O31 Percentage of tourism employees	0.1473	8	O13 Public toilets per 10,000 people	0.0961
4	O32 Percentage of education spending	0.1295	9	O11 Percentage of greenery coverage	0.0945
5	O14 Road area per capita	0.1230	10	D31 Per capita urban consumption expenditure	0.0710

All variables are significant at the 0.01 confidence level.

, the shortest travel time (OD_t, 0.2857) is the dominant driver, much higher than other factors, indicating that tourism cooperation in the Beijing–Tianjin–Hebei region is still of the locational transportation constraint type. Transportation is a prerequisite for promoting tourism cooperation, which influences the flow of human, material, information, and other factors between the cities. By 2020, the Beijing–Tianjin–Hebei region has basically formed a radial high-speed railroad network with Beijing as the center, and the time-space compression effect brought by the continuous construction of high-speed railways has met the efficient and convenient travel demand of tourists, reduced the resistance of tourism economic radiation, leading to a profound impact on the networked development of regional tourism cooperation. The domino effect is becoming more and more obvious. Other significant factors with q-value TOP 5 in order are the Number of scenic spots 4A and 5A per unit area (O₂₁, 0.1862) > Percentage of tourism employees (O₃₁, 0.1473) > Percentage of education spending (O₃₂, 0.1295) > Public buses and trams per 10,000 people (O₁₄, 0.1230), which are all destination market-related factors and can also be identified as the core influencing factors determining regional tourism cooperation. The influence of source market-related factors such as the Per capita urban disposable income (D₂₁), Urban registered employment rate (D₁₃), and Per capita urban consumption expenditure (D₃₁) should not be underestimated, with the explanatory power of 0.1001, 0.0994 and 0.0710. Furthermore, the influence of related indicators such as population status and economic level of rural residents of tourist origins is not very significant, and making up for the shortcomings of rural development is the key to further enhancing regional tourism cooperation.

From the study by period, shown in

Table 5. Geographical detection results of influencing factors in discrete steps.

	O21	O31	O32	O14	O41	D21	D31	ODt
2014	0.1861 ***	0.1864 ***	0.0358	0.1140 **	0.0194	0.1437 ***	0.1205 ***	0.3246 ***
2015	0.1902 ***	0.1902 ***	0.0774 **	0.0737	0.0259	0.1111 **	0.1010 **	0.3262 ***
2016	0.1837 ***	0.1417 ***	0.1420 ***	0.1245 ***	0.0940 **	0.0999 **	0.1248 ***	0.3018 ***
2017	0.1863 ***	0.1432 ***	0.1874 ***	0.1326 ***	0.0746	0.1369 ***	0.1107 **	0.2930 ***
2018	0.1958 ***	0.1987 ***	0.1959 ***	0.1558 ***	0.1335 ***	0.1328 ***	0.1045 **	0.2745 ***
2019	0.1980 ***	0.1979 ***	0.1084 ***	0.1617 ***	0.1414 ***	0.1315 ***	0.0931 **	0.2783 ***
2020	0.2165 ***	0.2163 ***	0.1361 ***	0.1591 ***	0.1555 ***	0.0642	0.0957 **	0.2848 ***

*** and ** indicate that the variables are significant at the 0.01 and 0.05 confidence levels, respectively.

, the three core influencing factors of the total study period, such as the Shortest travel time (OD_t), Number of scenic spots 4A and 5A per unit area (O₂₁), and Percentage of tourism employees (O₃₁), are still the core factors in each period with relative stability characteristics. Namely, destination factors such as transportation conditions, resource endowment, and hospitality capacity are bound to become the dominant factors for tourism cooperation between cities for a considerable period in the future. However, with the increasing improvement of the high-speed railway network in the Beijing–Tianjin–Hebei region, the degree of influence of transportation on regional tourism cooperation is gradually weakening, which is something that requires extra attention from tourism operators and managers to prepare for a rainy day and innovate the path of regional tourism development.

Specifically, the research period can be roughly divided into three stages. (1) From 2014 to 2015, while the core elements of each period existed stably, factors representing the economic strength of urban residents in source markets, such as the Per capita urban disposable income (D₂₁) and Per capita urban consumption expenditure (D₃₁), had a prominent impact on the spatial and temporal divergence of regional tourism cooperation. However, the impact values fluctuated from 0.1437 and 0.1205 in 2014, to 0.0642 and 0.0957 in 2020, respectively, and even the Disposable income of urban residents (D₂₁) does not pass the significance test in 2020. This result indicates that there is an income inflection point in the choice of travel behavior of urban residents in the source area, and when the income of residents reaches a certain inflection point, the influence of economic variables will no longer be significant, and travel decisions will be more flexible and independent. (2) From 2016 to 2017, similarly to the core influencing factors of the study period as a whole, the Percentage of education spending (O₃₂), i.e., the quality of residents in tourism destinations, on the influence of tourism cooperation, begins to appear and gradually increases. Combined with the Percentage of tourism employees (O₃₁), Road area per capita (O₁₄) and other indicators, this mainly reflects the destination tourism reception capacity, which determines the perceived experience of tourists and is an important condition for securing tourism sources, influencing the choice propensity of potential tourists, which in turn causes changes in the actual passenger flow. (3) From 2018 to 2020, the Road area per capita (O₁₄) and Excellent rate of air quality (O₄₁) rises. Among them, the significant influence of the Road area per capita (O₁₄) on the intensity of regional cooperation starts to appear in 2019. With the continued increase in the popularity of peripheral and self-driving tours in recent years, road congestion becomes one of the main decision factors for people to travel. On the other hand, accompanied by the arrival of the era of great health, good air quality has become one of the deciding factors guiding the choice of tourist destinations. However, the results of the influence of air quality conditions on the spatial layout for a considerable period from 2014 to 2017 are not significant, which may be because the areas with good air quality are mostly located in Chengde, Zhangjiakou and Hengshui, and the tourism market is small, which weakens the influence.

4.2.2. Expression of Factor Interactions

Given that the effects of different factors on tourism cooperation may not occur separately, this paper further conducts interaction detection analysis on the independent variables that pass the significance test (due to the limitation of space, the detection results only show the interaction between the core influencing factors), and the results are shown in Figure 4.

The detection results show that the core influencing factors interact with each other to produce a complementary enhancement effect of 1 + 1 > 2, including two-factor enhancement or non-linear enhancement, of which the non-linear enhancement effect is more prominent. It implies that the evolution of the cooperation network is not the result of uniform, independent and direct action of the influencing factors, but the product of the pairwise interaction of factors with spatial heterogeneity to form the enhancement effect. The future optimization of regional tourism cooperation needs to seek a multifactorial promotion path. It is noticed that the strongest interaction factor that determines the intensity of regional tourism cooperation is $O_{21}\cap O{D}_t$, which explains up to 0.7787, followed by the interaction factor $O_{21}\cap D_{21}$(0.7706), both of which include the Number of scenic spots 4A and 5A per unit area (O₂₁). The reason for this is that the spatial monopoly and immobility of tourism resources make their distribution shape the basic framework of regional tourism cooperation network, and also validate the biggest difference between the tourism industry and other industries. Namely, tourism activities are the flow of people rather than things, and consumption is not a single tourism product but a chain of products. Therefore, in addition to a single tourism resource, it is necessary to include such factors as the potential of the source market, traffic convenience, and other constraints to form a combination of regional tourism cooperation agglomeration advantage. In addition, the enhancement effect of the Shortest travel time (OD_t) on other factors is also particularly prominent, with a non-linear enhancement effect when acting together with any other independent variables on the dependent variable, playing an important role in the multifactorial interaction of the divergent pattern of the cooperation network. Particularly, it is noted that the Per capita urban disposable income (D₂₁), despite its weak explanatory power in the single-factor detection, increases significantly when superimposed with other factors, showing a significant jump in the interaction effect, indicating that the synergistic effect of D₂₁ with other factors rather than by itself is the dominant force in forming the existing network structure, which is mainly manifested as an indirect effect on tourism cooperation. So, when combined with factors such as transportation capacity and resource endowment, it can significantly promote tourism cooperation. This is due to the fact that the financial flow driven by people is an important lever to leverage the construction of tourism in the destination, providing sufficient funds for the development of tourism resources and the improvement of services and facilities, etc. If the financial flow is insufficient, tourism cooperation between cities will not be sustainable.

5. Discussion

In the post-COVID-19 era, tourism is facing unprecedented international competition and multiple challenges. It has become a consensus to build an interconnected and complete cross-administrative boundary tourism cooperation network to seek a synergistic, misaligned and win-win development of regional tourism. However, as a matter of priority, the understanding of the basic laws of regional tourism cooperation is limited, and quantitative analysis and quantitative attribution are still the weaknesses of existing studies. Therefore, taking the Beijing–Tianjin–Hebei region as an example, this paper measures the tourism cooperation intensity with the modified gravity model, on which the original, binary, and Top networks are generated to identify the spatiotemporal evolution characteristics from difference–association–agglomeration multi-dimension, and provide insight into the determinants by Geo-Detector model.

Here, we found that tourism cooperation network of Beijing–Tianjin–Hebei regional had initially formed, and cooperation links were gradually developing towards intensification and multi-district. In line with previous studies, this study again verified that regional tourism cooperation has developed from the traditional small-scale neighboring city cooperation to a cross-regional multi-level city cooperation network [2,7]. At the same time, the cities of the central region were more developed tourism economies of the cohesive type, the southern region contained more outward-looking cities with high external dependence, while the northern region had mostly the isolated type with little or no effective cooperation. Compared with previous studies that were mostly based on tourism economic linkages to construct tourism cooperation networks [42,43,54], weighted and undirected, and only explore the relationship characteristics [46,47,48], this study argues that tourism cooperation should also take into account the attribute characteristics of destinations themselves and identify the directional variability of cooperative linkages. Areas with better tourism resource endowment are more likely to attract tourists, which in turn will be tourist destinations to a greater extent in the cooperative relationship [52]. Therefore, this paper considers the tourist resource abundance as the moderating variable to construct a weighted directional tourism cooperation network, which deepens and expands the research content of tourism cooperation. Furthermore, the detection of the factors influencing the cooperation network evolution found that the Beijing–Tianjin–Hebei regional tourism cooperation was still of the transportation location constraint type. This result ties well with previous studies, wherein Salesi (2021) and Li (2015) point out that transportation and geographical location are the determining factors for tourists to choose tourism destinations, and the better the accessibility, the easier the flow of people, logistics, information and other factors is likely to occur, which has a potentially positive effect on tourism cooperation [24,25]. In addition, factors related to tourism origins such as demographic status and economic level of urban residents also had a significant positive effect on tourism cooperation, while indicators related to rural residents are not significant, whereas previous research has focused on the factor of regional economic conditions, without subdivision of towns and residents [51,54]. What is surprising is that tourism cooperation was not exclusively due to any of these factors, but selectively based on a combination of factors such as tourism access, resource endowment, and town spending power. These findings are deeper and more novel than previous studies and provide an important reference basis for the government to make tourism development decisions and tourist macro-regulation at the macro level.

Although there are important discoveries, there are also limitations. First, this paper takes the Beijing–Tianjin–Hebei region, which has the natural advantages of tourism cooperation, as an example to explore the evolution of structural characteristics and influencing factors of tourism cooperation networks, in an attempt to strengthen the overall regional tourism cooperation, but whether this is the best spatial scale for tourism cooperation still needs further verification in the future. Secondly, this study only explores cooperation characteristics after the basic formation of the Beijing–Tianjin–Hebei regional tourism cooperation area, however, tourism cooperation can be traced back to as early as the 1990s, and examining the historical experience and evolution will be more important for adjusting the structure of cooperation in the future. Thirdly, other driving factors of cooperation networks exist. We acknowledge the important influence of the factors that this study attempts to explore on tourism cooperation, but others are left for future research, such as political factors, security factors, cultural factors, etc.

6. Conclusions

6.1. Conclusions

Seeking win-win regional tourism cooperation is a key link to achieving high-quality regional tourism development. This paper systematically elucidates the spatiotemporal characteristics of the Beijing–Tianjin–Hebei regional tourism cooperation network and reveals the dominant driving factors of the network evolution, which can not only enrich the theoretical understanding of the cooperation network in the academic community but also has important practical values for promoting the linkage development of the Beijing–Tianjin–Hebei regional tourism cooperation network and thus accelerating the high-quality development of the regional tourism economy. The specific findings are as follows:

(1) The tourism cooperation intensity continues to strengthen, among which the three city combinations of Beijing ← Tianjin, Beijing ← Langfang, and Tianjin ← Langfang are significantly higher than other combinations. The city combinations with low cooperation intensity also show a catch-up effect on the higher combinations, and the overall differences tend to weaken. In terms of spatial structure, the cooperation network shows a spatially non-homogeneous feature of overall loosening and local clustering, which is manifested by the spreading trend of Beijing–Tianjin as the main axis and southward expansion. According to the spatial separation state of the cooperation line, the network is roughly divided into three large areas: the central compact area, the southwest development area, and the northern weak area, among which the central compact area is the key advantageous area, pulling the development and evolution of the cooperation network, while the northern weak area is still the crux of the current restriction on the development of regional tourism synergy and integration.

(2) The robustness of the association structure in the cooperation network gradually increases, but there is still plenty of room for improvement. Specifically, the outward equilibrium is improving, while the cohesiveness is gradually strengthening, and the overall network accessibility has changed very little. In terms of the power role of cities, Beijing is the core leader and the most important tourist destination and distribution center for tourists. The sub-core leaders are Tianjin, Shijiazhuang, and Tangshan, the secondary agglomeration centers in the middle, northeast and southwest directions. The general cooperators are usually adjacent to the leaders and have certain locational advantages, while the three cities of Chengde, Hengshui and Zhangjiakou, which are marginal cooperators, are in a disadvantageous position for cooperation. There is an urgent need to strengthen the exchange of tourism activities with other cities.

(3) Tourism cooperation has a spatial proximity effect, and the network spatial system presents the phenomenon of gradual stratified agglomeration with the expansion of the network scale. The connection between cities occurs not only between neighboring cities and neighboring agglomeration units, but also between core agglomeration units and core cities with greater connectivity in non-neighboring agglomeration units by the high-speed railway axis, reflecting the character of merit-based connection. The three associations of Beijing–Tianjin, Tangshan–Qinhuangdao and Shijiazhuang–Xingtai eventually evolved into Beijing–Tianjin cooperation circle and Shijiazhuang–Xingtai cooperation circle due to the integration of Beijing–Tianjin and Tangshan–Qinhuangdao. In the future, cooperation between these two camps should be strengthened.

(4) Tourism cooperation belongs to the locational traffic constraint type, and traffic is still a prerequisite for promoting tourism cooperation, followed by destination-related factors such as tourism resource endowment and hospitality capacity, while the influence on regional tourism cooperation is not very significant for the source market-related factors, especially with regards to the rural category indicators. There is an urgent need to make up for the shortcomings of rural development. Further detection found that the core factors interacted with each other to produce a 1 + 1 > 2 complementary enhancement effect, especially the urban per capita disposable income, which showed a significant jump in the interaction effect. The future optimization of regional tourism cooperation needs to consider the promotion path of multiple factors.

6.2. Theoretical Contributions

This study has three theoretical contributions: (1) it breaks through the traditional paradigm of individual attribute research, but based on holistic and structuralist ideas, it takes the interrelationship among system members as the main content, and selects the improved gravity model to construct the regional tourism cooperation network; (2) it explores the distribution characteristics from the difference–association–agglomeration multi-dimension in consecutive years, which is more comprehensive and accurate; (3) it carries out quantitative attribution for the evolution rules of the cooperation network, including driving factors and the interaction between factors.

6.3. Practical Implications

This paper not only provides an important reference basis for the government to make tourism development decisions and tourist macro-regulation at the macro level but is also beneficial to different cities for targeted follow-up tourism product development and tourism market cooperation. It is worth stating that tourism is facing unprecedented international competition and multiple challenges in the post-COVID-19 era, and the problem of unbalanced and insufficient development is still prominent. The optimization of the tourism cooperation network pattern with the core of city linkage development will become a key way to break the unbalanced development of tourism regions. Combining the research results, under the major demand of boosting the tourism market and promoting Beijing–Tianjin–Hebei regional tourism synergy development, the following countermeasures are proposed:

(1) The correlation between transportation network and tourism cooperation should be fully considered to provide targeted development countermeasures for cities with different power roles. Marginal cities such as Chengde and Hengshui should enhance the embedding of the high-speed rail network and strive to establish long-term tourism cooperation mechanisms with leader cities such as Beijing and Tianjin. Cities with advanced transportation development should give full play to their existing advantages, integrate tourism resources, extend tourism routes, expand tourism development markets, and then form advantageous industrial clusters and enhance tourism cooperation efficiency.

(2) Comprehensive association division results, for cities that belong to the same cooperative circle, we should give full play to the tourism economic spillover effect of advantageous cities, reduce the divergence of interests, establish the tourism development benefit sharing mechanism, promote interconnection within the cooperation circle, and create a new pattern of mutually beneficial tourism development. At the same time, the range of the Beijing–Tianjin main axis chain should be expanded, and the tourism connection between the Beijing–Tianjin cooperation circle and the Shijiazhuang–Xingtai cooperation circle strengthened to accelerate the realization of the new pattern of Beijing–Tianjin–Hebei tourism integration and cooperative development.