Search Results (5,140)

The remote sensing scene classification (RSSC) task plays a pivotal role in Earth observation missions, yet its progress remains constrained by the scarcity of high-quality labeled imagery. This article introduces a self-supervised learning (SSL) paradigm to address this challenge. First, for pseudo-label construction, a large set of long-interval satellite revisit imagery is collected and processed with pixel-level registration. The SIFT inliers retained during registration serve as saliency priors to guide asymmetric masking across views. This produces positive pairs that preserve global scene consistency while introducing controlled object-level ambiguities. Second, we propose a progressive layer-wise contrastive learning framework (MTC-Net) that couples the pseudo-label with the network’s representational hierarchy, forming a curriculum from local texture robustness to global semantic invariance. A dual-attention module with spatial–channel branches is further embedded to recalibrate intermediate features. The learning paradigm encourages the model to perform cross-view contextual reasoning rather than relying on pixel-wise correspondences. Experiments on three widely used datasets demonstrate that MTC-Net achieves competitive classification accuracy under limited-label settings, while ablation and visualization studies validate the effectiveness of establishing scene-level invariance through multi-temporal contrastive alignment. Full article

Software-defined satellite networks (SDSNs) enhance resource utilization and flexibility in space-based networks by leveraging a global view and programmability. A highly reliable control plane is essential to sustain network operations. However, the highly dynamic topology and physical failures in Low Earth Orbit (LEO) environments can cause satellite node outages or inter-satellite link disruptions, leading to control plane interruptions and local load imbalances. To address this, we propose a switch migration mechanism for failure recovery and establish a multi-objective migration model that jointly optimizes control link delay, controller load variance, and normalized migration ratio. To accommodate distinct dynamic characteristics such as frequent topology changes, failure-intensive periods, and stable periods, we design two algorithms: a robust migration algorithm, DNSGA-II, which features population diversity maintenance and environmental awareness, and an efficient migration algorithm, IHAOAVOA, which integrates strong global exploration with powerful local exploitation. Simulation results show that IHAOAVOA rapidly converges under large-scale failures, achieving millisecond-level delay recovery and low normalized migration ratio overhead during failure-intensive periods, while DNSGA-II focuses on long-term load balancing and system stability during stable periods, effectively suppressing localized controller overload. By adopting IHAOAVOA during topology fluctuations or high-failure phases to reduce delay, and switching to DNSGA-II during stable phases to optimize load distribution, the overall network robustness can be improved under the evaluated failure scenarios. This work provides effective support for achieving highly reliable control in SDSNs under failure scenarios. Full article

Accurate information on atmospheric temperature profiles is crucial for improving numerical weather prediction (NWP). However, the harsh environment of the Tibetan Plateau (TP) limits the availability of station observations, which thereby fail to meet the high spatial resolution required for NWP. In this study, we present an integrated framework as an engineering refinement combining the variation method with an artificial neural network (Var-ANN) to calibrate temperature profiles obtained from the Vertical Atmosphere Sounding System (VASS) aboard the polar-orbiting satellite FY-3C. The variation method is first applied to construct a spatially consistent reference field from available station observations, and this field is then used as the training target for a back-propagation neural network that learns the empirical relationship between satellite brightness temperatures and corrected atmospheric temperature. The calibrated temperature profiles were evaluated against independent radiosonde observations and further tested through assimilation into the Weather Research and Forecasting (WRF) model for precipitation simulation over the TP. Results indicate that the Var-ANN calibration reduces the root-mean-square error (RMSE) by approximately 60% and the mean bias from approximately −5 °C to −0.7 °C relative to radiosonde observations. In two WRF case studies, the calibrated profiles show potential for improving precipitation forecast skill, although the limited sample size precludes robust conclusions about operational forecast improvements. The Var-ANN framework provides a practical approach for enhancing the utility of FY-3C VASS temperature products for NWP applications over data-sparse complex terrain. Full article

Remote-sensing-derived spectral indices and in situ measurements provide complementary information for aquatic monitoring, but their practical integration is complicated by asynchronous observations, heterogeneous tables, missing acquisitions, and non-reproducible preprocessing. This study presents AI-Aquatica-RS, a modular Python framework for spectral-index calculation, station-aware nearest-neighbor temporal alignment, feature-set construction, regression benchmarking, command-line execution, and structured result export. The software was evaluated using a fully synthetic controlled benchmark; no real satellite scenes or field-monitoring measurements were used. The benchmark comprised 1080 daily in situ-like observations from six stations and 181 unique remote-sensing-like acquisitions generated as water-like surface-reflectance proxies. A ±3-day alignment tolerance produced a shared complete-case cohort of 954 records. To ensure a fair comparison, the in situ-only, spectral-index-only, and fused configurations were evaluated on exactly the same 667 training and 287 validation records. The fused configuration achieved the best performance using ridge regression (RMSE = 3.481 NTU, MAE = 2.768 NTU, R² = 0.729), compared with RMSE values of 5.086 NTU for the in situ-only configuration, and 5.538 NTU for the spectral-index-only configuration. The benchmark demonstrates reproducible execution and recovery of complementary information under controlled conditions; it does not constitute environmental validation. AI-Aquatica-RS provides an extensible software layer for future studies using real satellite products, monitoring networks, sensor-specific preprocessing, and spatially blocked validation. Full article

Effective conservation of the Near Threatened (NT) black-necked crane (Grus nigricollis) requires standardized frameworks for organizing multi-scale biodiversity data. This study proposes the Essential Biodiversity Variables (EBV) system as an optimal structure for archiving and sharing biodiversity data such as Habitat Suitability Index (HSI). Developed by the Group on Earth Observations Biodiversity Observation Network (GEO BON), the EBV framework is an emerging system offering a robust solution for standardizing data exchange. Based on 483,592 valid location records of 106 black-necked cranes using satellite telemetry, we apply species distribution models and demonstrate how the multi-dimensional EBV architecture accommodates distinct life-stage preferences: breeding sites favor mid-elevations modulated by temperature; migration staging relies on precipitation regimes; and wintering grounds are driven by moisture availability and the avoidance of human-modified landscapes. The EBV NetCDF (Network Common Data Form) format functions as a self-describing hypercube that captures spatial, temporal, and life-stage dimensions while ensuring metadata transparency. This integration facilitates critical applications, including the identification of priority conservation areas and climate vulnerability assessments, thereby bridging the gap between species-specific modeling and global biodiversity monitoring standards. Full article

Motivated by the increasing security requirements of next-generation satellite-terrestrial communication systems and the emergence of Open Radio Access Network (O-RAN) architectures, this paper presents a secrecy analysis of a novel reconfigurable intelligent surface (RIS)-assisted mixed free-space optical (FSO) and radio frequency (RF) satellite downlink transmission system within an O-RAN-enabled non-terrestrial network (NTN) framework. The inherent broadcast nature of RF transmissions presents significant eavesdropping risks, which serves as the primary impetus for this study. We analyze the combined effects of imperfect channel state information (CSI) and random link blockage within such integrated networks. The impact of discrete phase shift constraints at the RIS is also investigated. Closed-form expressions are derived for three key performance metrics: connection outage probability (COP), secrecy outage probability (SOP), and the probability of positive secrecy capacity (PPSC). Through high signal-to-noise ratio (SNR) asymptotic analysis, corresponding asymptotic expressions are obtained, and all analytical results are validated via extensive Monte Carlo simulations. Our findings demonstrate that: (i) Link blockage probability and channel estimation accuracy jointly govern the secrecy performance floor. (ii) Increasing the number of RIS elements enhances physical-layer security by driving both the COP and SOP toward their theoretical lower bounds. (iii) Improving channel estimation accuracy diminishes the eavesdropper’s channel advantage and improves the overall system security. These results offer valuable insights for designing secure mixed FSO/RF satellite-terrestrial systems within O-RAN-enabled NTN architectures that effectively balance connectivity and confidentiality. Full article

The southeastern Tibetan engineering corridor hosts the densest transportation network in Tibet, China, and is traversed by large-scale railway and power corridor projects under construction. This region is home to numerous glacial lakes, some of which are prone to glacial lake outburst floods (GLOFs), posing potential threats to the infrastructure. However, the spatiotemporal evolution and GLOF susceptibility of these lakes remain unclear. Using Landsat 5–9 and Sentinel-2 satellite imagery, we analyzed the spatiotemporal characteristics of glacial lakes from 1990 to 2020. Based on historical GLOF events, we established a susceptibility assessment criterion and determined the susceptibility levels of all glacial lakes in the study area. Results show that the number and area of glacial lakes increased by 40.4% and 26.2%, respectively, from 1990 to 2020, with expansion rates of 2.47 lakes/year and 0.26 km²/year. We identified 31 very highly and 48 highly susceptible lakes, mainly distributed along the Gongrigabu River and the Parlung Tsangpo River. Among them, 35 lakes are most likely to impact National Highways G219 and G318 within the study area. Additionally, three channels with glacial lake clustering amplification effects were found, which may lead to the superposition and amplification of flood impacts, significantly increasing GLOF risks and hazards. Our findings provide crucial references for ensuring the safe operation of local transportation networks and reducing GLOF risks in ongoing large-scale construction projects. Full article

Accurately predicting fine particulate matter (PM${}_{2.5}$) concentrations in regions with sparse monitoring networks remains a critical challenge for air quality management and public health. This study evaluates a machine learning (ML) data fusion approach that integrates daily federal regulatory observations, daily low-cost community sensor measurements, and monthly satellite-derived aerosol products (functioning as a regional background field) to improve PM${}_{2.5}$ prediction across under-monitored environments. Using a Long Short-Term Memory (LSTM) neural network architecture, the analysis examines how combining heterogeneous data sources influences predictions. Results show that pooled multi-source training was associated with higher holdout skill relative to some single-source configurations under this parsimonious baseline, though associations are city- and configuration-dependent and cannot be attributed solely to fusion because evaluation populations are not common. Comparisons against tree-based baselines (Random Forest, Gradient Boosting, XGBoost) indicate that overall predictive skill, not just the LSTM’s, is constrained by data availability, suggesting that data composition, rather than model choice, is the primary driver of the observed performance patterns. These findings highlight both the potential and the practical constraints of multi-source ML approaches for air quality prediction and exposure assessment, with implications for model design, monitoring strategy, and environmental equity. This study is intentionally scoped as an applied evaluation of data fusion performance rather than a comprehensive assessment of algorithmic optimality or operational forecasting readiness. The analysis focuses on daily PM${}_{2.5}$ prediction across a selected set of U.S. cities and does not address sub-daily variability, real-time deployment constraints, or event-specific model optimization. Model performance is therefore interpreted in the context of data availability, consistency, and representativeness, rather than as an upper bound on achievable predictive skill. Full article

Low Earth Orbit (LEO) communication networks are an important component of non-terrestrial networks (NTNs) in sixth-generation (6G) communication systems. LEO satellites are characterized by low propagation delay and highly time-varying topology. Space-based mobility management can effectively reduce transmission delay; however, rapid network variation makes space-based control node deployment and reconfiguration difficult to model and solve. Focusing on dual-layer LEO Walker constellations, this paper investigates the joint optimization of control node deployment and dynamic reconfiguration, and formulates a 0–1 mixed-integer linear programming model with multiple practical constraints, aiming to minimize the total handover and migration delay. The model incorporates practical constraints such as the CN resource budget, unique management of access layer satellites, inter-layer reachability, feeder link connectivity, non-empty control node (CN) management, and onboard resource capacity. To support online deployment, we propose a rolling-horizon migration-aware dynamic greedy control node placement algorithm (RH-MA-DGCNP), which updates the CN placement and the affiliation between access-layer satellites and CNs at each reconfiguration epoch while jointly considering the handover delay and the migration delay caused by transferring control-affiliation states from previous serving CNs to new serving CNs. A comparison with exact current-epoch MILP solutions obtained by CPLEX on validation instances shows that RH-MA-DGCNP achieves small optimality gaps with shorter computation time. Simulation results show that RH-MA-DGCNP achieves the lowest mean handover delay among all benchmark schemes and the lowest cumulative total delay cost among the quasi-dynamic and dynamic benchmark schemes. The CDF of handover delay further indicates that RH-MA-DGCNP has a higher proportion of low-delay handover events and effectively suppresses extremely high-delay handover cases. Sensitivity analyses under different elevation angle thresholds and ground station deployments further show that RH-MA-DGCNP maintains its performance advantage over the benchmark schemes under different network settings. Full article

Accurate extraction of road networks from high-resolution remote sensing imagery is a fundamental task underpinning autonomous-driving navigation, urban spatial planning, and the dynamic updating of geographic information databases. Although existing road extraction methods attain outstanding pixel-level segmentation accuracy and topological integrity, most follow an accuracy-first design paradigm that relies on heavyweight backbones and increasingly complex decoders, incurring a parameter volume and storage overhead that constitute the principal bottleneck for deploying them on resource-constrained edge platforms such as unmanned aerial vehicles, mobile terminals, and onboard satellite processors. Conversely, models that pursue extreme lightweighting often fail to preserve the thin, continuous, linear structure of roads, tending to produce topological breaks in the extracted road networks. To bridge the performance gap between segmentation accuracy and model size, we propose LOA-Net, a lightweight orientation-aware road extraction network. LOA-Net introduces a Road-Aligned Deformable Convolution (RA-DCN) that adaptively aligns the sampling region with the road geometry and explicitly supervises the predicted road orientation, thereby accurately capturing road connectivity while substantially reducing the parameter count. Experiments on the CHN6-CUG and DeepGlobe benchmarks show that LOA-Net surpasses representative state-of-the-art methods on both IoU and F1, while achieving the lowest parameter count of all compared models and a computational complexity comparable to its peers, striking an excellent trade-off between segmentation performance and a mobile-friendly footprint that makes it well suited for road extraction from remote sensing imagery in resource-constrained scenarios. Full article

Wildfires are a major environmental hazard with severe consequences for ecosystems, air quality, infrastructure, and public safety. The rising incidence and severity of wildfire events worldwide have increased the need for reliable early detection and monitoring systems. Remote sensing technologies, such as satellite and unmanned aerial vehicle (UAV) imagery, along with ground-based Closed-Circuit Television (CCTV) cameras, provide valuable geospatial data for large-scale wildfire monitoring. Recent advances in deep learning, particularly Convolutional Neural Networks (CNNs) and Transformer-based architectures, have significantly improved the accuracy of wildfire detection systems. Despite these advances, balancing local feature representation with global contextual modeling remains challenging. CNNs effectively capture local spatial features but have limited receptive fields, whereas Vision Transformers (ViTs) model long-range dependencies but often overlook fine-grained local details and require substantial computational resources. Consequently, accurately detecting small, occluded, and visually ambiguous fire regions remains difficult, particularly for real-time deployment on resource-constrained edge devices. To address these challenges, this study proposes EdgeNeXt-Attn, an enhanced EdgeNeXt-based framework that effectively integrates local feature learning and global contextual modeling through channel and spatial attention mechanisms. The proposed model improves the detection of small, occluded, and visually ambiguous fire regions while maintaining the computational efficiency required for real-time edge deployment. The proposed framework is evaluated on four multi-platform benchmarks spanning ground-based CCTV (DFAN, Complex-Fire), aerial drone (FLAME), and mixed drone–satellite (ADSF) imagery, achieving 92.09%, 95.16%, 96.65%, and 87.81% accuracy, respectively, and outperforming recent state-of-the-art baselines. With only 5.3M parameters, the model achieves real-time inference at 85.9, 27.3, and 8.4 FPS on GPU, CPU, and Raspberry Pi, respectively. Furthermore, ablation studies and Grad-CAM analysis validate its effectiveness and accurate fire localization. These results demonstrate an accurate and computationally efficient framework for real-time wildfire monitoring using multi-platform remote sensing and ground-based imaging systems. Full article

Accurate and reliable positioning is a fundamental requirement for marine applications, particularly in hydrographic surveying, where real-time data acquisition is essential. Despite recent advances in Global Navigation Satellite System (GNSS) technologies, the comparative performance of different real-time correction strategies under marine conditions remains insufficiently addressed. This study presents a comparative performance analysis of three real-time GNSS positioning solutions—Precise Point Positioning–Real-Time Kinematic (PPP-RTK), Network RTK (NRTK), and Single-Baseline RTK (SBL-RTK) with Post-Processed Kinematic (PPK) coordinates adopted as the reference. Performance was assessed using statistical accuracy metrics in accordance with the International Hydrographic Organization (IHO) S-44 standards. At the 95% confidence level, the Total Horizontal Uncertainty (THU) values were 5.6 cm for PPP-RTK and 4.5 cm for NRTK, while the Total Vertical Uncertainty (TVU) values were 3.6 cm and 6.8 cm, respectively. Both NRTK and PPP-RTK solutions met IHO S-44 standards. Under the test conditions considered in this study, the SBL-RTK solution exhibited lower performance than the other solutions, mainly due to the higher proportion of float and stand-alone solution epochs. The findings indicate that the NRTK solution is suitable for near-shore hydrographic surveys where cellular connectivity is available. PPP-RTK represents a reliable alternative for marine applications through satellite-based correction services. This study contributes to the evaluation and selection of appropriate GNSS-based positioning strategies for hydrographic surveying. Full article

High-precision frequency standards underpinning signal synchronization and data consistency are essential for satellites. Navigation satellites use atomic clocks as time and frequency standards; however, limitations in size, weight, and cost constrain their performance. Geostationary Earth orbit (GEO) communication satellites typically rely on onboard oscillators as references, whose performance is generally poor. Moreover, satellite–ground links can be used for frequency synchronization between GEO satellites and ground stations. However, phase drift in onboard oscillators restricts improvements in synchronization precision. This paper introduces a frequency synchronization method based on a beacon transponder that enables ground stations to calculate and eliminate satellite oscillator phase drift. This design achieves high-precision frequency synchronization through real-time carrier-phase compensation and replicates the ground station’s frequency standard onboard. Ground-based experiments validated the method using a beacon transponder prototype in motion. The standard deviation of the phase synchronization error between the replicated clock and ground-station frequency standard was 2.97 ps over 24 h. The stability was better than 2.61 × 10⁻¹² and 9.98 × 10⁻¹⁶ for averaging times of 1 s and 10,000 s, respectively. These results demonstrate this method’s potential for high-precision frequency synchronization between satellites and ground stations, crucial for establishing an integrated space–air–ground frequency synchronization network. Full article

Continuous Global Navigation Satellite System (GNSS) monitoring is essential for characterizing surface subsidence in underground coal-mining areas. However, monitoring-station site selection remains strongly dependent on empirical judgment. Furthermore, steep deformation gradients can cause interferometric synthetic aperture radar (InSAR) decorrelation, phase-unwrapping failure, and data gaps in areas where ground-based monitoring is most needed. This study develops a quantitative, multi-source remote-sensing-assisted framework for GNSS monitoring-station site selection under a short-baseline real-time kinematic (RTK) configuration. Sentinel-1A, Sentinel-2C, unmanned aerial vehicle (UAV) photogrammetric products, and road-network data were integrated to construct eight evaluation factors: normalized difference vegetation index (NDVI), slope, terrain ruggedness index, deformation intensity, cumulative subsidence, InSAR coverage, distance to the subsidence edge, and distance to roads. An analytic hierarchy process (AHP) was used as the primary suitability assessment framework, while a random forest (RF) model was introduced with a limited weight to provide auxiliary information only in InSAR data-sparse areas. Grid-level aggregation, three-component Gaussian mixture model (GMM) screening, minimum-distance thinning, field reconnaissance, and monitoring-network review were subsequently combined to convert the continuous suitability surface into deployable station candidates. The resulting high-suitability areas were concentrated mainly along subsidence margins and near InSAR coverage gaps, reflecting the combined effects of deformation representativeness, supplementary monitoring demand, observation conditions, and engineering accessibility. Sixteen candidate stations were identified, of which S12, S3, and S10 were finally recommended to improve the northern, southwestern, and southeastern coverage of the existing monitoring network, respectively. Sensitivity analyses confirmed that the overall suitability pattern and recommended-station rankings remained stable under moderate parameter perturbations. Comparison with two existing GNSS stations further showed that the station with the higher suitability score exhibited a higher fitted subsidence rate (0.438 versus 0.320 mm day⁻¹). Given the two-station and two-month validation dataset, this agreement represents preliminary consistency evidence rather than statistical proof of general effectiveness. The proposed framework links regional remote-sensing assessment with site-scale engineering review and monitoring-network optimization, providing practical decision support for GNSS deployment in underground mining-subsidence areas. Full article

Climatic changes in the Mediterranean area affect grapevine growth and productivity due to an increase in the magnitude and frequency of extreme climatic events. Functional traits of the vegetative stages of grapevines are drivers of stress tolerance, and their characterization may aid in the selection of cultivars showing greater resistance. Our aim was to evaluate and compare the adaptability to hot and arid climatic conditions of two varieties typical of small satellite islands of Sicily (Zibibbo and Corinto) with the two most common grape varieties grown on the main island (Catarratto and Nero d’Avola). The four selected varieties grew in a common rain-fed experimental germplasm repository. We followed seasonal changes in stomatal conductance, leaf water potential and leaf temperature, relating them to leaf water relation traits and xylem structural characteristics of the leaf vein network. We observed significant differences in functional traits and physiological behaviour of the four cultivars. Leaf water potential at turgor loss point was one of the most informative traits, together with midday leaf water potential and midday stomatal conductance. Nero d’Avola, Catarratto, and Zibibbo showed the highest adaptability to Mediterranean summer–autumn drought and heat stress through different physiological strategies; whereas, Corinto appeared more vulnerable. These findings highlight the importance of functional trait analysis for supporting cultivar selection, vineyard management, and future breeding strategies under climate change scenarios. Full article