Search Results (407)

Despite the promise of graph neural networks (GNNs) in hydrological forecasting, existing approaches face critical limitations in capturing dynamic spatiotemporal correlations and integrating physical interpretability. To bridge this gap, we propose a spatial-temporal graph neural network (ST-GNN) that addresses these challenges through three key innovations: dynamic graph construction for adaptive spatial correlation learning, a physically-informed feature enhancement layer for soil moisture and evaporation integration, and a hybrid Graph-LSTM module for synergistic spatiotemporal dependency modeling. The temporal and spatial modules of the spatio-temporal graph neural network exhibit a structural symmetry, which enhances the model’s representational capability. By integrating these components, the model effectively represents rainfall-runoff processes. Experimental results across four Chinese watersheds demonstrate ST-GNN’s superior performance, particularly in semi-arid regions where prediction accuracy shows significant improvement. Compared to the best-performing baseline model (ST-GCN), our ST-GNN achieved an average reduction in root mean square error (RMSE) of 6.5% and an average improvement in the coefficient of determination (R²) of 1.8% across 1–8 h forecast lead times. Notably, in the semi-arid Pingyao watershed, the improvements reached 13.3% in RMSE reduction and 2.5% in R² enhancement. The model incorporates watershed physical characteristics through a feature fusion layer while employing an adaptive mechanism to capture spatiotemporal dependencies, enabling robust watershed-scale forecasting across diverse hydrological conditions. Full article

Polar sea ice is undergoing rapid change, with recent record-low extents in both hemispheres, raising the demand for skillful predictions from days to seasons for navigation, ecosystem management, and climate risk assessment. Accurate sea ice prediction is essential for understanding coupled climate processes, supporting safe polar operations, and informing adaptation strategies. Physics-based numerical models remain the backbone of operational forecasting, but their skill is limited by uncertainties in coupled ocean–ice–atmosphere processes, parameterizations, and sparse observations, especially in the marginal ice zone and during melt seasons. Statistical and empirical models can provide useful baselines for low-dimensional indices or short lead times, yet they often struggle to represent high-dimensional, nonlinear interactions and regime shifts. This review synthesizes recent progress of DL for key sea ice prediction targets, including sea ice concentration/extent, thickness, and motion, and organizes methods into (i) sequential architectures (e.g., LSTM/GRU and temporal Transformers) for temporal dependencies, (ii) image-to-image and vision models (e.g., CNN/U-Net, vision Transformers, and diffusion or GAN-based generators) for spatial structures and downscaling, and (iii) spatiotemporal fusion frameworks that jointly model space–time dynamics. We further summarize hybrid strategies that integrate DL with numerical models through post-processing, emulation, and data assimilation, as well as physics-informed learning that embeds conservation laws or dynamical constraints. Despite rapid advances, challenges remain in generalization under non-stationary climate conditions, dataset shift, and physical consistency (e.g., mass/energy conservation), interpretability, and fair evaluation across regions and lead times. We conclude with practical recommendations for future research, including standardized benchmarks, uncertainty-aware probabilistic forecasting, physics-guided training and neural operators for long-range dynamics, and foundation models that leverage self-supervised pretraining on large-scale Earth observation archives. Full article

Traffic systems are quintessential complex systems, characterized by nonlinear interactions, multiscale dynamics, and emergent spatiotemporal patterns over complex networks. These properties make traffic prediction highly challenging, as it requires jointly modeling stable global topology and time-varying local dependencies. Existing graph neural networks often rely on predefined or static learnable graphs, overlooking hidden dynamic structures, while most RNN- or CNN-based approaches struggle with long-range temporal dependencies. This paper proposes a Spatiotemporal Causal–Trend Network (SCTN) tailored to complex transportation networks. First, we introduce a dual-path adaptive graph learning scheme: a static graph that captures global, topology-aligned dependencies of the complex network, and a dynamic graph that adapts to localized, time-varying interactions. Second, we design a Gated Temporal Attention Module (GTAM) with a causal–trend attention mechanism that integrates 1D and causal convolutions to reinforce temporal causality and local trend awareness while maintaining long-range attention. Extensive experiments on two real-world PeMS traffic flow datasets demonstrate that SCTN consistently achieves superior accuracy compared to strong baselines, reducing by 3.5–4.5% over the best-performing existing methods, highlighting its effectiveness for modeling the intrinsic complexity of urban traffic systems. Full article

Aiming at the challenges of complex spatial-temporal correlation and strong nonlinearity in the power prediction of large-scale wind farm clusters, this study proposes a short-term wind power prediction method that combines a dynamic graph structure and a Kolmogorov–Arnold Network (KAN) enhanced neural network. Firstly, a spectral embedding fuzzy C-means (FCM) cluster partition method combining geographic location and numerical weather prediction (NWP) is proposed to solve the problem of insufficient spatio-temporal representation ability of traditional methods. Secondly, a dynamic directed graph construction mechanism based on a stacked wind direction matrix and wind speed mutual information is designed to describe the directional correlation between stations with the evolution of meteorological conditions. Finally, a prediction model of dynamic graph convolution and Transformer based on KAN enhancement (DGTK-Net) is constructed to improve the fitting ability of complex nonlinear relationships. Based on the cluster data of 31 wind farms in Gansu Province of China and the cluster data of 70 wind farms in Inner Mongolia, a case study is carried out. The results show that the proposed model is significantly better than the comparison methods in terms of key evaluation indicators, and the root mean square error is reduced by about 1.16% on average. This method provides a prediction tool that can adapt to time and space changes for engineering practice, which is helpful to improve the wind power consumption capacity and operation economy of the power grid. Full article

In GPS-denied environments, the spatio-temporal coupling of errors caused by dynamic network topologies poses a fundamental challenge to cooperative localization, presenting existing methods with a dilemma: approaches pursuing global optimization lack dynamic adaptability, while those focusing on local adaptation struggle to guarantee global convergence. To address this challenge, this paper proposes ST-DCL, a cooperative localization framework based on a novel principle of closed-loop spatio-temporal decoupling. The core of ST-DCL comprises two modules: a Dynamic Weighted Multidimensional Scaling (DW-MDS) optimizer, responsible for providing a globally consistent coarse estimate with provable convergence, and a specially designed Spatio-Temporal Graph Neural Network (ST-GNN) corrector, tasked with compensating for local nonlinear errors. The DW-MDS effectively suppresses interference from historical errors via an adaptive sliding window and confidence weights derived from our error propagation model. The key innovation of the ST-GNN lies in its two newly designed components: a Dynamic Topological Attention Module for actively modulating neighbor aggregation to inhibit spatial error diffusion, and a Dilated Causal Convolution Module for modeling long-term temporal dependencies to curb error accumulation. These two modules form a closed loop via a confidence feedback mechanism, working in synergy to achieve continuous error suppression. Theoretical analysis indicates that the framework exhibits bounded-error convergence under dynamic topologies. In simulations involving 200 nodes, velocities up to 50 m/s, and 15% NLOS links, the ST-DCL achieves a normalized root mean square error (NRMSE) of 0.0068, representing a 21% performance improvement over state-of-the-art methods. The practical efficacy and real-time capability are further validated through real-world flight experiments with a 10-UAV swarm in complex, GPS-denied scenarios. Full article

Accurate estimation of soil moisture (SM) and vegetation optical depth (VOD) is essential for understanding land–atmosphere interactions, climate dynamics, and ecosystem processes. While passive microwave missions such as SMAP and SMOS provide reliable global SM and VOD products, they are limited by coarse spatial resolution and infrequent revisit times. Global Navigation Satellite System Reflectometry (GNSS-R) observations, particularly from the Cyclone GNSS (CYGNSS) mission, offer an improved spatiotemporal sampling rate. This study presents a deep learning framework based on an artificial neural network (ANN) for the simultaneous retrieval of SM and VOD from CYGNSS observations across the contiguous United States (CONUS). Ancillary input features, including specular point latitude and longitude (for spatial context), CYGNSS reflectivity and incidence angle (for surface signal characterization), total precipitation and soil temperature (for hydrological context), and soil clay content and surface roughness (for soil properties), are used to improve the estimates. Results demonstrate strong agreement between the predicted and reference values (SMAP SM and SMOS VOD), achieving correlation coefficients of R = 0.83 and 0.89 and RMSE values of 0.063 m³/m³ and 0.088 for SM and VOD, respectively. Temporal analyses show that the ANN accurately reproduces both seasonal and daily variations in SMAP SM and SMOS VOD (R ≈ 0.89). Moreover, the predicted SM and VOD maps show strong agreement with the reference SM and VOD maps (R ≈ 0.93). Additionally, ANN-derived VOD demonstrates strong consistency with above-ground biomass (R ≈ 0.77), canopy height (R ≈ 0.95), leaf area index (R = 96), and vegetation water content (R ≈ 0.90). These results demonstrate the generalizability of the approach and its applicability to broader environmental sensing tasks. Full article

Industrial aluminum-block heating processes exhibit nonlinear dynamics, substantial time delays, and stringent requirements for fault detection and diagnosis, especially in semiconductor manufacturing and other high-precision electronic processes, where slight temperature deviations can accelerate device degradation or even cause catastrophic failures. To address these challenges, this study presents a digital twin-based intelligent heating platform for aluminum blocks with a dual-artificial-intelligence framework (dual-AI) for control and diagnosis, which is applicable to multi-port aluminum-block heating systems. The system enables real-time observation and simulation of high-temperature operational conditions via virtual-real interaction. The platform precisely regulates a nonlinear temperature control system with a prolonged time delay by integrating a conventional proportional–integral–derivative (PID) controller with a Levenberg–Marquardt-optimized backpropagation (LM-optimized BP) neural network. Simultaneously, a relay is employed to sever the connection to the heater, thereby simulating an open-circuit fault. Throughout this procedure, sensor data are gathered simultaneously, facilitating the creation of a spatiotemporal time-series dataset under both normal and fault conditions. A one-dimensional convolutional neural network (1D-CNN) is trained to attain high-accuracy fault detection and localization. PID+LM-BP achieves a response time of about 200 s in simulation. In the 100 °C to 105 °C step experiment, it reaches a settling time of 6 min with a 3 °C overshoot. Fault detection uses a 0.38 °C threshold defined based on the absolute minute-to-minute change of the 1-min mean temperature. Full article

Accurate traffic flow prediction is a core component of intelligent transportation systems, supporting proactive traffic management, resource optimization, and sustainable urban mobility. However, urban traffic networks exhibit heterogeneous multi-scale periodic patterns and time-varying spatial interactions among road segments, which are not sufficiently captured by many existing spatio-temporal forecasting models. To address this limitation, this paper proposes PDR-STGCN (Periodicity-Aware Dynamic Relational Spatio-Temporal Graph Convolutional Network), an enhanced STGCN framework that jointly models multi-scale periodicity and dynamically evolving spatial dependencies for traffic flow prediction. Specifically, a periodicity-aware embedding module is designed to capture heterogeneous temporal cycles (e.g., daily and weekly patterns) and emphasize dominant social rhythms in traffic systems. In addition, a dynamic relational graph construction module adaptively learns time-varying spatial interactions among road nodes, enabling the model to reflect evolving traffic states. Spatio-temporal feature fusion and prediction are achieved through an attention-based Bidirectional Long Short-Term Memory (BiLSTM) network integrated with graph convolution operations. Extensive experiments are conducted on three datasets, including Metro Traffic Los Angeles (METR-LA), Performance Measurement System Bay Area (PEMS-BAY), and a real-world traffic dataset from Guizhou, China. Experimental results demonstrate that PDR-STGCN consistently outperforms state-of-the-art baseline models. For next-hour traffic forecasting, the proposed model achieves average reductions of 16.50% in RMSE, 9.00% in MAE, and 0.34% in MAPE compared with the second-best baseline. Beyond improved prediction accuracy, PDR-STGCN reveals latent spatio-temporal evolution patterns and dynamic interaction mechanisms, providing interpretable insights for traffic system analysis, simulation, and AI-driven decision-making in urban transportation networks. Full article

Forecasting the spatio-temporal evolutions of cellular traffic is crucial for urban management. However, achieving accurate forecasting is challenging due to “complex correlation modeling” and “model-blindness” issues. Specifically, cellular traffic is generated within complex urban systems characterized by an intricate structure and human mobility. Existing approaches, often based on proximity or attributes, struggle to learn the latent correlation matrix governing traffic evolution, which limits forecasting accuracy. Furthermore, while substantial knowledge about urban systems can supplement the modeling of correlations, existing methods for integrating this knowledge—typically via loss functions or embeddings—overlook the synergistic collaboration between data and knowledge, resulting in weak model robustness. To address these challenges, we develop a data–knowledge collaborative learning framework termed the knowledge-empowered spatio-temporal neural network (KESTNN). This framework first extracts knowledge triplets representing urban structures to construct a knowledge graph. Representation learning is then conducted to learn the correlation matrix. Throughout this process, data and knowledge are integrated collaboratively via backpropagation, contrasting with the forward feature injection methods typical of existing approaches. This mechanism ensures that data and knowledge directly guide the dynamic updating of model parameters through backpropagation, rather than merely serving as a static feature prompt, thereby fundamentally alleviating the “model-blindness” issue. Finally, the optimized matrix is embedded into a forecasting module. Experiments on the Milan dataset demonstrate that the KESTNN exhibits excellent forecast performance, reducing RMSE by up to 23.91%, 16.73%, and 10.40% for 3-, 6-, and 9-step forecasts, respectively, compared to the best baseline. Full article

Background/Objectives: Cognitive decline and compromised attention control serve as early indicators of neurodysfunction that manifest as broader psychopathological symptoms, yet conventional mental health assessment relies predominantly on subjective self-report measures lacking objectivity and temporal granularity. We propose Neurosense, an AI-driven brain health assessment framework using electroencephalography (EEG) to non-invasively capture neural dynamics. Methods: Our Dual-path Spatio-Temporal Adaptive Gated Encoder (D-STAGE) architecture processes temporal and spatial EEG features in parallel through Transformer-based and graph convolutional pathways, integrating them via adaptive gating mechanisms. We introduce a two-stage paradigm: first training on cognitive task EEG for reaction time prediction to acquire cognitive performance-related representations, then featuring parameter-efficient adapter-based transfer learning to estimate p-factor—a transdiagnostic psychopathology dimension. The adapter-based transfer achieves competitive performance using only 1.7% of parameters required for full fine-tuning. Results: The model achieves effective reaction time prediction from EEG signals. Transfer learning from cognitive tasks to mental health assessment demonstrates that cognitive efficiency representations can be adapted for p-factor prediction, outperforming direct training approaches while maintaining parameter efficiency. Conclusions: The Neurosense framework reveals hierarchical relationships between neural dynamics, cognitive efficiency, and mental health dimensions, establishing foundations for a promising computational framework for mental health assessment applications. Full article

Traditional routing algorithms optimizing for distance or travel time are inadequate for electric vehicles (EVs), which require energy-aware planning considering battery constraints and charging infrastructure. This work presents an energy-optimal routing system for EVs that integrates personalized consumption modeling with real-time environmental data. The system employs a Long Short-Term Memory (LSTM) neural network to predict State-of-Charge (SoC) consumption from real-world driving data, learning directly from spatiotemporal features including velocity, temperature, road inclination, and traveled distance. Unlike physics-based models requiring difficult-to-obtain parameters, this approach captures nonlinear dependencies and temporal patterns in energy consumption. The routing framework integrates static map data, dynamic traffic conditions, weather information, and charging station locations into a weighted graph representation. Edge costs reflect predicted SoC drops, while node penalties account for traffic congestion and charging opportunities. An enhanced A* algorithm finds optimal routes minimizing energy consumption. Experimental validation on a Nissan Leaf shows that the proposed end-to-end SoC estimator significantly outperforms traditional approaches. The model achieves an RMSE of 36.83 and an $R^2$ of 0.9374, corresponding to a 59.91% reduction in error compared to physics-based formulas. Real-world testing on various routes further confirms its accuracy, with a Mean Absolute Error in the total route SoC estimation of 2%, improving upon the 3.5% observed for commercial solutions. Full article

Spiking neural networks (SNNs) aim to simulate human neural networks with biologically plausible neurons. However, conventional SNNs based on point neurons ignore the inherent dendritic computation of biological neurons. Additionally, these point neurons usually employ single membrane potential decay and a fixed firing threshold, which is in contrast to the heterogeneity of real neural networks and limits the neuronal dynamic diversity needed when dealing with multi-scale sequential tasks. In this work, we propose a dendritic spiking neuron model with combined membrane potential decay and a dynamic firing threshold. Then, we extend the neuron model to the feedforward network level, termed dendritic spiking neural network with combined membrane potential decay and dynamic threshold (CD-DT-DSNN). By learning the heterogeneous neuronal decay factors, which combine two different membrane potential decay mechanisms, and learning adaptive factors, our networks can rapidly respond to input signals and dynamically regulate neuronal firing rates, which help the extraction of multi-scale spatio-temporal features. Experiments on four spike-based audio and image sequential datasets demonstrate that our CD-DT-DSNN outperformed state-of-the-art heterogeneous SNNs and dendritic compartment SNNs with higher classification accuracy and fewer parameters. This work suggests that heterogeneity in neuronal membrane potential decay and neural firing thresholds is a critical component in learning multi-timescale temporal dynamics and maintaining long-term memory, providing a novel perspective for constructing high biologically plausible neuromorphic computing models. It provides a solution for multi-timescale temporal sequential tasks, such as speech recognition, EEG signal recognition, and robot place recognition. Full article

Tropical rainforests play a vital role in maintaining global ecological balance, carbon cycling, and biodiversity conservation, making research on their biomass dynamics scientifically significant. This study integrates multi-source remote sensing data, including canopy height derived from GEDI and ICESat-2 satellite-borne lidar, Landsat imagery, and environmental variables, to estimate forest biomass dynamics in Hainan’s tropical rainforests at a 30 m spatial resolution, involving a correlation analysis of factors influencing spatiotemporal changes in Hainan Tropical Rainforest biomass. The research aims to investigate the spatiotemporal variations in forest biomass and identify key environmental drivers influencing biomass accumulation. Four machine learning algorithms—Backpropagation Neural Network (BP), Convolutional Neural Network (CNN), Random Forest (RF), and Gradient Boosting Decision Tree (GBDT)—were applied to estimate biomass across five forest types from 2003 to 2023. Results indicate the Random Forest model achieved the highest accuracy (R² = 0.82). Forest biomass and carbon stocks in Hainan Tropical Rainforest National Park increased significantly, with total carbon stocks rising from 29.03 million tons of carbon to 42.47 million tons of carbon—a 46.36% increase over 20 years. These findings demonstrate that integrating multimodal remote sensing data with advanced machine learning provides an effective approach for accurately assessing biomass dynamics, supporting forest management and carbon sink evaluations in tropical rainforest ecosystems. Full article

Against the backdrop of intensifying global climate change and expanding human encroachment into mountainous regions, landslides have increased markedly in both frequency and destructiveness, emerging as a key risk to socio-ecological security and development in mountain areas. Rigorous assessment and forward-looking prediction of landslide disaster vulnerability (LDV) are essential for targeted disaster risk reduction and regional sustainability. However, existing studies largely center on landslide susceptibility or risk, often overlooking the dynamic evolution of adaptive capacity within affected systems and its nonlinear responses across temporal and spatial scales, thereby obscuring the complex mechanisms underpinning LDV. To address this gap, we examine Yunnan Province, a landslide-prone region of China where intensified extreme rainfall and the expansion of human activities in recent years have exacerbated landslide risk. Drawing on the vulnerability scoping diagram (VSD), we construct an exposure–sensitivity–adaptive capacity assessment framework to characterize the spatiotemporal distribution of LDV during 2000–2020. We further develop a multi-model, multi-scale integrated prediction framework, benchmarking the predictive performance of four machine learning algorithms—backpropagation neural network (BPNN), support vector machine (SVM), random forest (RF), and XGBoost—across sample sizes ranging from 2500 to 360,000 to identify the optimal model–scale combination. From 2000 to 2020, LDV in Yunnan declined overall, exhibiting a spatial pattern of “higher in the northwest and lower in the southeast.” High-LDV areas decreased markedly, and sustained enhancement of adaptive capacity was the primary driver of the decline. At approximately the 90,000-cell grid scale, XGBoost performed best, robustly reproducing the observed spatiotemporal evolution and projecting continued declines in LDV during 2030–2050, albeit with decelerating improvement; low-LDV zones show phased fluctuations of “expansion followed by contraction”, whereas high-LDV zones continue to contract northwestward. The proposed multi-model, multi-scale fusion framework enhances the accuracy and robustness of LDV prediction, provides a scientific basis for precise disaster risk reduction strategies and resource optimization in Yunnan, and offers a quantitative reference for resilience building and policy design in analogous regions worldwide. Full article

With the proliferation of mobile internet and location-based services, location-based social networks (LBSNs) have accumulated extensive user check-in data, driving the advancement of next Point-of-Interest (POI) recommendation systems. Although existing approaches can model sequential dependencies and spatio-temporal patterns, they often fail to fully capture users’ dynamic preferences under varying spatio-temporal contexts and lack effective integration of fine-grained semantic information. To address these limitations, this paper proposes Language-Guided Spatio-Temporal Context Learning for Next POI Recommendation (LSCNP). It employs a pre-trained BERT model to encode multi-dimensional spatio-temporal context—including geographic coordinates, visiting hours, and surrounding POI categories—into structured textual sequences for semantic understanding; constructs dual-graph structures to model spatial constraints and user transition patterns; and introduces a contrastive learning module to align spatio-temporal context with POI features, enhancing the discriminability of representations. A Transformer-based sequential encoder is adopted to capture long-range dependencies, while a neural matrix factorization decoder generates final recommendations. Experiments on three real-world LBSN datasets demonstrate that LSCNP consistently outperforms state-of-the-art baselines. Ablation studies and hyperparameter analyses further validate the contribution of each component to the overall performance. Full article