Search Results (38,780)

As global climate change accelerates, the urban heat island (UHI) phenomenon has become increasingly pronounced, posing significant challenges to urban energy balance, atmospheric processes, and public health. This study used the Weather Research and Forecasting (WRF) model to dynamically downscale two CMIP6 scenarios—moderate forcing (SSP245) and high forcing (SSP585)—focusing on Zhengzhou, a rapidly urbanizing city in central China. High-resolution simulations captured fine-scale intra-urban temperature patterns and analyze the spatial and seasonal variations in UHI intensity in 2030 and 2060. The results demonstrated significant seasonal variations in UHI effects in Zhengzhou for both 2030 and 2060 under SSP245 and SSP585 scenarios, with the most pronounced warming in summer. Notably, under the SSP245 scenario, elevated autumn temperatures in suburban areas reduced the urban–rural temperature gradient, while intensified rural cooling during winter enhanced the UHI effect. These findings underscore the importance of integrating high-resolution climate modeling into urban planning and developing targeted adaptation strategies based on future UHI patterns to address climate challenges. Full article

Specific Emitter Identification (SEI) is a specialized technique for identifying different emitters by analyzing the unique characteristics embedded in received signals, known as Radio Frequency Fingerprints (RFFs), and SEI plays a crucial role in civilian applications. Recently, various SEI methods based on deep learning have been proposed. However, in real-world scenarios, the scarcity of accurately labeled data poses a significant challenge to these methods, which typically rely on large-scale supervised training. To address this issue, we propose a novel SEI framework based on self-supervised contrastive learning. Our approach comprises two stages: an unsupervised pretraining phase that uses contrastive loss to learn discriminative RFF representations from unlabeled data, and a supervised fine-tuning stage regularized through virtual adversarial training (VAT) to improve generalization under limited labels. This framework enables effective feature learning while mitigating overfitting. To validate the effectiveness of the proposed method, we collected real-world satellite navigation signals using a 40-meter antenna and conducted extensive experiments. The results demonstrate that our approach achieves outstanding SEI performance, significantly outperforming several mainstream SEI methods, thereby highlighting the practical potential of contrastive self-supervised learning in satellite transmitter identification. Full article

The fraction of absorbed photosynthetically active radiation (FAPAR) stands as a pivotal parameter within the Earth system, quantifying the energy exchange between vegetation and solar radiation. Accordingly, there is an urgent need for comprehensive validation studies to accurately quantify uncertainties and improve the reliability of FAPAR-based applications. This study validated five global FAPAR products, MOD15A2H, MYD15A2H, VNP15A2H, GEOV2, and GEOV3, over four boreal forest sites in North America. Qualitative quality flags (QQFs) and quantitative quality indicators (QQIs) of each product were analyzed. Time series high-resolution reference FAPAR maps were developed using the Harmonized Landsat and Sentinel-2 dataset. The reference FAPAR maps revealed a strong agreement with the in situ FAPAR from AmeriFlux (correlation coefficient (R) = 0.91; root mean square error (RMSE) = 0.06). The results revealed that global FAPAR products show similar uncertainties (RMSE: 0.16 ± 0.04) and moderate agreement with the reference FAPAR (R = 0.75 ± 0.10). On average, 34.47 ± 6.91% of the FAPAR data met the goal requirements of the Global Climate Observing System (GCOS), while 54.41 ± 6.89% met the threshold requirements of the GCOS. Deciduous forests perform better than evergreen forests, and the products tend to underestimate the reference data, especially for the beginning and end of growing seasons in evergreen forests. There are no obvious quality differences at different QQFs, and the relative QQI can be used to filter high-quality values. To enhance the regional applicability of global FAPAR products, further algorithm improvements and expanded validation efforts are essential. Full article

UAV platforms equipped with RTK positioning and LiDAR sensors are increasingly used for landslide monitoring, offering frequent, high-resolution surveys with broad spatial coverage. In this study, we applied high-frequency UAV-based monitoring to the active Baldiola earthflow (Northern Apennines, Italy), integrating 10 UAV–LiDAR and photogrammetric surveys, acquired at average intervals of 14 days over a four-month period. UAV-derived orthophotos and DEMs supported displacement analysis through homologous point tracking (HPT), with robotic total station measurements serving as ground-truth data for validation. DEMs were also used for multi-temporal DEM of Difference (DoD) analysis to assess elevation changes and identify depletion and accumulation patterns. Displacement trends derived from HPT showed strong agreement with RTS data in both horizontal (R² = 0.98) and vertical (R² = 0.94) components, with cumulative displacements ranging from 2 m to over 40 m between April and August 2024. DoD analysis further supported the interpretation of slope processes, revealing sector-specific reactivations and material redistribution. UAV-based monitoring provided accurate displacement measurements, operational flexibility, and spatially complete datasets, supporting its use as a reliable and scalable tool for landslide analysis. The results support its potential as a stand-alone solution for both monitoring and emergency response applications. Full article

LiDAR-based Simultaneous Localization and Mapping (SLAM) serves as a fundamental technology for autonomous navigation. However, in complex environments, LiDAR odometry often experience degraded localization accuracy and robustness. This paper proposes a computationally efficient enhancement strategy for LiDAR odometry, which improves system performance by reinforcing high-quality features throughout the optimization process. For non-ground features, the method employs statistical geometric analysis to identify stable points and incorporates a contribution-weighted optimization scheme to strengthen their impact in point-to-plane and point-to-line constraints. In parallel, for ground features, locally stable planar surfaces are fitted to replace discrete point correspondences, enabling more consistent point-to-plane constraint formulation during ground registration. Experimental results on the KITTI and M2DGR datasets demonstrated that the proposed method significantly improves localization accuracy and system robustness, while preserving real-time performance with minimal computational overhead. The performance gains were particularly notable in scenarios dominated by unstructured environments. Full article

In this study, we analyzed the lithospheric seismic background of the Tianshan and adjacent areas by combining various geophysical methods (effective elastic thickness, time-varying gravity, apparent density, and InSAR), and explored the genesis mechanism of the Wushi Ms7.1 earthquake as an example, which led to the following conclusions: (1) The effective elastic thickness (Te) of the Tianshan lithosphere is low (13–28 km) and weak, while the Tarim and Junggar basins have Te > 30 km with high intensity, and the loads are all mainly from the surface (F < 0.5). Earthquakes occur mostly in areas with low values of Te. (2) Medium and strong earthquakes are prone to occur in regions with alternating positive and negative changes in the gravity field during the stage of large-scale reverse adjustment. It is expected that the risk of a moderate-to-strong earthquake occurring again in the vicinity of the survey area between 2025 and 2026 is relatively high. (3) Before the Wushi earthquake, the positive and negative boundaries of the apparent density of the crust at 12 km shifted to be approximately parallel to the seismic fault, and the earthquake was triggered after undergoing a “solidification” process. (4) The Wushi earthquake is a leptokurtic strike-slip backwash type of earthquake; coseismic deformation shows that subsidence occurs in the high-visual-density zone, and vice versa for uplift. The results of this study reveal the lithosphere-conceiving environment of the Tianshan and adjacent areas and provide a basis for regional earthquake monitoring, early warning, and post-disaster disposal. Full article

The Xiong’an New Area, a newly established national-level zone in China, faces the threat of land subsidence and ground fissure due to groundwater overexploitation and geothermal extraction, threatening urban safety. This study integrates time-series InSAR and GNSS monitoring to analyze spatiotemporal deformation patterns from 2017/05 to 2025/03. The key results show: (1) Three subsidence hotspots, namely northern Xiongxian (max. cumulative subsidence: 591 mm; 70 mm/yr), Luzhuang, and Liulizhuang, strongly correlate with geothermal wells and F4/F5 fault zones; (2) GNSS baseline analysis (e.g., XA01-XA02) reveals fissure-induced differential deformation (max. horizontal/vertical rates: 40.04 mm/yr and 19.8 mm/yr); and (3) InSAR–GNSS cross-validation confirms the high consistency of the results (Pearson’s correlation coefficient = 0.86). Subsidence in Xiongxian is driven by geothermal/industrial groundwater use, without any seasonal variations, while Anxin exhibits agricultural pumping-linked seasonal fluctuations. The use of rooftop GNSS stations reduces multipath effects and improves urban monitoring accuracy. The spatiotemporal heterogeneity stems from coupled resource exploitation and tectonic activity. We propose prioritizing rooftop GNSS deployments to enhance east–west deformation monitoring. This framework balances regional and local-scale precision, offering a replicable solution for geological risk assessments in emerging cities. Full article

Vehicle re-identification (Re-ID) is a critical computer vision task that aims to match the same vehicle across spatially distributed cameras, especially in the context of remote sensing imagery. While prior research has primarily focused on Re-ID using remote sensing images captured from similar, typically elevated viewpoints, these settings do not fully reflect complex aerial-ground collaborative remote sensing scenarios. In this work, we introduce a novel and challenging task: aerial-ground cross-view vehicle Re-ID, which involves retrieving vehicles in ground-view image galleries using query images captured from aerial (top-down) perspectives. This task is increasingly relevant due to the integration of drone-based surveillance and ground-level monitoring in multi-source remote sensing systems, yet it poses substantial challenges due to significant appearance variations between aerial and ground views. To support this task, we present AGID (Aerial-Ground Vehicle Re-Identification), the first benchmark dataset specifically designed for aerial-ground cross-view vehicle Re-ID. AGID comprises 20,785 remote sensing images of 834 vehicle identities, collected using drones and fixed ground cameras. We further propose a novel method, Enhanced Self-Correlation Feature Computation (ESFC), which enhances spatial relationships between semantically similar regions and incorporates shape information to improve feature discrimination. Extensive experiments on the AGID dataset and three widely used vehicle Re-ID benchmarks validate the effectiveness of our method, which achieves a Rank-1 accuracy of 69.0% on AGID, surpassing state-of-the-art approaches by 2.1%. Full article

Geographic object-based image analysis (GEOBIA) has demonstrated substantial utility in remote sensing tasks. However, its integration with deep learning remains largely confined to image-level classification. This is primarily due to the irregular shapes and fragmented boundaries of segmented objects, which limit its applicability in semantic segmentation. While convolutional neural networks (CNNs) excel at local feature extraction, they inherently struggle to capture long-range dependencies. In contrast, Transformer-based models are well suited for global context modeling but often lack fine-grained local detail. To overcome these limitations, we propose GOFENet (Geo-Object Feature Enhanced Network)—a hybrid semantic segmentation architecture that effectively fuses object-level priors into deep feature representations. GOFENet employs a dual-encoder design combining CNN and Swin Transformer architectures, enabling multi-scale feature fusion through skip connections to preserve both local and global semantics. An auxiliary branch incorporating cascaded atrous convolutions is introduced to inject information of segmented objects into the learning process. Furthermore, we develop a cross-channel selection module (CSM) for refined channel-wise attention, a feature enhancement module (FEM) to merge global and local representations, and a shallow–deep feature fusion module (SDFM) to integrate pixel- and object-level cues across scales. Experimental results on the GID and LoveDA datasets demonstrate that GOFENet achieves superior segmentation performance, with 66.02% mIoU and 51.92% mIoU, respectively. The model exhibits strong capability in delineating large-scale land cover features, producing sharper object boundaries and reducing classification noise, while preserving the integrity and discriminability of land cover categories. Full article

Massachusetts Bay in the northeastern United States is highly vulnerable to ocean acidification (OA) due to reduced buffering capacity from significant freshwater inputs. We hypothesize that acidification varies across temporal and spatial scales, with short-term variability driven by seasonal biological respiration, precipitation–evaporation balance, and river discharge, and long-term changes linked to global warming and river flux shifts. These patterns arise from complex nonlinear interactions between physical and biogeochemical processes. To investigate OA variability, we applied the Northeast Biogeochemistry and Ecosystem Model (NeBEM), a fully coupled three-dimensional physical–biogeochemical system, to Massachusetts Bay and Boston Harbor. Numerical simulation was performed for 2016. Assimilating satellite-derived sea surface temperature and sea surface height improved NeBEM’s ability to reproduce observed seasonal and spatial variability in stratification, mixing, and circulation. The model accurately simulated seasonal changes in nutrients, chlorophyll-a, dissolved oxygen, and pH. The model results suggest that nearshore areas were consistently more susceptible to OA, especially during winter and spring. Mechanistic analysis revealed contrasting processes between shallow inner and deeper outer bay waters. In the inner bay, partial pressure of pCO₂ (pCO₂) and aragonite saturation (Ω_a) were influenced by sea temperature, dissolved inorganic carbon (DIC), and total alkalinity (TA). TA variability was driven by nitrification and denitrification, while DIC was shaped by advection and net community production (NCP). In the outer bay, pCO₂ was controlled by temperature and DIC, and Ω_a was primarily determined by DIC variability. TA changes were linked to NCP and nitrification–denitrification, with DIC also influenced by air–sea gas exchange. Full article

Multi-modal object detection that fuses RGB (Red-Green-Blue) and infrared (IR) data has emerged as an effective approach for addressing challenging visual conditions such as low illumination, occlusion, and adverse weather. However, most existing multi-modal detectors prioritize accuracy while neglecting computational efficiency, making them unsuitable for deployment on resource-constrained edge devices. To address this limitation, we propose PONet, a lightweight and efficient multi-modal vehicle detection network tailored for real-time edge inference. PONet incorporates Polarized Self-Attention to improve feature adaptability and representation with minimal computational overhead. In addition, a novel fusion module is introduced to effectively integrate RGB and IR modalities while preserving efficiency. Experimental results on the VEDAI dataset demonstrate that PONet achieves a competitive detection accuracy of 82.2% mAP@0.5 while sustaining a throughput of 34 FPS on the OrangePi AIpro 20T device. With only 3.76 M parameters and 10.2 GFLOPs (Giga Floating Point Operations), PONet offers a practical solution for edge-oriented remote sensing applications requiring a balance between detection precision and computational cost. Full article

Underground blasting vibrations are a critical factor influencing the stability of mine slopes. However, existing studies have yet to establish a quantitative relationship or clarify the underlying mechanisms linking blasting-induced vibrations and slope deformation. Taking the Shilu Iron Mine as a case study, this research develops a dynamic mechanical response model of slope stability that accounts for blasting loads. By integrating slope radar remote sensing data and applying the Pearson correlation coefficient, this study quantitatively evaluates—for the first time—the correlation between underground blasting activity and slope surface deformation. The results reveal that blasting vibrations are characterized by typical short-duration, high-amplitude pulse patterns, with horizontal shear stress identified as the primary trigger for slope shear failure. Both elevation and lithological conditions significantly influence the intensity of vibration responses: high-elevation areas and structurally loose rock masses exhibit greater dynamic sensitivity. A pronounced lag effect in slope deformation was observed following blasting, with cumulative displacements increasing by 10.13% and 34.06% at one and six hours post-blasting, respectively, showing a progressive intensification over time. Mechanistically, the impact of blasting on slope stability operates through three interrelated processes: abrupt perturbations in the stress environment, stress redistribution due to rock mass deformation, and the long-term accumulation of fatigue-induced damage. This integrated approach provides new insights into slope behavior under blasting disturbances and offers valuable guidance for slope stability assessment and hazard mitigation. Full article

Accurate estimation of Interrupted Sampling Repeater Jamming (ISRJ) sampling duration is essential for effective radar anti-jamming. However, in complex electromagnetic environments, the simultaneous presence of suppressive and deceptive jamming, coupled with significant signal overlap in the time–frequency domain, renders ISRJ separation and parameter estimation considerably challenging. To address this challenge, this paper proposes a method utilizing the Fractional Fourier Transform (FRFT) and a Hybrid Modality Fusion Network (HMFN) for ISRJ signal separation and sampling-duration estimation. The proposed method first employs FRFT and a time–frequency mask to separate the ISRJ and target echo from the mixed signal. This process effectively suppresses interference and extracts the ISRJ signal. Subsequently, an HMFN is employed for high-precision estimation of the ISRJ sampling duration, offering crucial parameter support for active electromagnetic countermeasures. Simulation results validate the performance of the proposed method. Specifically, even under strong interference conditions with a Signal-to-Jamming Ratio (SJR) of −5 dB for deceptive jamming and as low as −10 dB for suppressive jamming, the regression model’s coefficient of determination still reaches 0.91. This result clearly demonstrates the method’s robustness and effectiveness in complex electromagnetic environments. Full article

The NW–SE-trending Firenze-Pistoia Basin (FPB) is an intermontane tectonic depression in the Northern Apennines (Italy) bounded to the northeast by a SW-dipping normal fault system. Although it has moderate historical seismicity (maximum estimated Mw 5.5 in 1895), the FPB lacks detailed characterization of its recent tectonic structures, unlike those of nearby basins that have produced Mw > 6 events. This study focuses on the southeastern sector of the basin, including the urban area of Florence, using tectonic geomorphology derived from remote sensing, in particular LiDAR data, field verification, and high-resolution geophysical surveys such as electrical resistivity tomography and seismic reflection profiles. The integration of these techniques enabled interpretation of the subdued and anthropogenically masked tectonic structures, allowing the identification of Holocene activity and significant, although limited, surface vertical offset for three NE–SW-striking normal faults, the Peretola, Scandicci, and Maiano faults. The Scandicci and Maiano faults appear to segment the southeasternmost strand of the master fault of the FPB, the Fiesole Fault, which now shows activity only along isolated segments and cannot be considered a continuous active fault. From empirical relationships, the Scandicci Fault, the most relevant among the three active faults, ~9 km long within the basin and with an approximate Late Quaternary slip rate of ~0.2 mm/year, might source Mw > 5.5 earthquakes. These findings highlight the need to reassess the local seismic hazard for more informed urban planning and for better preservation of the cultural and architectural heritage of Florence and the other artistic towns located in the FPB. Full article

Unsupervised hyperspectral image (HSI) clustering is a fundamental yet challenging task due to high dimensionality and complex spectral–spatial characteristics. In this paper, we propose a novel and efficient clustering framework centered on adaptive and diverse anchor graph modeling. First, we introduce a parameter-free construction strategy that employs Entropy Rate Superpixel (ERS) segmentation to generate multiple anchor graphs of varying sizes from a single HSI, overcoming the limitation of fixed anchor quantities and enhancing structural expressiveness. Second, we propose an anchor-to-pixel label propagation mechanism to transfer anchor-level cluster labels back to the pixel level, reinforcing spatial coherence and spectral discriminability. Third, we perform clustering directly at the anchor level, which substantially reduces computational cost while retaining structure-aware accuracy. Extensive experiments on three benchmark datasets (Trento, Salinas, and Pavia Center) demonstrate the effectiveness and efficiency of our approach. Full article