Search Results (1,279)

The issue of water resources in a semi-arid country such as Morocco has been present for many years and is becoming increasingly critical. The droughts experienced over recent decades have demonstrated the country’s extreme vulnerability to any water deficit. In this context, the Berrechid plain represents a relevant case study illustrating both the practical and theoretical challenges of groundwater governance. The aquifer is heavily exploited to satisfy agricultural, industrial, and domestic needs. This study develops a hybrid “grey-box” modeling approach for predicting groundwater depth (GWD) fluctuations under climate change (CC). Unlike conventional black-box machine learning models, our framework combines a deterministic physical engine with a stochastic machine learning corrector. The physical component simulates aquifer mass balance using the Hargreaves method for evapotranspiration, linear drainage, climate memory via exponential decay, and an anthropogenic trend parameter (xi). The machine learning component—XGBoost with quantile regression—is trained exclusively on physical model residuals and predicts the 5th, 50th, and 95th percentiles, providing explicit 90% confidence intervals. Hydrological states (dry, normal, wet) are identified via K-means clustering for context-aware correction. The model is calibrated using historical data (1972–2019) and validated using blocked time-series cross-validation. Climate projections under the RCP 4.5 and RCP 8.5 scenarios were used to forecast GWD up to 2100. At piezometer 3933/20, the best performance was achieved, with an RMSE of 0.347 m and a KGE of 0.742 during the validation period. The proposed approach is suitable for seasonal GWD forecasting and offers practical value for water managers and decision-makers in the Berrechid region. Full article

Plunger lift systems are vital for optimizing production in gas wells, but their performance can be compromised by various operational anomalies. Traditional diagnostic methods and conventional convolutional neural network (CNN) approaches often struggle with the complex, transient data from these systems, particularly in capturing long-range temporal dependencies and generalizing from limited, imbalanced datasets. This study presents an enhanced diagnostic framework for plunger lift anomaly detection by leveraging the strengths of a pre-trained Vision Transformer (ViT). The methodology transforms one-dimensional time-series pressure data into two-dimensional image representations using the element-wise summation of Gramian Angular Summation Field (GASF) and Gramian Angular Difference Field (GADF), which simultaneously preserves global operational trends and local transient dynamics for vision model analysis. The ViT model, initialized with pre-trained weights, is further optimized using Bayesian optimization (BO) for hyperparameter tuning, and a tailored data augmentation pipeline is employed to improve robustness. Comparative evaluations demonstrate that the proposed ViT-based approach, particularly the ViT + GAF + BO model, significantly outperforms baseline CNN models and their optimized variants, achieving the highest Precision, Recall, and F1-score, with an F1-score of 0.93. Visualizations using t-SNE confirm the ViT’s superior capability in learning discriminative features, showcasing well-separated clusters for different operational conditions compared to CNNs. This research underscores the potential of pre-trained ViTs combined with appropriate data representation and optimization techniques for achieving accurate and reliable anomaly detection in plunger lift systems. Full article

Coarsely gridded Land Surface Models (LSMs) often smooth over sub-grid spatial heterogeneity and non-linear surface soil moisture dynamics during extreme-precipitation events. This study introduces a clustering-based Soil Moisture Active Passive (SMAP) residual framework, evaluating the spatiotemporal discrepancies between 3 km SMAP Level 2 (SMAP-L2) retrievals and 9 km SMAP Level 4 (SMAP-L4) data-assimilation products within the Yanco study region during the extreme March 2021 floods in New South Wales, Australia. By applying k-means clustering to the residual time series, we partitioned the landscape into three distinct hydrological response patterns: a Low-Residual Baseline ($64.5\%$), a Persistent Positive Anomaly ($20.7\%$) indicative of unmodeled inundation, and a Transient Negative Anomaly ($14.8\%$) representing rapid drainage. Consequently, $35.5\%$ of the usable analysis area exhibited temporal trajectories that diverged significantly from model expectations, highlighting profound geographic heterogeneity in surface wetting and retention that cannot be captured by uniform precipitation inputs alone. Benchmarking the satellite-derived time series against the Yanco in situ network provided critical context for cross-scale variations, illustrating general agreement in overarching temporal trends despite the inherent scale mismatch. Ultimately, this approach leverages residual dynamics as a scalable spatial diagnostic, offering a robust, data-driven method to map localized flood responses that are typically obscured by broad-scale model parameters. Full article

This paper extends a previous classification study by examining clustering methods on the same synthetic datasets and comparing their behavior with the previously obtained classification results. This study investigates the performance of selected time series clustering methods under controlled changes in noise level and class complexity. Six clustering methods representing distance-based, feature-based, and deep learning approaches were evaluated on 82 balanced synthetic datasets. The datasets contained from two to six classes, different levels of additive Gaussian noise, 200 time series per dataset, and 1000 observations per time series. The analysis focused on clustering quality, comparative behavior with classification models, and computational cost in terms of training time and peak memory usage. Clustering quality was assessed mainly using Adjusted Rand Index and V-measure, while accuracy after Hungarian label matching was used as an auxiliary measure for comparison with classification models. The results show that distance-based methods, and particularly TimeSeriesKMedoids, achieved the most robust and consistent clustering performance across the considered settings. Clustering quality decreased with both the number of classes and the noise level, but the effect of noise was clearly stronger. Feature-based and deep learning-based clustering methods were generally more sensitive to noise, while deep models were also associated with substantially higher computational cost. In terms of memory usage, classical clustering methods remained below 50 MiB, whereas deep learning-based clustering methods required substantially more memory. This study further shows that accuracy computed after Hungarian label matching may provide an overly optimistic view of clustering quality. Accuracy after Hungarian label matching is reported only as an auxiliary metric, while the main interpretation of clustering quality is based on structure-sensitive measures such as Adjusted Rand Index and V-measure. Overall, the findings highlight the importance of robust distance-based approaches and of using structure-sensitive evaluation measures when analyzing time series clustering. Full article

Precision agriculture increasingly relies on high-resolution, long-term remote sensing to delineate sub-field management zones. However, traditional spatial zonation assumes temporal stationarity, utilizing seasonal aggregates that obscure transient, intra-annual stress signals. This study develops a data-driven framework to characterize both persistent and non-stationary crop water use dynamics by integrating monthly, 30-m evapotranspiration (ET) data from OpenET (2000–2025) with zero-shot temporal anomaly detection. A pre-trained time-series foundation model (Chronos-T5-Small) generated counterfactual expectations for sub-field ET, quantifying deviations using a mean absolute error-based anomaly score. Unsupervised clustering of these anomaly scores with longitudinal ET metrics partitioned the landscape into dynamic biophysical regimes. Cross-registered against legacy persistence mapping based on seasonal totals, the foundation model showed strong directional agreement (86.1%, Cohen’s Kappa = 0.716) in identifying chronically constrained zones across 869 shared active pixels. Crucially, the framework identified 966 historically persistent pixels undergoing stability decay, of which 95.3% were statistically verified via paired t-tests to have collapsed into the field’s baseline variance pool. Furthermore, counterfactual anomaly detection isolated zones of recent acute divergence, differentiating enduring edaphic constraints from sudden system disruptions. This approach demonstrates how foundation models can transition from purely predictive engines to diagnostic instruments, advancing operational precision agriculture. Full article

To address the challenges of complex multi-channel signal coupling and insufficient long-term temporal dependency characterization in remaining useful life (RUL) prediction of rotating machinery, this paper proposes a multivariate time series forecasting framework integrating multi-channel information fusion and a self-attention gated augmentation unit (SGAU). First, a multilayer perceptron (MLP) explicitly models nonlinear coupling among channels; SGAU replaces the conventional feed-forward network in the Transformer encoder, using multi-head self-attention outputs as gating signals to adaptively regulate feature transformation. Second, multi-channel signals are predicted via this framework; high-dimensional feature vectors are extracted to construct multi-channel Gaussian mixture models (GMMs). Third, Jensen–Shannon divergence (JSD) quantifies deviations between the target and initial data clusters; centroid distance evolutionary trajectory is fused with JSD to construct the health indicator (HI). Continuous HI predictions yield the RUL prediction curve. Experiments on a self-designed wind turbine gearbox platform and the XJTU-SY bearing dataset demonstrate that the proposed framework outperforms baseline methods on Mean Square (MS), Root Mean Square (RMS), and Energy metrics, with average error reductions of 6.6% and 12.1% in the horizontal and vertical directions on the gearbox dataset and 20.9% and 32.3% on the bearing dataset, confirming its effectiveness and generalization capability. Full article

Long-term ichthyoplankton time series provide an effective framework for understanding how marine communities respond to environmental variability across temporal scales. We analyzed larval fish assemblage dynamics in the central Mexican Pacific under contrasting seasonal hydrographic conditions and ENSO phases using multivariate analyses, indicator species analysis, clustering, and generalized additive models. Environmental variability exhibited a hierarchical structure, with recurrent seasonal changes in sea surface temperature (SST) and coastal upwelling intensity (CUI), whereas the Oceanic Niño Index (ONI) varied mainly at the interannual scale. Significant differences in assemblage composition were detected among ENSO–seasonality regimes. Distance-based redundancy analysis showed that the primary compositional gradient was associated with seasonal hydrography, while secondary variation reflected ENSO-related interannual shifts. Species responses were expressed primarily through shifts in relative dominance rather than wholesale species replacement, indicating that assemblage reorganization was largely driven by changes in the relative contribution of recurrent taxa. This pattern highlights the role of seasonal hydrography as the primary environmental filter structuring the assemblage, whereas ENSO variability acts mainly as a secondary modulator of species dominance and community trajectories. Consequently, interannual climate anomalies influenced the relative importance of species without substantially redefining the underlying species pool. These findings improve the understanding of plankton community responses to climate variability in the tropical eastern Pacific. Full article

Low back pain (LBP) is a major global health problem and can result in a variety of movement impairments. Advances in smart technology have enabled the collection of novel streams of movement data, and machine learning (ML) methods have been increasingly used for data analysis. However, many existing technologies remain expensive and unsuitable for widespread clinical use, and ML approaches have largely focused on distinguishing people with LBP from healthy controls rather than identifying meaningful subgroups within the LBP population. Motion Tape (MT) is a recently developed wearable strain sensor that translates skin deformation from underlying movement and muscle engagement into electrical signals. In this exploratory study involving 10 participants with LBP, we demonstrate that MT data from six sensors applied on the lower back capture rich movement information capable of characterizing movement patterns among participants with LBP. We propose a feature engineering approach based on biomechanical features as well as time-series causal discovery applied to multivariate sensor time-series data to extract directed inter-segment coordination patterns. We further develop an exploratory subgroup discovery pipeline by aggregating clustering coassociation information across diverse movement tasks. Our causal coordination features show promising discriminative information across several movement types, capturing aspects of motor control not reflected in amplitude-based or embedding-based features alone, such as asymmetries and movement restrictions. Preliminary ensemble clustering analysis indicates three potential LBP subgroups distinguished by biomechanical and inter-segment coordination patterns, which may reflect varied strategies under different movement demands. We investigate the differences in clinical characteristics among these LBP subgroups. We show that time-series foundation models are not well suited for LBP subgrouping due to their uninterpretability, which is improved in our feature engineering pipeline. This framework could reveal additional subgroups with larger cohorts and may generalize to other sensor modalities. Full article

Vibration-based Structural Health Monitoring (SHM) exploits automated Operational Modal Analysis (OMA) to track changes in modal parameters over time for subsequent statistical pattern recognition and anomaly detection. However, weak excitation, measurement noise, non-stationarities, non-linearities, and model inaccuracies can jeopardize the reliability of automated OMA and pollute the modal parameter time series with a number of outliers or spurious estimates. These have an impact on statistical pattern recognition and consequently, the anomaly detection accuracy. Thus, a preliminary data cleansing to enhance the robustness of modal parameter tracking is imperative to ensure the reliability of SHM outcomes. Clustering techniques represent an attractive solution to automatically identify underlying data patterns and discriminate possible spurious results. However, the curse of dimensionality is often an issue in the application of such techniques to time series of experimentally identified modal parameters. To mitigate this issue and, at the same time, the computational efforts, the present study proposes an innovative approach leveraging clustering techniques coupled with mode-pairing constraints for robust and automatic tracking of modal parameters in the context of vibration-based SHM applications. Different clustering algorithms have been embedded in the proposed data processing strategy and applied to a real dataset collected on a full-scale structure under operational conditions. The comparative performance assessment demonstrated how DBSCAN outperforms other clustering methods in the context of the proposed approach, allowing the effective separation of the physical poles from the spurious ones even in the presence of closely spaced modes and highly polluted feature space. Full article

Presently, video communities such as YouTube, bilibili and TikTok have emerged as core fields for information dissemination and public opinion generation. Their embedded user dynamic interaction data support research on public cognitive behavior and content dissemination laws. This study used web crawling technology to construct a complete dataset including 367 video metadata and 2.39 million comment records from Luo Xiang Speaks on Criminal Law—a prominent legal popularization account on the bilibili platform—and systematically explored the temporal evolution patterns of comment interactions in video communities. By establishing a four-dimensional feature system alongside the k-means++ clustering algorithm, this study successfully identified three distinct comment growth patterns (p < 0.001): the burst–decay, the multi-wave oscillation, and the delayed peak. The results of non-parametric tests showed that these three patterns have significant differences in core features (e.g., peak delay time, skewness) and are systematically related to user grade structure, content interaction depth, and release timing. In addition, the user interaction networks of different videos demonstrate significant structural heterogeneity and disassortative mixing, characterized by a highly active minority dominating the discourse, while peripheral nodes gravitate toward high-profile hubs. These findings offer researchers deeper insights into the micro-mechanisms of information dissemination. Full article

Volatility is a fundamental indicator for assessing the risk of financial assets. By integrating unstructured data, such as earnings call transcripts, the limitations of traditional time series data can be transcended, enabling collaborative forecasting from multiple data sources, enhancing the robustness of volatility prediction, and improving the efficiency of risk management. Although current research has effectively utilized earnings call data to predict asset volatility, price trends, and stock correlations, it often overlooks the inherent challenges of integrating textual and time series data, as well as the self-exciting and clustering characteristics of financial events. While conventional Long Short-Term Memory (LSTM) networks excel in processing fused data, they lack the structural capacity to explicitly model event-driven temporal decay, often failing to differentiate the varying influence of historical shocks over time. To surmount this limitation, we have significantly enhanced the predictive model by focusing on extracting salient information and integrating temporal dependency modeling with dynamic state adjustment mechanisms. The core innovation is introducing the Hawkes process to explicitly capture the self-exciting effect of financial events, which is the key to modeling volatility clustering around earnings releases. The proposed Hawkes LSTM model introduces a decay gating module and a textual information knowledge enhancement module. The decay gating module is specifically designed to more effectively capture the temporal dependencies between events within an event sequence. This allows the model to focus more on recent significant events, with the influence of an event on subsequent events typically diminishing as the temporal interval between them increases. By integrating temporal dependency modeling, the model is enabled to utilize historical data in a more flexible manner. The dynamic state adjustment mechanism further enhances its capacity to capture dynamically changing characteristics. Together, these features provide a more robust and precise solution for volatility prediction. Experimental results on two real-world earnings call datasets show that this approach significantly outperforms existing benchmark models on most prediction horizons, achieving competitive and superior performance and verifying its effectiveness and robustness. Full article

The three-dimensional (3D) built environment encodes critical information about urban form intensity, environmental exposure, and resource consumption. However, previous studies have often overlooked the integration of long-term analyses of both horizontal expansion and vertical growth. This study aims to identify the spatial differentiation, morphology types, and vertical growth trajectories of major Chinese megacities over the past three decades. Using high-resolution GABLE building data and time-series GAIA impervious surface data, we examine the evolution of urban 3D morphology across six major Chinese megacities from 1991 to 2023 through a retrospective analysis of building construction years combined with spatial gradient analysis. The results reveal that although the megacities exhibit distinct differences in vertical structure, shape complexity, and spatial compactness, they share a consistent center-to-periphery gradient across most 3D indicators. The most active volumetric growth was concentrated in a zone 8–14 km from city centers, which accounted for 23.6% of total new development, whereas the inner core within 6 km contributed less than 2.68%. In terms of temporal dynamics, Beijing, Shanghai and Guangzhou follow an inverted-V-shaped 3D expansion trajectory driven by mid-rise construction; Tianjin and Hangzhou show accelerated growth with a higher proportion of high-rise clusters; while Shenzhen demonstrates an early peak and a decelerated growth rate, accompanied by a pronounced polycentric pattern. While recent global-scale studies have suggested a shift from outward urban sprawl to vertical development, our findings indicate that horizontal expansion still dominates in the selected Chinese megacities, with outward sprawl exceeding vertical densification during the study period. The integrated approach provides a robust framework for mapping 3D urbanization and offers practical insights for policymakers seeking to manage horizontal expansion, guide vertical intensification, and optimize land-use efficiency in rapidly urbanizing megacities. Full article

Accurate time series forecasting is fundamental to geohazard early warning, yet remains a major challenge. Conventional in situ geotechnical monitoring remains costly and spatially constrained, whereas deep learning applied to remote sensing data has become increasingly prevalent but often suffers from opacity of model decision-making. To address this issue, we propose a Transformer-based forecasting framework, namely PatchTST-Fusion, adapted for multivariate Multi-Temporal Interferometric Synthetic Aperture Radar (MT-InSAR) time series. The framework integrates model interpretability analysis through TimeSHAP, providing temporal and feature-level attributions across the input sequence. Landslide deformation time series are first derived from Copernicus Sentinel-1 SAR data. Variational Mode Decomposition is then applied to decompose the non-linear signals into trend, seasonal, and noise components. The denoised displacement series are modeled and forecast using the proposed PatchTST-Fusion, which incorporates rainfall and reservoir water level fluctuations as feature-level drivers. Application to the Daping landslide cluster in the Three Gorges Reservoir Area in China demonstrates that our method captures both the long-term and transient non-linear coupling between deformation and its triggers, surpassing state-of-the-art models including CNN-BiGRU-Attention, Informer and original PatchTST with 7–55% improvements in MAE and 10–52% improvements in RMSE. Beyond predictive gains, feature attribution of environmental triggers via TimeSHAP reveals that rainfall and reservoir regulation exert temporally distinct influences on slope kinematics, with high relative importance concentrated in specific periods and characteristic lagged responses. This interpretable framework provides both enhanced forecasting accuracy and process-based insights, offering a broadly applicable tool for landslide early warning in reservoir regions. Full article

Autonomous navigation on unknown uneven terrain remains a critical challenge for unmanned ground vehicle (UGV) deployed in unstructured environments such as disaster relief, wilderness exploration, and off-road logistics. Existing motion planning methods for such environments suffer from three key limitations: under-utilization of the solution space due to discretized terrain assessment, difficulty in transforming complex terrain safety constraints into optimization-compatible forms, and the inherent trade-off between environmental modeling accuracy and real-time performance. This paper presents a hierarchical motion planning framework that enables safe and fast navigation of UGV on unknown uneven terrain. We first construct a traversability map based on terrain slope, roughness, and sparsity extracted from ground point cloud clusters. Non-traversable points are then transformed via spherical inversion and inverse mapping to generate terrain safety corridors composed of a series of convex polygons. The geometric containment relationship between the vehicle’s convex hull and the corridor is reformulated as continuously differentiable Control Barrier Function (CBF) constraints to ensure driving safety. The front-end employs a kinodynamic Hybrid A* algorithm with a traversability-aware node pruning strategy, while the back-end trajectory optimization embeds the CBF constraints as hard constraints within the optimization loop to guarantee forward invariance of the safety set under the linearized dynamics. The proposed framework achieves full-shape collision avoidance without sacrificing the solution space, while maintaining real-time performance for autonomous navigation on complex terrain. Full article

Interpreting complex, non-linear Interferometric Synthetic Aperture Radar (InSAR) displacement time-series data for infrastructure risk assessment remains a significant geotechnical challenge. This is particularly evident in regions with established road and railway infrastructures, where the primary objective is monitoring the entire network to ensure safety and operational continuity. Because landslide displacement is a highly complex process affected by a combination of internal geological conditions and external triggers, time-series data inherently encode non-linear trends and periodic fluctuations. To address this, a data-driven framework utilizing a sliding-window transformation to engineer temporal-kinematic features is proposed, providing a broader framework for the contextualization of slope stability assessment from a network perspective. This is paired with Principal Component Analysis (PCA) for dimensionality reduction and evaluated across four unsupervised architectures: K-means, Self-Organising Maps (SOMs), Hybrid SOM-K-means, and PCA-K-means. The comparative evaluation reveals that the PCA-K-means pipeline performed best, offering a highly efficient and scalable workflow. The analysis revealed that the optimized PCA-K-means architecture successfully captured 79.20% of the kinematic variance across the first two principal components. Furthermore, it achieved a robust Between-Cluster-to-Total-Sum-of-Squares (BCSS/TSS) ratio of 71.70%, an optimal Silhouette Score of 0.320, and a low Quantisation Error (QE) of 0.90, demonstrating superior spatial separation and geometric accuracy compared to traditional heuristic methods. When cross-validated against static topographic susceptibility models, the dynamic kinematic clusters exhibited a 23% spatial convergence at the polar bounds of risk, successfully grounding the algorithm’s predictions in physically verified geomorphological features. Relying on the statistical volatility of displacements, this optimal model successfully partitioned the data into five distinct geotechnical risk classes, ranging from stable (Class A) to extreme risk (Class E). The results demonstrate that the developed dynamic framework provides a highly reliable, actionable tool for proactive, large-scale slope stability and infrastructure risk assessment. Full article