In the context of China's national strategy to “strengthen the overall planning of major bays and promote high-quality development of the marine economy”, the bay economy has emerged as a critical regional economic form. However, existing theoretical frameworks, which are primarily centered on the “bay-area economy,” tend to emphasize land-based development while neglecting marine coordination and focus on phenomenological descriptions rather than internal mechanisms. This study constructed a theoretical framework for the bay economy, distinguishing it from related notions such as the marine economy and bay area economy, and explored its evolutionary mechanisms and development models. A theoretical construction approach combined with an empirical case study was employed. The study developed a triple-helix evolutionary mechanism model structured around three core elements (space, industry, and institution), emphasizing their coupling and dynamic interactions. Within this framework, space serves as the physical carrier; industry functions as the endogenous driving force; and institutions act as governance guarantees. The interplay among these three elements determines the developmental trajectory and evolutionary stage of the bay economy. The “Golden Inner Bay” of the Guangdong-Hong Kong-Macao Greater Bay Area was selected as a representative sample for the empirical investigation. The results indicate that the evolution of the bay economy follow a spiral and upward trajectory driven by the coupling of three core elements and progresses through four distinctive stages: port-led agglomeration, port-city coordination driven by spatial restructuring to resolve port-city conflicts, regional coordination defined by multicenter networking and institutional collaboration, and land-marine coordination, representing the advanced stage of integrated development. Empirical analysis of Golden Inner Bay identified a pattern of non-equilibrium transformations. Its evolution demonstrates a clear progressive logic in the interaction of the three elements: in the initial stage, industry serves as the primary driver, while space and institutions passively adapt; in the regional coordination stage, institutions take on an active rule-setting role. The Golden Inner Bay has largely achieved port-city coordination and is transitioning toward regional coordination, whereas land-marine coordination remains at a nascent, exploratory stage. However, structural asymmetry, in which spatial expansion and industrial upgrading outpace institutional innovation, has led to a lag in systematic land-marine coordination governance. Specifically, the region exhibits strong momentum in port-based economic agglomeration and cross-city infrastructure connectivity yet lacks integrated policy instruments that span both terrestrial and marine administrative jurisdictions. The study concludes that the essence of the bay economy is a fundamental paradigm shift from port-city synergy to comprehensive land-marine coordination. This study fills the theoretical gap in the systematic analysis of land-marine coordination within coastal regions by providing a novel space-industry-institution analytical paradigm. Furthermore, it offers targeted strategies for the Golden Inner Bay, including strengthening the networked spatial allocation from land to sea, building a modern land-marine integration industrial cluster by guiding the sea-bound transfer of innovation, and innovating institutional designs, notably by proposing the establishment of a Lingdingyang Rim Special Cooperation Zone to bridge the governance gap and serve as an institutional springboard for mature development. These strategies are intended to provide actionable pathways for other coastal regions to pursue high-quality bay economic development.

Concert tourism supports Macao’s tourism diversification and sustainable tourism development, yet the visitor experience goes beyond the performance itself, involving also ticketing, cross-border mobility, accommodation, venue operations, destination services, and fan interactions. This study identifies the full-journey tourism experience attributes of Chinese Mainland concert tourists visiting Macao, examines their asymmetric functions in relation to satisfaction and dissatisfaction, and compares their classifications and priorities across consumption-level groups. An exploratory sequential mixed-methods design was adopted. Study 1 used 23 semi-structured interviews to identify 22 attributes spanning pre-trip preparation, ticketing, cross-border mobility, accommodation, on-site attendance, destination activities, and fan interactions. Study 2 converted these attributes into paired functional and dysfunctional KANO questions, and then analyzed 415 valid questionnaires retained from 506 responses. Eligibility screening verified attendance within the previous 12 months; data-quality checks excluded responses with implausibly short completion times, duplicate IP addresses, or patterned answers. Responses were assigned using the KANO evaluation matrix. Traditional KANO classifications were determined by modal frequency, after which Better and Worse coefficients were calculated to estimate the satisfaction-gain potential and dissatisfaction risk. Quadrant classifications used the mean better and absolute worse coefficients of the relevant sample/group as cutoffs. The low-, medium-, and high-consumption groups were compared, and the attributes within each category were ranked by sensitivity. The findings indicate that concert tourists evaluate an interconnected experience comprising performance content, journey connections, and destination services. These attributes form a layered asymmetric functional structure. Ticketing rules, accommodations, and ground support services are classified as Must-be attributes that serve as safeguards against dissatisfaction. Venue visual and sound quality, mobile network stability, and recording convenience are One-dimensional attributes sensitive to both satisfaction gains and dissatisfaction risk. Performance surprises and special concert occasions are Attractive attributes that offer high satisfaction-gain potential, with limited dissatisfaction risk. Across consumption levels, the classifications contained both a stable core and a boundary that varied across groups. The stable core combines Must-be attributes, One-dimensional venue attributes, and Attractive value-added elements. Other attributes, including interpersonal interaction, cross-border and destination connections, fan-oriented services, and ground support, change categories across groups. In the high-consumption group, artist-audience interaction enters the Must-be category, whereas luggage storage, post-event traffic dispersal, venue guidance, and convenience facilities leave it. Official merchandise, Mandarin-language services, and fan activity support also move from Indifferent to Attractive in this group. These shifts do not constitute a uniform progression toward more demanding categories. They show that group differences concern how the boundary of what is regarded as indispensable is configured. Accommodation and venue audiovisual conditions remain common priorities, whereas interaction, customs-clearance efficiency, and fan services require differentiated attention. By locating these asymmetric functions within the full visitor journey, this study extends concert-tourism experience research beyond on-site performance content, thereby clarifying the distinct functional positions of performance content and venue presentation conditions. It also provides exploratory evidence that the KANO classifications can combine a cross-group functional core with group-specific category boundaries. The findings support safeguarding common service foundations, identifying priorities shared across consumption groups, and differentiating experience-resource allocation across segments in Macao.

The Guangdong-Hong Kong-Macao Greater Bay Area (GBA) has been positioned in the national strategy as an international science and technology innovation center. As innovation resources are unevenly distributed across the 11 cities, technology transfer has become a key channel through which cities acquire external technology, break path dependence and technological lock-in, and upgrade their industrial structures. However, existing studies have paid more attention to the volume of technology transfer than to its direction; thus, the spatiotemporal evolution and driving mechanisms of the directed technology transfer network in GBA remain insufficiently understood. To fill this gap, this study constructed a directed, weighted technology transfer network for the eleven GBA cities from the perspective of the "space of flows." Based on invention patent transfer records from 2005 to 2024, together with statistical data from provincial and municipal yearbooks, this study employed social network analysis (network density, hierarchical stratification, degree centrality, and betweenness centrality) and stage-specific negative binomial regression to characterize the spatiotemporal evolution of the network across four stages (2005-2009, 2010-2014, 2015-2019, and 2020-2024) and identify the mechanisms underlying its influencing factors. At the network-wide level, network density increased steadily, and technology transfer conformed to a significant rank-size rule, with transfers preferentially occurring among core cities. The network also exhibited a pronounced Matthew effect, whereby a small number of high-value city pairs accounted for a large share of the total transfer volume. At the channel level, both the number and scale of transfer channels grew continuously. Shenzhen and Guangzhou acted as radiation (outflow) centers; Huizhou, Jiangmen, Zhaoqing, and Zhongshan served as convergence (inflow) centers; Hong Kong shifted from a convergence to a radiation role; and Foshan and Zhuhai shifted from radiation to convergence. The direction of transfer shifted from mutual exchanges among core cities toward diffusion to surrounding cities, and the diffusion mode evolved from hierarchical diffusion to a mixed mode combining hierarchical, contact, and jump diffusion. In terms of advantageous flows, Guangzhou and Shenzhen exhibited both agglomeration and spillover effects. Shenzhen-Dongguan and Guangzhou-Foshan constituted the largest mutual flow pairs, and Hong Kong's advantageous flow pointed stably toward Shenzhen. At the node level, the network exhibited a polycentric structure: Shenzhen functioned as the first-tier core, with Guangzhou, Zhongshan, Foshan, and Dongguan as secondary centers, and the network evolved from a Guangzhou-Shenzhen "dual core" toward a multi-center pattern; Shenzhen, Guangzhou, and Zhongshan remained the key nodes, while Dongguan, Shenzhen, Guangzhou, and Hong Kong maintained relatively high betweenness centrality. Negative binomial regression showed that geographic distance, industrial similarity, economic development level, R&D personnel, imports, and technological absorption capacity significantly influenced technology transfer, whereas R&D expenditure and FDI were significant only at specific stages. Among them, geographic distance and economic development level exerted negative effects, confirming their constraining roles, while industrial similarity exerted a robustly positive but gradually declining effect because within the GBA, similar industrial structures reflect collaborative division of labor along the same industrial chain rather than homogeneous competition. This study constructed a directed technology transfer network from a "space of flows" perspective and revealed the spatiotemporal evolution and multi-factor driving mechanisms of technology transfer in the GBA, uncovering its evolutionary patterns across multiple scales, namely the overall network, transfer channels, and node functions. By incorporating both dyadic and node-attribute variables, this study further revealed the dynamic evolution of influencing factor effects through four stage-specific regressions. The findings provide empirical evidence and policy implications for optimizing regional technology transfer and advancing the construction of the GBA as an international science and innovation center.

Against the backdrop of global climate governance being severely disrupted by strategic competition between major powers, the United States is experiencing a significant shift in its climate aid from a global public good to a geostrategic tool. How this shift occurs and the mechanisms driving it are urgent questions that require scholarly attention. This study employed a combination of grounded theory coding, spatial visualization and two-layer policy analysis. It systematically examined U.S. climate aid practices across the Obama, Trump, and Biden administrations from three analytical dimensions: discourse narrative, spatiotemporal evolution, and policy instruments. The findings reveal that the geostrategic shift in U.S. climate aid advances synergistically across these three dimensions. In terms of discourse narrative, U.S. climate aid discourse has undergone a phased evolution from a “strategic altruism” narrative to an “America First” narrative and then to a “great-power competition” narrative, progressively constructing China as a “partner” in climate governance and subsequently as a “systemic rival” in the domain of climate aid. In terms of spatiotemporal evolution, the allocation of climate funds follows the logic of “strategic value screening,” with resources shifting from counterterrorism pivot states to Indo-Pacific strategic pivots, whereas regions with high climate vulnerability but limited strategic value are systematically marginalized. In terms of policy instruments, the shift has been institutionally consolidated through a two-tiered linkage of “domestic empowerment-external implementation” and the triple synergy of institutional rules, security alliances, and funding programs. These three dimensions do not operate in isolation but constitute an integrated whole: discourse shapes cognitive legitimacy, spatiotemporal patterns provide orientation for resource allocation, and policy instruments provide institutional support. The external drivers of this shift stem from the disruptive impact of China’s Belt and Road Initiative and the rising discursive power of the Global South, while the internal drivers originate from the bipartisan consensus on hegemonic preservation amid political polarization and anticipated economic returns from the green industrial sector. This study demonstrates that the geostrategic shift in U.S. climate aid not only undermines the fairness and effectiveness of climate aid mechanisms but also poses a serious challenge to the global climate governance system. The academic contribution of this study lies in the construction of a comprehensive three-dimensional analytical framework that reveals the logical mechanism underlying the synergistic evolution of the geostrategic shift at the levels of intention, behavior, and institutions and provides a methodological reference for interdisciplinary research on climate politics.

Armed conflicts in the India-Pakistan region are a core security issue in South Asia, affecting the stability and security of China’s southwestern border and posing a threat to the implementation of the Belt and Road Initiative in the region. Studying this conflict is necessary not only to secure peace, stability, security, and development in China’s southwestern border areas, but also as an important practical step in China’s engagement in global security governance and its efforts to promote the building of a community with a shared future for humankind. These examinations hold significant geostrategic, security, economic, and diplomatic implications for China. Spatiotemporal patterns represent a key focus of geographical inquiry, yet existing research has devoted insufficient attention to the characteristics of the temporal evolution of conflicts. This study systematically analyses the temporal characteristics, cyclical patterns, and self-exciting and decay effects of the evolution of armed conflicts in the India-Pakistan region between 1989 and 2025. It conducts its investigation based on five key dimensions, as follows: conflict between the Indian and Pakistani governments; conflict between the Indian government and armed groups; conflict between the Pakistani government and armed groups; conflict between armed groups; one-sided violence. The results indicate that, first, different types of conflicts exhibit significant differences in their evolutionary stages and characteristics, which in turn are reflected in the number of phases, intensity levels, and developmental trends. Second, weekly periodic effects are not evident, whereas monthly and quarterly periodic effects are pronounced. Monthly peaks are concentrated from May to October, and quarterly peaks are concentrated in the second and third quarters (troughs occurr in December and the first quarter, respectively). The periodic effects explain the existence of temporal dependence. Third, the self-exciting effect is present in all five categories of conflict but exhibits heterogeneity, with the probability of a conflict recurring rising markedly within a short period after an initial outbreak. The self-exciting effect serves as the primary driver of conflict between the Indian and Pakistani governments, among armed groups, and between the Pakistani government and armed groups. Although these effects are also present in one-sided violence, their contributions remain relatively limited, and such violence is still more influenced by historical factors. Fourth, the self-exciting effect exhibits a robust net impact, reflecting genuine temporal dependence among conflict events rather than being driven by hot months or commemorative dates. Except for one-sided violence, hot months generally increase conflict intensity. Political festivals have a more pronounced influence than celebratory festivals do. This study advances the literature by moving from descriptive analysis to mechanistic analysis, quantifying conflict dynamics, and comparing conflict types to support targeted interventions. Future research may (1) refine covariate specifications to improve model accuracy and explanatory power; (2) analyze, from a temporal perspective, the evolution from conflict to negotiation and trigger effects; (3) incorporate a spatial dimension to capture the heterogeneity and geographical diffusion patterns of conflicts; (4) compare the goodness-of-fit of different decay kernels; (5) further explore mutual excitation effects in India-Pakistan conflicts.

This paper takes 30 provinces in China from 2012 to 2023 as research samples. By constructing a comprehensive evaluation index system for data elements and the innovative development of the circulation industry. and by characterizing their spatiotemporal evolution characteristics, this study adopts fixed-effect and interaction-term models to explore the contextual impacts of data elements on the innovative development of the circulation industry, and identify the boundary conditions of government behavior. The empirical results show that, first, the development levels of both data elements and circulation industry innovation achieved steady growth during the research period. Regional disparities tend to converge under the guidance of national strategies, while their spatial distribution still presents an unbalanced pattern of “the East taking the lead, the Central region keeping pace, and the West lagging behind.” Second, data elements serve as a driving force empowering innovative development in the circulation industry, and their empowerment effect exhibits a spatial gradient pattern: gradually increasing from the West to the East of China. Third, the empowerment effect of the data elements has a prominent contextual dependence. Human capital and industrial upgrading exert negative moderating effects, and the coefficient signs of their interaction terms conform to the theoretical characteristics of substitution moderation. Innovation output contrasts these negative effects by generating a significant positive moderating effect, indicating that higher levels of innovation output strengthen the marginal contribution of data elements. Fourth, the moderating effect of government behavior presents remarkable regional heterogeneity; excessive government intervention restrains the empowerment effect of data elements in Eastern China, whereas it has a positive synergistic relationship with data elements in Western China. Given the imperfect market mechanisms in the Western region, moderate government intervention effectively magnifies the innovation-driven effects of data elements. These findings provide an important geographical interpretation for understanding the regional adaptability of government-market relations in the digital economy era. The conclusions offer policy references for advancing the market-oriented allocation of data elements, facilitating the digital and intelligent transformation of the circulation industry, and boosting coordinated regional development. Regarding contributions, first, from a research perspective, it regards the incorporation of government behavior (into the analytical framework of data elements empowering circulation industry innovation) as an important institutional pathway; overcomes the bias in existing literature that overemphasizes “effective markets” while ignoring “proactive governments”; constructs a three-dimensional analytical framework of “data element allocation-circulation industry innovation-government behavior boundaries”; and expands the institutional analysis dimension for interdisciplinary research linking data elements and circulation economics. Second, regarding indicator construction, the study systematically constructs an evaluation system for data elements covering five dimensions (data infrastructure, development ecosystem, application level, financial support, and R&D investment), and an evaluation system for the innovative development of the circulation industry covering six dimensions (innovation, coordination, green development, openness, sharing, and security). It more comprehensively depicts the multi-dimensional characteristics and internal correlation between data elements and the innovative development of the circulation industry. Third, in terms of research methodology, this study adopts fixed-effects and interaction-term models to effectively identify the enabling effects of data elements, further explores the contingency of these effects from the perspective of three contextual dimensions (human capital, industrial upgrading, and innovation output), and reveals the regional heterogeneity embedded in the moderating effects of government behavior by incorporating spatial friction theory. Thus, it uncovers the gradient pattern, contextual dependence, and spatial adaptation mechanisms of data elements empowering the innovative development of the circulation industry from the dual dimensions of temporal evolution and spatial correlation.

The green assessment policy integrates the effectiveness of environmental governance with traditional GDP assessment, which is an important lever for the government to promote sustainable economic growth. Whether a green assessment policy can break the "carbon curse" effect of resource-based cities is an urgent research question. Existing research has laid the theoretical foundation for analyzing the intrinsic relationship between green assessment policies and the "carbon curse" of resource-based cities. However, resource-based cities are unique in terms of dominant resources and development stages, and the underlying mechanism of the "carbon curse" breaking effect requires further exploration. Therefore, this study focuses on resource-based cities in China and aims to explore new paths for resource-based cities to break through the "carbon curse" development dilemma within the context of green assessment policy implementation. At the theoretical level, an attempt is made to incorporate green assessment policies into the analytical framework of the "carbon curse" breaking mechanism, clarifying the mechanism of action of green assessment policies. At the empirical level, by collecting and calculating energy-saving and emission reduction target assessment data as a proxy for green assessment policies and using variable coefficient, mediation effect, and spatial Durbin models, this study examines the effect of green assessment policies on breaking the "carbon curse" of resource-based cities. This study draws the following research conclusions: (1) A "carbon curse" effect existed in resource-based cities in China, where the degree of impact of the "carbon curse" on mature, growing, and declining cities decreased in order, and the impact on oil and gas-, coal-, and metal-based cities also showed a decreasing trend in order. (2) Green assessment policies could effectively reduce the carbon emission intensity of resource-based cities, which is conducive to unleashing their "carbon curse" breaking effects. This conclusion held true after a series of robustness tests. (3) The breaking effect of green assessment policies was achieved by enhancing the environmental attention of local governments, driving green technological progress, and promoting the optimization of production factor allocation. (4) The breaking effect of green assessment policies exhibited heterogeneity, particularly evident in declining, coal-based, and southern resource-based cities; the green assessment policy has a spatial "local neighbor" spillover effect. The green assessment policy not only significantly reduced the local carbon emission intensity but also had a breaking effect on the "carbon curse" of surrounding resource-based cities through spatial spillover effects. Based on the research findings, this study suggests three policy implications: (1) At the central level, we should focus on the long-term and construct a targeted green assessment policy system to break the "carbon curse" according to local conditions. (2) The transmission effect of local government's attention to environmental protection, green technology progress, and optimized allocation of production factors should be enhanced. (3) Attention should be paid to the spatial spillover effects of green assessment policies, and a new pattern of green and low-carbon coordinated development should be constructed. Overall, this study empirically tests the existence and heterogeneity of the "carbon curse" effect in China's resource-based cities and examines the breaking effect of green assessment policies on this "carbon curse." This assists in comprehensively understanding the effects of green assessment policies, providing decision-making references for refining policy intervention priorities and improving targeting and implementation efficiency.

Digital media are becoming increasingly involved in the production and reproduction of rural spaces by reshaping how rural places are made visible, visited, and consumed. Existing studies have examined rural tourism, media communication, and new consumption spaces but have rarely explained how online visibility turns into a sustained operation and restructures rural spaces. Taking Shecun Village in Nanjing, China, as a case, this study constructs a “social media-actors-rural space” analytical framework from the perspective of virtual-real interaction. This study draws on field observations, interviews with operators of 23 rural cafés, and social media data from Xiaohongshu and Douyin. It traces the evolution of rural cafés in Shecun from 2020 to 2025, unveiling three evolution stages based on physical expansion and online communication intensity. In the initial stage (2020-2021), rural cafés were embedded as functional nodes that served cycling communities. Information circulated mainly through WeChat groups, WeChat Moments, and specialized outdoor sports applications. Although the scope of communication was limited, interest-based ties supported repeated visits. Rural café operators provided rest, supplies, and places to socialize, which led cyclists to begin to stop and gather in Shecun rather than just pass through. The spatial changes were gradual, and local residents participated mainly through house rental and basic services. In the acceleration stage (from late 2021 to 2024), peri-urban leisure demand increased, platforms recommended rural cafés more actively, the village became highly visible online, and cafés rapidly multiplied. Visual elements that were easy to recognize and circulate became increasingly important in space-making. More rural cafés started posting online content, and more content attracted more visitors. The online and physical growth fed each other. This process also made cafés look increasingly alike, brought tidal visitor flows, and created parking and environmental pressures, with the tensions between visitors and residents growing as well. This highlights the limitations of relying heavily on external visitors. In the adjustment and differentiation stage (from late 2024 onward), online traffic growth stabilized and competition intensified. Rural café operators shifted from chasing broad exposure to identifying target customers, retaining them, and maintaining long-term relationships. Operators increasingly organized recurring activities and used specific labels to attract target groups; some one-time visitors were gradually converted into repeated participants through continuous interactions. Collaboration among cafés and linkages with homestays, farm restaurants, handicraft shops and other local businesses expanded, forming a micro-level collaborative network. Rural cafés were no longer just places to consume, but began to organize customers, activities, and local resources. Their revenue sources diversified through catering, cultural and creative products, handicraft experiences, team-building packages, and space rental. Overall, rural spatial restructuring driven by rural cafés is not a simple response to online popularity but a staged process shaped by the interactions among media visibility, actors’ organizational practices, and existing rural spatial conditions. Digital media play different roles across stages, while their sustained operation increasingly depends on actors’ capacities to organize customers, relationships, and local resources. Studies have often separated online communication from offline restructuring, whereas this study considers them as one continuous process. The study demonstrates how actors turn online attention into visits, repeated visits, cooperation, and spatial changes. For planning and governance, this means moving beyond fixed, one-time control, and since rural cafés evolve in stages, regulations should adapt to changes in online visibility, actor coordination, and spatial pressures.

In China, urban agglomerations are key spatial carriers of new urbanization and coordinated regional development. Existing studies on urban agglomerations have mainly focused on morphological structure, functional linkage, economic networks, and transportation connectivity, leaving the socio-spatial dimension relatively unattended. Studies on socio-spatial differentiation have also failed to address the cross-city distribution and hierarchical differentiation of socioeconomic groups within large urban agglomerations, mostly because these studies have largely concentrated on individual cities or neighborhoods. To address this gap, this paper takes the Beijing-Tianjin-Hebei (BTH) region, the Yangtze River Delta (YRD), and the Guangdong-Hong Kong-Macao Greater Bay Area (GBA) as comparative cases. Based on the 2023 Baidu Huiyan population profile data, this study constructs a Socioeconomic Status Index (SESI) that integrates four indicators, namely, the proportions of high-income, highly educated, white-collar occupational, and high-consumption groups. Using standardized grid units, entropy weighting, natural breaks classification, and spatial visualization, this study compares the socio-spatial hierarchy, spatial distribution characteristics, and differentiation patterns of the three major urban agglomerations. The results show that socioeconomic attributes in all three agglomerations are concentrated in core cities, but their concentration patterns differ. The BTH features a high concentration of high-value groups in Beijing, the YRD exhibits a polycentric distribution, and the GBA is characterized by concentration around the Pearl River Estuary. The socio-spatial hierarchy also varies across the three regions. The BTH features a “middle-raised and two-end asymmetric” structure dominated by the middle level and with relatively prominent high and low levels, depicting a relatively strong degree of differentiation. The YRD forms an “olive-shaped” structure dominated by middle-level social spaces, reflecting relatively smoother hierarchical transitions. The GBA shows a structure dominated by high-level social spaces and a very small proportion of low-level spaces. This paper identifies three differentiated patterns: “Core Concentration and Stratified Disparity” in the BTH, “Multi-core Support and Axial Extension” in the YRD, and “Bay-area Concentration and High-level Equilibrium” in the GBA. It also develops a multidimensional explanatory framework that includes administrative resources and urban hierarchy, industrial division and employment opportunities, housing costs and livelihood opportunities, transport accessibility and spatial foundations, and openness and institutional coordination. This study extends socio-spatial research from the intra-urban scale to the urban agglomeration scale, integrates population big data with socioeconomic indicators, and offers policy implications for identifying socially disadvantaged spaces, stabilizing the middle strata, and promoting more equitable access to education, employment, housing, healthcare, and public services.

Urban thermal environments are jointly shaped by urban form, land-surface conditions, functional organization, and human activity. Taking the main urban area of Hefei, China, as a case study, this study examines the daytime-nighttime differentiation in summer Land Surface Temperature (LST) across urban functional zones. Daytime and nighttime LST data were derived from the MODIS MOD11A1 product for the summer of 2024, and multi-source spatial data were used to construct 11 indicators representing building morphology, ecological and surface-cover characteristics, and activity functions. In total, 1,217 road-bounded blocks were classified into different functional zones. Random forest models, permutation importance, kernel-smoothed response curves, and bootstrap confidence intervals were employed to identify dominant factors, nonlinear responses, and candidate thresholds. The study unveils pronounced temporal and functional differentiation in the urban thermal environment. Daytime hotspots were mainly concentrated in industrial and high-intensity mixed-use zones. Industrial zones recorded the highest mean daytime LST (36.7 °C) and the largest day-night temperature difference (9.5 °C). In contrast, nighttime hotspots shifted toward residential, mixed-use, and commercial zones, with residential zones exhibiting the highest mean nighttime LST (27.8 °C). Agricultural zones maintained the lowest mean LST during both daytime and nighttime (at 32.7 °C and 26.1 °C, respectively), indicating their cool-background and ecological-buffering functions. The nighttime random forest model showed substantially greater explanatory power than the daytime model (R² values: 0.725 and 0.415, respectively). The driving structures differed markedly between daytime and nighttime. At night, the activity-function variables accounted for 72.6% of the total feature importance, with Population Activity Intensity (PAI) alone contributing 55.6%, whereas building morphology and surface-cover characteristics contributed 16.6% and 10.8%, respectively. During daytime, the contributions were more balanced among building morphology, activity functions, and surface cover, which accounted for 38.8%, 36.0%, and 25.2%, respectively. The PAI, impervious surface area ratio (ISA), and building height were the leading daytime predictors. These findings suggest that daytime heat is jointly associated with development intensity, surface hardening, and activity concentration, whereas nighttime heat is more strongly related to persistent activity, facility agglomeration, stored-heat release, and restricted radiative and ventilative cooling. PAI should be interpreted as an integrated proxy for human activity and correlated with built-environment intensity, not as an independent causal factor. Clear, non-linear, and function-specific thresholds were identified. Across all samples, daytime PAI and ISA changed the effect direction by nearly 7.46 and 33.14%, respectively; nighttime PAI changed near 10.29%. The daytime ISA threshold increased to 52.53% in industrial zones. Residential nighttime LST became more sensitive near a PAI of 11.20 and a point of interest density of 197.12 km^-2. The nighttime response of commercial zones changed near a land-use-mix value of 0.31, whereas the daytime NDVI threshold in green zones was 0.49,the nighttime SVF threshold in green zones was approximately 0.97. The principal contribution of this study lies in its replacement of spatially uniform analytical units with functionally meaningful urban zones, integration of daytime and nighttime thermal processes within a unified framework, and translation of explainable machine learning results into quantitative and function-specific planning thresholds. Industrial and high-intensity mixed-use zones should prioritize reducing continuous impervious surfaces, applying cool roofs and high-albedo materials, and improving shade and ventilation. Residential and commercial zones should enhance nocturnal ventilation and permeable green spaces, as well as manage nighttime lighting and energy-intensive operations. The proposed framework provides a quantitative basis for time-sensitive and function-specific urban heat governance, although finer-resolution and multi-temporal data are required to test the stability and applicability of the identified mechanisms.

Existing studies of running behavior have largely focused on runners’ use of individual sites, with limited attention to runners who rotate among multiple parks over extended periods. Taking Guangzhou as the study area, this study aims to examine cross-park rotational running in terms of spatial expansion, combinational hierarchy, and park-node use. Running records covering 43 months (from January 2022 to July 2025) were obtained from the JOYRUN app. After spatial matching, distance-threshold filtering, and behavioral consistency screening, 1,968 valid activity records in 43 parks from 452 runners were retained. Parks were defined as nodes, and the complete set of parks used by each runner was defined as a hyperedge. A hypergraph network comprising 43 nodes and 452 hyperedges was constructed. Spatial mapping, standard deviational ellipse analysis, multi-order structural analysis, and indicators including active layer count and participation coefficient were used to examine cross-park rotational running. The results showed that cross-park rotational running exhibited citywide spatial expansion. Inter-park associations displayed a spatial pattern dominated by central areas, supplemented by peripheral areas, and linked across administrative districts. Based on the spatial span of hyperedges, three patterns were identified: short-distance connection, medium-distance extension, and long-distance crossover. This classification indicates that runners’ use of parks was not confined to nearby spaces but extended across multiple spatial scales. Cross-park rotational running also exhibited a combinational hierarchy characterized by core stability and flexible expansion. Second- and third-order hyperedges together accounted for 89.823% of all hyperedges, indicating that most runners used two to three parks during the study period. Fourth- to sixth-order hyperedges accounted for 8.849%, while seventh- to ninth-order hyperedges accounted for 1.327%. Layer redundancy rates ranged from 0.60 to 0.94, and 16 core anchor combinations that persisted across multiple orders were identified. These results indicated that higher-order combinations generally formed by adding new park nodes while retaining major existing park connections, allowing the spatial organization of running to maintain stability while accommodating flexible expansion. In addition, cross-park rotational running demonstrated a tendency toward balanced resource utilization in park-node use. The global network density was 0.35, while the mean active layer count and mean participation coefficient of park nodes were 4 and 0.40, respectively, indicating that most parks were involved in cross-park combinations of different sizes. Through multi-location rotation, runners reduced concentrated dependence on a limited number of local health resources and increased opportunities to incorporate parks into their running-place systems, thereby contributing to more balanced park utilization. Overall, by transforming both the analytical perspective and organization of behavioral data, this study extends running-behavior research from isolated-site use to multi-location associations. The introduction of hypergraph analysis enabled the representation of higher-order co-occurrence relationships among multiple parks that are difficult to preserve in pairwise networks. The study therefore complements existing research on intraurban health mobility and the spatial organization of multi-location physical activity. The findings also provide a demand-side reference for organizing cross-setting exercise spaces and supporting the planning and development of urban park systems.

High-density urban parks focus on recreation, tourism, and daily movement within limited spaces; however, concentrated use does not always produce balanced spatial vitality. This study examined how human activity demand matched the environmental supply in Xiaoyaojin Park, Hefei, and how mismatched conditions explained differences in spatial vitality. A total of 303 valid global positioning system trajectories from residents and visitors were collected from May 2024 to May 2025 and combined with multisource environmental data. The park was divided into 17 spatial units. Use intensity was calculated from the equally weighted standardized visit frequency and cumulative dwell time and was used to represent activity demand. The environmental supply was measured through facilities, plant and green-space characteristics, internal built environment, surrounding built environment, and climatic conditions. Spatial autocorrelation, ordinary least squares regression, geographically weighted regression, and nonlinear response analyses were used to identify spatial clustering, global relationships, local heterogeneity, and marginal changes. Demand-supply differences, ratios, and four-quadrant classifications were then used to identify mismatch types and assess conditions for spatial vitality formation. The results revealed the following four main findings. First, both groups displayed a clear “core concentration-peripheral decline” pattern. High use was clustered along the southern entrance, core recreational area, and eastern waterfront, whereas the northern and western peripheral units remained weaker. Resident use remained relatively stable, whereas visitor use exhibited stronger seasonal fluctuations and greater spatial selectivity. Second, the associations between environmental supply factors and use intensity differed between the two groups and exhibited spatial non-stationarity and nonlinear patterns. Residents’ use intensity was positively associated with service facilities, plant richness, seasonal vegetation variation, and accessibility. Sports facilities and the surrounding population density showed negative associations in some units. Visitors’ use intensity was positively associated with the surrounding building density, service facilities, recreational facilities, and landscape features, whereas sanitation facilities and accessibility showed negative local associations. These negative relationships represent local statistical patterns rather than direct adverse effects. The geographically weighted regression models provided a better fit than the corresponding ordinary least squares models, and several indicators showed diminishing marginal gains or potential saturation as the supply increased. Third, both groups formed four relative demand-supply states: high demand-low supply, high demand-high supply, low demand-high supply, and low demand-low supply. The spatial distribution and degree of mismatch are clearly different. Recreational lawns showed the strongest demand, exceeding the supply for residents, while Xiaoyao Pavilion showed the strongest mismatch for visitors. Several units showed reversals in mismatch type between the two groups. Fourth, the mismatch diagnosis indicated that a high use intensity mainly indicated activity concentration. High demand-low supply indicated constraints on vitality transformation, low demand-high supply indicated insufficient vitality activation, high demand-high supply required attention to potential pressure, and low demand-low supply reflected weak peripheral vitality. These findings show that high use intensity alone does not necessarily indicate balanced spatial vitality. Demand-supply matching provides a stronger basis for explaining vitality formation and internal differences. This study contributes a fine-scale framework that links the global positioning system-based activity demand, multidimensional environmental supply, spatially varying responses, mismatch diagnosis, and vitality interpretation. This also demonstrates the need to assess residents and visitors separately. Park management should shift from uniform supply expansion to differentiated and dynamic adjustments based on user groups and mismatch states, thereby improving demand-supply coordination and supporting more balanced spatial vitality.

As urban renewal increasingly shifts from physical reconstruction toward environmental quality and human-centered experience improvement, street shading has become an important spatial resource for mitigating summer heat exposure and supporting commercial outdoor activities. Existing studies have, however, largely relied on single-year or short-term cross-sectional data and have rarely examined whether the effects of different shading types remain stable over time or operate through street-user aggregation. Taking the six central urban districts of Nanjing, China, as the study area, this study develops a longitudinal analytical framework using street-view imagery and historical points of interest for eight years between 2014 and 2022. To reduce phenological inconsistencies among images acquired in different seasons, non-summer streetscapes were transformed into summer vegetation conditions using a CycleGAN-based unpaired image-to-image translation model. DeepLabV3+ was then applied to quantify the natural, artificial, and integrated shading resources. Pedestrians and riders were detected using Faster R-CNN to characterize street-user aggregation, whereas commercial vitality was represented by the number of commercial points of interest within 20-m street buffers. Structural equation models were estimated separately for each year to distinguish the direct from the indirect effects mediated by street-user aggregation. Our study reveals three major findings. First, all three shading resource types generally exhibited higher levels in the central areas and lower levels in the peripheral areas. Natural shading remained dominated by the very-low and low categories, which together accounted for more than 70% of street segments throughout the study period. Commercial vitality became increasingly concentrated in the very-low category (rising from 85.73% in 2014 to 90.28% in 2022), whereas high-vitality clusters became more spatially fragmented. Second, natural shading showed negative direct associations with commercial vitality (standardized path coefficient range, -0.28 to -0.09), artificial shading showed positive direct associations (standardized path coefficient range, 0.10 to 0.24), and integrated shading exhibited mostly negative direct associations (standardized path coefficient range, -0.05 to 0.07). Third, all three shading types exerted positive indirect effects through street-user aggregation. The path from street-user aggregation to commercial vitality remained comparatively stable across the years, with standardized coefficients of 0.27 to 0.66. Therefore, natural and integrated shading displayed competitive mediation, characterized by negative direct but positive indirect effects, whereas artificial shading displayed complementary mediation because its direct and indirect effects were both positive. These findings demonstrate that the influence of street shading on commercial vitality depends not only on shading quantity but also on shading type, street function, and the capacity of shaded environments to attract and retain pedestrian activity. Methodologically, this study extends shading-vitality research from static assessment to longitudinal mechanism identification by integrating seasonal correction, semantic segmentation, object detection, and year-specific structural equation modeling. These findings provide evidence for differentiated street-renewal strategies to improve the compatibility between summer thermal adaptation and commercial vitality: artificial shading should be coordinated with commercial frontages, while natural shading should be configured according to street function and pedestrian use.

Food security is a fundamental cornerstone of national security. With the accelerating pace of urbanization and upgrading of residents’ consumption structures, the patterns of China’s grain demand and supply have undergone pronounced changes. In major consumption zones, there has been a continual decline in the rate of self-sufficiency, whereas fluctuations in production-marketing balance zones have intensified, thereby contributing to an increasing intensification of regional supply-demand imbalances. Consequently, there is an urgent need to scientifically identify the spatial patterns and evolutionary characteristics of such grain supply-demand mis-matches at the county scale. In this study, based on data obtained from 2406 counties in China regarding permanent population size and grain output in 2000, 2010, and 2020, we used the grain production-demand spatial mis-match index, Markov chains, and first-difference OLS with regional interactions to examine the grain supply-demand associations in China’s three major grain production-consumption zones, as well as the regional heterogeneity in the effects of driving mechanisms. The findings revealed that overall, although county-level grain supply-demand relationships have become more sufficient nationwide, there has been an intensification of regional disparities. In terms of grain supply-demand relationships, the main production zone has shifted toward a looser balance, the main consumption zone has experienced acute tensions, and there has been only a limited improvement in the balance zone. Moreover, among the three zones, there has been an increase the degree of mis-match between grain production and demand, with the major consumption zone being characterized by the highest mis-match at the three assessed time points. Consequently, this reveals that the key to efficient food security governance lies in addressing structural mis-matches and regional imbalances. Whereas within the main production zone (e.g., the Northeast region and Huaihai Plain) and the balance zone (e.g., the Northern Arid and Semi-Arid Region, Yunnan-Guizhou Plateau), there have been expansions in the counties with surplus supply, there have been expansions of deficit counties into the consumption zone, particularly in South China and Zhejiang Province. Although balanced counties have contracted in spatial scope, they remain concentrated in the provinces of Guizhou and Guangxi. Nationwide, the disparities between the demand for and supply of grain can mainly be attributed intra-zonal county-level imbalances, rather than overall differences among the three major zones. This implies that in the future, the governance of food security should not only focus on macro-level balance across zones but also target intra-zonal county-level disparities. Our findings revealed that grain supply-demand relationships are characterized by a strong inertia. The production zone still has room for improvement, whereas the consumption zone faces entrenched scarcity risks, and the balance zone shows greater volatility. Cropland expansion, the input of agricultural labor, adjustments in cropping structure, and improvements in hydrothermal conditions were found to be positively associated with increases in grain self-sufficiency, and whereas population growth exerts a negative pressure on the rate of grain self-sufficiency, overall, rising grain prices and technological progress have had no significant positive effects in this regard. Consequently, the mechanisms associated with the aforementioned eight variables can be categorized into two types, the first of which is the production capacity support mechanism, which centers on the expansion of cultivated land, input of agricultural labor, adjustments of cropping structure, and improvements in hydrothermal conditions. The second mechanism is that of constraint pressure, which is driven by population pressure and return-risk constraints. Moreover, the impacts of these eight factors on grain self-sufficiency are characterized by pronounced temporal stages and regional heterogeneity in both direction and magnitude. These findings can provide a more refined regional explanation for understanding the changes in China’s self-sufficiency in grain, and also provide important references for optimizing the governance of regional food security.

In the context of climate change, the systematic identification and quantitative assessment of suitable cultivation areas for ancient tea trees in Yunnan Province can help strengthen resource conservation and promote the development of regional specialty industries. This study aimed to characterize the spatial distribution of suitable cultivation areas for ancient tea trees in Yunnan Province and their responses to future climate change. Filtered and spatially processed occurrence records of ancient tea trees, together with climatic, soil, and topographic variables, were used with the kuenm package in R was used to optimize two key settings of the Maximum Entropy (MaxEnt) model: Feature Classes (FC) and the Regularization Multiplier (RM). The optimized models were then used to project suitable cultivation areas for all ancient tea trees and two representative tea taxa, Pu’er tea and Dali tea, under current and future climate scenarios and to analyze the key environmental variables influencing the distribution of suitable areas and the migration characteristics of their centroids. The results showed that (1) compared with the models using default settings, the omission rates of the optimized models for the three modeling targets decreased by 31.93%, 38.11%, and 70.10%. The optimized models achieved high mean Area Under the Curve (AUC) and True Skill Statistic (TSS) values, indicating good predictive performance. (2) The spatial distribution of suitable areas for the three modeling targets was primarily influenced by temperature and precipitation. Elevation and slope showed relatively high contributions to some models, whereas soil variables generally showed lower contributions. (3) Under the current climate scenario, the three modeling targets differed in the extent and area of their suitable areas; however, all were primarily concentrated in southwestern Yunnan. The suitable cultivation areas for the three modeling targets accounted for 17.15%, 20.46%, and 10.76% of the total land area of Yunnan Province, respectively. (4) Under future climate scenarios, suitable areas for the three modeling targets were projected to shift overall from southwestern to central-western Yunnan and contract to varying degrees with later projection periods and higher greenhouse gas emission levels. The migration distances and directional shifts of the centroids of the suitable areas varied among the climate scenarios; however, they generally shifted northward, with their spatial locations remaining relatively clustered. This study provides a scientific basis for the conservation and sustainable utilization of ancient tea tree resources in Yunnan Province, strengthens the capacity of the ancient tea industry to respond to climate change, and promotes long-term, stable, and high-quality development.

Riparian zones are among the most sensitive interfaces where watershed ecological processes interact with human activity. It remains poorly understood whether riparian ecological structure and ecological function respond synchronously (or exhibit structure-function decoupling) under cascade hydropower development. To reveal the ecological effects of cascade hydropower development on riparian ecosystems, this study selected a 100-m river buffer zone and nine typical cascade hydropower stations in the Guizhou section of the Wujiang River Basin, China, as the research objects. It used 30-m land-use data from 1990 to 2025 and Landsat growing-season imagery from 1990 to 2023, and conducted land-use reclassification, buffer overlay analysis, the Normalized Difference Vegetation Index (NDVI) pixel dichotomy model, linear trend analysis, and correlation analysis of corresponding years. The goal was investigating the spatiotemporal dynamics of riparian ecological structure (natural shoreline rate) and ecological function (riparian vegetation cover), revealing their structure-function response relationship, and identifying ecological response modes among different cascade hydropower stations. The results showed that, first, the natural shoreline rate in the Guizhou section of the Wujiang River Basin exhibited an overall declining trend from 1990 to 2025, remaining relatively stable during the early stage but decreasing markedly after 2005, particularly around 2010. Significant differences were observed among hydropower stations, with the largest declines occurring at the Goupitan, Dongfeng, and Hongjiadu stations, indicating that cascade hydropower development and reservoir impoundment continuously compressed the natural riparian space. Second, riparian vegetation cover showed an overall increasing trend at a rate of 0.36%/a from 1990 to 2023, with more pronounced improvements after 2010, and vegetation responses differed significantly among hydropower stations. Most stations experienced a decline during the construction period followed by a gradual recovery during the operation period, whereas Goupitan exhibited continuous degradation throughout both stages. Third, the riparian ecosystem exhibited a clear structure-function response decoupling under cascade hydropower development. During 1990-2020, the natural shoreline rate and riparian vegetation cover showed an overall negative correlation, indicating a long-term trajectory characterized by continuous structural degradation and gradual functional recovery at the basin scale. Three typical ecological response modes were identified among hydropower stations, namely synchronous degradation, structure-damaged but function-recovering, and relatively stable. These findings reveal the long-term structure-function response decoupling of riparian ecosystems under cascade hydropower development, delivering a structure-function perspective for evaluating the ecological effects of cascade hydropower development. This study offers scientific support for natural shoreline conservation, riparian ecological restoration, and the differentiated ecological management of cascade-regulated river basins.