Evaluating street interface morphology is essential for urban design, yet existing approaches often struggle to combine large-scale applicability with higher-level morphological interpretation. This study proposes a scalable framework for assessing street interface morphology using an automated multimodal large language model (MLLM) agent. Using street view imagery (SVI), the framework evaluates four core morphological dimensions—enclosure, continuity, transparency, and roughness–through two complementary analytical streams: objective geometric measurement and subjective morphological assessment. To support reliable evaluation, the framework incorporates a dual-benchmark strategy consisting of manually derived geometric measurements and expert-consensus ratings for calibration and validation. Applied in Shanghai, the framework demonstrated reliable performance across the evaluated dimensions. The optimized agent was further extended to continuous street-segment analysis, demonstrating its applicability to large-scale urban assessment. By integrating objective and subjective evaluation within a scalable and interpretable workflow, the proposed methodology provides a practical tool for street interface morphology analysis and urban design assessment.

Walkability has become a critical issue in contemporary urban design and renewal to build sustainable cities, as decades of rapid urban expansion have undermined pedestrian experiences. Despite considerable scholarly discussion, urban designers still lack practical, morphology-based tools to guide the enhancement of walkability. In response, we distilled 12 walkability- and design-oriented urban indicators from 78 candidates using large language models (LLMs) and expert validation. Using a suite of morphological analysis and computer vision techniques, we quantitatively examined similar to 4600 urban blocks across 50 pedestrian-friendly districts worldwide and identified their common features through the lens of five design factors derived from PCA. Based on these findings, we developed a personalized recommendation approach that provides reference indicator value ranges using multilabel semantic matching and relevance-weighted calculation. Subsequently, URBANPEDIA, an interactive and collaboratively extensible design dictionary built on crowdsourced data, was developed. The platform provides design guidance for practitioners to enhance walkability while serving as a data infrastructure for the research community. This study demonstrates a scalable methodological framework that integrates LLMs, spatial analysis, and crowdsourced data to bridge the disconnect between research and practice, contributing to a paradigm shift toward computational urban science with human concerns.

This study presents a perception-informed urban comfort assessment framework. By systematically quantifying multidimensional interactions between people and their urban environment, it functions as a policy tool for evaluating and advancing livable city development. However, due to the lack of human-centered assessments, livability studies in China often remain at the macro-policy level, overlooking intrinsic influencing factors and their underlying mechanisms. This study develops a multidimensional urban comfort framework consisting of 34 indicators across four dimensions: socio-economic factors, three-dimensional urban morphology, environment factor, and human visual perception. Urban comfort is measured using the entropy-weighted TOPSIS method, the obstacle degree model, and the optimal parameter geographic detector. The study further examines internal obstacles and external drivers to reveal the possible mechanisms shaping urban comfort. The results show that: (1) Yantai’s urban comfort displays clear spatial heterogeneity. High-comfort areas are concentrated in flat inland zones, commercial centers, and environmentally favorable coastal areas. Distinct local clusters form a pronounced core–periphery structure. (2) Key internal constraints include the public transit station density, road network density, temperature-humidity index, and green space accessibility. (3) External factors—development motivation, cultural base, well-being of people, ecological foundation—significantly influence urban comfort. Resident activity level shows strong interaction effects with other variables. By integrating multidimensional urban characteristics and incorporating human perception, this study fills key gaps in urban comfort assessment. It provides theoretical support for improving regional livability and offers scientifically grounded pathways for urban renewal. The findings contribute to fostering harmonious coexistence between people and urban systems.

Urbanization exerts unprecedented pressure on the microbial community, with far-reaching implications for human health. While previous studies have documented microbial variations across the geographic urban-rural gradient, the temporal dynamics of microbial communities in response to urbanization remain poorly understood. Leveraging global datasets of 2,110 metagenomes spanning 45 cities on 6 continents and annual artificial impervious surface data (1985-2018), we investigated how microbial composition, diversity and functions correlate with urbanization age. Our results reveal that, with increasing urbanization age, there is a significant decline in microbial alpha diversity (P < 0.05), a pronounced compositional shift from nature-associated to human-associated microorganisms (P < 0.001), a rise in the total relative abundance of opportunistic human pathogens (P < 0.05) and a functional restructuring with potential human health implications. These findings provide empirical evidence of microbial community dynamics across the temporal urbanization process, offering actionable insights for urban planning (such as urban regeneration and renovation) and stage-targeted public health policies aimed at mitigating potential microbiome-related health risks in an increasingly urbanizing world.

Townscape continuity, which integrates modern architecture with traditional building features, is vital to preserving urban identity. Nevertheless, current architectural practice, particularly façade design, which is the key element in maintaining townscape continuity, remains labor-intensive due to fragmented understanding and inadequate tool support. This study proposes a knowledge-guided AI-generated content (AIGC) approach to improve the efficiency of architectural design. Jiangnan water town in China was adopted as a case study. Specifically, a comprehensive design knowledge repository was developed by integrating academic literature and high-quality design cases. In this process, a large language model (LLM) and a convolutional neural network (CNN) were used to distill the professional achievement, while eye-tracking experiments captured the public perception. This repository was then used to fine-tune a Stable Diffusion model and informed prompt engineering to improve generative quality, thereby improving generative controllability and semantic reliability. A custom evaluation model, trained on preliminary knowledge-guided AIGC, further filtered suboptimal outputs. The final generation results demonstrated strong townscape continuity and adaptability across various architectural scenarios. Finally, expert evaluations confirmed that the proposed approach outperformed baseline models in the design quality. In short, this study offers a systematic knowledge repository for leveraging AIGC in architectural design, supporting practitioners in achieving townscape continuity more effectively.

Urban renewal is increasingly implemented through heterogeneous micro-scale interventions, yet monitoring these changes and assessing their equity remains methodologically elusive. We introduce an expert knowledge-guided vision-language model (VLM) framework using 105,332 paired street view images from Shanghai (2017–2022) to detect, classify, and interpret renewal measures. Through domain knowledge injection and negative constraints, our model achieves high-precision classification across six substantive renewal types, revealing multi-dimensional renewal disparities. a) Allocation: Mantel tests show renewal configuration is more strongly associated with functional services and perceptual characteristics than with housing prices, with most renewal types concentrated in moderately served transitional zones rather than the most disadvantaged communities. b) In terms of measures, semantic analysis reveals qualitative measure inequity—high-value areas receive aesthetic interventions, while low-value areas receive pragmatic maintenance. c) Capitalization: exploratory hierarchical regression models show that renewal composition is associated with housing-price growth in both high- and low-price areas, and that the direction of the association differs by renewal type and price zone: Greenery & Landscaping is positively associated with housing-price growth in low-price areas but negatively in high-price areas, while Road & Traffic Infrastructure and Storefront & Signage are negatively associated with price growth in low-price areas. This study demonstrates VLMs can unveil justice issues beyond spatial allocation.

Urban vitality serves as a crucial indicator for evaluating the enhancement and transformation of the human environment within the framework of stock planning. Accurately identifying urban spatial patterns at a human scale and uncovering the spatiotemporal heterogeneity of how urban spatial structure influences urban vitality can provide scientific guidance for fostering vibrant and sustainable urban environments. Leveraging multi-source geospatial big data and a multi-scale geographically weighted regression model, this study integrates three-dimensional space and human perception to construct an urban spatial structure indicator system and reveals the spatiotemporal heterogeneity of the relationship between urban spatial structure and urban vitality across different dimensions. Additionally, it compares variations between weekdays and weekends to capture dynamic changes. The research results indicate that: (1) Urban vitality in Yantai central urban area exhibits a typical core–edge structure with strong spatial autocorrelation, suggesting the absence of an orderly and well-planned spatial framework. (2) The spatial distribution of urban vitality is closely linked to urban spatial structure, with notable differences in the influence of specific spatial structure variables between weekdays and weekends, highlighting the necessity of integrating temporal factors in urban vitality assessments. (3) The effects of spatial structure variables on urban vitality demonstrate spatial attenuation or amplification, emphasizing the importance of adopting a context-sensitive approach that considers both global and local perspectives in urban planning. The research findings serve as a valuable foundation for urban planning and design in Yantai central urban area, aiming to improve the precision and effectiveness of response strategies.

Under a human-centered approach, accurately identifying the spatial patterns of urban vitality and revealing the mechanisms through which the built environment affects it can scientifically guide the organic cultivation of urban vitality. In light of this, the main urban area of Yantai City is taken as a case study, utilizing multi-source geographic big data to conduct both theoretical and empirical research. An index system for the urban built environment is established based on four dimensions: human perception, functional, accessibility, and building form. Advanced methods, including Deep Fully Convolutional Neural Networks (SegNet), Random Forest Regression (RFR), and Spatial Lag Regression (SLR), are employed to explore the impact of the built environment on urban vitality. The research findings indicate that: (1) Urban vitality presents a composite spatial structure that embodies both "multi-center" and "clustered" characteristics, exhibiting two primary types of local spatial autocorrelation: "high-high" clustering and "low-low" clustering. (2) The disparities in urban vitality reflect an imbalance in the distribution of functional, accessibility, building form, and human perception, with functional playing a more critical role in nighttime and daytime urban vitality than other dimensions. (3) The effects of the built environment on daytime and nighttime urban vitality show varying degrees of heterogeneity regarding significance and direction. Factors such as BPOI(Commercial Points of Interest), integration, accessibility, and vibrancy have a substantial positive impact on vitality clustering, while human perception becomes increasingly important for enhancing nighttime vitality. These results provide refined technical support for urban micro-renewal, enhancing the relevance and effectiveness of response strategies.

This study proposes a multidimensional framework for measuring sense of place continuity in traditional settlements—a quality defined as the persistence of place identity cues that enable recognition and meaning making amid urban transformation. Under rapid urbanization, sense of place in traditional townscape is increasingly at risk of erosion. As a form of “living heritage”, it plays a vital role in sustaining social interaction and cultural memory. By leveraging machine learning algorithms with large language models, the study develops an AI-enhanced methodology to advance the computational understanding of “place”, enabling both diagnosis of current conditions and evaluation of continuity. Jiangnan water-town was selected as the study area known for its strong place identity. The framework utilizes multi-source urban data and develops a series of python-based tools to extract spatial, visual, and semantic features using deep convolutional models and natural language processing techniques. These are then compared against reference profiles derived from preserved environments to assess the continuity of sense of place. This study offers a novel pathway for quantifying previously intangible qualities, supporting data-informed urban design and assessment in the context of urbanization. It also contributes to a theoretical reconsideration of authenticity in place-making, shifting emphasis from formal imitation to continuity rooted in local context and lived experience.

The role of artificial intelligence (AI) in urban science has evolved from early-stage analytical support to active involvement in generative design, transforming how designers conceptualize and iterate spatial solutions. Benefitting from the rapid growth of generative urban design, urban designers can now quickly receive many design solutions, but selecting the best among numerous results remains challenging. In response, this study attempts to develop an intelligent method capable of measuring a key performance in design evaluation, i.e., the aesthetic performance of two-dimensional urban textures. Specifically, we first apply large language models (LLMs) to extract and refine key evaluative dimensions, providing a theoretical basis for aesthetic assessment. Four dimensions related to urban morphological aesthetic performance were selected based on classical urban theory: orderliness, richness, imaginability, and building density. Using urban texture data from OpenStreetMap and expert ratings collected through questionnaires, we built a training dataset of paired images and scores. The evaluation method was then developed through a multimodal approach via the Contrastive Language-Image Pre-Training (CLIP), which can identify numerical aesthetic performance of urban textures and has demonstrated accuracy comparable to expert evaluations. Moreover, using LLMs within a transformer-based architecture, the interactive interface provides textual feedback based on the evaluation results, improving interpretability and usability for designers. In short, this study contributes to generative urban design and quantitative urban morphology by providing a multi-model approach to measure morphological aesthetic performance.

The growing emphasis on walkable environments underscores the need to understand how streetscapes influence pedestrian perception, yet the specific threshold intervals at which these effects become significant remain underexplored. This study proposes a workable framework that integrates immersive VR, wearable biosensors, and machine-learning algorithms to identify refined thresholds and further advance evidence-based urban design. Initial steps involved identifying key streetscape elements across different street types, for example, sidewalk width, interface permeability, utility area width, and cycle parking, as informed by literature and existing guidelines. This was followed by constructing 251 VR streetscape scenes, through which subjective evaluations were collected from 185 participants to establish streetscapes' preliminary thresholds. Subsequently, wearable biosensors and a window-based change point detection algorithm were employed to determine refined threshold intervals, the results of which informed actionable guidelines for street design. The framework integrates human-centered analytical tools to systematically quantify refined streetscapes' thresholds and translates them into evidence-based guidelines for urban design, offering a replicable pathway for optimizing streetscape quality. Overall, the study provides quantitative insights and actionable design guidance that advance the broader agenda of evidence-based and human-centered urban design.

Urbanization and low-carbon development are critical issues of global concern. As urbanization has reached its middle to late stages, cities face the dual pressures of development and environmental challenges. This study constructed a theoretical framework for urban vitality in six dimensions: social, economic, cultural, environmental, spatial, and perceptual. Using methods such as spatial syntax, entropy-weighted TOPSIS, deep learning models, and geographic detectors, we analysed the distribution characteristics of urban vitality in Yantai’s central area, explored how vitality-contributing factors influenced carbon emissions, and elucidated the association of urban vitality with carbon emissions. The results indicated that (1) urban vitality exhibited a multicentred distribution pattern of “low in the hinterland—high along the coast”; (2) significant differences existed in the impacts of various vitality dimensions on urban carbon emissions; (3) different urban vitality factors have varying levels of explanatory power regarding the spatial distribution of carbon emissions, with maximum building height exhibiting the strongest explanatory power, while the selection degree shows the weakest; and (4) the interactions between these factors typically demonstrate a two-factor enhancement, with the interaction between maximum building height and integration having the most significant effect on urban carbon emissions. This study innovatively integrates three-dimensional spatial and cultural perception perspectives, addressing the biases found in previous research that represented urban vitality from a singular viewpoint. It provides a more comprehensive framework and methodology for evaluating urban vitality, and the findings can offer recommendations for building low-carbon, high-vitality, and sustainable urban environments.

The integration of generative artificial intelligence (GenAI) into urban planning and design has rapidly advanced as a key research frontier in recent years. This study reviews the application and emerging trends of GenAI in different stages of planning and design, including theoretical understanding, spatial analysis, and generation and evaluation of planning and design. Specifically, ① theoretical understanding: GenAI can construct multimodal knowledge graphs that support a more systematic understanding of fragmented knowledge in planning and design by integrating heterogeneous textual, visual, and spatial data. ② Urban spatial analysis: GenAI can enhance analytic capacity and inclusiveness in spatial analysis. It enables efficient interpretation of current socioeconomic and spatial conditions from multimodal data and can simulate the reasoning of multiple stakeholders, including experts and the public. Although limited in mechanistic, rule-based analyses, it can be extended via prompt engineering and tool use, lowering technical barriers. ③ Planning and design generation: GenAI can assist practitioners in drafting text, generating design images, and producing simple three-dimensional models. However, it is not yet capable of independently producing comprehensive, regulation-compliant planning documents or spatial layouts. Thus, it should be regarded as a supplementary tool rather than a replacement for human expertise. ④ Planning and design evaluation: Through domain adaptation and knowledge integration, GenAI can evaluate planning texts, spatial performance, ecological performance, and other multicriteria dimensions. It also supports multistakeholder assessments via large language model-based agentic workflows, but does not yet reliably automate the iterative optimization of proposals. Although the use of GenAI in this field is still in its early stages, it demonstrates cross-process potential across the entire planning and design workflow. It is expected to accelerate the shift toward computational urban science, move practice from experience-oriented to engineering-oriented approaches, and enhance the efficiency and quality of public participation, thereby better aligning planning outcomes with diverse societal needs.

Recent advances in large language models have transformed urban planning from passive tool-assisted workflows to active human–AI collaborative partnerships, enabling natural language-driven design generation, multi-agent stakeholder simulation, and intelligent decision support. This survey systematically examines the integration of LLMs in urban planning, establishing a comprehensive taxonomy covering task categories, technical paradigms, and collaboration patterns. Furthermore, the survey identifies critical evaluation frameworks and benchmark datasets while examining implementation challenges, including domain knowledge integration, scalability constraints, and ethical implications. The work bridges theoretical advances with practical deployment considerations, providing guidance for selecting appropriate LLM approaches across different urban planning contexts and scales.

With the global trend of population aging, human-centered development that integrates medical convenience with daily life quality has become a critical necessity. However, conceptual frameworks, evaluation methods, and spatial prototypes for such ‘healthcare–daily-life’ development remain limited. This study proposes Hospital-Oriented Development (HOD) as a framework to promote collaborative development by considering both hospital accessibility and urban development intensity, derived from multi-sourced urban data. First, a conceptual framework was established, consisting of three dimensions, i.e., network accessibility, facility completeness, and environmental comfort, which was then characterized by twelve indicators based on urban morphological features. Second, these indicators were quantitatively evaluated through detailed values measured among 20 exemplary hospitals in Shanghai selected via user-generated content. Finally, HOD performance and morphology informed the spatial prototype. The results reveal confidence intervals for each indicator and recommended spatial features. Numerically, there was a positive correlation between facility completeness and network accessibility, but a negative correlation with environmental comfort. Spatially, a context-specific HOD prototype for China was developed. This study proposes the concept of HOD, delivers quantitative measurements, and develops a spatial prototype via empirical research, providing theoretical insights and evidence to support the improvement in healthcare environments from a human-centered perspective.

Digital twins, having gained prominence in industrial sectors, are emerging in construction for enhancing intelligent management through real-time monitoring, performance simulation, and data integration. Despite their potential, systematic analysis of frameworks and enabling technologies remains lacking. This review systematically investigates over 150 studies to: (1) analyze the general framework and its extensions for building digital twins, and (2) evaluate enabling technologies and tools based on the modeling procedure. We propose a general structure for building digital twins and analyze frameworks centered on four data types derived from different sources. Our analysis indicates that different types of frameworks exhibit distinct characteristics and inherent limitations, yet a standardized framework for integrating heterogeneous data is still lacking. Through systematic analysis of enabling technologies across four key aspects of the modeling procedure, we investigate building digital twin cases regarding their modeling procedures, technologies employed, and common issues. We identify four key challenges: (1) limited prediction data integration and data analysis, restricting frameworks to monitoring rather than decision-supporting; (2) multi-source data heterogeneity and poor tool interoperability; (3) complex and non-standardized data integration procedures; and (4) low automation in model development. Future research should focus on: (1) standardizing data formats and interoperability tools, (2) developing unified platforms for multi-source data integration, and (3) integrating predictive analytics to enhance decision-making. This study establishes connections between frameworks and enabling technologies, identifies existing problems, and provides actionable recommendations to accelerate the adoption of building digital twins.

Urban building energy modeling (UBEM) empowers the construction of green and low-carbon cities. However, its development is hindered by the uncertainty of data inputs, including inherent uncertain data (IUD), measurement uncertain data (MUD) and scenario uncertain data (SUD). This paper employed one typical MUD, namely, the thermal parameters of construction assemblies, as the study object to analyze the accuracy and stability of UBEM using four different approaches, i.e., archetypes built with standards, archetypes built with local datasets, probabilistic models and urban factor methods. The results showed that when focusing solely on thermal parameters, the RE values could reach 500 % at the building level but tended to converge to less than 90 % at the district level. In addition, the mean of relative errors at the building level influenced the accuracy at the district level as well as the rate of mean convergence. However, this metric did not affect the threshold to attain range convergence, since its number was fixed, neither related to the sample size nor to the calculation accuracy. This study emphasized that using real data could enhance the accuracy of UBEM, regardless of the archetype or the stochastic approach used, but the distinctions mainly occurred at the building level. Moreover, the large-scale simulation work could be transformed into the task of calculating energy use data for several convergence units, each consisting of dozens of buildings, since these units were able to exhibit stability on their own.

ObjectiveWalking space is an important part of urban public space, and its spatial quality is a basis for the construction of walkable cities, which directly affects residents’ travel willingness and walking experience.MethodsThis research takes relevant literature on walking safety screened from Web of Science and CNKI databases spanning the period from 1975 to 2022 as research object, and analyzes the distribution characteristics of literature related to the current research on walking safety and environmental influencing factors thereof, as well as such contents as research context and research hotspots.ResultsFirstly, from the perspective of literature distribution, a large number of works of scholars in such fields as traffic engineering, public health and sociology have been cited in literature on walking safety, reflecting that pedestrians’ activity safety, traffic safety and security needs are important directions to promote the research on walking safety. At present, the research on walking safety in developed countries is more active and fruitful. The advanced urbanization process and prominent social contradictions make developed countries and regions more capable of investing energy and funds in walking safety research, exploring the safety issues faced by vulnerable road users, and deepening the degree of research refinement. Secondly, from the perspective of research evolution, the research has the following findings. 1) The analysis results of highly cited literature concerns show that walking safety and walkers’ safety perception dominate the development direction of relevant research. 2) Based on WoS database, the analysis of theme words shows that taking accident analysis as the core, taking pedestrian characteristics as the perspective and taking new technology as the guidance are the three major topics in the field of walking safety research. New data and technologies support the analysis of core environmental factors influencing walking safety. 3) Based on CNKI database, the analysis of keywords shows that domestic research focuses on the spatial quality of walking environment and the special needs of vulnerable walkers, and it is urgent to establish a walking safety environment evaluation system in China. In addition, the keyword analysis also shows that the impact of built environment on pedestrian safety accidents and pedestrian safety perception has become a focus of attention. Thirdly, from the perspective of research hotspot, it can be seen form the three dimensions of activity safety, traffic safety and security that, different levels of environmental elements such as neighborhoods, streets and intersections, have important influence on walking safety. In recent years, the extensive application of quantitative analysis and data calculation methods has brought new ideas and technologies to the research. Especially in the field of data mining regarding urban form and walkers’ psychological perception, new technologies are helpful for in-depth analysis of the influence of built environment factors on walking safety. Finally, the research trend of walking safety is reflected in three aspects: Research object, content and method. Research object has expanded from collision accident to walkers’ psychological perception, and researchers increasingly pay attention to the influence of the safety quality of walking environment on pedestrian psychology. As the urban built environment is facing more and more complex conditions, the research focus extends from the analysis of individual collision accidents to the comprehensive urban environment construction. Research methods have also been developed from single-specialty technical analysis to multi-dimensional integrated research, which promotes the comprehensive application of research methods from such fields as environmental psychology, traffic safety and road engineering in the research on walking safety.ConclusionThis research reveals the development direction, main context and research hotspots on walking safety and environmental influencing factors thereof, highlights the significant influence of multi-level built environment factors on walking safety and walkers' safety perception, and provides theoretical support for empirical researches on walking safety environment. In future research, it is necessary to conduct collaborative thinking on walking safety and pedestrian safety perception, promote the applicability of relevant research technologies, and pay attention to the influence of the development of emerging transportation technologies on the safety quality of walking environment.

In recent years, digital twin city platforms often encounter issues such as emphasizing the physical model’s accuracy over social cognition and specialized applications over a comprehensive data system, hindering the fulfilment of refined urban management’s real needs. Therefore, it is essential to define the characteristics of an urban management-oriented digital twin platform and construct a detailed evaluation mechanism. This study examines the framework for evaluating mapping rates, introducing three indicators: data resolution, data freshness, and data relevance. We developed a quantifiable and replicable evaluation model to assess data completeness, update timeliness, and network correlation degree. Using Shanghai’s Huamu digital twin platform as a case study, we calculated each indicator and formed a comprehensive mapping rate evaluation. This research achieves a quantitative analysis of digital twin city platforms’ development quality which was previously unmeasurable. Additionally, this study aids in advancing digital twin city platforms to facilitate the development of a “bottom-up” refined urban management approach.

Urbanization has a profound impact on the global carbon cycle and climate change, posing dual challenges of development and environmental sustainability for cities. Accurately identifying urban vitality and exploring the influence of its constituent factors on urban carbon emissions can provide scientific guidance for the sustainable construction of low-carbon, high-vitality cities. This study focuses on the central urban area of Yantai as a case study, conducting both theoretical and empirical research based on multi-source big data. The main findings are as follows: (1) Urban vitality exhibits a multi-centered pattern characterized by "low vitality in the hinterland and high vitality along the coast," with dimensions such as social welfare and economic development demonstrating similar spatial distributions, while the environmental atmosphere dimension shows a contrasting distribution; (2) The strongest correlation is found between the built environment and carbon emissions, whereas the environmental atmosphere demonstrates a significant negative correlation; (3) Spatial scale differences exist in the relationship between urban vitality and carbon emissions; (4) The development of urban vitality factors should be tailored to local conditions, ensuring an appropriate balance between overall and local spatial dynamics. This study integrates three-dimensional spatial and perceptual perspectives, establishing a systematic theoretical framework for urban vitality and employing Multi-scale Geographically Weighted Regression (MGWR) to investigate the spatial heterogeneity of the impact of various vitality factors on carbon emissions, thereby offering insights for sustainable urban development.