Search Results (4,483)

River water is a vital component of the hydrological cycle, sustaining ecosystems and serving as the most accessible freshwater resource for human use. Beyond elemental concentrations, isotopic tracers such as boron (δ¹¹B) and radiogenic strontium (⁸⁷Sr/⁸⁶Sr) provide insights into weathering processes and anthropogenic impacts. This study examines spatial and temporal variations in the chemical composition of the Erren River to distinguish natural contributions from human-derived inputs and assess recent pollution. Samples collected from upstream to downstream were processed by micro-sublimation or column chromatography, with isotopes measured using MC-ICP-MS. Results show δ¹¹B values from +4.8‰ to +30.4‰ (variation ~26‰) and ⁸⁷Sr/⁸⁶Sr ratios from 0.709679 to 0.710446. Major ion and isotopic data indicate upstream waters are dominated by silicate weathering, while downstream areas reflect seawater and salt spray influence, consistent with regional geology and hydrology. Furthermore, δ¹¹B patterns combined with Cl⁻/Na and NO₃⁻/B ratios suggest that tributaries in the mid-to-lower basin remain affected by anthropogenic pollution, likely linked to agricultural and urban activities. These findings highlight both natural controls and ongoing human impacts on the Erren River system. Full article

To host the 2022 Winter Olympics, Beijing and Zhangjiakou implemented extensive ecological restoration projects, improving the ecological quality of the region. However, detailed evidence of long-term spatiotemporal dynamics in vegetation productivity remains limited. This study employed the Carnegie–Ames–Stanford Approach (CASA) to estimate the vegetation Net Primary Productivity (NPP) in the Beijing–Zhangjiakou region from 2004 to 2023, utilizing 250 m monthly NDVI data. The 30 m resolution China Land Cover Dataset (CLCD) was incorporated to mask non-vegetated pixels and refine the vegetation mask, reducing mixed-pixel effects. Spatiotemporal variations, seasonal change-point detection, interannual stability, and trend persistence were analyzed across administrative regions and land cover types. Results indicate pronounced spatial heterogeneity in NPP, with persistently high values in forest-dominated western and northern Beijing and northeastern Zhangjiakou, and lower values concentrated in Beijing’s built-up and cropland-dominated southeastern plain. Pixel-level boxplots suggest stronger intra-regional variability in Beijing than in Zhangjiakou. Across landcover types, forests generally maintain the highest NPP, while grasslands are relatively lower. Boxplots further show that shrubs exhibit the highest variability, with all types showing right-skewed distributions. Annual mean NPP increased significantly for the entire region, Beijing, and Zhangjiakou, with interannual increase rates of 3.57, 1.56, and 4.53 gC·m⁻²·yr⁻², respectively; the lowest values occurred in 2007 and the highest in 2022. Trend maps and category statistics consistently suggest that positive trends dominate most of the region and expanded slightly during 2014–2023. BEAST analysis suggests a stable seasonal NPP cycle with no significant seasonal change points. CV-based assessment indicates generally high to extremely high stability, whereas low-stability zones are mainly associated with urban expansion areas, surrounding croplands, and parts of Zhangjiakou grasslands. Hurst results suggest that persistently increasing trends cover more than 90% of the study area, while persistently decreasing trends account for about 5.25% and are primarily linked to Beijing’s expansion zones. Full article

Background/Objectives: Accurate sleep stage classification is essential for evaluating sleep quality and diagnosing sleep disorders. Despite recent advances in deep learning, existing models inadequately represent complex brain dynamics, particularly the time-lag effects inherent in neural signal propagation and regional variations in cortical activation patterns. Methods: We propose the MFST-GCN, a graph-based deep learning framework that models these neurobiological phenomena through three complementary modules. The Dynamic Dual-Scale Functional Connectivity Modeling (DDFCM) module constructs time-varying adjacency matrices using Pearson correlation across 1 s and 5 s windows, capturing both transient signal transmission and sustained connectivity states. This dual-scale approach reflects the biological reality that neural information propagates with measurable delays across brain regions. The Multi-Scale Morphological Feature Extraction Network (MMFEN) employs parallel convolutional branches with varying kernel sizes to extract frequency-specific features corresponding to different EEG rhythms, addressing regional heterogeneity in neural activation. The Adaptive Spatio-Temporal Graph Convolutional Network (ASTGCN) integrates spatial and temporal features through Chebyshev graph convolutions with attention mechanisms, encoding evolving functional dependencies across sleep stages. Results: Evaluation on ISRUC-S1 and ISRUC-S3 datasets demonstrates F1-scores of 0.823 and 0.835, respectively, outperforming state-of-the-art methods. Conclusions: Ablation studies confirm that explicit time-lag modeling contributes substantially to performance gains, particularly in discriminating transitional sleep stages. Full article

Emergency Medical Services (EMS) require timely and equitable ambulance allocation supported by accurate demand estimation. In our prior work, we developed a statistical forecasting module based on Overall Smoothed Average Demand (OSAD) and Average Maximum (AMX) to estimate proportional EMS demand across spatial zones. Although this approach was interpretable and computationally efficient, it was limited in modeling nonlinear spatiotemporal dependencies and adapting to dynamic demand variations. This paper presents a unified deep learning-based EMS planning framework that integrates spatiotemporal demand forecasting with adaptive ambulance allocation. Specifically, the statistical OSAD/AMX estimators are replaced by graph-based spatiotemporal forecasting models capable of capturing spatial interactions and temporal dynamics. The predicted demand is then incorporated into a reinforcement learning-based allocator that dynamically optimizes ambulance placement under fairness, coverage, and operational constraints. Experiments conducted on real-world EMS datasets demonstrate that the proposed end-to-end framework not only improves demand forecasting accuracy but also translates these improvements into tangible operational benefits, including enhanced equity in resource distribution and reduced response distance. Compared with traditional statistical and heuristic-based baselines, the proposed approach provides a more adaptive and decision-aware solution for EMS planning. Full article

Global change represents one of the most pressing threats to ecosystems, profoundly influencing habitats and redefining management and conservation priorities. Rising temperatures, altered precipitation regimes, invasive species and the increasing frequency of extreme events, such as prolonged droughts and wildfires, are modifying the composition, structure, and resilience of forests. Often, these changes result in habitat fragmentation, which isolates populations and diminishes their ability to adapt. This situation calls for an urgent reassessment of traditional protected area management practices. In response to climate change, it is essential to prioritize conservation strategies that focus on adaptation and maintaining biodiversity, while combating the spread of invasive species. For this reason, this study aims to analyze the impact of global changes on forest vegetation within protected areas, using Aspromonte National Park, a highly biodiverse region, as a case study. It evaluates the transformations in habitat cover and fragmentation over twenty years by comparing the 2001 vegetation map of Aspromonte National Park with the Map of Nature of the Calabria region, to quantify spatial and temporal habitat variations using QGIS 3.42.3 software. Morphological Spatial Pattern Analysis (MSPA) and FRAGSTATS v4.2 were employed to evaluate habitat fragmentation. The results indicate that most forest habitats have experienced a slight increase in area over the past 20 years. However, the area occupied by Pinus nigra subsp. laricio forests (Habitat 42.65) has decreased significantly, most likely due to repeated fires in previous years. In conclusion, this study establishes a scientific foundation for guiding conservation policies in the protected area and promoting the resilience of native plant communities against global change. This is essential for ensuring their survival for future generations while mitigating both habitat fragmentation and the introduction and spread of non-native species. Full article

Grassland productivity faces heightened uncertainty under nonlinear climatic forcing. This study characterizes the spatial heterogeneity of nonlinear variations and nonstationary climate sensitivities across the Eurasian Steppe Region (EASR) to provide a scientific basis for its adaptive management. Using the aboveground net primary productivity (ANPP) and climate datasets (1982–2015), we employed piecewise linear regression, LOWESS, and sliding window partial correlation analysis to identify temporal turning points and dynamic climate–productivity relationships. We identified distinct turning points in 1994 and 2008, revealing a phased “Increasing–Decreasing–Increasing” trajectory. A key novelty is the mapping of eight phased trajectory patterns, illustrating significant spatial heterogeneity in productivity trends. Furthermore, we demonstrate temporally reversed climate sensitivities. Notably, the sensitivity of ANPP to temperature shifted from positive to negative as warming-induced water stress intensified. While precipitation remains the dominant driver (68% of the region), its influence is nonstationary and state-dependent. In the Qinghai–Tibet Plateau, the limiting factor transitioned from thermal to water availability. Overall, productivity in the EASR appears to undergo phased reorganization under shifting climatic baselines. Our findings suggest that future ecosystem models should incorporate time-varying sensitivity parameters to account for nonlinear dynamics and potential trend reversals in grassland ecosystems. Full article

This study constructs a time-dependent model to predict the nighttime suspended sediment concentration near Wenzhou based on the convolutional neural network U-Net, which integrates the high-resolution Delft3D (version 4.03.01) hydrodynamic model and GOCI satellite observation data. The model’s prediction accuracy is significantly improved by replacing the original tide level with the tide level variation and increasing the temporal resolution of the flow field to 15 min via sensitivity analysis of the model’s input parameters. The validation results show that the model can maintain high consistency with GOCI observations in short-term prediction, with a structural similarity index (SSIM) of 0.82. For multi-hour continuous nighttime predictions, while quantitative uncertainty increases with the forecast horizon, the model successfully captures the spatial evolution patterns and maintains stable structural characteristics. The model effectively provides missing remote sensing nighttime observations as well as a new method for full-cycle prediction of nearshore SSC. Full article

Human activity recognition in privacy-sensitive indoor environments requires sensing modalities that remain robust under illumination variation and background clutter while preserving user anonymity. To this end, this study proposes a lightweight radar–camera fusion deep learning model that integrates motion signatures from FMCW radar with coarse spatial cues from ultra-low-resolution camera frames. The radar stream is processed as a Range–Doppler–Time cube, where each frame is flattened and sequentially encoded using a Transformer-based temporal model to capture fine-grained micro-Doppler patterns. The visual stream employs a privacy-preserving $4\times 5$-pixel camera input, from which a temporal sequence of difference frames is extracted and modeled with a dedicated camera Transformer encoder. The two modality-specific feature vectors—each representing the temporal dynamics of motion—are concatenated and passed through a lightweight fully connected classifier to predict human activity categories. A multimodal dataset of synchronized radar cubes and ultra-low-resolution camera sequences across 15 activity classes was constructed for evaluation. Experimental results show that the proposed fusion model achieves 98.74% classification accuracy, significantly outperforming single-modality baselines (single-radar and single-camera). Despite its performance, the entire model requires only 11 million floating-point operations (11 MFLOPs), making it highly efficient for deployment on embedded or edge devices. Full article

The development of urban green infrastructures (UGI) is considered among the main nature-based solutions for climate mitigation in cities; however, the role of soils in the carbon (C) balance of UGI ecosystems remains largely overlooked. Urban green spaces are typically dominated by constructed Technosols, created by adding organic materials on top of former natural or agricultural subsoils. The combined effects of land-use history and current UGI management result in a high spatial variation of soil organic carbon (SOC) stocks and soil CO₂ emissions. Our study aimed to explore this variation for the case of Wageningen University campus. Developed on a former agricultural land, the campus area includes green spaces dominated by trees, shrubs, lawns, and herbs, with well-documented management practices for each vegetation type. Across the campus area (~32 ha), a random stratified topsoil sampling (n = 90) was conducted to map the spatial variation of topsoil (0–10 cm) SOC stocks. At the key sites (n = 8), representing different vegetation types and time of development (old, intermediate, and recent), SOC profile distribution was analyzed including SOC fractionation in surface and subsequent horizons, as well as the dynamics in soil CO₂ emissions, temperature, and moisture. Topsoil SOC contents on campus ranged from 1.1 to 5.5% (95% confidence interval). On average, SOC stocks under trees and shrubs were 10–15% higher than those under lawns and herbs. The highest CO₂ emissions were observed from soil under lawns and coincided with a high proportion of labile SOC fraction. Temporal dynamics in soil CO₂ emissions were mainly driven by soil temperature, with the strongest relation (R² = 0.71–0.88) observed for lawns. Extrapolating this relationship to the calendar year and across the campus area using high-resolution remote sensing data on surface temperatures resulted in a map of the CO₂ emissions/SOC stocks ratio, used as a spatial proxy for C turnover. Areas dominated by recent and intermediate lawns emerged as hotspots of rapid C turnover, highlighting important differences in the role of various UGI types in the C balance of urban green spaces. Full article

The root system provides the interface between the plant and the soil that is responsible for water and nutrient uptake and transport. We hypothesized that almond trees in the commercial production environment could adjust their root acquisitive traits with distance vertically and horizontally from driplines as adaptive responses to within-orchard reductions in irrigation and nitrogen inputs. We compared the responses of root acquisitive traits under four years of treatments ranging from +W+N (15 ML ha⁻¹ water and 300 kg ha⁻¹ nitrogen per season) to −W−N (10.5 ML ha⁻¹ water and 160 kg ha⁻¹ nitrogen per season, with −W involving a 30% reduction in irrigation and −N involving a 46% reduction in nitrogen). Roots (<3 mm) were sampled through soil coring in the winters of 2017, 2018, and 2019. Root sampling was conducted along the vertical gradient and along the horizonal gradient (0 cm, 80 cm, and 240 cm from the dripline). Four years of treatments highlighted that the data variation was driven mainly by the difference between the +W and −W treatments (along PC1). Further, the difference between −W−N (combined resource reduction) and the other three treatments (+W+N, +W−N, and −W+N) contributed to the data variation (along PC2). Also, the temporal dynamics of treatment effects over 2017, 2018, and 2019 suggested a temporally strengthened +W−N effect to increase root biomass, average root diameter, specific root surface area (SRA), and specific root length (SRL) at deeper soil depths and at greater soil distances from driplines. These findings on the spatial and temporal plasticity of traits representing root resource acquisition capabilities highlighted the important role of root systems in maintaining crop productivity under reduced irrigation and nitrogen inputs. Full article

Interferometric Synthetic Aperture Radar (InSAR) is an advanced imaging geodesy technique for detecting and characterizing surface deformation with high spatial resolution and broad spatial coverage. However, as an inherently post-event observation method, InSAR suffers from limited capability for near-real-time and short-term updates of deformation time series. In this paper, we proposed a data-driven adaptive framework for deformation prediction based on a hybrid deep learning method to accurately predict the InSAR-derived deformation time series and take the Xi’erguazi−Mawo landslide complex (XMLC) as a case study. The InSAR-derived time series was initially decomposed into trend and periodic components with a two-step decomposition process, which were thereafter modeled separately to enhance the characterization of motion kinematics and prediction accuracy. After retrieving the observations from the multi-temporal InSAR method, two-step signal decomposition was then performed using the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) and Variational Mode Decomposition (VMD). The decomposed trend and periodic components were further evaluated using statistical hypothesis testing to verify their significance and reliability. Compared with the single-decomposition model, the further decomposition leads to an overall improvement in prediction accuracy, i.e., the Mean Absolute Errors (MAEs) and the Root Mean Square Errors (RMSEs) are reduced by 40–49% and 36–42%, respectively. Subsequently, the Radial Basis Function (RBF) neural network and the proposed CNN-BiLSTM-SelfAttention (CBS) models were constructed to predict the trend and periodic variations, respectively. The CNN and self-attention help to extract local features in time series and strengthen the ability to capture global dependencies and key fluctuation patterns. Compared with the single network model in prediction, the MAEs and RMSEs are reduced by 22–57% and 4–33%, respectively. Finally, the two predicted components were integrated to generate the fused deformation prediction results. Ablation experiments and comparative experiments show that the proposed method has superior ability. Through rapid and accurate prediction of InSAR-derived deformation time series, this research could contribute to the early-warning systems of slope instabilities. Full article

Camera, lidar, and wearable-based gesture recognition technologies face practical limitations such as lighting sensitivity, occlusion, hardware cost, and user inconvenience. Wi-Fi channel state information (CSI) can be used as a contactless alternative to capture subtle signal variations caused by human motion. However, existing CSI-based methods are highly sensitive to domain shifts and often suffer notable performance degradation when applied to environments different from the training conditions. To address this issue, we propose a domain-robust touchless finger location sensing framework that operates reliably even in a single-link environment composed of commercial Wi-Fi devices. The proposed system applies preprocessing procedures to reduce noise and variability introduced by environmental factors and introduces a multi-domain segment combination strategy to increase the domain diversity during training. In addition, the dual-axis transformer learns temporal and spatial features independently, and the GNN-based integration module incorporates relationships among segments originating from different domains to produce more generalized representations. The proposed model is evaluated using CSI data collected from various users and days; experimental results show that the proposed method achieves an in-domain accuracy of 99.31% and outperforms the best baseline by approximately 4% and 3% in cross-user and cross-day evaluation settings, respectively, even in a single-link setting. Our work demonstrates a viable path for robust, calibration-free finger-level interaction using ubiquitous single-link Wi-Fi in real-world and constrained environments, providing a foundation for more reliable contactless interaction systems. Full article

The Venice lagoon is the largest in the Mediterranean Sea. The historic city of Venice, located on a cluster of islands in the centre of this lagoon, is an enchanting and iconic destination for national and international tourists. The historical centre of Venice and the other islands of the lagoon, such as Burano, Murano and Torcello, attract crowds of tourists every year. Transportation is provided by boats navigating the lagoon along a network of canals. The lagoon itself attracts visitors who enjoy various outdoor recreational activities in the open air, such as fishing and sunbathing. While statistics are available for the activities targeting the islands, no information is currently available on the spatio-temporal distribution of recreational activities across the lagoon waters. This study explores the feasibility of using Sentinel-2 satellite images to assess and map the spatio-temporal distribution of boats in the Venice Lagoon. Cloud-free Level-2A images have been selected to study seasonal (summer vs. winter) and weekly (weekends vs. weekdays) variabilities in 2023, 2024, and 2025. The RGB threshold filtering and the U-Net Semantic Segmentation were applied to the Sentinel-2 images to ensure reliable results. Two spatial indices were produced: (i) a Water Recreation Index (WRI), identifying standing boats in areas attractive for recreation; and (ii) a Water Transportation Index (WTI), mapping moving boats along the canals. Multi-temporal WRI maps allow areas with recurring recreational activities—that are significantly higher in the summer compared to winter, and on weekends compared to other weekdays—to be identified. The WTI identifies canal paths with higher traffic intensity with seasonal and weekly variations. The latter should be targeted by measures for traffic control to limit wave induced erosion, while the first could be subject to protection or development strategies. Full article

Existing studies on urban public space vitality predominantly focus on single temporal scales or macro-urban levels, lacking a systematic understanding of day–night and weekday–weekend differentiation patterns at the meso-scale. This study examines 149 public spaces in the Yuexiu District, Guangzhou, employing Baidu heatmap data and the geographically weighted random forest (GWRF) model to analyze built environment impacts across four temporal scenarios. The SHAP interaction analysis is incorporated to quantitatively evaluate factor interdependencies and their temporal variations. Findings reveal significant spatiotemporal heterogeneity. Building density shows greater night-time importance while residential density exhibits enhanced daytime importance, particularly on weekend. Weekday–weekend comparison demonstrates contrasting spatial reorganization patterns, with weekday showing divergence and weekend showing convergence in factor importance distributions. The factor interaction analysis highlights stable synergistic relationships between density and diversity, alongside temporal transitions in density–residential density interactions from competitive to synergistic during night-time. Low-vitality public spaces are concentrated in peripheral areas with high building density but insufficient commercial facilities and functional mix. These findings deepen our understanding of the spatiotemporal mechanisms underlying public space vitality generation and the interaction effects among built environment factors, thereby providing an empirical foundation for the formulation of temporally adaptive planning strategies. Full article

The construction of reservoirs has undeniably provided numerous conveniences and benefits to human societies. However, it has also markedly altered downstream flow regimes, leading to essential fish habitat loss that directly undermines the ecosystem services provided by fish populations, thereby jeopardizing the long-term sustainability of fishery resources. Existing assessments of spawning suitability largely concentrate on static characteristics of available spawning grounds, while the dynamics of habitat suitability migration and contraction in response to changing environmental flows remain poorly understood. To address this gap, we classified hydrological years into wet, flat, and dry categories to reflect the varying environmental flow requirements during the fish-spawning period. Using the Mike21 hydraulic model together with a spatial suitability analysis for spawning habitats, we quantified spawning ground suitability from both temporal and spatial perspectives. Taking the four major Chinese carps (FMCC) and the Dongta spawning ground in the Pearl River as a case study, our findings reveal that the proportion of highly suitable habitats closely tracks the environmental-flow trajectories. Throughout the FMCC spawning period, the spatial pattern of high suitability zones undergoes a marked migration in response to flow variations across wet, flat, and dry years, consistently shifting upstream. Specifically, as discharge rises from low-flow to high-flow events, the most suitable areas move from downstream deep-pool sections toward upstream shallow riffle zones, which is crucial for the sustainable development of fishery resources. Full article