TerraMosaic Daily Digest: September 6, 2026
Daily Summary
Three studies sharpen how high-mountain cascades should be reconstructed and prioritized. For the 26 August Gyirong mixed rock-ice disaster, multisensor imagery constrains a preferred changed source area of 1.009 km², a 21.843 km route descending about 3396 m, and a 37.353 km² downstream disturbance footprint. The analysis finds no unambiguous large precursor and no exceptional precipitation signal in five tested windows; instead, the preceding seven-day temperature and positive-degree-day totals exceed all 25 matched years, providing thermal context without proving causation. In Yunnan, an auditable knowledge graph reduces 10,842 mapped landslides to 193 river-blockage assessment units while preserving eligibility rules, provenance and interpretation limits. Complementary L-band and C-band InSAR tests in Hanyuan show why sensor physics matters: Lutan-1 maintains substantially higher coherence and valid coverage under dense vegetation, identifying 74 of 77 interpreted potential hazards, whereas Sentinel-1 contributes denser temporal verification.
Process-resolving studies move from susceptibility toward coupled consequences. Deb2L explicitly represents erosion, entrainment and deposition in debris-flow-generated reservoir waves; analytical and laboratory tests achieve R² above 0.93 and RMSE of 0.004-0.06 m, and coupling with TiVaSS demonstrates scenario transfer to the 2020 Sanyang Reservoir event. A separate Eulerian multiphase formulation addresses waves generated by bidisperse granular landslides, while the Luding study extends analysis to a watershed-scale earthquake-triggered landslide-debris-flow chain. In the Xining Basin, root reinforcement is demonstrably nonlinear across scales: multiroot strength falls 14.86-48.13% below predictions from isolated roots, yet root-soil cohesion rises by 78.34-157.99% across three herbaceous species. Yigong paleoflood deposits record three Holocene outbursts with reconstructed peak discharges of 2.3-4.4 × 10⁵ m³ s⁻¹, placing modern dam-break risk within a much longer history of catastrophic river reorganization.
The wider literature emphasizes uncertainty that enters through forcing, observation and decision design. The westerly stratospheric QBO nearly doubles tropical-cyclone extreme precipitation along southeastern China by reorganizing BSISO2, adding seasonal predictive information beyond ENSO. Conversely, a Bay of Bengal emulator performs worse when prescribed cyclone tracks appear on only 7.9% of training days; removing the storm map at inference improves held-out forecasts by 7.5-16.4%. Flood-mapping research evaluates whether GeoAI explanations agree with spectral domain knowledge, while city-scale ground-motion mapping uses building response sensors to recover spatial shaking fields. New lake, peatland, forest-disturbance and isotope data products expand environmental baselines. Across remote sensing, hydrology and infrastructure, the common advance is not a larger model alone but an explicit account of when observations are missing, when priors fail to transfer and which uncertainties alter decisions.
Key Trends
The dominant methodological direction is to preserve the physical and observational links that turn an initiating event into a cascading consequence.
- Cascade reconstruction is becoming uncertainty-explicit: Gyirong is reconstructed as a source-pathway-receptor system with bounded source area, route, thermal context and mapped exposure, while unresolved volume and initiation mechanics remain explicit. Luding, reservoir impulsive waves and Yigong paleofloods extend this approach across earthquake, debris-flow and dam-break cascades.
- Landslide inventories are being converted into auditable decisions: Knowledge-graph screening links terrain, waterways, material supply and provenance to inspection priority rather than claiming blockage probability. Lutan-1/Sentinel-1 comparison similarly separates visibility, coherence and temporal sampling, clarifying what each sensor can support in vegetated mountains.
- Erosion, entrainment and material interaction are entering operational models: Deb2L shows that omitting erosion and entrainment understates debris-flow wave magnitude. Root tests reject linear scaling from single roots to root networks, and fault-slip coalburst experiments connect segmented motion, static concentration and dynamic loading. These studies replace additive approximations with interacting process descriptions.
- Prediction is being judged against transfer and decision performance: Cyclone-track conditioning fails when a rare prescribed input activates outside its training distribution, whereas QBO-BSISO2 coupling adds physically interpretable seasonal information. Related studies examine forecast value, model shortcuts, uncertainty-aware regression and the alignment of GeoAI explanations with domain knowledge.
- Risk analysis is coupling physical recovery with social and network constraints: Mountain-road recovery integrates topology with time-cost trade-offs, railway analysis adds fairness to flood resilience, and seismic infrastructure studies combine structural mitigation, financial transfer and building-scale sensing. Compound-event and hydrosocial perspectives broaden the decision boundary beyond a single hazard metric.
Selected Papers
The collection is anchored by direct studies of rock-ice and landslide-debris-flow cascades, river blockage, vegetated-slope detection, impulsive waves, paleofloods and earthquake risk. A second group advances physically constrained remote sensing, uncertainty-aware forecasting, infrastructure resilience and reusable environmental data. Together, the papers show that credible geohazard inference depends on retaining process coupling, sensor limitations and decision context from observation through action.
1. When a high-mountain slope failure cascades downstream: reconstructing the 26 August 2026 Gyirong mixed rock-ice disaster
Core Problem: Rapidly constraining the source, route, environmental context and exposed infrastructure of a transboundary rock-ice cascade despite sparse pre-event observations.
Key Innovation: Integrates Sentinel-1/2, Landsat-9, PlanetScope, terrain analysis and rapid mapping in an uncertainty-explicit source-pathway-receptor reconstruction, while separating exceptional thermal context from an unproven trigger.
2. Development and application of a two-dimensional numerical model for debris flow-induced impulsive wave considering debris flow erosion-entrainment process
Core Problem: Reservoir-wave models can underestimate compound hazard when debris-flow erosion and entrainment are omitted from the mass and momentum balance.
Key Innovation: Deb2L couples two-layer shallow-water equations with erosion, entrainment and deposition, validates against analytical and laboratory tests, and links TiVaSS landslide prediction to field-scale wave scenarios.
3. Auditable Knowledge-Graph Screening of Landslides for River Blockage and Dammed-Lake Assessment
Core Problem: Large landslide inventories require defensible prioritization for river-blockage inspection without presenting a heuristic ranking as event probability.
Key Innovation: A mechanism-oriented knowledge graph preserves eligibility, provenance and score components while reducing 10,842 polygons to 193 auditable assessment units.
4. Applicability Assessment of Lutan-1 and Sentinel-1 for Potential Landslide Identification in Densely Vegetated Mountainous Areas: A Case Study of Hanyuan County, Sichuan Province, China
Core Problem: Vegetation decorrelation and radar geometry limit the reliability of regional InSAR screening for slow slope deformation.
Key Innovation: A unified visibility-coherence-validation framework shows Lutan-1 retaining much greater valid coverage than Sentinel-1 while their joint use balances spatial detection and temporal verification.
5. Mechanism of coalburst triggered by thrust fault slipping based on field monitoring, indoor experiments and numerical simulation
Core Problem: Mining-induced segmented fault slip can concentrate static stress and generate dynamic disturbances, but the causal sequence leading to roadway failure is poorly resolved.
Key Innovation: Links opposed slip of upper and lower fault segments to a clamping effect, then uses hollow-specimen static-dynamic tests to connect concentrated stress, reflected tensile waves and coalburst damage.
6. Multiscale mechanical characteristics and slope protection effects of herbaceous roots in the loess region of the Qinghai-Tibet Plateau
Core Problem: Slope-protection design cannot infer root-network reinforcement reliably by linearly scaling single-root tensile properties.
Key Innovation: Combines single-root, multiroot and root-soil direct-shear tests for three grasses, quantifying nonlinear root interaction and species-specific cohesion gains of 78.34-157.99%.
7. Outburst floods records in the Yigong River since holocene, southeastern Tibet
Core Problem: Prehistoric outburst frequency and magnitude in the Yigong River remain poorly constrained by studies focused on the 1902 and 2000 events.
Key Innovation: Integrates valley-fill, giant-bar and slackwater deposits with chronology and hydraulic scenarios to identify three Holocene floods and reconstruct peak discharges up to 4.4 × 10⁵ m³ s⁻¹.
8. Comparative seismic hazard assessment using PSHA and AHP-based susceptibility mapping in northeastern to southeastern Bangladesh
Core Problem: Regional seismic planning requires reconciling probabilistic hazard estimates with spatial susceptibility indicators across data-variable terrain.
Key Innovation: Compares PSHA with AHP-based susceptibility mapping from northeastern to southeastern Bangladesh to expose how methodological choice changes the mapped hazard picture.
9. Earthquake risk assessment of the city of Zagreb, Croatia: recent advances
Core Problem: City-scale risk management needs an updated representation of exposure, vulnerability and shaking across Zagreb.
Key Innovation: The study consolidates recent advances in Zagreb earthquake-risk assessment into a city-scale evidence base for mitigation and recovery planning.
10. Watershed-scale high-position landslide-debris flow hazard chain triggered by the Luding Earthquake: Evolution mechanism and dynamic processes
Core Problem: Earthquake-triggered high-position failures can transform into long-runout debris flows whose watershed-scale evolution is not captured by isolated slope analysis.
Key Innovation: The verified record combines evolution-mechanism and dynamic-process analysis for the Luding hazard chain; the accessible publisher record does not expose quantitative results.
11. Glacial lake changes and outburst flood risk assessment in the North American Coast Mountains
Core Problem: Rapidly changing glacial lakes in the North American Coast Mountains require consistent mapping and downstream outburst-risk evaluation.
Key Innovation: The verified record links multitemporal lake change with GLOF risk assessment across a regional mountain system; detailed scenarios and results are not exposed in the accessible abstract.
12. A New National Scale Inventory of Natural Hazards and Transport Disruptions in Taiwan Derived from News Media Reports
Core Problem: National transport-risk analysis lacks consistent event records that connect hazard occurrence with documented network disruption.
Key Innovation: Derives a Taiwan-scale hazard and transport-disruption inventory from news media reports, creating an event evidence base for cross-hazard infrastructure analysis.
13. Assessment of Resilience Recovery Strategies for Mountainous Road Networks under Multiple Landslides Considering Time-Cost Trade-Off
Core Problem: Post-landslide road recovery is often sequenced without coupling restored topology to repair time and cost under constrained resources.
Key Innovation: Integrates complex-network resilience with a time-cost trade-off mechanism; the Fengjie case reaches Re = 0.9777 by prioritizing bottlenecks and strategically scheduling repairs.
14. Eulerian multiphase modeling of landslide-generated waves from bidisperse granular mixtures
Core Problem: Wave generation by compositionally heterogeneous granular landslides requires resolving interactions between solid size classes and water.
Key Innovation: Applies an Eulerian multiphase formulation to bidisperse granular mixtures, extending landslide-wave simulation beyond single-phase or monodisperse assumptions.
15. Stratospheric Quasi-Biennial Oscillation Modulates Tropical Cyclone Extreme Rainfall Over East Asia via Intraseasonal Oscillation
Core Problem: Seasonal prediction of cyclone extreme rainfall lacks a resolved pathway linking stratospheric variability to regional storm rainfall.
Key Innovation: Identifies BSISO2 as the mediator through which the QBO reorganizes cyclone activity and nearly doubles extreme precipitation along southeastern China during the westerly phase.
16. Resilience assessment method for deepwater moored platform-riser coupling system under typhoon conditions
Core Problem: Platform and riser responses interact under typhoon loading, complicating system-level resilience assessment.
Key Innovation: The verified record evaluates the coupled moored-platform-riser system rather than isolated components; the accessible source does not expose response metrics.
17. Evaluating the Alignment Between GeoAI Explanations and Domain Knowledge in Satellite-Based Flood Mapping
Core Problem: Accurate flood maps remain difficult to trust operationally when model explanations are not tested against established spectral knowledge.
Key Innovation: ADAGE groups channels for Shapley attribution and scores agreement with reference domain explanations, revealing explanation differences even among models trained with the same data and configuration.
18. Invited perspectives: reframing transboundary flood vulnerability through “Hydrosocial Connectivity” for just and adaptive governance
Core Problem: Basin-wide vulnerability persists when data sharing, formal agreements and local knowledge remain disconnected from the movement of water.
Key Innovation: Hydrosocial connectivity links physical flows, social relations and governance to center justice, participation and adaptive coordination in the Ganges-Brahmaputra Basin.
19. Invited perspectives: Towards usable compound event research
Core Problem: Compound-event science can remain difficult to use when research questions and outputs are weakly connected to real decisions.
Key Innovation: Synthesizes pathways for bridging compound-event modeling and forecasting with operational applications and stakeholder needs.
20. Seismic performance of RC frames with burnt and unburnt brickwork masonry infills: experimental and macro-model evaluation
Core Problem: Masonry type changes stiffness loss, strength degradation and energy dissipation, yet common macro-models may not reproduce these differences reliably.
Key Innovation: Compares burnt-brick and unburnt-sandcrete infills experimentally and numerically, exposing a capacity-ductility trade-off and limitations in empirical macro-model calibration.
21. Towards resilient precast concrete frame featuring a hybrid system with replaceable ductile steel fuses
Core Problem: Brittle precast beam-column connections concentrate damage and provide limited ductility during earthquakes.
Key Innovation: Confines yielding to a bolted replaceable I-shaped steel fuse inside a jacketed joint, protecting primary members while maintaining stable hysteretic response.
22. Ground motion duration effects on the response of buildings experiencing earthquake-induced pounding
Core Problem: Duration effects on adjacent buildings may depend strongly on their period mismatch but are rarely isolated from amplitude effects.
Key Innovation: Tests 77 long- and short-duration motion ensembles and identifies a period-ratio threshold beyond which short motions sharply amplify acceleration, displacement, base shear and pounding force.
23. City-scale ground motion mapping from building-embedded structural response sensors using building-specific hybrid deep learning models
Core Problem: Sparse free-field stations cannot resolve spatial ground-motion variability densely enough for rapid urban assessment.
Key Innovation: Fuses response features from building-embedded sensors with building-specific CNN-LSTM models, recovering city-scale shaking fields and validating on unseen records with correlations up to 0.97.
24. Disaster risk management in informal settlements: Household evidence from KwaZulu-Natal, South Africa
Core Problem: Households in informal settlements face hazard exposure and recovery constraints that formal risk programs may not capture.
Key Innovation: Uses household evidence from KwaZulu-Natal to connect lived risk, coping capacity and practical disaster-management needs.
25. A hybrid probabilistic framework for integrating structural mitigation and financial transfer to enhance the seismic risk management of critical transportation infrastructure
Core Problem: Structural strengthening and financial transfer are often optimized separately despite jointly controlling recovery and loss.
Key Innovation: The verified record integrates probabilistic structural mitigation and financial risk transfer for critical transport assets; quantitative outcomes are not exposed in the accessible abstract.
26. What Is the Cost of Urban Flooding? Changes in Travel and Emission Costs of Łódź’s Transport System in Poland
Core Problem: Flood impacts on travel time and emissions are rarely monetized together at the urban transport-system scale.
Key Innovation: The Łódź study connects inundation-driven network disruption to changes in travel and emission costs, translating hydraulic consequences into mobility and environmental burdens.
27. Joint resilience-fairness optimization of a multi-commodity railway network under flood disruptions
Core Problem: Recovery plans that maximize aggregate railway resilience can distribute service losses and restoration benefits inequitably.
Key Innovation: Jointly optimizes resilience and fairness for a multi-commodity railway network under flood disruption, making distributional consequences part of restoration design.
28. Melt-Regulated Crustal Kinematics in the Pamir, NW Tibetan Plateau: Insights From Surface-Wave Tomography
Core Problem: The Pamir, located at the northwestern margin of the Tibetan Plateau, is undergoing E-W extension within its interior, accommodated by micro-block motions along major fault systems; however, the mechanisms governing this modern block kinematics remain elusive.
Key Innovation: Here, we present an updated crustal shear-wave velocity model of the region, developed via adjoint-state surface-wave traveltime tomography. The resulting model reveals distinct middle-to-lower crustal low-velocity zones (LVZs) that spatially align with the major active faults bounding the micro-blocks.
29. Hydrologic Retention and Sediment Redistribution Following Beaver Dam Analog (BDA) Installation in a Flashy, Ephemeral, Stormwater-Impacted Stream
Core Problem: In many watersheds, urbanization and agriculture have altered stream hydrology and morphology, increasing flow variability, and sediment transport imbalances.
Key Innovation: This study describes the geomorphic and hydrologic responses 1 year after installing eight BDAs in an ephemeral, stormwater-driven stream within a mixed-land use watershed (68% agricultural, 28% developed, 4% forested) in Selinsgrove, Pennsylvania. Water-level loggers showed significantly longer retention time in upstream BDA pools than downstream pools ( p < 0.01), suggesting that BDAs moderate flow during storm events along.
30. Mapping Groundwater-Dependent Ecosystems for Contiguous China With Selected Hydroclimatic Windows
Core Problem: The distribution of GDEs remains largely unknown across contiguous China.
Key Innovation: In this study, remote sensing data, soil data, and modeled groundwater depth (GWD) data are integrated to map GDEs across both arid and humid regions of contiguous China. The results indicate that high-, moderate-, and low-potential GDEs account for 1.6%, 7.7%, and 22.1% of the mapped area, respectively.
31. Spectral-Target Physical Latent Structuring for JEPA-Style World Models
Core Problem: Even with such regularization preventing representation collapse, we identify a new world model failure mode of \textit{physical representation laziness}, particularly noted in highly dynamic environments.
Key Innovation: To resolve this issue, we propose training-time auxiliary supervision with a lightweight "Fourier auxiliary head", which enforces physically-informed structuring of the latent space with no additional inference-time cost and can be generalized to any environment. Experimentally, we show that the auxiliary head substantially improves planning success rates in dynamic environments where the baseline LeWM exhibits physical.
32. Disentangling Attention in Deep Operator Learning: A Controlled Study of Data-Driven and Physics-Informed Architectures
Core Problem: Deep neural operators learn mappings between input functions and complete PDE solution fields, enabling forward evaluations of new problem instances orders of magnitude faster than conventional numerical solvers.
Key Innovation: Attention mechanisms have recently been introduced into neural operators, but most studies change several architectural components at once, making it difficult to identify what actually improves accuracy.
33. Too Rare to Learn: Prescribed Cyclone Tracks Degrade a Bay of Bengal Ocean Emulator
Core Problem: Neural ocean emulators are being proposed for regional forecasting in cyclone-exposed coastal seas, and a natural design choice is to hand the network the cyclone as a prescribed input.
Key Innovation: We test that choice in the Bay of Bengal and find it harmful.
34. Bridging Modalities and Tasks: A Unified Hierarchical ViT for SAR-to-Optical Translation and Semantic Segmentation
Core Problem: However, compared with optical images, their speckle noise and non-intuitive scattering mechanism limit the interpretability of the images.
Key Innovation: We propose a unified collaborative dual-task learning framework, termed BMT (Bridging Modalities and Tasks), that jointly optimizes S2O image translation and semantic segmentation through a shared hierarchical Vision Transformer. The experimental results show that the proposed method achieves competitive S2O translation quality and semantic segmentation performance.
35. Weather-Conditioned Depth Anything
Core Problem: However, they still suffer from critical failures under adverse weather conditions, such as fog, rain, snow, or at night.
Key Innovation: To address this, we present Weather-Conditioned Depth Anything (DA-W), a framework that explicitly disentangles style from content for weather-robust depth estimation. Our comprehensive experiments demonstrate that our proposed DA-W achieves state-of-the-art robust depth estimation, improving AbsRel by an average of 3.7% on our curated weather benchmarks, while matching or slightly outperforming performance on standard clean.
36. Bayesian inversion of multilayer CO₂ migration from seismic plume observations using a graph-based finite-rate invasion-percolation model
Core Problem: Vertical migration of CO₂ in layered sandstone reservoirs is controlled by thin shale barriers whose properties are often poorly known.
Key Innovation: We develop a Bayesian framework that uses time-lapse seismic plume observations to estimate effective parameters governing lateral and vertical CO₂ migration. These experiments show that the information gained from monitoring depends on the migration events captured, with breakthrough and post-breach redistribution providing particularly strong constraints.
37. Learning Spatial-Spectral Refinement and Calibrating Complementary Observations for Hyperspectral Image Super-Resolution
Core Problem: Recent advances in implicit neural representations (INRs) have enabled flexible coordinate-based modeling for HMIF; however, existing INR-based approaches may not fully capture fine-grained spatial structures and rich spectral dependencies.
Key Innovation: To address these limitations, we propose Two-Stage Reconstruction with Implicit Tensor Neural Representation (TSR-ITNR), a unified self-supervised framework integrating representation refinement and observation-guided calibration. Extensive experiments on multiple benchmark datasets demonstrate strong quantitative, visual, and spectral reconstruction performance without ground-truth HR-HSI supervision.
38. MEOX: Compact Multimodal Mixture-of-Experts for Earth Observation
Core Problem: Sensor-specific adapters, explicit validity signals, and a shared sparse-expert block preserve modality-dependent processing before a learned patch-wise fusion.
Key Innovation: We present MEOX (Multimodal Earth Observation with eXperts), a multimodal masked autoencoder with a 2.939 million-parameter encoder and 3.115 million parameters in total. The model reaches 64.42% mean intersection-over-union on cashew segmentation at 64 pixels and 90.56% average accuracy on EuroSAT at 224 pixels, exceeding the corresponding reported CSMoE results.
39. Integrated Experimental and Numerical Investigations on the Thermo-Hydro-Mechanical Behavior of Clays and Argillaceous Rocks: A Perspective
Core Problem: This paper synthesizes nearly a decade of research on the coupled thermo-hydro-mechanical (THM) behavior of clays and argillaceous rocks.
Key Innovation: Key findings are drawn from constitutive modeling, in situ tests, and energy geostructure applications, offering a practical THM framework for nuclear waste repositories and climate-resilient infrastructure.
40. Forecast Skill Is Not Decision Skill: Evidence from Weather-Dependent Decision Tasks
Core Problem: In practice, however, forecasts are used to make decisions, so it seems natural to take the decision-maker's perspective and quantify the value of a forecast by its ability to improve decision-making.
Key Innovation: Decision calibration provides a novel framework for evaluating probabilistic forecast performance at the decision level rather than the forecast level. We find that model performance at the forecast level does not reliably translate to performance in downstream decision-making: some performance differences only become apparent at the decision level, and even among seemingly similar decision tasks, model rankings can change.
41. WaveGRL: A wavelet-graph reinforcement learning framework for sequential imputation of ocean time series
Core Problem: However, most existing imputation methods rely solely on time-domain modeling, fail to capture such coupling patterns, and thus cannot decouple valid physical signals from redundant observation noise.
Key Innovation: To address these limitations, this study proposes a wavelet-graph reinforcement learning framework (WaveGRL) that reformulates OTS imputation as a multi-scale wavelet coefficient correction task. Finally, the corrected coefficients are reconstructed via the inverse discrete wavelet transform to generate imputation results.
42. Rock-socketed offshore wind turbine monopile foundations: A transfer-learning-enhanced modular “Digital-mechanical” twin method for full-field strain prediction
Core Problem: The verified record addresses rock-socketed offshore wind turbine monopile foundations: A transfer-learning-enhanced modular “Digital-mechanical” twin method for full-field strain prediction, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
43. Prediction of the Bohai Sea and northern Yellow Sea wave-current system based on physics-informed dynamic graph attention network
Core Problem: The verified record addresses prediction of the Bohai Sea and northern Yellow Sea wave-current system based on physics-informed dynamic graph attention network, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
44. Joint Inversion of Gravity Anomaly and Seismic Data for Improved Crustal Imaging in the Guangdong-Hong Kong-Macao Greater Bay Area
Core Problem: Satellite-derived gravity observations provide important constraints on subsurface structure variations, but are limited by poor depth resolution and strong nonuniqueness.
Key Innovation: In this study, we built a crustal S-wave velocity and density model of the GBA by joint inversion of gravity anomaly and seismic data. The results show that the crustal structure is very complex.
45. Millimeter-Wave Radar-Based Three-Dimensional Sea-Surface Wavefield Reconstruction From Sparse Point Clouds
Core Problem: To address the challenge of sparse distribution in the sea-surface point clouds, which are derived from sea clutter obtained by mmWave radar, linear wave theory and sparse representation are integrated to facilitate the extraction of wave model parameters.
Key Innovation: This technology also serves as a key generic technology for optimizing various maritime operations. Accordingly, this article presents a novel research on employing 4-D millimeter-wave (mmWave) radar mounted on uncrewed aerial vehicle (UAV) to achieve high-resolution 3-D wavefield detection.
46. A High-Efficiency Diffusion Model-Inspired Network for Pixel-Level Self-Supervised Hyperspectral Anomaly Change Detection
Core Problem: To address these challenges, we propose a novel hyperspectral anomaly CD method named efficient latent denoising-inspired network (ELDI-Net).
Key Innovation: To address these challenges, we propose a novel hyperspectral anomaly CD method named efficient latent denoising-inspired network (ELDI-Net). It employs a one-step manifold projection paradigm, achieving high computational efficiency while preserving the noise-robustness advantages of generative models.
47. Unrolled Low-Rank Tensor Completion for SAR-Guided Cloud Removal in Multispectral Images
Core Problem: We address the challenge of cloud removal in multispectral satellite images (MSIs), where clouds obscure critical spatial and spectral information.
Key Innovation: In this work, we propose a deep unrolled tensor completion framework with synthetic aperture radar (SAR)-guided detail injection for MSI cloud removal, effectively integrating model- and data-driven approaches to leverage their complementary strengths.
48. An AI-driven reconstruction of global surface temperature with emphasis on refining the Antarctic record
Core Problem: However, substantial missing information in global ST datasets, remains a major source of uncertainty in estimating global or regional temperature changes.
Key Innovation: In this study, partial convolutional neural network (PConv) models were trained using the 20CR reanalysis data and CMIP6 climate model outputs as training samples, with the aim of achieving a proper reconstruction of the global surface temperature dataset.
49. A Monthly 30 m Dataset of Lakes Larger than 1 km² on the Tibetan Plateau from 2001 to 2025
Core Problem: However, existing products are constrained by factors such as cloud cover, snow and ice contamination, terrain shadow, observational gaps, and scale-related bias.
Key Innovation: Here we present a new monthly dataset of lakes larger than 1 km² across the TP for 2001-2025 at 30 m spatial resolution. Technical validation against three reference datasets showed strong spatial agreement (IoU = 0.80-0.92) and very high consistency in lake area estimates (R² > 0.99).
50. A European forest disturbance database for robust area estimation and map validation
Core Problem: Reliable data on disturbance-induced tree cover change are essential to address the challenges faces by Europe's forests.
Key Innovation: Europe's forests play a critical role as carbon sink, timber production and for the societal well-being, yet their capacity to maintain these functions is increasingly threatened by increasing disturbance rates and a growing demand for wood. Our results underscore the need for consistent, transparent, and independent reference data for understanding disturbance change in Europe and for validation and uncertainty.
51. Development and testing of ensemble-variational data assimilation capabilities for radar data within JEDI coupled with FV3-LAM model
Core Problem: Abstract.
Key Innovation: This study presents the first implementation and evaluation of radar reflectivity data assimilation capabilities within the ensemble three-dimensional variational (En3DVar) data assimilation (DA) system of the Joint Effort for Data assimilation Integration (JEDI) framework.
52. GWSWEX v1.0: a dual-solver 1D unsaturated zone model for mass-conservative groundwater recharge and runoff computation in distributed hydrological modelling
Core Problem: The faithful numerical representation of the coupled dynamics between groundwater (GW), the unsaturated zone (UZ) and surface water (SW) remains one of the more persistent challenges of regional-scale integrated hydrological modelling.
Key Innovation: This paper introduces GWSWEX (Groundwater-Surface Water EXchange), a vertically resolved process-based modelling package designed to occupy the middle ground between these two extremes and to act as a UZ coupler between an external GW model and an external SW model within an integrated modelling chain.
53. Numerical simulation of P-wave propagation in 2D VTI media with a topographic free surface by a curvilinear-grid finite-difference method
Core Problem: However, conventional finite-difference methods struggle to accurately simulate both topographic free surface and anisotropy in first-order acoustic velocity-stress equations, failing to capture complex wavefield responses.
Key Innovation: With increasing demands for high-precision subsurface imaging, advanced seismic prospecting and seismological imaging methods require simultaneous consideration of topographic free surface and anisotropy effects on P-wave propagation. Numerical experiments on flat-surface, Gaussian-topography, and BP benchmark models demonstrate stable performance under the tested conditions, while near free-surface waveforms computed by our.
54. A Change Detection Network for Heterogeneous Remote Sensing Images Based on Decoupled Differential Architecture Search
Core Problem: However, existing heterogeneous remote sensing image change detection methods rely largely on fixed network architectures, which makes adapting to complex modal differences and severe noise interference difficult.
Key Innovation: Furthermore, to suppress erroneous graph connections caused by noise, structural consistency and smooth denoising (SCSD) loss is introduced, and deep semantic feedback and graph smoothing regularization constraints are collaboratively used to dynamically generate graphs, thereby effectively increasing the internal consistency of the transformed graph and suppressing misconnection noise.
55. A Unified Framework for Individual Tree Segmentation and Forest Biometrics Derivation from LiDAR Point Clouds Captured by Different Platforms in Diverse Forest Environments
Core Problem: However, differences in point density, viewing geometry, and occlusions among these acquisition systems pose challenges for processing heterogeneous LiDAR datasets using a common workflow.
Key Innovation: This study proposes a forest inventory pipeline for individual tree segmentation and the derivation of key forest biometrics including tree location and diameter at breast height (DBH) across heterogeneous LiDAR datasets. The proposed tree detection pipeline achieved Precision ranging from 86.44% to 100%, Recall from 74.17% to 100%, and F1-scores from 81.82% to 100% across the evaluated datasets.
56. DualGLEAN: Dual Allocation for VLM-Guided Generalized Category Discovery in Remote Sensing Images
Core Problem: Generalized category discovery (GCD) aims to classify known categories while discovering novel ones in unlabeled data, yet existing methods lack mechanisms to correct boundary-ambiguous samples that receive noisy pseudo-labels, as they primarily rely on visual feature learning without external semantic guidance.
Key Innovation: To resolve this, we propose DualGLEAN, a framework that addresses the dual allocation challenge through two coupled mechanisms: decoupled contrastive alignment (DCA), which routes the VLM-guided neighbor contrastive loss to a dedicated projector space while preserving the backbone space for global clustering, and compound uncertainty querying (CUQ), a three-stage filtering metric that jointly evaluates predictive entropy.
57. Hierarchical Fusion Method for SAR-Based Coastal Bathymetric Inversion in Short-Period Wave-Dominated Areas: A Case Study of the Wengtian Coast, Hainan Island
Core Problem: However, in short-period wave-dominated areas, this method suffers from large errors and poor stability, limiting its practical application.
Key Innovation: To address this problem, this study proposes a hierarchical fusion method for SAR-based coastal bathymetric inversion in short-period wave-dominated areas. The proposed method significantly reduces bathymetric inversion errors by hierarchically fusing multi-angle wavelength estimation results and multi-temporal SAR bathymetry inversion results, thereby achieving more robust underwater topography mapping.
58. Soil Moisture Retrieval Based on Multi-Temporal Dual-Polarization Brightness Temperature Parameterization
Core Problem: However, most existing passive microwave soil moisture retrieval methods rely on fixed empirical parameters to characterize vegetation single-scattering albedo and soil surface roughness, which may not fully account for variations in surface conditions across different regions and seasons, thereby affecting retrieval accuracy.
Key Innovation: To address this issue, this study proposes a method for jointly constraining key parameters of the forward model for passive microwave soil moisture retrieval using multi-temporal brightness temperature observations. The retrieval results were evaluated using ground-based observations and compared with existing soil moisture products.
59. Evaluating the effects of visual AI measurement uncertainty on Bayesian structural durability assessment
Core Problem: The verified record addresses evaluating the effects of visual AI measurement uncertainty on Bayesian structural durability assessment, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
60. A reliability-informed stochastic hybrid graph attention network (SHGAT) framework for uncertainty-aware road operation-maintenance risk inference
Core Problem: Direct maintenance records and condition measurements are often sparse, whereas traffic trajectories, road topology, camera deployment, and geometric attributes are continuously available but unevenly reliable.
Key Innovation: This study proposes a reliability-informed stochastic hybrid graph attention network (SHGAT) framework for inferring latent road operation-maintenance pressure as a reliability-risk proxy. In the Shenzhen case, SHGAT achieved a predictive correlation of r = 0.992.
61. Robust optimization of emergency evacuation and supply allocation considering personnel demand heterogeneity and temporary shelters
Core Problem: The verified record addresses robust optimization of emergency evacuation and supply allocation considering personnel demand heterogeneity and temporary shelters, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
62. Fast Resilience Enhancement Strategy of Distribution Networks under Extreme Natural Hazards via Coordinated Deployment of Switch Resources
Core Problem: The verified record addresses fast Resilience Enhancement Strategy of Distribution Networks under Extreme Natural Hazards via Coordinated Deployment of Switch Resources, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
63. A Bayesian Network Framework for Multi-Hazard Quantitative Risk Assessment of Hydrocarbon Facilities Undergoing Degradation
Core Problem: The verified record addresses a Bayesian Network Framework for Multi-Hazard Quantitative Risk Assessment of Hydrocarbon Facilities Undergoing Degradation, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
64. Glaciers of Ecuador: a review of current scientific knowledge
Core Problem: Glaciers cover ∼10% of the Earth's land surface, but they are shrinking rapidly across most parts of the world, leading to cascading impacts on downstream systems.
Key Innovation: Glaciers impart unique footprints on river flow at times when other water sources are low.
65. GFS-Net: Geometry-frequency synergistic fusion for accurate optical-SAR land-cover classification
Core Problem: The verified record addresses gFS-Net: Geometry-frequency synergistic fusion for accurate optical-SAR land-cover classification, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
66. Monitoring forest growth dynamics from harmonized multi-source time series of canopy height
Core Problem: Most remote sensing assessments of forest change focus on canopy cover, whereas the dynamics of vertical growth remain less investigated despite their sensitivity to external forcings.
Key Innovation: Here, we present a transferable workflow to derive plot-level indicators of vertical growth dynamics from mixed-source canopy height model (CHM) time series, harmonized against forest inventory plots at ~1000 m² resolution.
67. Macro-scale freezing deformation performance and micro-mechanism of soilbag-reinforced expansive soils: perspective from a single soilbag unit
Core Problem: The verified record addresses macro-scale freezing deformation performance and micro-mechanism of soilbag-reinforced expansive soils: perspective from a single soilbag unit, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
68. Damage characterization of deep Jinping marble under pre-static loading and multi-period dynamic impacts: Insights from acoustic emission and electromagnetic radiation
Core Problem: The verified record addresses damage characterization of deep Jinping marble under pre-static loading and multi-period dynamic impacts: Insights from acoustic emission and electromagnetic radiation, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
69. Impacts of climate change on rain-on-snow events and spring agricultural drought in Northeast China
Core Problem: The verified record addresses impacts of climate change on rain-on-snow events and spring agricultural drought in Northeast China, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
70. Asynchronous response relationships between groundwater recharge and river-aquifer exchange in the urbanized karst river basin
Core Problem: The verified record addresses asynchronous response relationships between groundwater recharge and river-aquifer exchange in the urbanized karst river basin, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
71. Grounding large language models in hydrologic modelling
Core Problem: Ablation studies have revealed that removing the expert task workflow causes complete system failure owing to planning chaos, plummeting the success rate to 0%.
Key Innovation: To address this limitation, we present an LLM-based agentic intelligent modelling approach for hydrological time-series forecasting (HydroAIM). Comprehensive experiments demonstrated that across 100 modelling tasks of varying types utilising four LLMs, HydroAIM achieved an execution success rate of 91%.
72. Why Does the Broken Hill Deposit Sit in Resistive Crust? Magnetotelluric Evidence for Metamorphic Decoupling of a World-Class Mineral System
Core Problem: The spatial association between electrical conductors and ore deposits underpins the widespread use of magnetotellurics (MT) in mineral exploration, yet not all mineral systems preserve a diagnostic conductive signature.
Key Innovation: Here, we present a high-resolution MT study of the region hosting the world-class Broken Hill Pb-Zn-Ag deposit (NSW, Australia). Our MT model reveals a predominantly resistive upper crust containing spatially discrete conductive anomalies that coincide with the Broken Hill lode and the sulphide-rich Broken Hill Group (specifically the Hores Gneiss).
73. Strategic Deep-Water Observations Enhance Probabilistic Parameter Estimation of Lake Hydrodynamic Models
Core Problem: Calibration of physics-based hydrodynamic models of lake water temperature often faces challenges including limited observational data and parameter equifinality, that is, many parameterizations yield similar goodness-of-fit to observations.
Key Innovation: This study presents a framework to investigate how the number and location of observation depths used in calibration influence temperature prediction accuracy and parameter equifinality. Results show that single-depth deep-water observations consistently outperform shallower depths and sometimes even full-profile observations, significantly reducing equifinality and achieving accurate match throughout the water column.
74. Data-Driven Learning of Unknown Nonlinear Differential Equations Using Functional Analysis
Core Problem: In this paper, the problem of data-driven discovery of nonlinear ordinary differential equations (ODEs) is recast, and a new interpretable machine learning (ML) method is proposed.
Key Innovation: The proposed method aims to learn the unknown vector field of nonlinear dynamics without prior knowledge of the system's physics from only one single state trajectory's data. Finally, numerical examples are given to demonstrate the advantages of the proposed method.
75. On-board ML for Trace Gas detection in Imaging Spectroscopy data
Core Problem: However, current processing pipelines depend on slow, on-the-ground processing, which delays the time to information of each detected event and prohibits immediate follow-up actions.
Key Innovation: However, current processing pipelines depend on slow, on-the-ground processing, which delays the time to information of each detected event and prohibits immediate follow-up actions. We show the first on-board detection of methane point source emission with Imaging Spectroscopy data using Edge ML.
76. HiSfM: Disambiguating Structure-from-Motion via Scaffold-Anchored Hierarchical Reconstruction
Core Problem: Structure-from-Motion (SfM) is a fundamental tool for sparse 3D reconstruction with broad impact in robotics and vision, supporting mapping, localization, and large-scale scene modeling.
Key Innovation: We present HiSfM, a hierarchical coarse-to-fine SfM framework that improves robustness and efficiency through scaffold construction. Experiments on ambiguity-focused benchmarks and general datasets show that HiSfM prevents ambiguity-induced failures while substantially reducing runtime compared to previous methods, and improves completeness over aggressive sparsification methods.
77. SimFuse3D: Source-Guided Target Simulation and Confidence-Guided Multi-Stage Localization Reweighting for Cross-Platform 3D Object Detection
Core Problem: Changes in sensor height and viewpoint alter object-level point distributions, making cross-platform LiDAR unsupervised domain adaptation (UDA) difficult.
Key Innovation: We introduce SimFuse3D, which preserves the target placement and repairs the associated pseudo-object using measured geometry from labeled source scans.
78. Methane Detection On Board Satellites from Unorthorectified Imagery
Core Problem: Conventional detection methods rely on orthorectification to correct geometric distortions and matched filters to enhance plume signals, which are steps designed for ground processing and poorly suited to onboard execution.
Key Innovation: We introduce UnorthoDOS, a dataset and approach for training machine learning models directly on unorthorectified hyperspectral imagery, bypassing both orthorectification and matched-filter products. 18.47% on all plumes), while both substantially outperform the mag1c matched-filter baseline (IoU 4.76%).
79. Beyond Homoscedasticity: Decoupled Uncertainty Optimization for Deep Imbalanced Regression
Core Problem: Deep Imbalanced Regression (DIR) is pervasive in continuous prediction tasks across diverse modalities, such as age estimation, depth prediction, and protein mutation activity prediction, where label-scarce tail samples often carry higher practical value.
Key Innovation: To address this, we propose DUO, an uncertainty-aware long-tailed regression framework. Across visual and biological DIR benchmarks, DUO achieves the best few-shot bMAE and GM on IMDB-WIKI-DIR, AgeDB-DIR, and AAV2-DIR while remaining competitive on few-shot MAE.
80. Training-Free Logical and Structural Anomaly Detection via Calibrated Fusion
Core Problem: In particular, training-free methods effectively exploit frozen representations but lack an explicit notion of object count, while methods that reason about counts usually rely on category-specific component modeling.
Key Innovation: We show that counting ability can be introduced into training-free anomaly detection without additional training or part-level supervision.
81. Beyond Stationarity in Time Series: Discovering Causal Structures and Latent Regimes via Markov Blankets
Core Problem: Time series are typically observed at discrete time points and often exhibit regime changes that challenge the assumption of a static causal structure, a limitation in many real-world dynamic systems.
Key Innovation: This paper introduces Regime-aware Constraint-Based and Noise-Based causal discovery with Markov Blankets (RCBNB-MB), a novel causal discovery algorithm for time series that relaxes the common assumption of a single, time-consistent causal structure. Empirical results show that RCBNB-MB systematically outperforms baseline approaches in accurately detecting regime changes and their associated causal graphs, positioning it as a.
82. BLASt3R: Bundle Adjustment of Any Image Set with Multi-View Matching and Monocular Priors
Core Problem: They are usually the best performing methods however their scalability and usability remains limited since estimating dense correspondences between views is prohibitively costly, especially considering time constraints inherent to online applications like Visual SLAM (VSLAM).
Key Innovation: In this paper, we introduce a regularized BA framework that leverages a fast multi-view matcher and monocular priors for initialization and regularization. Extensive experiments across both domains demonstrate improved performance and speed tradeoffs over traditional, feed-forward, and hybrid baselines.
83. CrossDepth: Geometry-Constrained Attention for Generalizable Multi-View Surround Depth Estimation
Core Problem: We target two main sources of cross-image inconsistency: differences in camera intrinsics and the limited receptive field of each image.
Key Innovation: Multi-view surround camera rigs provide broad scene coverage, but the spatially adjacent images typically overlap only minimally. Evaluations on DDAD and nuScenes show improved overall depth accuracy and cross-image depth consistency over state-of-the-art self-supervised methods under in-domain and cross-domain evaluation.
84. Out-of-Distribution Semantic Occupancy Prediction
Core Problem: However, existing methods focus on in-distribution scenes, making them susceptible to Out-of-Distribution (OoD) objects and long-tail distributions, which increase the risk of undetected anomalies and misinterpretations, posing safety hazards.
Key Innovation: To address these challenges, we introduce the task of Out-of-Distribution Semantic Occupancy Prediction, targeting OoD detection in 3D voxel space. Experimental results demonstrate that OccOoD achieves an AuROC of 65.50% and an AuPRCr of 31.83% within a 1.2m radius, while maintaining competitive semantic occupancy prediction accuracy, significantly improving detection sensitivity for unknown obstacles, and validating strong.
85. Gradient-based Model Shortcut Detection for Time Series Classification
Core Problem: However, the shortcut behavior of DNNs in time series remain under-explored.
Key Innovation: Recently, deep neural networks (DNN) have surpassed classical distance-based methods and achieved state-of-the-art performance.
86. Omega-N: Interpretable Structural Node Descriptors and Their Applicability Domain
Core Problem: A composite structural index summarises a network in one number, and for a triangle-based index it is spectrally redundant: Tr(A³) is the third moment of the adjacency spectrum.
Key Innovation: The non-redundant content sits one level down, in diag(A³), which depends on eigenvectors and is not spectrally determined.
87. Field Observations of Tidal- and Thermal-Induced Deformations and Anchor Forces in Quay Walls
Core Problem: This case study investigates the mechanical and thermal response of anchored quay walls in the Port of Rotterdam through continuous field monitoring.
Key Innovation: Three structurally distinct combined quay walls, consisting of steel tubular piles with intermediate sheet piles, were examined. The results show that daily lateral wall displacement fluctuations of approximately 1.5 mm (about 10% of initial displacement) are primarily driven by tidal water-level variations.
88. Rock Type Identification From AVIRIS-NG Hyperspectral Data Using a Bidirectional Gated Attention Convolutional Transformer
Core Problem: Mineral exploration using hyperspectral data that has high spectral resolution across visible and shortwave infrared wavelengths become an advanced method for recognizing potential mineral-rich zones.
Key Innovation: This study introduces a novel deep-learning model for identifying rocks that have associated with iron (banded iron formation) and chromite minerals in the Sittampundi region of Tamil Nadu state, India by using airborne visible infrared spectrometer (AVIRIS-NG) air borne hyperspectral data. Finally, the results are explored in two different ways to compare the designed model.
89. C-PEAT's Global Peatland Carbon Database (v.2025)
Core Problem: Abstract.
Key Innovation: Field-based measurements are foundational to the study of short- and long-term peatland carbon dynamics.
90. Artificial Intelligence for Stable Isotope Tracers (AISIT): A Database of δ¹⁸O and ancillary data in the Pan-Arctic
Core Problem: Abstract.
Key Innovation: Here, we present a unified pan-Arctic seawater δ¹⁸O database comprising observations collected north of 60 °N, compiled from public repositories, peer-reviewed literature, and datasets contributed directly by colleagues across the scientific community. To unify these records, we harmonised metadata, applied standardised quality-control procedures, and integrated records into a single Artificial Intelligence (AI) compatible.
91. Past abrupt changes in Atlantic Ocean currents controlled the Earth’s energy balance
Core Problem: Nature Geoscience, Published online: 07 September 2026; doi:10.1038/s41561-026-02086-y During the last ice age, the Atlantic Meridional Overturning Circulation (AMOC) underwent a series of abrupt reorganizations.
Key Innovation: The global impact of these events reflects the accumulation of heat in the ocean interior during the weak AMOC mode and its loss during the strong mode, facilitated by radiative feedbacks that altered the planetary energy balance.
92. Confusion-Resistant Learning for Few-Shot Oriented Object Detection in Aerial Images
Core Problem: This study is devoted to few-shot oriented object detection in aerial images, aiming to enhance detection performance for novel object classes, using only limited supervised samples.
Key Innovation: This study is devoted to few-shot oriented object detection in aerial images, aiming to enhance detection performance for novel object classes, using only limited supervised samples.
93. A Geometry-Constrained Framework for Automatic Geometric Positioning Accuracy Assessment of Large-Scale Satellite Imagery
Core Problem: However, conventional GCP-free inspection methods based on local feature matching often exhibit limited robustness under large initial positioning errors, weak-texture regions, cloud contamination, and temporal appearance variations, resulting in poor generalization across large-scale production scenarios.
Key Innovation: To address these challenges, this paper proposes a geometry-constrained framework that integrates Rational Polynomial Coefficient (RPC) prior constraints, coarse-to-fine registration, adaptive match-density-based block selection, hierarchical geometric verification, and a geolocation residual confidence measure into a unified automatic quality inspection pipeline.
94. Improving Cross-River Turbidity Retrieval by Incorporating Environmental Variables: When and Why It Works
Core Problem: However, strong seasonal variability, high sensitivity to disturbances, and pronounced spatial heterogeneity make turbidity patterns difficult to characterize and generalize across river systems.
Key Innovation: However, strong seasonal variability, high sensitivity to disturbances, and pronounced spatial heterogeneity make turbidity patterns difficult to characterize and generalize across river systems. These results demonstrate that integrating spectral and environmental information improves both the accuracy and generalizability of turbidity retrieval and helps clarify when and why environmental context benefits water-quality.
95. Scattering-Aware Latent Field Modulation for Synthetic Aperture Radar Object Detection
Core Problem: Synthetic aperture radar (SAR) object detection remains challenging, as target evidence is often sparse, discontinuous, and heavily influenced by speckle noise, sidelobes, shadows, and clutter-like background scattering.
Key Innovation: To address this issue, we propose a SAR-inspired Scattering-Center Field (SCF) modulation framework for multi-scale dense object detection. The proposed detector achieves mAP50/mAP50-95 scores of 0.954/0.682 on SSDD, 0.930/0.661 on SAR-AIRcraft-1.0, 0.971/0.676 on SAR-Ship, and 0.688/0.482 on MSAR-1.0.
96. Retrospective Forest Volume Estimation in Southern Chile Using ALOS-PALSAR for Carbon MRV Applications
Core Problem: Accurate historical estimates of forest carbon stocks are essential for greenhouse gas inventories and REDD+ Measurement, Reporting and Verification (MRV) systems, particularly in remote and persistently cloudy regions where field inventories and optical remote sensing are limited.
Key Innovation: This study presents a retrospective mapping framework with potential relevance for Tier-3 forest carbon estimation of forest volume and carbon stocks in the temperate forests of southern Chile using historical ALOS PALSAR L-band SAR data integrated with Chile’s Continuous National Forest Inventory (CNFI). Results demonstrate the potential of combining historical ALOS PALSAR archives with national forest inventories to support.
97. CPD-FCOS: A Scale-Isolated P2 Pathway for UAV Small-Object Detection
Core Problem: Small objects in UAV imagery fail for two coupled reasons.
Key Innovation: Repeated downsampling destroys their evidence-60.5% of VisDrone instances are COCO-small and the median object spans 26 pixels, so a typical target covers fewer than four cells at stride 8-and the in-box assignment rule then places most of their positive locations on object borders, where supervision is unreliable.
98. Analytical framework for coupled hydromechanical modeling of bearing capacity of stone column-supported footings in unsaturated soils
Core Problem: Traditional analyses of stone column (SC) bearing capacity typically assume saturated soil conditions, overlooking the unsaturated states that SCs may experience during their service life.
Key Innovation: To address this limitation, this study presents a novel analytical framework for evaluating the bearing capacity of SCs embedded in unsaturated soils. Model validation against a published case study shows strong agreement between predicted and observed values.
99. Investigation on long-term hydraulic behaviour and microstructure evolution of alkaline solution-treated bentonite pellet mixtures
Core Problem: Long-term hydraulic characteristics of bentonite pellet mixtures impacted by alkaline solution released via concrete degradation need accurate characterization, which is indispensable for safety design of high-level radioactive waste (HLW) deep geological repositories.
Key Innovation: This research focused on permeability variations of compacted bentonite pellet mixtures subjected to varied particle size distribution, alkaline solution concentration and aging time, with the aid of the mercury intrusion porosimetry and thermogravimetric tests. The results are as follows.
100. Study on the Thermal Aging Damage Mechanism of Calcitic Marble Based on Coupled Analysis of Multi-Scale Indicators
Core Problem: Calcitic marble undergoes simultaneous microstructural and macroscopic changes under high temperatures.
Key Innovation: To elucidate its thermal aging damage mechanism, this study subjected specimens to cyclic thermal aging at 400 °C with varying cycle counts. The results show a stage-dependent evolutionary trend (initial rapid degradation followed by stabilization) for ultrasonic wave velocity, compressive strength, pore-fracture network parameters, and fractal dimension under cyclic thermal aging.
101. Role of Loading Rate in Energy Evolution and Fatigue Damage of Salt Rock under Cyclic Loading
Core Problem: Ensuring the mechanical stability of salt caverns for deep underground energy storage is currently challenged by an insufficient understanding of the intrinsic mechanisms governing energy evolution and damage accumulation under varying loading rates.
Key Innovation: This study clarifies the loading-rate-dependent energy evolution and fatigue failure behavior of salt rock through uniaxial cyclic loading-unloading tests. The findings demonstrate a rate-induced energy retardation effect, indicating that high-rate loading improves elastic storage while inhibiting viscous dissipation.
102. A Simple Elastoplastic Bounding Surface Model for Drained and Undrained Loading of Saturated Soils
Core Problem: Predicting the stress-strain behavior of overconsolidated or normally consolidated soils by applying different stress or strain paths is a challenging task.
Key Innovation: Engineers seek the safety of structures, and predicting the behavior of the soil is a cornerstone task. The numerical and experimental comparisons show the pertinence of the model.
103. Assessment of Collapse Behavior and Post-Collapse Compressibility Response of Soft Clay under Repeated Drying-Wetting Cycles
Core Problem: While extensive research exists on collapsible silts and loess, the collapse behavior of soft clays subjected to repeated drying and wetting remains under-researched.
Key Innovation: While extensive research exists on collapsible silts and loess, the collapse behavior of soft clays subjected to repeated drying and wetting remains under-researched. Results show that collapse susceptibility, ranging from moderate to severe, can be caused by repeated drying cycles, with the effect more pronounced under 200 kPa vertical stress and in specimens that underwent 1-day cycles that retained higher water contents.
104. Filtering GEDI footprints based on combined spatial-spectral characteristics for individual building height mapping in 12 major global cities
Core Problem: The verified record addresses filtering GEDI footprints based on combined spatial-spectral characteristics for individual building height mapping in 12 major global cities, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
105. Bearing capacity of transmission towers with anti-icing conductor under combined ice-wind loads
Core Problem: The verified record addresses bearing capacity of transmission towers with anti-icing conductor under combined ice-wind loads, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
106. Numerical investigation of repurposing oil storage caverns for hydrogen storage using lined rock cavern (LRC) technology
Core Problem: The verified record addresses numerical investigation of repurposing oil storage caverns for hydrogen storage using lined rock cavern (LRC) technology, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
107. Global mapping of potential hydropolitical system archetypes in international transboundary basins: a shift in perspective from geopolitical temperature into structural process
Core Problem: The verified record addresses global mapping of potential hydropolitical system archetypes in international transboundary basins: a shift in perspective from geopolitical temperature into structural process, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
108. Hydrodynamic reconstruction of tidal river networks based on sensitivity analysis of responses to sluice-pump scheduling: a case study of episodic pollution control
Core Problem: The verified record addresses hydrodynamic reconstruction of tidal river networks based on sensitivity analysis of responses to sluice-pump scheduling: a case study of episodic pollution control, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
109. Experimental investigation on dynamic and energy dissipation characteristics of fluid viscous dampers under low temperatures
Core Problem: The verified record addresses experimental investigation on dynamic and energy dissipation characteristics of fluid viscous dampers under low temperatures, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
110. Dynamic confluence behavior and borehole-layout optimization of self-expanding polymer grout during multi-source injection for pavement sub-slab void remediation
Core Problem: The dynamic confluence of self-expanding polymer grout during multi-source injection remains insufficiently quantified for pavement sub-slab void remediation.
Key Innovation: This study combines visualized horizontal-void experiments with Volume-of-Fluid (VOF) simulations to investigate the effects of injection mass, source spacing, void condition, borehole layout, and injection sequence on grout diffusion and confluence. An empirical power-law model was established for intersected grout bodies, achieving an in-sample coefficient of determination of 0.899.
111. Performance assessment of crushed rock road base stabilised with enzyme-induced carbonate precipitation under cyclic loading
Core Problem: The verified record addresses performance assessment of crushed rock road base stabilised with enzyme-induced carbonate precipitation under cyclic loading, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
112. Lightweight Vision Transformer Compression for On-Device Plant Disease Detection in Resource-Constrained Agricultural Field Conditions
Core Problem: Chilli (Capsicum annuum) is one of India's most economically significant crops, yet its productivity is persistently threatened by diseases that are difficult to identify without expert intervention.
Key Innovation: We propose a unified Vision Transformer compression framework that combines Hessian-Balanced Adaptive Block Pruning (H-BAC), guided by second-order sensitivity estimation, with quantization and attention-based knowledge distillation. While Vision Transformers (ViTs) have achieved high classification accuracy, their large computational footprint makes deployment on resource constrained devices challenging.
113. Collaborative On-Sensor Array Cameras
Core Problem: However, due to the inherent wavelength dependence of metalenses, in practice, these cameras do not match their refractive counterparts in image quality for broadband imaging, and may even suffer from hallucinations when relying on generative reconstruction methods.
Key Innovation: We introduce a distributed meta-optics learning method to tackle this challenge.
114. Two-phase Temperature Reconstruction in Ice-Water Systems
Core Problem: Yet experimental progress remains limited by the lack of a minimally invasive methodology for simultaneously resolving temperature fields in coupled liquid water and non-isothermal ice systems.
Key Innovation: Here, we introduce a physics-based data-assimilation method for reconstructing temperature fields in buoyancy-driven flows interacting with non-isothermal ice. We demonstrate the method in laboratory experiments in which the temperature range across the liquid water in contact with ice drives cabbeling-induced convection.
115. E-RGB-D: Real-Time Event-Based Perception with Structured Light
Core Problem: Despite their high dynamic range, temporal resolution, low power consumption, and computational simplicity, traditional monochrome ECs face limitations in detecting static or slowly moving objects and lack color information essential for certain applications.
Key Innovation: To address these challenges, we present a novel approach that integrates a Digital Light Processing (DLP) projector, forming Active Structured Light (ASL) for RGB-D sensing. This integration, facilitated by a commercial TI LightCrafter 4500 projector and a monocular monochrome EC, not only enables frameless RGB-D sensing applications but also achieves remarkable performance milestones.
116. Water Reflection Detection Using Symmetric Attention
Core Problem: Reflections of water pose a significant challenge for computer vision systems, as standard deep learning models frequently confuse objects with their mirror images, producing spurious false positives and negatives in tasks such as object detection and semantic segmentation.
Key Innovation: To mitigate this issue, we leverage the intrinsic imperfect reflective symmetry of water and introduce a Symmetry-Aware Water Reflection Detection Network, namely, SAWRD-Net, that couples dihedral group-equivariant convolutions with a matrix-decomposition decoder in an end-to-end framework. Evaluated on the largest available water reflection scene data set, SAWRD-Net achieves a true-positive rate of 0.890 against human.
117. HRN: A Hybrid Recall Network for Crop Classification Using Multisource Time-Series Remote Sensing Data
Core Problem: However, this field still faces several challenges, including the low classification accuracy of single-source remote sensing datasets (SSRSDs) and the limited generalization capability of existing DL methods.
Key Innovation: To address these issues, this study constructs three multisource remote sensing datasets (MSRSDs). Experimental results show that, first, HRN achieves the best classification performance among all compared methods, with Overall Accuracy (OA), Average Accuracy (AA), and Kappa Coefficient (κ) improved by 0.11%-9.75%, 0.45%-37.31%, and 0.18%-16.09%, respectively.
118. An Interpretable BO-TCBDA Deep Learning Framework for Winter Wheat Yield Estimation Using Multi-Source Remote Sensing Data
Core Problem: However, current deep learning models still face limitations in selecting and integrating multi-source features, and their high predictive accuracy is often accompanied by limited interpretability.
Key Innovation: This study introduces a Bayesian Optimization-Temporal Convolutional Network-Bidirectional Long Short-Term Memory-Dual Attention (BO-TCBDA) deep learning framework for winter wheat yield estimation. BO-TCBDA achieved the best performance, with an R² of 0.823 and an RMSE of 561.26 kg/ha.
119. DMDNet: Decoupled Multimodal Detection Network for Fine-Grained Ulva Prolifera Segmentation
Core Problem: To overcome these challenges, we develop a decoupled multimodal detection network (DMDNet) for fine-grained Ulva prolifera segmentation.
Key Innovation: To overcome these challenges, we develop a decoupled multimodal detection network (DMDNet) for fine-grained Ulva prolifera segmentation. Equipped with a shared encoder and independent task-specific decoder heads, DMDNet accepts either a single optical image or a single SAR image as input and generates stable and reliable segmentation results.
120. Exact decision-diagram fault tree quantification: Aralia benchmarks, simulation baselines, and independent module extraction safety audit
Core Problem: The verified record addresses exact decision-diagram fault tree quantification: Aralia benchmarks, simulation baselines, and independent module extraction safety audit, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
121. Infrastructure Availability and Urban Railway Performability: A Stochastic Petri Net Approach to Resilience, Congestion, and Service Reliability
Core Problem: The verified record addresses infrastructure Availability and Urban Railway Performability: A Stochastic Petri Net Approach to Resilience, Congestion, and Service Reliability, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.
122. Portable X-ray fluorescence spectrometry library for Brazilian soils: chemical characterization and prediction of soil particle size fractions across all biomes
Core Problem: The verified record addresses portable X-ray fluorescence spectrometry library for Brazilian soils: chemical characterization and prediction of soil particle size fractions across all biomes, but the publisher feed exposes bibliographic metadata rather than the study motivation.
Key Innovation: The title and bibliographic record establish the study scope; methods and quantitative results are not asserted because the correct abstract was not exposed.