The Report to the 20th National Congress of the CPC (2022) emphasized that high-quality development is the primary task of current China. As the basic platform for Chinese urbanization and socio-economic development, mega-regions play a crucial role in promoting the realization of high-quality development in the new era, driven by new quality productive forces. However, as “unevenness” has become the main contradiction in contemporary development in China, the disparities between large, medium, and small cities in mega-regions, especially between the core cities and smaller cities, have hindered the leap forward regarding new quality productive forces. Therefore, this paper takes the coordination of inter-city relations as the basic perspective and conceptualizes smaller cities in mega-regions as “secondary cities” as opposed to core cities. In the uneven mega-regional system of China, secondary cities play a crucial role in promoting new quality productive forces, namely, new platforms for multidimensional technological innovation, new pivots for polycentric industrial distribution, and new drivers for multidirectional talent flow. However, this paper also argues that secondary cities face the practical obstacles of prioritizing the cores, endowments lacking, and institutional conflicts. Based on the conceptual framework of coordinated development between mega-regional core and secondary cities, with coexistence, connectivity, and cooperation as the main components, this paper proposes emerging pathways for secondary cities in terms of technological innovation, industrial transformation, and talent cultivation, aiming to utilize the potentials of these cities to promote the development of new quality productive forces from three aspects: upgrading and innovation of resources for productivity, objects of productivity, and laborers of productivity, providing theoretical references and practical foundations for the design of development pathways of new qualtiy productive forces and the formulation of policies for the coordinated development of large, medium, and small cities in Chinese mega-regional system.

As China's urban system transforms from a hierarchical structure to a networked one, the spatial organization and underlying mechanisms of intercity co-opetition within urban regions have become central to promoting regional coordination and high-quality development. Drawing on complex network theory, this study constructs a two-dimensional measurement framework of intercity cooperation and competition based on geographic proximity and technological relatedness, capturing the evolutionary characteristics of collaboration and competition within urban regions. Using data on co-applied invention patents from 2005 to 2017 across 19 urban regions in China, the study quantitatively identifies the intensity and temporal dynamics of intercity cooperation and competition. It further investigates how morphological polycentricity and functional polycentricity influence co-opetition dynamics. The findings reveal that functional polycentricity tends to simultaneously strengthen both cooperation and competition among cities, forming a pattern of “positive-sum co-opetition”. In contrast, morphological polycentricity may suppress overall intercity interaction by weakening network activity. Moreover, the influence of polycentric structures exhibits clear hierarchical heterogeneity: functional polycentricity has a significantly stronger effect on cooperation in national-level regions than in regional or local-level ones, while morphological polycentricity shows consistently negative effects across all levels, especially pronounced in regional-level regions. This study contributes to the measurement and mechanism identification of urban co-opetition and offers theoretical and policy insights for fostering complementary advantages, healthy competition, and innovation-driven collaboration within urban regions.

To explore whether polycentric urban agglomerations can serve as a spatial foundation for the economic development of cities of different sizes, this paper utilizes population remote sensing data to evaluate the polycentric structure of urban agglomerations. Employing instrumental variable estimation and mediation effect models, this study systematically investigates the impact of the polycentric structure of urban agglomerations on the economic performance of cities of different sizes within the urban agglomeration, along with the mechanisms and limiting factors involved. This research finds that there exists a significant inverted “U”-shaped relationship between the polycentric structure of China's urban agglomerations and the economic performance of large, medium, and small cities within them, with the relationship remaining predominantly promoting at the current stage. The mechanism analysis reveals that compared with monocentric urban agglomerations, polycentric urban agglomerations can more effectively promote inter-city functional specialization. However, this effect exhibits heterogeneity depending on the resource endowments of urban agglomerations, the distance between cities, the stage of development, and locational conditions. The findings of this study offer important implications for optimizing the spatial organization of urban agglomerations in China.

Inter-regional collaboration plays a crucial role in promoting innovation develop-ment in peripheral regions. Based on patent and economic development data from cities in the Yangtze River Delta from 2005 to 2020, this study employs social network analysis methods to examine the characteristics of innovation collaboration among peripheral cities across four spatial scales: local, provincial, regional (within Yangtze River Delta), and national. Using an OLS panel regression model, this study further evaluates the effect of various types of collaboration on the innovation performance of peripheral cities. The results show that: (1) The key participants in innovation collaboration of peripheral cities vary significantly across spatial scales. Local collaboration is predominantly driven by traditional manufacturing firms and local universities. Within the provincial and Yangtze River Delta scales, collaboration is mainly dominated by “Double First Class” universities and high-tech enterprises, while national-level collaboration is primarily led by state-owned electric power corporations. (2) Peripheral cities exhibit weak local innovation network connectivity, while cross-regional collaboration is characterized by distance decay and preferential attachment. Innovation collaboration patterns can be classified into two types: the isolated type, which is characterized by weak connections in both local and cross-regional cooperation, and the connected type, which shows relatively strong linkages both internally and externally. (3) Among the four spatial scales, provincial collaboration has the strongest positive effect on innovation performance, followed by regional collaboration within the Yangtze River Delta, whereas local and national collaboration show no significant impact. Isolated cities benefit primarily from provincial-scale cooperation, while connected cities rely more on regional-scale collaboration. These findings contribute to the understanding of innovation development in peripheral areas and provide insights for policymakers, aiming to strengthen multi-scalar innovation networks.

With the rapid global expansion of international hotel groups, the hospitality industry has emerged as one of the most globalized sectors. However, existing studies have yet to answer the causal mechanisms and potential impacts of the dynamic interactions between international hotel groups and local actors. Drawing on the global production network (GPN) theory, this study develops an analytical framework for examining the strategic coupling of international hotel groups. Using high-end international branded hotels in Zhuhai, Guangdong Province, as a case study, the paper investigates their expansion strategies and underlying causal mechanisms. The findings provide three key insights. First, the globalization process of international hotel groups is a strategic coupling rather than a market-driven process. This strategic coupling process is deeply embedded in local governments' solid developmental imperatives and somewhat distorted developmental behavior under China's high-powered political-economic development system. Second, market imperatives and brand value play a critical role in determining the bargaining power of the global and local actors in the coupling, which in turn influence the evolution of coupling models. Third, strategic coupling distorts the market behavior of international hotel groups, leading to sustainability risks within their business models. However, these risks can be mitigated at the urban scale through regional economic systems. This research makes two important contributions. Firstly, it discloses the intricate interrelationships and mechanisms between international hotel groups and local actors by constructing the framework of strategic coupling within the global hotel industry. In doing so, this study addresses the lack of geographical perspective on existing hotel globalization studies. Secondly, this research innovatively applies the concept of strategic coupling, which originated from manufacturing to services. It refines this concept based on the characteristics of services, thereby enhancing its applicability to service industries.

The global value chain (GVC) network represents a key organizational form of international division of labor. Illuminating its sectoral differentiation and the evolution of China's position is critical for advancing China's global competitiveness. Drawing on multi-regional input-output data, this study constructs the GVC network from 2000 to 2023, focusing on major sectors, and systematically analyzes sectoral differentiation in terms of spatiotemporal patterns, topological features, and core nodes, while comprehensively evaluating the evolution of China's position in the GVC network from both network and value chain perspectives. The study finds that: (1) From an overall perspective, sectoral disparities in value scale are pronounced, with the top 8 sectors consistently accounting for over 60% of total value. Most sectoral networks exhibit a polarized structure supported by a few key nodes and backbone flows. (2) From a sectoral perspective, the mining GVC network is sparse and strongly constrained by resource endowments. The GVC networks of chemicals and chemical products, and basic metals and fabricated metal show increasingly regionalized characteristics, while machinery, electrical and optical equipment, and transport equipment exhibit globalization. The wholesale and commission trade GVC network is centered on Hong Kong (China) and Singapore, involving many developing economies with generally shallow participation. The GVC network of business activities has formed a backbone value flow linking Ireland, the U.S., and India. (3) China holds a relatively stronger position in manufacturing GVCs compared to mining and services. Over time, it has moved toward higher value-added segments, improved its control over raw materials, and enhanced service provision capacity. However, challenges persist, including intensifying competition in traditional manufacturing, lack of absolute advantage in high-tech sectors, and limited influence in services. Based on these findings, the paper offers policy insights to support China's deeper integration and functional upgrading in global value chains.

Internet Data Center (IDC) serve as critical infrastructure for the digital transformation of society. Revealing the spatiotemporal evolution of their spatial configuration and dynamically assessing their enabling role in advancing common prosperity hold significant academic and practical implications for promoting coordinated regional development and optimizing major productive forces. Based on data of server racks in China's IDC from 2004 to 2024, this study employs spatial statistics and geographic concentration index methods for empirical analysis. The findings reveal that: (1) China's IDCs exhibit a spatial pattern of large scale in the eastern and western regions and relatively small scale in the central region, with the center of gravity of their layout shifting significantly westward. (2) The spatial layout is driven by multiple factors including market scale, cost factors, green development, and operational environment. The dominant mode has transitioned from a demand-cost orientation to a comprehensive demand-cost-green-and-low-carbon orientation. Energy transition and technological innovation have become crucial factors influencing the current layout of IDC. (3) The construction of IDC demonstrates significant economic and social effects, particularly in substantially boosting the economic and social development of less developed regions. This effectively narrows the development gap between the eastern and western regions, contributing to the realization of common prosperity. This research aims to provide a decision-making foundation for optimizing the spatial allocation of data resources, analyze pathways for balanced regional development under the major national strategy of “East Data, West Computing”, and help all people share in the digital dividends of the era.

Developers are the “city builders”, boosting urban and regional development. Studying their investment location choices is crucial for understanding the evolutionary mechanisms of the built environment in China's urbanization process. Based on the perspective of new institutionalism, this study integrates elements such as local markets, corporate characteristics, and institutional structures to construct a systematic analytical framework for developers' residential project location (city) choices. Using 290 cities in China as spatial units, a database was established for the residential land purchases of 58 publicly listed real estate development companies (A-share and H-share). K-means clustering and Gradient Boosting Decision Tree (GBDT) algorithms were employed to quantitatively analyze the spatial differences and influencing mechanisms of developers' investment location choices. The research identifies four spatial strategic models in developers' city choices for residential projects: (1) National-Central City Model, represented by Vanke, which places greater emphasis on factors such as economic scale and rail transit in city selection; (2) National-Non-Central City Model, represented by Rongsheng and Poly, which shows a decreased emphasis on economic scale and rail transit, focusing instead on the population agglomeration of selected cities; (3) Local-Central City Model, represented by Shoukai, which concentrates on specific local markets and relies more on local governments with higher degrees of fiscal decentralization; (4) Local-Non-Central City Model, represented by Jianye, which similarly focuses on specific local markets but places greater emphasis on local trust. The study indicates that the behavioral patterns of developers' investment location choices are difficult to alter through singular policy interventions, and a stable transformation of the real estate industry necessitates deeper modernization of local governance.

Industrial relocation is a fundamental law governing the global distribution of production and a significant theme in global economic history. However, previous research has primarily focused on the phenomenon of outward industrial migration, paying insufficient attention to the relocation of production capacity back home. Since China proposed its “dual circulation” strategy, it has become an urgent theoretical question and practical imperative to understand how companies represented by Foxconn are relocating production back to China and how this relocation interacts with local and regional development. Based on this, this paper conducts in-depth research on the economic geography logic of Foxconn's production reshoring in Zhengzhou through on-site investigations, and explores how it can achieve a new strategic coupling with the region after the reshoring. The research findings are as follows: (1) The essence of Foxconn's production reshoring is the re-optimization of economic geography factors. Based on factors such as the supply-chain system and risk assessment, labor cost and quality advantages, policy-strategy alignment and infrastructure advantages, and regional industrial upgrading and synergistic effects, it is revealed that the production capacity layout of multinational enterprises increasingly depends on the dynamic adaptation ability of regional resource endowments, rather than simple competition for low-cost locations. (2) Foxconn's production reshoring is not only a physical return of the supply chain but also a paradigm shift in its relationship with regional development. Through joint technology research, industrial co- construction, and risk sharing between Henan Province and Foxconn, the relationship between them is evolving from a one-way dependent coupling relationship of “enterprise-led, local- dependent” to a reciprocal coupling stage of “innovation cooperation, benefit sharing”. This transformation is not only a response to the reconstruction of the global industrial chain but also provides a Chinese model for the transformation and upgrading of labor-intensive multinational companies. (3) Following Foxconn's production capacity relocation, its strategic coupling model with regional development has evolved. This evolution, driven by factors including the dynamic game theory of power relations, interest interactions within the geopolitical environment, leaps in value capture hierarchies, and mechanisms reinforcing spatial stickiness, reveals the interplay between “first-nature advantages” (supply chains) and “second-nature advantages” (policy) as theorized in the “New Economic Geography”.

The digital innovation network enhanced the efficiency of factor flow, generated new growth momentum, and became a key force in overcoming the economic development dilemma, promoting economic speed and facilitating a gear shift. This study constructed the structural characteristics of the digital innovation network based on patent cooperation data in the digital economy sector from 2012 to 2022, and empirically examined the impact of digital innovation network embeddedness on urban economic development. The study found that: ① The digital innovation network structure of Chinese cities was shifting to a “diamond structure”, meaning the four-legged digital innovation network comprising Beijing-Tianjin-Hebei, the Yangtze River Delta, the Pearl River Delta, and Chengdu-Chongqing is gradually becoming clearer. ② Digital innovation network embeddedness significantly promoted urban economic development, primarily by enhancing cities' technological innovation capacity and alleviating capital factor distortions. However, it exacerbated labor factor distortions, which may stem from a mismatch between low-skilled labor and the evolving structure of digital innovation networks. ③ The heterogeneity analysis indicated that the positive effect of digital innovation network embeddedness on urban economic development was more pronounced in the Five Major Urban Agglomerations, in regions with stronger intellectual property protection, and in cities located southeast of the Hu Huanyong Line. This suggested that amplifying the growth-enhancing effect of digital innovation networks required strengthening intellectual property protection, improving the efficiency of capital and labor allocation, and accelerating the construction of digital innovation networks within urban agglomerations. From a network perspective, this study highlighted the critical role of digital innovation networks in urban economic development, providing both theoretical and empirical evidence for deepening cities' embeddedness in digital innovation networks and breaking down geographical and administrative barriers between cities.

In the digital era, multi-dimensional coordinated development is a critical pathway to achieving high-quality development of the digital economy. This study examines the coupling coordination characteristics and spatiotemporal evolution patterns of multi-dimensional development in the digital economy using panel data from 279 Chinese cities during 2011-2021, employing the coupling coordination degree model, Dagum Gini coefficient, and obstacle degree model. The findings reveal three key insights. First, the multi-dimensional development of the digital economy demonstrates significant differentiated evolution characteristics: while the coupling coordination degree between digital infrastructure and digital technology shows continuous growth yet remains insufficient, the coupling degree between digital technology and digital industry exhibits a declining trend, while the coupling coordination degree shows an overall upward trend. Meanwhile, the coupling coordination degree between digital infrastructure and digital industry maintains a slow upward trajectory. These patterns indicate the staged characteristics of digital economic development. Second, regional disparities in multi-dimensional development constitute the primary source of unbalanced development, manifesting in a spatial gradient pattern where eastern coastal regions take the lead in overall coordination levels while the central and western regions generally lag behind, highlighting the persistent “digital divide”. Third, obstacle factor analysis reveals that insufficient digital technology innovation capability continues to constrain the coordinated development of digital infrastructure and technology, while shortages of key factors in the digital industry significantly affect coordination involving the industrial dimension, with the intensifying trend of these obstacle factors indicating ongoing structural challenges in digital economic development. These findings provide theoretical and practical implications for promoting coordinated and high-quality development of China's digital economy.

With the rapid development of China's online retail market, a thorough understanding of the spatial characteristics and influencing factors of online retail development level contributes to the sustainable and healthy development of online retail, optimizing industrial distribution, and advancing the coordinated development of large, medium, and small cities. Based on prefecture-level online retail data from the Ministry of Commerce and Fudan University's Consumer Market Big Data Lab, this study employs Exploratory Spatial Data Analysis (ESDA), a polycentric index, and a multidimensional fixed effects model to examine the spatial characteristics and influencing factors of online retail in China. The main findings are as follows: (1) From 2019 to 2022, the overall scale of online retail has grown, characterized by a spatial distribution radiating from megacities to surrounding small and medium-sized cities, with a concentration in the southeastern coastal regions. High-high agglomeration areas are mainly concentrated in the Yangtze River Delta, Pearl River Delta, and Beijing-Tianjin-Hebei urban agglomeration, while low-low agglomeration areas are widely distributed in western and northeastern regions. (2) There are significant differences in the spatial structure of urban agglomerations based on online retail. The Yangtze River Delta and Pearl River Delta urban agglomerations exhibit a dual-core and polycentric structure, whereas the Beijing-Tianjin-Hebei, Chengdu-Chongqing, and Middle Yangtze River urban agglomerations display a predominantly monocentric structure. Cities with higher levels of socio-economic development occupy core positions, while the rapid rise of certain cities has also led to adjustments in the internal spatial structures of urban agglomerations. (3) The city's production conditions, consumer demand, transportation logistics, and government support are important influencing factors driving online retail development. In large cities, industrial structure, technological investment, and urban household income are more influential. In medium-sized cities, population density and demonstration city policies play greater roles. In small cities, economic growth, rural income increases, better road infrastructure, and the influence of geographical indication products are especially important.

Scale construction theory posits that scales are socially constructed. It primarily examines the social relations among different actors, including production relations and consumption relations, as well as the resulting scale-political practices, in which scale production serves as both the center and the outcome of practice. It provides a unique perspective for the formation of scale space. The existing geo-setting research, especially the regional geo-setting, lacks the analytical ideas and methods for the formation of this “space”. Therefore, this paper draws on the theory of scale construction and constructs an analytical framework for regional geo-setting based on the ideas of scale and space production. Taking the Transcaucasia from 1991 to 2024 as an example, it demonstrates the formation and evolution process of a multi-country regional geo-setting and explores the mechanisms behind its different stages of development. The findings indicate that: (1) Within certain geographical contexts, actors, the scale political practice of actors or multiple actors produces a new “national” scale, and then shapes a new regional geo-structure, which is the direct cause of the formation of regional geogrageo-setting. (2) After the formation of regional geo-setting, further scale political practice among related actors is carried out under the influence of complex social relations such as international social relations, production relations and consumption relations, The cross-scale coupling of these complex social relations is the intrinsic driver of the evolution of the regional geo-setting. (3) The independence of the three countries following under the collapse of the Soviet Union in 1991, specifically its up-scaling effect, is the direct cause for the formation of “Transcaucasia” as a regional geo-setting. Over the subsequent over 30 years, the geo-setting of Transcaucasia has undergone three distinct development stages, characterized by geo-structure of“disorder”, “imbalance”, and “balance oriented” types. The article concludes with three prospects for future regional geo-setting research.

It is of significant importance to conduct a quantitative assessment of China's neighborhood geo-setting from the dual perspectives of risk and problem orientation. In this paper, geo-setting risk is theorized as the structural instability of geo-institutional power game in big countries with global scale, the interaction of threat diffusion and vulnerability in neighboring regions with regional scale, and the comprehensive integration of risks in natural, political, economic and social fields within countries with unit national scale, and the theoretical framework and quantitative model of geo-setting risk with multi-dimensional geo-institutional game, cross-scale coupling and cross-domain interaction are constructed, and the geo-setting risk around China is systematically studied. The results show that: (1) Multi-scale geographical scenarios strengthen the cognitive basis of risks, and the surrounding geo-setting risks show a dynamic trend of “being the same but not harmonious”. On the global scale, the power game between China and the United States in the surrounding areas tends to be fierce, and the interaction between threat diffusion and vulnerability in the neighboring areas on the regional scale weakens, and the integration of risks in natural, political, economic and social fields within the unit national scale increases, and risks are obviously affected by sudden and sporadic factors. (2) Geo-setting risks of different scales have obvious spatial heterogeneity, obvious spatial dislocation and lack of obvious spatial dependence and spatio-temporal correlation; There is no obvious law to follow in the spatial and temporal evolutions of multi-scale risks. The geo-setting risks of different countries under different scales generate their own systems, among which the unit national scale risk is the main source of the geo-setting risks of integrated countries, and the domestic geo-setting risks mainly come from the political dimension. (3) The same neighboring country is simultaneously affected by risks of different scales such as global, regional and unit national levels. Multi-scale risk develops in disorder around China, and the multi-scale risk coupling shows a low level of dynamic correlation coupling, with limited interaction and a downward trend. The multi-scale risk coordination has been in a state of serious imbalance for a long time, but the degree of mutual restraint of different scale risks has increased slightly. At the end of the paper, the authors appeal to the academic circles to put situational into the geopolitical research agenda and regard it as a reasonable and commonly used research method.

The dynamic changes in the coastal zone, such as the frequent flow of population, capital, and technology between different spaces, cause the resilience of the coastal human‑nature system to constantly face new pressures and challenges. Clarifying the interaction mechanism between coastal system resilience and elements flow is helpful to strengthen the resilience level and resource utilization efficiency, and further enrich the research on the coastal zone. Based on the panel data of 51 coastal cities in China from 2010 to 2021, this study used the variable fuzzy identification method to measure the level of coastal human-nature systems resilience, constructed the comprehensive strength matrix of “flow space” network, analyzed the spatiotemporal evolution pattern of coastal human-nature system resilience and “flow space” network, and applied the PVAR model to quantitatively analyze the interaction mechanism between coastal human-nature system resilience and “flow space” network in China. Research shows that: (1) The coastal human-nature system resilience in China shows an increasing trend as a whole, but there are obvious differences between cities. The comprehensive strength of the “flow space” network is increasing, but the flow of elements has obvious geographical adjacency. (2) The coupling and coordination level of the coastal human-nature system resilience and the “flow space” network in China showed a stable and improving trend, and the coupling coordination degree generally showed a pattern of the eastern > the northern > the southern marine economic circles. At the municipal level, Shanghai and Guangzhou have always maintained a high level of coordination, whereas some cities in Guangxi and Guangdong regions have experienced mild imbalances. (3) Through GMM estimation, impulse response, and forecast error variance decomposition based on a PVAR model, the coastal human‑nature system resilience and the “flow space” network in coastal cities of China exhibit inertial development characteristics. The impact on themselves gradually decreases, and the impact of the “flow space” network on the coastal human‑nature system resilience is greater. This study is of great significance for clarifying the mutual feeding mechanism of coastal resilience and elements flow, ensuring the stability and security of land space in coastal areas, and realizing the sustainable and high-quality development of coastal areas.

With the acceleration of global extreme climate change and urbanization, urban flooding disasters have become increasingly frequent, causing not only substantial economic losses but also severely threatening the mental health and well-being of urban residents. This study examines the impact of perceived differences in flood experiences before and after disasters on residents' happiness, introducing the moderating role of built environment elements in the relationship between flooding and well-being. Using the 2021 Zhengzhou “7·20” mega-flood as a quasi-natural experiment, the research employs a Difference-in-Differences-Perspective Score Matching (PSM-DID) model based on one-year follow-up survey data before and after the disaster. Findings reveal that: (1) Residents who perceived flooding as significantly “worse” in the post-disaster year showed a marked decline in subjective well-being; (2) Built environment elements play a critical heterogeneous moderating role: convenience store density and bus stop density significantly mitigate the negative impact of perceived differences on well-being, while subway station density, road intersection density, and population density exacerbate this effect. The study provides a subjective perception perspective on the medium-to-long-term impact of flooding on residents' well-being, clarifying the differentiated moderating role of built environment in post-disaster recovery. The conclusions not only offer new theoretical insights into understanding the psychosocial consequences of disasters but also provide empirical evidence for promoting refined, space-oriented resilient urban planning and post-disaster psychological interventions.

It is of great significance to conduct an in-depth exploration of the relationship between urban population distribution and heat exposure risk for improving the health level of urban residents and enhancing their well-being. Based on the LandScan population distribution data and the Chinese apparent temperature dataset, this study constructs a polycentric index and an indicator system for heat exposure risk. A spatial econometric model is adopted to analyze the spatio-temporal evolution patterns of the population distribution and heat exposure risk in prefecture-level cities in the Yangtze River Delta from 2006 to 2020, and to explore the influence and mechanism of the polycentric distribution of urban population on heat exposure risk. The results show that: (1) During the study period, the heat exposure risk showed a fluctuating upward trend, the polycentric index remained relatively stable, while the population aggregation degree exhibited a fluctuating characteristic of first increasing and then decreasing. (2) High-value areas of both heat exposure risk and the polycentric index were persistently concentrated in Shanghai, southern Jiangsu and northern Zhejiang, while low-value areas were stably distributed in northern Anhui and northern Jiangsu, with significant heterogeneity in the spatial correlation between the two variables. (3) Compared to the compact monocentric model, an increase in the polycentricity index can effectively disrupt the spatial continuity of the thermal environment by reducing the urban agglomeration index, thereby significantly decreasing both the heat-exposed population and the heat exposure index. This inhibitory effect is significantly and negatively moderated by urban population size; specifically, the continuous expansion of population scale attenuates the thermal environment improvements yielded by the polycentri-city index.The research can provide planning and decision-making guidance for coordinating the relationship between urban spatial form and heat exposure and promoting the healthy grow-th of urban space.