Spatio-temporal Framework Research Articles

Abstract The COVID-19 pandemic has sparked intense global discussions about vaccine safety, efficacy, and distribution on social media. It underscored the need to analyze how vaccine-related sentiments propagate across social media and interact with news media articles. Despite extensive research on COVID-19 vaccines, most existing studies examine the sentiment of the COVID-19 vaccine by focusing on social media or news articles in isolation. This study bridges the gap by exploring correlations between these sources through a hierarchical spatiotemporal sentiment analysis framework that integrates social media discussions and mainstream news across global, national (US), and regional (Pennsylvania and Philadelphia) scales. Leveraging over 7 million English tweets and 6,500 news articles alongside physical events, official government records, and demographic data collected between January 2020 and June 2022, we introduce a user location inference method to approximate geographic context. Our approach leverages TriLex, a multi-lexicon sentiment method, and BERTopic to extract nuanced topics, further refined by ChatGPT for enhanced interpretability. The study period was divided into six key intervals, ranging from the beginning of the COVID-19 pandemic to the emergence of the Delta and Omicron variants. The results indicate distinct sentiment patterns in different regions and periods, partially aligning with the NYT’s vaccine-related articles. Although no causal link has been established, our findings highlight the value of correlating multi-scale social media analysis with news articles to address vaccine hesitancy, refine public health messaging, and guide future research on information diffusion in global crises.

Tropical cyclones represent one of the most destructive natural phenomena, causing massive loss of life, property damage, and disruption to coastal ecosystems [7][18]. Accurate forecasting of cyclone formation, path, and intensity remains a complex challenge due to the nonlinear and dynamic nature of atmospheric systems [9][24]. Traditional numerical weather prediction (NWP) models often require immense computational resources and are limited in their ability to capture intricate spatial–temporal dependencies [24]. To overcome these challenges, this research proposes a deep spatiotemporal learning framework that integrates Convolutional Long Short-Term Memory (ConvLSTM) networks for cyclone forecasting [2][10][19]. The ConvLSTM model combines the spatial feature extraction capabilities of convolutional layers with the temporal learning strength of LSTM units [1][6], allowing it to effectively learn from historical climate patterns [15][16]. The system utilizes multi-source meteorological data, including satellite imagery, sea surface temperature (SST), air pressure, wind velocity, and relative humidity [12][18][25]. These diverse climate variables are synchronized, normalized, and preprocessed to ensure temporal alignment and noise reduction [7][13]. Through a series of experiments, the proposed ConvLSTM framework demonstrates its ability to accurately predict cyclone tracks and intensity progression over time [19][20]. Comparative analysis with baseline models, such as CNNs and standalone LSTM architectures [4][6][11], shows significant improvement in mean squared error (MSE) and prediction stability [14][15]. The proposed model captures complex atmospheric interactions that influence cyclone evolution, enabling early warning generation with improved precision [17][21]. Furthermore, it minimizes false alarm rates and enhances spatial resolution in cyclone path estimation [9][22]. This research also emphasizes model interpretability by visualizing learned spatial features and temporal transitions within the ConvLSTM layers [3][5][10]. Such visual insights contribute to a better understanding of the underlying meteorological processes [23]. The overall system provides an efficient, datadriven approach for short-term and medium-range cyclone forecasting [8][24]. Integration with real-time climate data sources can enable continuous updates and adaptive learning [20][25]. Results from this study suggest that deep spatiotemporal learning, supported by diverse climate inputs, can complement existing meteorological forecasting systems [9][11]. The proposed ConvLSTM framework offers a scalable, accurate, and cost-effective solution that strengthens disaster preparedness and management strategies [18][19]. In conclusion, this work highlights the potential of hybrid deep learning architectures in enhancing the reliability of climate prediction models [10][17], marking a step forward toward intelligent and automated weather forecasting solutions [15][25].

The rapid expansion of mobile user behavior data has made next-Point-of-Interest (POI) recommendation increasingly vital for enhancing personalized location-based services. However, the non-uniform spatio-temporal distribution of user behavior poses significant challenges to recommendation performance. Most existing methods neglect this fundamental issue at the distribution level, while conventional data augmentation strategies fall short in optimizing spatio-temporal distribution properties. To tackle this problem, we propose a spatio-temporal Distribution Calibration framework for next-POI Recommendation (DCal-Rec), which optimizes behavioral sequence distributions through disentangled spatial and temporal operator pools. This is combined with a dual-constraint mechanism that incorporates both distribution and interest information to maintain semantic consistency. Furthermore, a multi-channel contrastive learning paradigm is introduced to jointly optimize the recommendation and contrastive tasks under a unified training objective, thereby improving the model’s generalization capability. Experimental results on three public real-world datasets demonstrate that DCal-Rec significantly outperforms baseline methods across various evaluation metrics.

A proactive traffic safety approach provides a forward-looking method for managing traffic and preventing accidents by identifying high-risk conditions before they occur. Previous studies have often focused on historical crash data or demographic factors, relying on limited single-source inputs and neglecting spatial, temporal, and environmental interactions. This study develops a multimodal spatiotemporal deep fusion framework for predicting traffic accidents in Toronto, Canada, by integrating spatial, temporal, environmental, and lighting features within a proactive modeling structure. Three fusion approaches were investigated: (1) environmental feature fusion, (2) extended fusion incorporating lighting and road surface conditions, and (3) a double-stage fusion combining all feature types. The double-stage fusion achieved the best performance, reducing RMSE from 0.50 to 0.41 and outperforming conventional models across multiple error metrics. The framework supports fine-grained hotspot analysis, improves proactive traffic safety management, and provides a transferable roadmap for applying deep fusion in real-world intelligent transportation and urban planning systems.

Solar-induced chlorophyll fluorescence (SIF) is a direct quantitative indicator of photosynthesis. High-precision modeling and cross-scale prediction of SIF are crucial for deciphering multiscale ecological response mechanisms and assessing regional carbon sinks. However, existing SIF prediction models rely excessively on the simple stacking of spectral and meteorological variables, neglecting the process-based response mechanisms to stress factors. This limitation results in difficulties effectively capturing the nonlinear interactions among environmental factors and their compound effects on SIF within areas of high landscape heterogeneity. To address these limitations, this study taking the grasslands of Northern China as a case study, developed a multi-source data-driven spatiotemporal prediction framework for SIF. Through a synergistic multistep mechanism involving feature selection, spatiotemporal decoupling, hybrid modeling, and surface extension, this framework couples geographical spatiotemporal heterogeneity with ecological process-driven mechanisms. It achieved the high-precision simulation and dynamic prediction of SIF distribution across multiscale grassland areas. The results demonstrated that the hybrid LSTM + Transformer model significantly outperformed traditional architectures. Its gated memory units and self-attention modules performed synergistically to effectively capture cross-climatic zone photosynthetic synergy and the fluorescence suppression effect under low-temperature stress. Further analysis revealed that the fraction of absorbed photosynthetically active radiation (FPAR), the leaf area index (LAI), and surface solar radiation (SSR) predominantly govern SIF variations by regulating canopy light energy allocation and phenological rhythms. Future projections indicate that grasslands in Northern China will exhibit a trend of slow growth but enhanced spatial heterogeneity. This research not only provides critical methodological support for carbon sink function assessment and degradation restoration monitoring in Northern China’s grasslands and similar large-scale regions but also offers a novel perspective for advancing ecosystem process modeling under global change.

High-voltage ceramic insulators are routinely exposed to short-duration overvoltages such as lightning impulses, switching surges, and partial discharges. These events occur on microsecond to millisecond timescales and can produce highly localized thermal spikes that are difficult to measure directly but may compromise long-term material integrity. This paper addresses the estimation of the internal temperature distribution immediately after a lightning impulse by solving a three-dimensional inverse heat conduction problem (IHCP). The forward problem is modeled by the transient heat diffusion equation with constant thermal diffusivity, discretized using the finite element method (FEM). Surface temperature measurements are assumed available from a 12 kV ceramic post insulator and are used to reconstruct the unknown initial condition. To address the ill-posedness of the IHCP, a spatio-temporal regularization framework is introduced and compared against spatial-only regularization. Numerical experiments investigate the effect of measurement time (T=60 s, 600 s, and 1800 s), mesh resolution (element sizes of 20 mm, 15 mm, and 10 mm), and measurement noise (σ=1 K and 5 K). The results show that spatio-temporal regularization significantly improves reconstruction accuracy and robustness to noise, particularly when early-time measurements are available. Moreover, it is observed that mesh refinement enhances accuracy but yields diminishing returns when measurements are delayed. These findings demonstrate the potential of spatio-temporal IHCP methods as a diagnostic tool for the condition monitoring of ceramic insulators subjected to transient electrical stresses.

In image-guided radiotherapy (IGRT), four-dimensional cone-beam computed tomography (4D-CBCT) is critical for assessing tumor motion during a patient's breathing cycle prior to beam delivery. However, generating 4D-CBCT images with sufficient quality requires significantly more projection images than a standard 3D-CBCT scan, leading to extended scanning times and increased imaging dose to the patient. To introduce a novel spatiotemporal Gaussian neural representation framework to reconstruct high-temporal dynamic CBCT images from 1-minute acquisition, preserving motion dynamics and fine spatial details without relying on prior images or motion models. Our framework employs a differentiable 4D Gaussian representation initialized from average CBCT images. Gaussian points are characterized by position, covariance, rotation, and density, offering a compact and dynamic model for CBCT scenes. A Gaussian deformation network, incorporating a HexPlane encoder and multi-head decoder, predicts Gaussian deformations to minimize L1 and structural similarity index measure (SSIM) losses between rendered and measured projections. Adaptive Gaussian control refines the representation by pruning underutilized Gaussians and densifying points in high-gradient regions. The method was benchmarked on the AAPM SPARE challenge datasets and further validated with clinical CBCT scans from a Varian TrueBeam system. For the AAPM SPARE challenge datasets, the performance of the proposed method was evaluated using the root-mean-squared-error (RMSE) and the structural similarity index (SSIM) in the four regions of interest: Body, Lung, PTV, and Bone. The geometric accuracy was evaluated by calculating the registration error when aligning the tumor to the ground truth using the Elastix package, focusing on pixels within the planning target volume (PTV). To demonstrate our method's capability in high-temporal motion dynamic modeling using extremely undersampled projections, the clinical half-fan projections from a 1-minute Varian TrueBeam acquisition were sorted into 50 phases with approximately 18 projections per phase, significantly finer than the commonly used 10-phase binning. Compared to the AAPM SPARE challenge participant methods, our method achieved superior geometric accuracy in terms of PTV alignment error, and comparable RMSE and SSIM when no prior 4DCT or motion model is used for our reconstruction. For PTV alignment, our method achieved translational and rotational errors of 0.54mm (LR), 0.76mm (SI), 1.36mm (AP), 0.55° (rAP), and 0.93° (rSI), and 1.31° (rLR), respectively. For high temporal dynamic CBCT reconstruction, our method successfully reconstructed a 50-phase CBCT from a 1-minute Varian Truebeam half-fan scan, demonstrating effective streak artifact suppression, respiratory motion preservation, and fine detail restoration. Reconstruction on a single NVIDIA RTX A6000 GPU required approximately 30-80 min, depending on the number of Gaussian points used (ranging from 50 to 400K), to reconstruct CBCT from 680 projections acquired with a 30 × 40cm detector. Our code and reconstruction results can be found at: https://github.com/fuyabo/4DGS_for_4DCBCT/tree/main. The spatiotemporal Gaussian framework is a novel data-driven dynamic CBCT reconstruction technique that features excellent geometric accuracy in terms of PTV alignment and high-temporal motion modeling, indicating promise for tumor motion assessment and high-temporal respiratory motion modeling based on a 1-minute half-fan scan prior to beam delivery.

Real-time reconstruction of fragmented trajectories: An integrated machine learning and behavior-based spatiotemporal framework

Objective.Functional magnetic resonance imaging (fMRI) is crucial for identifying neurological disorder biomarkers, but current deep learning methods face some limitations. Template-dependent methods reliant on fixed brain atlases lack inter-subject specificity and generalizability due to fixed anatomical priors. Emerging template-free models, which process raw data directly, often separate spatial and temporal processing. This approach discards temporal continuity, which encompasses key characteristics such as the smooth and correlated nature of neural dynamics over time. To address these limitations, we propose a novel axial slice-centric model that jointly models spatiotemporal representations through end-to-end processing of native 4D fMRI data. This eliminates template dependency while preserving intrinsic brain activity patterns.Approach.Our framework redefines 4D fMRI analysis by decomposing it into 3D spatiotemporal manifolds along the axial axis, enabling joint learning of spatial and temporal features and preserving individualized structure organization. A hierarchical encoder extracts local spatiotemporal interactions within each slice, progressively aggregating information to capture multi-granularity neural patterns. To maintain temporal continuity and computational efficiency, a differentiable TopK operation adaptively selects informative slices and time points, balancing computational demands with long-range temporal dependencies.Main results.Experimental results on the ADNI dataset (324 subjects, for classifying early mild cognitive impairment and normal controls )) and a private disorder of consciousness dataset (164 subjects) demonstrate the effectiveness of our 4D fMRI framework in classifying both neurodegenerative and consciousness disorders. Specifically, on the ADNI dataset, our proposed model achieves 97% classification accuracy with over 25% reduction in floating-point operations compared to baseline methods. On the private dataset, our model outperforms state-of-the-art approaches by 5% accuracy. Visualization of slice-level attention maps identify biomarkers consistent with previous research, demonstrating that our template-free framework can discover biomarkers comparable to those identified by template-dependent methods.Significance.Our joint spatiotemporal modeling framework, enabled by axial slice-centric decomposition of 4D fMRI data while preserving temporal continuity, achieves excellent complexity-accuracy trade-offs for brain disorder analysis. Biomarker visualization confirms its template-free capability to identify clinically-relevant neural patterns, offering an efficient and interpretable solution for 4D fMRI-based diagnosis.

Inverse Synthetic Aperture Radar (ISAR) has recently advanced to volumetric 3D-ISAR imaging, creating new opportunities and challenges for automatic target recognition (ATR). This work proposes a spatiotemporal deep learning framework that jointly learns target structure and motion dynamics from high-resolution 3D-ISAR sequences. A CNN backbone (ResNet) extracts per-frame spatial features, which are fed to temporal models Bidirectional LSTM and/or ConvLSTM to capture micro-Doppler cues and aspect-dependent scattering over time; the pipeline is supported by physics-aware formation and backprojection-style 3D reconstruction. We evaluate on a four-class dataset (aircraft, helicopter, drone, tank) comprising 400 labeled samples drawn from MSTAR and simulated 3D-ISAR sequences, with standard train/validation/test partitions and targeted denoising, normalization, and augmentation to enhance robustness. The proposed model achieves strong performance across metrics: an overall accuracy of 95% on the final evaluation set with near-ideal class separability (AUC ≈ 0.98–1.00), and a best accuracy of 96.7% when all preprocessing and geometric/data-level augmentations are enabled. Ablation and robustness studies show consistent gains from motion-aware temporal modeling and the preprocessing stack under low-SNR and distortion conditions, while confusion is largely confined to visually and dynamically similar aerial classes. These results demonstrate that coupling modern spatiotemporal architectures with principled ISAR signal processing yields reliable, accurate, and deployment-oriented ATR for 3D-ISAR systems.

As urban carbon neutrality initiatives accelerate, green spaces in cities are playing an increasingly critical role as natural carbon sinks in mitigating greenhouse gas emissions. However, conventional carbon estimation approaches struggle with spatial fragmentation and temporal variability in urban green areas, resulting in limited accuracy and poor adaptability. To address this challenge, this study proposes a deep spatiotemporal modeling framework combining Convolutional Neural Networks (CNN) and Temporal Convolutional Networks (TCN), integrating multi-source remote sensing data from Landsat-8, Sentinel-2, and MODIS to estimate carbon storage in Guangzhous green spaces from 2018 to 2023. Experimental results demonstrate that the model achieves robust performance across diverse land types and seasonal conditions, with an overall RMSE of 2.71 tC/ha, R of 0.926, and SSIM of 0.841, significantly outperforming traditional statistical and machine learning methods. The study confirms the effectiveness of deep fusion modeling in urban carbon sink estimation and offers a scalable technical pathway to support carbon asset management, green space planning, and low-carbon policy development in complex urban contexts.

A consistent deposition trend across reservoirs was identified using parametric and nonparametric methods at the 5% significance level, corroborated by elevated Hurst coefficients reflecting long-term temporal persistence. These insights informed the formulation of a Linear Regression Trend Model (LRTM) within a spatio-temporal framework for sediment distribution prediction. The LRTM exhibited strong goodness-of-fit across reservoirs—mimicking natural deposition behaviour—with Nash–Sutcliffe efficiency (0.99–1.00), standard error of estimate (1.53–19.20), and relative error (−0.35% to 3.52%), outperforming benchmark methods: Empirical Area Reduction and Area Increment approaches. To address limitations in sedimentation modelling under parametric uncertainty, a dual-layered framework was developed integrating regression diagnostics, perturbation-based sensitivity analysis, and elevation-specific coherence metrics. Applied across reservoirs with varying sensitivity regimes, it revealed an inverse relationship between persistence and diagnostic weight. This coherence-sensitive extension enhances regime classification and forecasting precision, offering a scalable, empirically defensible tool for long-term sediment prediction across diverse reservoir contexts.

ABSTRACT The Southern Silk Road is an important carrier for trade, ethnic exchange and integration, cultural interaction, and political communication, connecting the southwest frontier with the Central Plains, as well as China with foreign regions. Reconstructing its historical route network is of great significance for understanding the history of China's border governance and transportation development. Based on literature review, field surveys, and modern information technologies (including GPS, GIS, and AI), this study addressed three key issues in historical route reconstruction: defining the spatiotemporal start-stop framework, geographically locating traffic nodes, and extracting and reconstructing linear features. Adopting a ‘point-line-network' method, the study reconstructs the traffic route network of the Sichuan section during the Han, Tang, and Qing dynasties. This enables the visualization of the complex evolutionary pattern of the Southern Silk Road's traffic network in Sichuan, which transformed from a ‘two vertical and one horizontal' structure to a ‘three vertical and three horizontal’ framework, with land and water routes operating in parallel.

This study proposes and empirically validates a spatiotemporal life cycle assessment (LCA) framework for hydroelectric power generation applied to the Sinop Hydroelectric Power Plant in Brazil. Unlike conventional LCA, which assumes spatial and temporal homogeneity, the framework incorporates annual temporal discretisation and geographically differentiated impacts across all phases of assessment. The methodology combines the Enhanced Structural Path Analysis (ESPA) method with temporal modeling and region-specific inventory data. The results indicate that environmental impacts peak in the fourth year of the ‘Construction and Assembly’ stage, primarily due to the intensive production of concrete and steel. A spatial analysis shows that these impacts extend beyond Brazil, with notable contributions from international supply chains. By identifying temporal and geographical hotspots, the framework offers a refined understanding of impact dynamics and drivers. Uncertainty analysis further demonstrates that temporal discretisation significantly affects impact attribution, with the ‘Construction and Assembly’ stage results varying by up to ±15%, depending on scheduling assumptions. Overall, the study advances the LCA methodology while offering robust empirical evidence to guide sustainable decision-making in Brazil’s power sector and to inform global debates on low-carbon energy transitions.

Melatonin, a hormone secreted by the pineal gland, regulates circadian rhythms and exhibits strong anticancer potential through its antioxidant, immunomodulatory, and hormonal effects. This study develops a novel reaction–diffusion mathematical model to describe multiscale interactions among melatonin, breast cancer cells, and immune responses, emphasizing blind women who sustain high melatonin levels due to a lack of light perception. The model uniquely integrates hormonal, oxidative, and immune processes within a unified spatio-temporal framework, enabling joint analysis of tumor proliferation, melatonin-induced inhibition, immune cytotoxicity, and fibroblast-mediated tumor stimulation. Epidemiological evidence indicates that blind women show a markedly lower incidence of hormone-dependent cancers, attributed to continuous melatonin activity. Numerical simulations demonstrate that elevated melatonin concentrations suppress tumor growth, strengthen immune activity, and reduce fibroblast-driven promotion. Moreover, stress-related melatonin depletion is shown to disrupt tumor–immune balance, supporting the hypothesis that circadian rhythm disturbance accelerates tumor progression. The findings offer mechanistic insight into melatonin’s dual preventive and therapeutic roles and establish a quantitative link between biochemical regulation and tumor dynamics. By coupling biological data with mathematical rigor, the proposed framework advances mathematical oncology by uniting circadian biology, immunology, and tumor modeling within a single analytical structure.

Human motion exhibits high-dimensional and stochastic characteristics, posing significant challenges for modeling and prediction. Existing approaches typically employ coupled spatiotemporal frameworks to generate future poses. However, the intrinsic nonlinearity of joint interactions over time, compounded by high-dimensional noise, often obscures meaningful motion features. Notably, while adjacent joints demonstrate strong spatial correlations, their temporal trajectories frequently remain independent, adding further complexity to modeling efforts. To address these issues, we propose a novel framework for human motion prediction via the decoupled spatiotemporal clue (DSC), which explicitly disentangles and models spatial and temporal dependencies. Specifically, DSC comprises two core components: (i) a spatiotemporal decoupling module that dynamically identifies critical joints and their hierarchical relationships using graph attention combined with separable convolutions for efficient motion decomposition; and (ii) a pose generation module that integrates local motion denoising with global dynamics modeling through a spatiotemporal transformer that independently processes spatial and temporal correlations. Experiments on the widely used human motion datasets H3.6M and AMASS demonstrate the superiority of DSC, which achieves 13% average improvement in long-term prediction over state-of-the-art methods.

In-depth exploration of the spatial heterogeneity patterns of urban carbon emissions holds significant scientific importance for regional sustainable development. However, few scholars have examined the spatiotemporal characteristics of county-level carbon emissions in Inner Mongolia. This study focuses on the three major cities of Hohhot, Baotou, and Ordos in Inner Mongolia. By integrating NPP-VIIRS nighttime light data, the CLCD (China Land Cover Dataset) dataset, and statistical yearbooks, it quantifies county-level carbon emissions and establishes a spatiotemporal analysis framework of urban morphology–carbon emissions from 2013 to 2021. Six morphological indicators—Class Area (CA), Landscape Shape Index (LSI), Largest Patch Index (LPI), Patch Cohesion Index (COHESION), Patch Density (PD), and Interspersion Juxtaposition Index (IJI)—are employed to represent urban scale, complexity, centrality, compactness, fragmentation, and adjacency, respectively, and their impacts on regional carbon emissions are examined. Using a geographically and temporally weighted regression (GTWR) model, the results indicate the following: (1) from 2013 to 2021, The high-value areas of carbon emissions in the three cities show a clustered distribution centered on the urban districts. The total carbon emissions increased from 20,670 (104 t/CO2) to 37,788 (104 t/CO2). The overall spatial pattern exhibits a north-to-south increasing gradient, and most areas are projected to experience accelerated carbon emission growth in the future; (2) the global Moran’s I values were all greater than zero and passed the significance tests, indicating that carbon emissions exhibit clustering characteristics; (3) the GTWR analysis revealed significant spatiotemporal heterogeneity in influencing factors, with different cities exhibiting varying directions and strengths of influence at different development stages. The ranking of influencing factors by degree of impact is: CA > LSI > COHESION > LPI > IJI > PD. This study explores urban carbon emissions and their heterogeneity from both temporal and spatial dimensions, providing a novel, more detailed regional perspective for urban carbon emission analysis. The findings enrich research on carbon emissions in Inner Mongolia and offer theoretical support for regional carbon reduction strategies.

A physically informed domain-independent data-driven inundation forecast model.

PM2.5 pollution events evolve continuously through spatiotemporal diffusion. However, their three-dimensional spatiotemporal variation characteristics are often overlooked, and the interactions among key characteristics (e.g., duration, maximum concentration) have not yet been systematically analyzed. This study established a three-dimensional (longitude, latitude, and time) spatiotemporal framework for identifying contiguous PM2.5 pollution events based on the high-resolution ChinaHighAirPollutants (CHAP) dataset (1 km spatial and 1-day temporal resolution). The framework applied the meteorological event tracking algorithm (i.e., the Forward-in-Time method) to track PM2.5 pollution events. Based on this framework, we systematically tracked and characterized the spatiotemporal evolution of PM2.5 events across China from 2013 to 2021, quantified the relationships among key event characteristics, and tracked their transport pathways. The results show that: (1) The combination of the FiT algorithm and CHAP dataset enables effective tracking and identification of the three-dimensional spatiotemporal evolution of PM2.5 pollution events across China. (2) Event PM2.5 totals, average totals per event and pollution events exhibit a distinct right-inclined “T”-shaped pattern, with hotspots located in Xinjiang, the Beijing-Tianjin-Hebei (BTH) region, Shandong, and Henan, where annual event frequency exceeds 15. (3) Event PM2.5 totals show strong correlations with average duration per event and average maximum concentration per event, particularly in heavily polluted areas where the Pearson correlation coefficient is close to 1. (4) PM2.5 pollution events are mainly characterized by short durations of 1 day or 2–3 days, accounting for over 80% of occurrences. Long-duration events are mostly concentrated in areas with severe pollution problems, and their persistence is closely linked to spatial coverage, terrain barrier effects, and meteorological conditions. (5) PM2.5 pollution events consistently exhibit a west-to-east transport pattern. Short-duration events propagate slower across the inland northwest, whereas long-duration events show a pronounced increase in meridional transport speeds along the eastern coastal areas. This study elucidates the continuous spatiotemporal evolution and intrinsic drivers of PM2.5 pollution events, offering scientific insights to support air quality improvement and the development of targeted management strategies.

Syndemics involving Human immunodeficiency virus (HIV) and other sexually transmitted infections (STIs) remain a major public health challenge in sub-Saharan Africa, and understanding their spatial and temporal dynamics is critical for effective interventions. Using data from two consecutive, population-based cross-sectional surveys conducted in 2014 and 2015 under the HIV Incidence Provincial Surveillance System (HIPSS) in KwaZulu-Natal, South Africa, we applied a Bayesian spatio-temporal framework grounded in latent variable modeling to quantify and map the syndemic burden of HIV and other STIs. A confirmatory factor analysis constructed a continuous latent syndemic score from four binary indicators (HIV diagnosis, HIV testing, STI diagnosis, and STI symptoms), which was modeled using Bayesian hierarchical spatial methods via Integrated Nested Laplace Approximation (INLA), incorporating spatial random effects through the Stochastic Partial Differential Equation (SPDE) approach and temporal effects via a first-order random walk. Local spatial autocorrelation, assessed using Local Moran's I and Getis-Ord Gi* statistics, revealed consistent hotspots and coldspots. Syndemic burden of HIV and other STIs was higher among younger adults (20–49 years), women, individuals with incomplete secondary education, those engaging in sexual risk behaviors or reporting forced sexual debut, and those facing socioeconomic vulnerabilities such as food insecurity. Access to healthcare and treatment for depression were also positively associated, likely reflecting increased detection. Local Moran's I identified 11 significant clusters (three hotspots, eight coldspots), and Getis-Ord Gi* identified 32 (17 hotspots, 15 coldspots), with hotspot patterns persisting across both years, indicating temporal stability. These findings highlight the utility of Bayesian latent variable and spatio-temporal modeling in integrating multiple co-occurring health conditions into a single spatial framework, providing actionable evidence to support geographically targeted, multi-sectoral interventions that address structural and behavioral drivers of co-epidemics in resource-limited settings.