Atmospheric and Oceanic Sciences

High-resolution multispectral images are of great value in various fields. However, the physical limitations of satellite sensors hinder the simultaneous acquisition of high spatial resolution and high spectral resolution images. Deep learning has become a powerful tool for remote sensing image fusion, but its full potential has not been fully utilized. In order to maximize the quality of the high-resolution multispectral images generated by the improved model, this paper proposes a module called EBIFusion, which introduces an error backtracking mechanism to improve fusion performance. The module uses the intermediate results of the deep learning model to capture the information lost during the high-resolution image generation process, thereby guiding the optimization of model training. The experimental results on the GF-6 dataset show that the QNR index is increased by 2.22% after the introduction of the module. In addition, the spatial and spectral quality has been continuously improved on multiple datasets, including QuickBird and GF-2. The optimized models, such as PSGAN, PNN, GPPNN, and UCGAN, show stronger spatial details and spectral fidelity. The main contribution of this paper is to propose an error backtracking framework for recovering and compensating for the lost information in the process of remote sensing image fusion based on deep learning. Based on this, a lightweight and model-independent enhancement module EBIFusion is designed, which can be integrated into different deep learning fusion architectures. At the same time, the generalization ability of the module in multiple datasets and network paradigms is verified. In summary, the error backtracking module enhances the quality of the generated high-resolution multispectral images. In addition, it is not limited to specific models and data and can be used as the basis for a versatile and effective optimization component to improve the availability of high-resolution multispectral images.

Large-scale 3D reconstruction from UAV imagery is essential for remote sensing and environmental monitoring, yet modern neural rendering methods such as 3D Gaussian Splatting (3DGS) are computationally intensive due to redundant views. Reducing input images while preserving reconstruction accuracy and completeness remains a practical challenge. We present a scalable image subset selection framework that integrates transformer-based visual place recognition (PairVPR) with a facility-location selection strategy to identify and remove redundant views while maintaining spatial coverage. We first validate the approach on six standard benchmark scenes (five Mip-NeRF360 scenes and the Tanks&Temples Truck scene): under a fixed image budget, PairVPR-based selection attains the best average PSNR, SSIM, and LPIPS among all evaluated VPR selectors. Building on this, we present a case study on a real-world glacier UAV survey of 589 images, in which pruning to a compact subset reduces overall runtime by more than half at the most aggressive setting, while incurring only a modest loss in reconstruction quality, and in which PairVPR-based selection again attains the best reconstruction quality among the evaluated VPR selectors. By combining learned similarity with a coverage-driven selection process, the method operates with linear memory complexity and adapts to large image collections, offering a practical solution for large-scale geospatial mapping and monitoring workflows.

Closing the nutrient loop between food consumption and crop production is essential for keeping nitrogen (N) and phosphorus (P) flows within safe planetary boundaries. Separately collecting and treating sanitary streams to produce human excreta-derived fertilizers (HEDF), such as nitrified urine fertilizer (NUF) and fecal compost (FC), could partly replace fossil-based mineral fertilizers while reducing nutrient pollution from N and P discharges to aquatic systems. However, broader HEDF adoption is constrained by limited knowledge of crop-specific performance and the fate of potential contaminants. This study evaluated the combined use of NUF from source-separated urine and FC from dry toilet contents for greenhouse cultivation of kohlrabi ( Brassica oleracea var. gongylodes L.) in comparison with a synthetic multi-nutrient mineral fertilizer. In addition to crop performance, plant N uptake, soil N dynamics, and the fate of doxycycline as an indicator pharmaceutical were assessed to provide an integrated evaluation of agronomic and environmental implications. HEDF and mineral fertilizer treatments produced comparable yields, with no significant differences in leaf, bulb, or root biomass. However, total plant N uptake was 13% higher under HEDF fertilization, driven by increased N contents in leaves (+30%), bulbs (+29%), and roots (+13%). At harvest, soil mineral N concentrations were on average five times higher in the mineral fertilizer treatment than in the HEDF treatment, whereas soil organic N concentrations showed an opposite trend and were approximately 25% lower under mineral fertilization. Doxycycline, detected in the applied FC, remained below the limit of quantitation in both kohlrabi bulbs and soil at harvest. These findings indicate that the combined application of NUF and FC can effectively substitute mineral fertilizer in kohlrabi cultivation. The results further suggest that HEDF application may modify short-term soil N cycling by increasing N immobilization and reducing soil mineral N accumulation compared with mineral fertilization. Long-term studies are needed to evaluate the effects of repeated FC application on soil organic carbon, N leaching, and potential contaminant uptake by crops.

Abstract Quartz grain surface microtextures serve as robust records of sedimentary history and depositional environment, yet a coherent framework quantitatively linking specific microtextures to stream power remains underexplored. This study integrates scanning electron microscopy (SEM) with quantitative hydrodynamic analysis along the Kinu River, Japan, to elucidate the sensitivity of quartz microtextures to unit stream power ( ω ). Analysis reveals that microtexture frequencies vary distinctly along the river, yet correlations with unit stream power ( ω ) are insignificant at the overall catchment scale. However, within low‐gradient reaches ( S < 0.001), we identified a strong positive correlation between unit stream power ( ω ) and the frequency of crescentic percussion marks. As crescentic percussion marks require high‐magnitude impacts to fracture the crystal lattice, they function as sensitive, high‐threshold indicators of local collisional energy. In contrast, chemical microtextures show no significant relationship with the unit stream power ( ω ) in the Kinu River data set, consistent with previous interpretations that they primarily reflect cumulative residence in low‐energy or organic‐rich zones rather than instantaneous flow intensity. Furthermore, comparative analysis between a channel deposit and an adjacent dune deposit formed directly by wind action suggests that warm‐humid riverine dunes can retain strong microtextural inheritance from their fluvial source, distinguishable mainly by incipient aeolian overprinting (e.g., bulbous edges and flat cleavage surfaces). Ultimately, building upon qualitative discrimination, this study establishes a framework relating grain‐scale surface microtextures to fluvial energy heterogeneity.

Abstract Coronal mass ejections (CMEs) are among the key solar eruptive activities, triggering space weather disturbances. Thus, forecasting their geoeffectiveness has become a research focus. This study constructs a model to recommend similar events for forecasting the geoeffectiveness of CMEs. The input parameters are optimized via feature dimensionality reduction, while the cosine similarity algorithm, combined with logistic regression, is employed to perform four tasks: binary classification of whether a CME will reach Earth, prediction of travel time, assessment of the intensity of geomagnetic disturbances, and matching of historical similar events. Experimental results show that the model achieves an F1‐score of 0.43 for binary classification, with a mean absolute error (MAE) of 13.75 hr for travel time prediction and a MAE of 1.68 for the maximum Kp value ( Kp max) on the test set. The findings indicate that the model not only achieves staged multi‐task forecasting for the geoeffectiveness of CMEs but also enhances the interpretability of forecasting results through similar event recommendation, providing pragmatic decision support for space weather warnings.

Despite growing concern over per- and polyfluoroalkyl substances (PFAS), evidence for packaging-driven dietary exposure under hot-contact scenarios remains limited, especially in China’s rapidly expanding takeout market. Here, we conducted a survey of fast foods ( n = 150) and corresponding food-contact materials (FCMs, n = 150) across Chinese markets, quantifying 32 legacy and emerging PFAS. PFAS were detected in all foods (range: 0.01–79.4 ng/g; median: 3.45 ng/g) and FCMs (0.01–205 ng/g; 3.88 ng/g). Paper-FCMs showed the highest PFAS levels among different FCM categories and exhibited industry-specific patterns. Food PFAS profiles mirrored those of paired FCMs, with significant correlations for several dominant congeners ( p < 0.05), indicating migration contributions. Moreover, migration patterns were jointly influenced by PFAS physicochemical properties (e.g., log K OW, functional moieties), FCM texture, and food composition. Clear differences between leading and smaller brands were observed, with smaller brands exhibiting higher PFAS burdens in fast foods ( p < 0.05). Notably higher PFAS concentrations were identified in soup-based instant foods. A controlled experiment demonstrated 3.1–26-fold higher PFAS concentrations in instant foods prepared in original paper packaging than in glass containers, largely driven by 6:2 FTS, which accounted for 88.3% of the total ∑ 32 PFAS increment. Integrated multifactor prioritization revealed several emerging PFAS of particular concern, especially given their expanding use in FCMs. These findings highlight hot-contact packaging as a non-negligible exposure source and call for enhanced surveillance of emerging PFAS in fast-food supply chains.

Trihalomethanes (THMs) are suspected neurotoxicants, yet their relationships with depression remain unclear. This study examined the associations between blood THM concentrations and depressive symptoms, assessed by the Patient Health Questionnaire-9 (PHQ-9), in 2,130 postmenopausal U.S. women. Blood concentrations of bromodichloromethane (BDCM), dibromochloromethane (DBCM), and brominated trihalomethanes (Br-THMs) showed positive dose–response associations with depressive symptoms (PHQ-9 ≥ 5; all P for trend <0.05). Compared with the lowest exposure category, participants in the highest category of BDCM (T3), DBCM (≥75th percentile), and Br-THMs (Q4) had increased odds ratios (ORs) for depressive symptoms of 1.38 (95% confidence interval: 1.04–1.82), 1.62 (1.25–2.09), and 1.46 (1.02–2.09), respectively. Further mechanistic experiments in a human neuroimmune organ-on-a-chip model (SH-SY5Y cells cocultured with THP-1 cells) showed that environmentally relevant exposure (0.001 mM BDCM and 0.0005 mM DBCM) increased interleukin-1 beta (IL-1β) levels and reduced 5-hydroxytryptamine (5-HT) levels. Low-dose BDCM exposure (0.001 mM) also induced neuronal cytoskeletal changes in SH-SY5Y cells, as reflected by changes in mean fluorescence intensity and skeletonized area. Together, our findings suggest that exposure to Br-THMs may be associated with depressive symptoms in postmenopausal women, potentially via IL-1β-mediated neuroinflammation and neurotransmitter imbalance.

Ultra-shallow and confined water environments are challenging to survey with conventional towed side-scan sonar (SSS) due to limited access and positioning uncertainties. This study introduces a portable, battery-powered acoustic survey system that integrates a pole-mounted dual-frequency side-scan sonar (600/1600 kHz) with RTK GNSS (Real-Time Kinematic Global Navigation Satellite System), deployable from a small inflatable boat. The system was validated in two settings: an inland lake and a marina. Field trials demonstrated reliable acquisition of high-resolution sonar imagery and effective detection of both natural and anthropogenic seabed features, including small and low-reflectivity objects. The high-frequency channel (1600 kHz) produced superior image quality and interpretability compared to the lower frequency. While there are limitations associated with fixed sonar mounting and limited altitude control, the system offers high mobility, rapid deployment, and operational safety. This approach represents a practical, cost-effective solution for high-resolution acoustic remote sensing in ultra-shallow water settings where traditional survey methods are ineffective or impractical.

Abstract The Gaggiano High, buried beneath the sediments of the Western Po Plain (northern Italy), represents one of the few examples of Mesozoic structural highs largely preserved within a foreland–foredeep system. Unlike the adjacent Lacchiarella Basin, which underwent strong positive inversion during the Neogene, the Gaggiano High experienced only mild contractional reactivation, allowing its original extensional framework to be preserved. By integrating 3D seismic interpretation, isochore maps, and well data, we reconstruct the tectono‐stratigraphic evolution of the Gaggiano High from Middle Triassic rifting to Neogene burial. Isochore maps and stratigraphic geometries reveal major thickness variations and westward tilting along N‐S faults during Early to Middle Jurassic extension, followed by long‐lasting structural uplift and reduced sedimentation during Jurassic‐Cretaceous times. In the Paleogene, renewed extension along NW–SE faults produced local depocentres within the Scaglia Fm., while subsequent Oligo‐Miocene compression resulted in gentle folding, minor inversion of inherited normal faults, and regional tilting. Comparison with the exhumed Lugano Swell highlights the long‐term persistence of rift‐related highs as mechanically rigid domains within evolving foreland basins. The Gaggiano High thus provides a well‐preserved subsurface analogue illustrating how inherited normal faults and long‐lived structural highs may constrain, modulate, or even inhibit subsequent deformation, stratigraphic architecture and basin evolution across the Alpine‐Apennine system.

Abstract In the northwestern Gulf of Mexico, gravity‐driven contractional deformation is recorded across the Salina del Bravo, the Perdido and Peripheral fold belts, and the Mexican Cordilleras. Integrating seismic interpretation, structural mapping, and analysis, we examine the spatial and temporal partitioning of shortening above three principal detachments: the Louann Salt, lower Paleocene shales, and middle Eocene to lower Oligocene shales. The Louann Salt is the master detachment in most of the area. Louann‐detached deformation transitions from fold‐dominated in the northern Perdido Fold Belt to thrust‐dominated southward and beneath the Salina del Bravo, becoming absent in the Mexican Cordilleras. The lower Paleocene detachment is locally restricted, acting as a minor intermediate flat above the Louann level. The mid‐Eocene to lower Oligocene shales serve varying structural roles: they underlie a proximal perched fold belt and provide distal decoupling in the Salina del Bravo, constitute the upper flat to deep thrusts in the Peripheral Fold Belt, and represent the primary detachment for gravity‐linked systems in the Mexican Cordilleras. Deformation propagated basinward and southward in three distinct phases: (a) pre–mid Eocene Louann‐detached shortening that generated salt‐cored folds, thrusts, and incipient salt sheets; (b) Oligocene strain partitioning between the Louann Salt and overpressured shale decoupling levels; and (c) Neogene reactivation of ramp‐flat configurations and development of the Mexican Cordilleras.

Abstract Reconstructing the deformation and topographic growth history of Southeast Tibet is critical for understanding continental lithospheric dynamics driving Tibetan Plateau development. However, the spatiotemporal pattern of plateau growth and underlying mechanisms remain debated. Here, we investigate the northern Jinsha River fold‐and‐thrust belt, an inherited structure of the Jinsha suture zone, using low‐temperature thermochronology and thermal‐history modeling. Our results reveal three episodes of accelerated exhumation: during the Late Cretaceous (ca. 80–68 Ma), Late Eocene (ca. 38–34 Ma), and since the Early Miocene (ca. 19 Ma). The Late‐Cretaceous phase is contemporaneous with syn‐tectonic sedimentation in the Gonjo Basin, suggesting a coupled tectono‐sedimentary system. A regional synthesis of thermochronology data reveals widespread Late‐Cretaceous to Early‐Cenozoic rapid exhumation in eastern and central Tibet, highlighting the role of Neo‐Tethyan subduction and initial India‐Asia collision in shaping early topographic relief. The Late‐Eocene pulse coincides with magmatism along the reactivated Jinsha suture zone and regional uplift, reflecting a major compressional deformation event. Post‐19 Ma exhumation at rates of 150–300 m/Myr is likely linked to Jinsha River incision, although age patterns across the Zigasi‐Deqin fault suggest that thrusting also contributed to Early–Middle Miocene exhumation. Integrating our results with published records of deformation, we identify southeastward propagation of compressional deformation since the Late Eocene in northern Southeast Tibet, supporting stepwise plateau growth through crustal shortening. We also infer a possible second phase of southeastward‐propagating deformation since the Early Miocene, although this interpretation remains uncertain. Our multi‐stage tectonic model provides new insights into the stepwise plateau growth of Southeast Tibet.

Abstract The NW‐SE‐oriented Eastern Talesh and Western Alborz Mountains of the Arabia‐Eurasia collision zone have constituted an efficient orographic barrier to northerly Caspian Sea moisture transport since the Early–Middle Miocene, resulting in pronounced rainfall and surface‐process gradients. Nevertheless, low‐temperature thermochronology indicates disparate orogen‐parallel and orogen‐perpendicular patterns of long‐term exhumation that do not follow the expected climatic gradients. West of the Lahijan lineament (LL), a suspected transverse NNE‐SSW‐striking, left‐lateral strike‐slip fault, south‐vergent thrusting dominates with &lt;4 km of exhumation across the wetter northern flanks and &gt;4 km across the arid southern flank. This reverse behavior with respect to the precipitation gradient, suggests that orographic precipitation alone cannot lead to the growth of new tectonic structures. Conversely, east of the LL, north‐vergent thrusting and superposed erosional processes resulted in &gt;4 km of exhumation along the wetter northern orogenic flank. There, the lack of outward orogenic growth, together with protracted Miocene thrusting, suggests a balance between erosion and uplift. This structural setting (north‐dipping faults to the west and inward‐dipping faults to the east) may be inherited from the opening of the South Caspian Basin. Importantly, the structural boundary between the two areas (LL) has been exploited by deep incision and headward erosion by the Sefid Rud river. Thermo‐kinematic Pecube modeling supports a causal link notion between the formation of the Sefid Rud valley and base‐level fall in the Caspian Sea (∼5.5–3.2 Ma). Sea‐level lowering promoted fluvial incision that breached the Alborz‐Talesh range and integrated upstream basins into the regional drainage network.