Achieving real-time indoor 3D reconstruction using low-cost devices in GNSS-denied environments is highly valuable for applications such as indoor navigation, indoor spatial measurement, and augmented reality guidance. However, vision–IMU-based real-time indoor 3D reconstruction inevitably involves repeated scanning of the same scene regions. When pose inconsistency exists among repeated observations, the same structures may be incorrectly fused into different locations, leading to artifacts such as surface layering and ghosting, which are further amplified by accumulated estimation errors. These artifacts are not caused by errors in a single module, but are closely related to pose estimation, topological relationships, and reconstruction error propagation, which severely degrade the geometric reliability, measurability, and task usability of indoor maps. To address these issues, this paper proposes an end-to-end real-time 3D reconstruction framework that reduces artifacts from three aspects: encoder enhancement, structural relationship constraints, and adaptive reconstruction. Specifically, IMU measurements are encoded into tokens and injected into a Vision Transformer to improve pose estimation accuracy. Then, a Fiedler-vector-based topological region partitioning strategy is introduced to incorporate global scene connectivity into the end-to-end pipeline, thereby limiting artifact propagation across different regions. Finally, each segmented region is reconstructed through adaptive neural point-cloud reconstruction and global optimization to further improve local geometric consistency. Extensive quantitative, qualitative, and ablation experiments are conducted on multiple self-collected and public datasets, together with a comprehensive comparison against existing state-of-the-art vision-based end-to-end and 3DGS-based indoor reconstruction methods. The experimental results demonstrate that the proposed method achieves clear advantages in artifact suppression, geometric accuracy, reconstruction quality, and real-time performance. This method provides a new solution for improving the reliability and practical usability of real-time indoor 3D reconstruction in GNSS-denied environments. The source code and supplementary data are available at: https://github.com/starlingonearth/end-to-end-artifact-suppression-real-time-3d-reconstruction.

Urbanization and rapid economic growth have exacerbated urban heat effects, increasing the frequency of heat-related disasters and intensifying human health risks. Urban traffic generates substantial carbon emissions and associated heat, which significantly alter roadside thermal environments and impact human activities. Numerous previous studies have investigated urban thermal environments and their influencing mechanisms. However, the relationships between road-level traffic carbon emission (TCE) and road surface temperature (RST) remain insufficiently explored. In this study, roadway segment-based TCE and RST were acquired by integrating hourly traffic flow information, localized vehicle carbon emission factors, high-resolution Landsat-8 remote sensing datasets, and the road network. Three commonly used linear regression models and an improved Random Forest (RF) model were utilized to assess the impact of TCE on RST for different grades of roads. The study showed that carbon emissions from road traffic exhibit a locally focused distribution pattern in space. Compared to other grades of roads, higher levels of TCE were observed in urban main roads. In summer, roads (e.g., minor arterials) with lower grades tended to have a higher thermal risk, with freeways having the lowest TCE and urban expressways experiencing the greatest TCE fluctuations. An improved RF model integrating the spatial weight matrix and Gaussian process could more efficiently identify the nonlinear effects of TCE on RST. The contributions of TCE to summer RST were 0.4, 0.37, 0.54, and 0.56 for freeways, urban expressways, main roads, and minor arterials, respectively. The relative impact of road TCE with lower grades on RST becomes more significant, while the impact of surrounding buildings and green areas tends to decrease. Our findings provide valuable insights for reducing urban carbon emissions and thermal risks.

Understanding microclimatic variations across Local Climate Zones (LCZs) is crucial for optimizing urban morphology to enhance human thermal comfort and promote sustainable urban environments. While numerous studies have examined the spatio-temporal patterns and underlying mechanisms of thermal environments in different LCZ types, temperature variability within identical LCZs remain insufficiently explored. Moreover, research on other microclimate factors-such as relative humidity (RH) and wind speed (WS)-at both inter- and intra-LCZ scales is limited. In this study, a hybrid modeling framework based on the Weather Research and Forecast (WRF) model was proposed to accurately predict near-surface meteorological fields. It was achieved by refining the Urban Canopy Model (UCM) during the preprocessing stage of the WRF model and then integrating ML algorithms during its postprocessing stage. The predictive performance of four WRF-based datasets was evaluated and compared under relatively stable weather conditions across four seasons. The best-performing ML(XGBoost)-enhanced model dataset was applied to identify multivariate microclimate variations both inter- and intra-LCZs. The analysis was conducted in Shenzhen, a coastal hilly city in southern China. The results revealed that: (1) the WRF-ML hybrid models performed significantly better than the Standard and UCM-refined WRF models, with the optimal XGBoost-enhanced model achieving hourly average RMSE values of 0.613 K for AT, 1.131 % for RH, and 0.207 m/s for WS; (2) unique local geographic conditions, such as coastal surroundings and continuous natural landscapes, significantly influence microclimate variations within urban built-up areas; (3) inter-LCZ microclimate differences were generally within 1.5 K for AT, 9 % for RH, and 0.7 m/s for WS, exceeding intra-LCZ differences; and (4) built-up LCZs could be further divided into 2 similar to 4 subcategories with distinct microclimate conditions, some of which showed relatively favorable microclimate environments during the high temperature period.

Cities and social systems are undergoing fundamental transformations driven by digitalization, as technology interacts with and integrates into traditional urban social processes in complex ways, significantly reshaping spatial structures. While prior research has yielded valuable theoretical insights and case studies, a broadly comparable and replicable research paradigm remains elusive. This gap underscores the need to explore and construct an effective framework that connects historical and contemporary research streams to promote the systematic development of the field. This study addresses this gap by reporting relevant knowledge through quantitative and qualitative analyses and integrating a complex network theoretical framework. The results indicate that 2022 marked a research inflection point; 50% of studies are concentrated in two countries; urban studies journals are the primary contributors; and 76% of the research supports the technology embedding perspective. Based on these findings, this study explores the "space-element-structure" systemic composition of urban spatial structures in the digital era. It further constructs a research roadmap centered on four dimensions-"node-link-network-dynamic"-and proposes key future research topics and directions. This study aims to provide systematic knowledge and establish a foundational framework and pathway for emerging research needs, thereby advancing the field toward a data- and model-driven research paradigm.

Current fire safety protocols in nursing homes face critical challenges due to elderly residents’ physiological vulnerabilities and limited empirical evidence on caregiver decision-making during multi-floor fire emergencies. Traditional drills and conventional computational simulations often fail to capture vertical evacuation complexity, elderly resident heterogeneity, and fine-grained caregiver-resident interactions. To address these issues, this research develops a desktop-based Virtual Reality (VR) evacuation simulation model. The simulation model enables safe and repeatable interactions between real caregivers and virtual elderly residents, while automatically recording high-resolution behavioral trajectories under controlled conditions to simulate and analyze the effects of caregivers’ initial floor location, the proportion of self-evacuating elderly residents, and the behavioral patterns of high-performance caregivers on evacuation performance. Based on extensive behavioral data generated from 270 simulation trials, multidimensional performance metrics are constructed to distinguish high- and low-performance caregivers. Subsequently, multiple analytical methods are employed to analyze these simulation data and reveal the mechanisms driving performance differences. The simulation results reveal significant floor-level differences in evacuation efficiency, indicating that the second floor (2F) provides the most effective initial deployment location under the spatial configuration examined in this research. Furthermore, a potential threshold-like pattern is observed: when the proportion reaches approximately 30%, caregivers’ cognitive and physical workloads are substantially alleviated within the observed experimental range. Spatiotemporal analysis of trajectories shows that high-performance caregivers adopt proactive call mobilization and systematic, low-tortuosity search paths. The Bayesian causal mediation model further shows that technical proficiency improves evacuation outcomes primarily through indirect pathways, particularly time-cost reduction and efficient return-to-assist organization. These findings provide quantitative and data-driven support for optimizing staffing strategies and emergency response planning, as well as for incorporating behavior-informed rules into evacuation simulation models. Collectively, they highlight the value of the proposed VR-based framework for evacuation behavior analysis and supporting emergency planning in nursing homes.

Urban streets form the backbone of cities, and variations in street-junction spatial configuration provide valuable insights into urban form and function, both of which exhibit fractal and scaling characteristics in their structural and statistical properties. However, these patterns emerge from the interaction of multiple spatial processes operating at different scales, making it difficult to capture and compare their form-function relationships using single-scale analyses. To address this challenge, this study proposes a new α–β framework for multi-scale urban form comparison, grounded in fractal theory. Using a progressive clustering method, we construct scaling profiles for multiple cities, deriving the α (form) and β (function) parameters to systematically compare urban structures with differing morphologies and to identify urban clusters at their characteristic scales. This approach transforms the subjective perception of morphology into an actionable analytical tool, enabling integrated form-function assessment and overcoming the challenges of comparative studies from different urban systems. Overall, the proposed α–β framework effectively deconstructs urban complexity, as evidenced by its application to a comparative analysis of nine global metropolises.

High-fidelity 3D tree mesh models have broad applications in virtual geographic environments, forest ecology, and digital entertainment. Although laser point clouds provide precise data, existing point cloud-based methods struggle with accuracy and completeness due to tree structural complexity, leaf occlusion, and data gaps. To address these challenges, we propose a novel framework that integrates data-driven and morphology-knowledge-driven techniques for generating high-fidelity tree mesh models with accurate trunk and overall crown morphology. The proposed framework consists of three main stages. First, in the trunk separation stage, non-scattering feature clustering is employed to extract the main trunk points. Second, in the robust trunk mesh generation and optimization stage, a coarse-to-fine skeleton extraction and refinement method is proposed, followed by weighted Levenberg-Marquardt cylindrical fitting to generate the trunk geometry. Finally, we propose a novel method that integrates improved space colonization with Alpha Shape constraints to overcome challenges of reconstructing fine branches within the crown, thereby generating complete tree mesh models with accurate trunk and crown morphology. The method was evaluated on laser-scanned, 3D Gaussian, and synthetic point clouds, and benchmarked against state-of-the-art techniques. Results demonstrate superior accuracy and robustness across diverse point cloud types, with strong resilience to noise and incomplete data.

With the growing importance of urban ecosystem services and carbon neutrality, accurately estimating carbon stocks in urban trees has become essential for low-carbon urban transformation and enhanced carbon-sink capacity. However, existing studies mainly rely on tree structural parameters or remote-sensing features, with limited attention to the joint effects of biological attributes, environmental context, and spatial neighborhood relationships at the individual-tree scale. To address this gap, this study proposes a fine-grained urban tree carbon-stock estimation method that integrates biological, environmental, and spatial topological features. Biological variables include species, tree height, DBH, and canopy structure, while local habitat features incorporate vegetation activity, background productivity, land-use condition, and topographic context. Spatial proximity and bio-environment interaction topology are further modeled using a Graph Attention Network (GAT) for cross-channel feature aggregation and carbon-stock regression. Results from approximately 140,000 trees in Shenzhen Bay Park show that the Bio + Env + Spatial model achieves the best performance at a 15 m radius and 2-hop propagation (RMSE = 0.3443, MAE = 0.1927, R2 = 0.8360), improving R2 by over 0.21 relative to the biology-only baseline. SHAP results indicate that biological structural traits dominate prediction, while local habitat and spatial features provide complementary refinements by correcting estimation deviations associated with neighborhood competition, habitat heterogeneity, and spatial context.

This study presents a comparative scientometric analysis of 526 publications from the Web of Science Core Collection and China National Knowledge Infrastructure (CNKI) database to decode the global research trajectory on urban green space (UGS) carbon sinks and to delineate China’s distinctive role. Analysis reveals a clear evolution in research focus and output. Early work (before 2011) established foundational concepts. The subsequent period (2011–2015) witnessed significant methodological diversification, driven by advances in remote sensing and quantitative carbon flux assessment. Following the Paris Agreement in 2015, research emphasis shifted markedly towards policy-integrated applications and the exploration of synergies between carbon sequestration and sustainable urban development. A key finding is the rapid ascendancy of Chinese research output post-2020, a surge that has established China as a leading contributor to the field and is closely aligned with the national “dual carbon” goals. The synthesis identifies a critical methodological shortcoming: an “ecological analogy trap,” in which an over-reliance on forest-derived biomass equations systematically overlooks urban-specific anthropogenic factors. The methodological frontier is increasingly defined by the integration of multi-source remote sensing (e.g., LiDAR-optical fusion) and machine learning to overcome urban heterogeneity, though challenges in model transferability persist. The study concludes that while international scholarship increasingly emphasizes the coupling of carbon sink analysis with socio-ecological equity, China’s policy-driven research has prioritized the optimization of technological pathways for carbon accounting. To advance the field, future research must develop urban-specific protocols, leverage advanced computational techniques, and foster interdisciplinary, multi-scale frameworks to provide robust decision-support for achieving urban carbon neutrality.

High-resolution historical gridded population data are essential for quantifying long-term anthropogenic impacts. However, existing datasets often lack regional accuracy or continuity due to the scarcity of historical records and methodological limitations in spatial disaggregation. Here, we present a spatiotemporal population dataset for the Yellow River Basin spanning 1000–2000 AD at a 10-km spatial resolution. The dataset was reconstructed by harmonizing demographic statistics from The Population History of China with digitized historical administrative boundaries. We employed a hierarchical interpolation framework and an elevation-based spatial allocation model constrained by historical population quotas. The resulting dataset provides grids at centennial intervals for the pre-modern era (1000–1800 AD) and 50-year intervals for 1800–2000 AD. Technical validation indicates finer regional spatial details compared with HYDE 3.1, particularly in reflecting historical administrative shifts and reducing the overestimation of urban agglomeration in pre-industrial reconstructions. Serving as a high-resolution regional supplement to global archives, this product enhances both regional consistency and temporal granularity. It provides data support for deciphering human-environment coupling mechanisms in the Yellow River Basin—a paradigmatic region for long-term anthropogenic activity.

3D building models play a critical role in smart cities and strongly support applications in urban planning, augmented reality and urban event simulation. Urban scale city modelling with City Geography Markup Language (CityGML) LOD2 building models have been constructed in over several developed cities due to their significant role and relatively high cost. However, existing single-building reconstruction methods for LOD2 models are unsatisfactory in preserving roof details, and large-scale 3D building reconstruction still requires extensive manual editing. This paper proposes a fully automated framework for generating CityGML LOD2 building models with preferred roof details from photogrammetric point clouds from aerial oblique images, aiming to address two key challenges: (1) difficulties in LOD2 building model generation caused by missing facade photogrammetric point clouds, and (2) insufficient fidelity of building roof details. Based on the observation that buildings have typical “roof-vertical walls-ground” structures, this paper infers facade areas by height maps generated from roof point clouds. Besides, the Hypothesis-Selection-Based (HSB) polygon surface reconstruction frameworks are extended by introducing a novel voxel depth index to measure the importance of each candidate planar unit in preserving roof details. Experimental comparison with existing HSB methods and deep-learning-based methods revealed that the reconstruction of proposed methods achieves the best geometry accuracy in Root Mean Squared Error (RMSE) ranging from 0.157m to 0.660m, and also achieves the best model coverage that is between 75.14

Studying the long-term migration network within cities effectively reveals changes in urban spatial structures. This paper focuses on intra-city migration activities, introducing a method for dynamic network construction and analysis based on spatiotemporal data, offering a new perspective for examining the dynamic characteristics of residential migration within cities and their relationship to urban development. Migration behaviors are aggregated at the community level in the form of migration flows, and a time-varying network model of migration flows is constructed. Community detection and dynamic evolution analyses are conducted to uncover migration behavior characteristics. Using Shenzhen, a megacity in China, as a case study, a migration network spanning six consecutive years (2018–2023) was constructed using mobile phone signaling data. Our findings reveal that the gravitational pull of communities in the northern and eastern sub-center areas has increased. The dynamic evolution of community structures is closely related to the city's polycentric structure. Smaller network communities are more susceptible to changes, while strengthened regional connections have led to community mergers. The time-varying migration network proposed in this study provides an effective method for observing the development patterns of urban spatial structures from a mesoscopic perspective by tracking and analyzing migration flow patterns between different areas within a city.

Fire accidents in metro systems posed severe threats to passenger safety, particularly in tunnel environments characterized by confined spaces and poor visibility. However, revealing evacuation behaviors in such contexts remains challenging, primarily constrained by safety and ethical considerations, the inability to replicate realistic fire propagation, and the difficulty of capturing multimode evacuation data precisely. To address these limitations, recent advances in Intelligent Transportation Systems (ITS), and in specific Virtual Reality (VR) technologies, have made it feasible to reproduce and capture evacuation behaviors with high fidelity. Nevertheless, existing VR-based studies on metro tunnel fire evacuation remain limited, which often fail to address three critical aspects: realistic mapping of physical motion, effective measurement of dynamic decision-making processes, and the causal relationship between decision outcomes and physical motions. To overcome these limitations, this study proposes a VR-based framework that: i) establishes a synchronized multimodal data pipeline utilizing an omnidirectional treadmill and VR devices, which moves beyond basic physical replication to extract high-fidelity, spatiotemporally aligned data of evacuation dynamics; ii) captures dynamic eye-tracking data spatiotemporally synchronized with physical motion as a real-time indicator of decision-making, enabling comprehensive analysis of spatiotemporal distributions and evolution via diverse metrics; iii) employs explainable machine learning to systematically decode the interdependent relationships between smoke concentrations, crowd conditions, visual perception and physical movement outcomes, etc. Applying the proposed framework to a representative metro tunnel under varied fire emergencies yields several key findings: first, reduced visibility caused by smoke significantly delays and destabilizes evacuation movements, while the presence of NPCs can direct participants toward safer routes at the expense of increased motion disturbance; second, visual attention during evacuation dynamically adapts to environmental constraints, with smoke and spatial configurations narrowing attentional focus and intensifying cognitively driven visual scanning; third, the coupling between visual attention and movement behavior progressively shapes evacuation decisions, revealing a causal perception-behavior-choice pathway characterized by distinct evolutionary patterns of speed and density. Practically, our findings address critical empirical gaps regarding metro tunnel emergencies and provide robust evidence to support the development of effective evacuation management strategies.

This study quantifies the synergistic relationship between transport flow and economic activity, as conventional methods like road accessibility and GDP are insufficient for highly mobile megacities. Synergy measures connections in complex systems, with flow data capturing the interplay between transport and economic elements. The study proposes a framework based on flow data and synergy effect theory to analyze this relationship in network space. Specifically, using highway traffic flow data and network analysis, we constructed the synergetic effects model to analyze the interplay between highway traffic flow and economy in the Greater Bay Area (GBA). The results showed that highway traffic flow exhibited significant spatial discrepancy, and primary highway traffic flow was located in the GBA in Shenzhen-Dongguan and Guangzhou-Foshan. The higher comprehensive economic indices were located on the Guangzhou-Foshan-Macao and Guangzhou-Shenzhen-Hong Kong. The potential economic connectivity strength of Guangzhou, Foshan and Shenzhen was significantly higher than that of other cities. The synergistic effect between highway traffic flow and potential economic connectivity in the GBA was highly related to the strength of highway traffic flow, and it was significantly higher in GuangzhouDongguan-Shenzhen than in other regions. This study can support sustained economic development and the implementation of optimized transportation layouts.

Understanding the supply-demand relationship of medical services is essential for regional planning. Existing city-scale studies typically exclude cross-city flows, whereas national-scale studies often overlook intra-city heterogeneity. In urban agglomerations, healthcare resources and transport infrastructure are usually planned by cities, although patients may cross city boundaries to seek care. The implications of cross-city trips for regional medical services remain insufficiently understood. Taking the Pearl River Delta as a case, this study investigates cross-city hospital visiting trips and their implications for medical service evaluation. Using 91.2 million automobile navigation records collected in 2019, 1.37 million hospital visiting trips to Grade 3 hospitals were identified through a modified spatial join method. A population–hospital bipartite network and a multi-scale analytical framework were constructed. Cross-city demand and supply indices were developed at the city, subdistrict, and hospital scales to characterize cross-city medical service patterns and influencing factors. Accessibility and Gini coefficients were computed under intra-city and regional evaluation scenarios to assess how incorporating cross-city hospital visiting trips affects medical service evaluation. Based on automobile navigation data, 9.1

Realistic 3D scenes enhance spatial comprehension and support decision-making by creating intuitive and immersive representations of the physical world. However, achieving realism in natural scenes is challenging because of the complexity of microscale details influenced by natural disturbance events such as weathering, aging, and decay. This study introduces a cartography-informed parametric method for constructing 3D natural scenes with fine-scale details, including three components: a semantic tree, a spatial layout, and a model placement. The method simultaneously considers global entities and local details, where the generation of the latter is spatially constrained by the context of the former. Experiments are conducted across five natural scenes, and two types of user studies are carried out to evaluate visual perception. The results demonstrate the adaptability and effectiveness of the proposed method in scenes comprising various types of entities and details. Compared with scenes without details, scenes containing details exhibit superior visual realism, as evidenced by an average user rating scores of 3.51 versus 2.78 and a user selection count of 5.12 versus 0.88. This work provides a new methodological perspective on natural scene modeling and expands upon potential interdisciplinary applications that span dynamics and aesthetics.

Realistic road modeling is a core component of three-dimensional geographic information systems. Traditional methods usually emphasize geometric precision and rarely consider the effects of environmental conditions and natural aging. Thus, they generate visually clean and static road models that fail to dynamically represent the evolution of road states. To overcome this limitation, we propose a novel method for generating realistic 3D road models with multiple styles. The method abstracts a road structure as a top-down directed acyclic graph, decomposing geometry into semantically independent but functionally interrelated nodes. Coupled with a parameter propagation mechanism and a physically based decal library, the method enables the automatic generation of 3D road models that adhere to user-defined styles across temporal, environmental, and functional factors. We evaluated the method through experiments covering nine style categories and two user studies, validating its effectiveness in generating realistic road representations with diverse influencing factors. Compared with common generative AI and procedural methods, the method exhibits superior structural controllability and environmental interactivity. This work advances road modeling by introducing a parametric scheme that explicitly maps the semantic context into realistic representations, providing practical support for applications such as landscape design and traffic simulation.

In high-density crowds, close proximity between pedestrians makes the steady state highly vulnerable to disruption by pushing behaviors, potentially leading to serious accidents. However, the scarcity of experimental data on pushing behaviors has hindered systematic investigations into the underlying mechanisms and the development of accurate models. Using behavioral data from bottleneck experiments, we analyze the heterogeneity of pedestrians' internal pushing tendencies, revealing that pedestrians tend to push under high-motivation conditions and in wider corridors. In addition, we introduce a spatial discretization method to encode the state of pedestrian neighbors into feature vectors, serving together with pedestrian internal pushing tendency as the input of random forest classifiers to predict whether a pedestrian would engage in pushing behaviors. By analyzing speed-headway relationships, we reveal that pushing behaviors correspond to an aggressive space-utilization movement strategy. Consequently, we propose a hybrid machine learning and physics-based model integrating the heterogeneity of internal pushing tendencies, the random forest-based prediction of pushing behaviors, and multiple movement strategies associated with pushing and non-pushing behaviors. The proposed model is calibrated using experimental data, and parameter sensitivity analysis is conducted. Validation results demonstrate that the hybrid model effectively reproduces experimental crowd dynamics, particularly in high-motivation scenarios. Moreover, the hybrid structure of the proposed model is suitable for incorporating additional behaviors, providing a solid foundation for advancing the understanding and simulation of complex pedestrian dynamics.