TerraMosaic Daily Digest: September 11, 2026
Daily Summary
A coupled reconstruction of a 25 km³ submarine landslide in the East Sea shows that mass failure can generate coastal waves of 9.3 m in Korea and 16.8 m in Japan, with arrival times of only 25-50 min at major cities. The result exposes a hazard systematically understated by earthquake-centred tsunami assessments. On land, dynamic rainfall-driven susceptibility updates, physics-informed coseismic prediction and a Himalayan rock-ice avalanche synthesis connect changing slope state to regional forecasting and long-runout cascading failure.
Case studies sharpen both susceptibility mapping and process diagnosis. Along Pakistan's Sheringal-Kumrat Road, three interpretable models are tested against a field-checked inventory; in Recife, a decade of civil-defence records resolves persistent urban landslide clusters. Field monitoring, dynamic analysis and machine learning quantify blast-induced stability of jointed mine slopes, while recurrent debris flows in the Yizhong Basin and extreme-rainfall bioengineered slopes link material weakening, sediment recharge and hydrological forcing to failure.
Seismic monitoring advances from wave propagation to catalog reliability. Three-dimensional basin-mountain simulations reveal geometry-dependent amplification; non-detections are converted into likelihood features that remove most false automatic bulletin events while preserving analyst oversight; and a five-year Lacq catalog combines manual relocation with deep learning to constrain induced seismicity. Coastal recovery research complements these physical models by showing how institutional inertia and place-based risk perception can obstruct post-tsunami adaptation.
Key Trends
The day's strongest work couples source processes, spatial validation and operational consequences across slope, tsunami and seismic hazards.
- Mass movement is being coupled to downstream hydrodynamics: Dynamic landslide-tsunami simulation connects reconstructed failure geometry to coastal amplitude, inundation and warning time instead of prescribing an isolated wave source.
- Susceptibility is becoming dynamic and spatially accountable: Rainfall-driven updates, field-checked inventories and spatial validation replace static rankings and geographically optimistic performance estimates.
- Process constraints are moving inside learning systems: Coseismic landslide prediction, blast-affected slope stability and hydro-mechanical reinforcement combine physical response variables with data-driven inference.
- Event catalogs are treated as uncertain observations: Likelihood scoring uses both detections and non-detections, while manual and deep-learning catalogs are compared to improve location accuracy and completeness.
- Risk assessment is extending from hazard intensity to institutional response: Infrastructure, exposure and post-tsunami governance studies examine whether warnings, services and recovery policies remain functional after physical thresholds are exceeded.
Selected Papers
The September 11 collection is led by coupled submarine-landslide tsunami modelling, dynamic rainfall-induced landslide prediction, physics-informed coseismic susceptibility and climate-driven rock-ice avalanche chains. New case studies resolve road-corridor susceptibility in Pakistan, persistent urban landslide clusters in Recife and blast-induced instability of jointed mine slopes. Companion work advances seismic ground-motion modelling, automatic bulletin quality, induced-seismicity catalogs, GLOF assessment, recurrent debris-flow analysis, post-tsunami governance and physically constrained Earth observation.
1. An interpretable spatiotemporal deep learning framework for rainfall-induced landslide prediction based on susceptibility dynamic updates
Core Problem: Static susceptibility maps cannot represent short-term rainfall-driven changes in slope state.
Key Innovation: Couples interpretable spatiotemporal learning with dynamic susceptibility updates to forecast rainfall-induced landslide conditions.
2. Prediction of regional co-seismic landslide susceptibility: Physics-informed and deep learning models
Core Problem: Regional coseismic inventories require models that retain physical controls while learning nonlinear spatial relations.
Key Innovation: Combines physical constraints and deep learning for regional coseismic landslide susceptibility prediction.
3. Climate-driven rock-ice avalanche disaster chains in remote high-mountain regions
Core Problem: Climate-driven ice and rock degradation can propagate through multiple poorly connected hazard stages.
Key Innovation: Organizes climate forcing, slope failure, mixed-material mobility and downstream consequences within a cross-scale disaster-chain framework.
4. Submarine Landslides Could Cause Wide-Spread Tsunami Impact in the East Sea (Sea of Japan)
Core Problem: Regional tsunami assessments underrepresent large submarine mass failures and their short warning times.
Key Innovation: Reconstructs a 25 km³ prehistoric slide and dynamically couples runout, wave propagation and inundation, showing impacts exceeding regional earthquake scenarios.
5. Spatial pattern and driving mechanisms of geohazards in the lower yarlung zangbo suture zone: implications for engineering construction in extreme mountainous environments
Core Problem: Engineering development in the lower Yarlung Zangbo suture zone requires spatially resolved understanding of multiple hazard controls.
Key Innovation: Links regional geohazard patterns to tectonic, terrain and environmental drivers in an extreme mountain corridor.
6. LMAF-Net: Lightweight Multi-Attention Fusion Landslide Identification Network
Core Problem: Operational landslide mapping must retain accuracy under limited computing resources and complex backgrounds.
Key Innovation: Uses multi-attention fusion in a lightweight network for rapid remote-sensing landslide identification.
7. Recurrent Landslide-Derived Debris Flows in the Yizhong River Basin: Geomorphic Controls, Structural Predisposition, and Rainfall-Driven Instability
Core Problem: Repeated debris flows require separating persistent geomorphic predisposition from event-scale rainfall forcing.
Key Innovation: Integrates geomorphic, structural and rainfall evidence to explain recurrent debris-flow activity in the Yizhong basin.
8. Hydro-mechanical performance of granite residual soil slopes reinforced by combined biocementation-vegetation treatment under extreme rainfall
Core Problem: Nature-based reinforcement must maintain hydraulic and mechanical performance under extreme rainfall.
Key Innovation: Tests combined biocementation and vegetation in granite residual soil slopes under coupled infiltration and loading.
9. Predicting blast-induced slope stability using a hybrid field-monitoring and machine learning surrogate model
Core Problem: Repeated mine blasting requires rapid stability estimates grounded in measured vibration and jointed-slope mechanics.
Key Innovation: Combines field-monitored acceleration, FLAC/SLOPE simulations and machine-learning surrogates to predict blast-dependent factor of safety.
10. Landslide susceptibility assessment along Sheringal Road, Upper Dir Northern Pakistan using AHP and bivariate statistical models
Core Problem: Data-constrained mountain corridors need susceptibility maps that are transparent, field anchored and comparable across modelling assumptions.
Key Innovation: Evaluates AHP, Frequency Ratio and Statistical Index models against the same field-checked inventory, identifying high-priority Sheringal-Kumrat Road sections.
11. Spatiotemporal autocorrelation of landslides in neighborhoods of Recife, Brazil (2015-2024) based on the Moran and LISA index
Core Problem: Annual event totals obscure whether urban landslide hotspots persist spatially through time.
Key Innovation: Uses Global Moran's I and LISA on a decade of civil-defence records to identify chronic and episodic landslide clusters across Recife.
12. Aseismic Slip Along the Evaporite-Rich Strike-Slip Katouna-Stamna Fault in Greece
Core Problem: Aseismic slip has been observed on many active faults globally, but its controlling mechanisms remain unclear.
Key Innovation: This study focuses on the NW-striking, left-lateral Katouna-Stamna Fault in Western Greece.
13. Structural Heterogeneity Shapes Segmentation of Megathrust Earthquakes and Volcanism in Central America
Core Problem: The mechanisms underlying the patchy distribution of forearc megathrust earthquakes and arc volcanoes in subduction zones remain poorly understood.
Key Innovation: Here we present a high-resolution seismic tomographic model of the Central American subduction zone, derived from a joint inversion of local body-wave and teleseismic Rayleigh-wave data. Our model reveals significant along-trench structural heterogeneities in the crust and upper mantle, closely matching segmentation patterns in the forearc megathrust and arc volcanism.
14. Crustal Shear-Wave Velocity and Azimuthal Anisotropy Imaging Reveals the Merapi-Merbabu-Telomoyo Volcanic Chain, Central Java
Core Problem: Volcanic-chain structure and local deformation are difficult to resolve where regional plate motion and shallow magmatism interact.
Key Innovation: Combines ambient-noise tomography with azimuthal anisotropy to distinguish crustal structure, fluid-rich zones and locally generated stress beneath the Merapi-Merbabu-Telomoyo chain.
15. Spatial heterogeneity of green infrastructure performance in flood mitigation under urban expansion
Core Problem: Green infrastructure (GI) mitigates these impacts by improving infiltration, storage, and surface roughness; however, the pathways linking land-use patterns to hydrological responses remain insufficiently quantified.
Key Innovation: This study investigates the Poyang Lake Basin by constructing three 2044 urban expansion scenarios, including a no-GI scenario, a storage-oriented GI scenario, and a permeability-oriented GI scenario. Results show that urban expansion increases impervious surface coverage from 3.93 to 7.37%, while GI partially mitigates this trend, with permeability-oriented GI achieving the strongest reduction (6.68%).
16. Assessing Meteo-HySEA performance for Adriatic meteotsunami events
Core Problem: Operational meteotsunami forecasts require rapid inundation modelling without losing sensitivity to atmospheric forcing and harbor resonance.
Key Innovation: Benchmarks the GPU-based Meteo-HySEA system against AdriSC-ADCIRC and observations across three Adriatic events, isolating atmospheric forcing as the dominant uncertainty.
17. Seismogenic structures of the 2022 M S 6.1 Lushan earthquake and the M S 6.0 Maerkang earthquake swarm in western Sichuan, China
Core Problem: The Markang M S 6.0 earthquake, on the other hand, took place in the northwestern segment of the Songgang Fault, an area with limited research and varying interpretations of fault activity.
Key Innovation: These events are associated with the complex tectonic setting of the area where multiple active faults are distributed.
18. Bridging the divide: an integrated risk-resilience coupling framework to decode spatial mismatches for precision flood management
Core Problem: Climate change and rapid urbanization have intensified the frequency and severity of flood disasters, posing substantial challenges to regional resilience.
Key Innovation: To address this gap, this study develops an integrated assessment framework that incorporates flood resilience into full-cycle flood risk management. The results indicate that flood risk increases from west to east, with high- and highest-risk areas accounting for 58.0% of the YRD, while high- and highest-resilience areas account for 73.9% and show a similar eastward increase.
19. Tsunami hazard in the Anguilla Bank Archipelago, Lesser Antilles: historical data, fault characterization and numerical simulation
Core Problem: Sparse historical observations leave tsunami exposure around the low-lying Anguilla Bank poorly constrained.
Key Innovation: Simulates local, regional and far-field earthquake sources to quantify coastal amplification and identify bathymetric controls on tsunami impact.
20. Land subsidence and infrastructure failure risk assessment in karst and collapsible soils for climate-resilient infrastructure
Core Problem: Infrastructure decisions require evidence that connects karst and collapsible-soil deformation mechanisms to asset performance.
Key Innovation: Synthesizes mechanism-to-asset observations across karst, evaporite and collapsible-soil settings and identifies the validation needed for risk-informed mitigation.
21. A Unified Path-Dependent Risk Index for CO₂ Leakage via Fault Slip and Caprock Fracture
Core Problem: CO₂ storage risk cannot be represented by independent static thresholds when slip and fracture alter the subsequent stress path.
Key Innovation: Derives a physics-based leakage-risk index from coupled flow-geomechanics simulations and validates its sequence-dependent behavior against two storage projects.
22. Rockfall Impact Response and Graded Design of CPRC Structures using Physics-constrained Machine Learning
Core Problem: Protective structures require impact-response predictions that remain mechanically admissible across rockfall loading regimes.
Key Innovation: Uses physics-constrained machine learning to predict CPRC impact response and translate it into graded protective-structure design.
23. Equity-aware interpretable policy rules for personalized hurricane risk communication
Core Problem: Risk messages must remain interpretable while responding to heterogeneous experience, trust and baseline risk perception.
Key Innovation: Distills causal-forest treatment effects into a shallow fairness-constrained policy tree that retains predictive value while reducing demographic assignment disparity.
24. Next frontier of Earth observation: Quantum-enhanced multispectral remote sensing for ultra-low signal environments
Core Problem: This limitation is not primarily technological but statistical, arising from the assumptions underlying classical photon detection, which fail in ultra-low-light regimes.
Key Innovation: This study introduces quantum-enhanced multispectral remote sensing (QEMRS) as a fundamentally new observational framework that exploits non-classical photon correlations to extract information below the classical noise floor. Primarily focusing on advances in quantum photonics, we synthesize experimental evidence demonstrating sub-shot-noise detection, coincidence-based measurements, and quantum illumination that remains.
25. Integrating hydrogeomorphic changes into flood hazard assessment: approach for an eco-touristic mountainous river
Core Problem: Traditional flood risk assessments often overlook the hydrogeomorphic changes that occur during extreme events, despite their significant impact on subsequent flood behavior.
Key Innovation: This study investigates the influence of riverbed evolution on flood hazards in the Boi River catchment, located in a mountainous region in southern Brazil characterized by canyons, intense orographic precipitation, and ecotourism activities, through the application of the CAESAR-Lisflood Landscape Evolution Model. Findings reveal that more extreme events cause significant bed aggradation and hydraulic spreading, altering.
26. Three-dimensional basin-mountain coupling effects on ground motion amplification in sedimentary basins
Core Problem: Basin response cannot be isolated from surrounding mountain topography when wavelength and geometry interact.
Key Innovation: Resolves frequency-dependent basin-mountain coupling and resonance-like amplification with a multi-domain three-dimensional boundary-element model.
27. Improved Automatic Seismic Bulletins via Likelihood-Based Model Fit Scores for Classification
Core Problem: Automatic bulletins form false events when non-detecting stations are treated as missing rather than informative evidence.
Key Innovation: Transforms detection and non-detection likelihoods into interpretable classifier features that reject most false events while retaining analyst oversight.
28. Governing risk in Anthropocene coasts: institutional frictions and social resilience in post-tsunami Chile
Core Problem: Formal coastal planning can remain disconnected from community risk perception and socioeconomic recovery needs.
Key Innovation: Identifies institutional friction after the 2015 Illapel earthquake-tsunami and proposes closed-loop governance with mixed-resilience zoning.
29. Analysis of 5 years of continuous monitoring (2021-2026) of the Lacq induced seismicity (Southwestern France)
Core Problem: Sparse historical monitoring leaves Lacq earthquake locations and their relationship to the depleted reservoir uncertain.
Key Innovation: Combines a temporary network, a local 3D velocity model, manual relocation and deep-learning detection to produce a more complete five-year catalog.
30. The Natural Range of Variability of Floodplain Spatial Heterogeneity
Core Problem: We use remotely sensed data to quantify the spatial heterogeneity of floodplains along 33 rivers in the United States based on the distribution of floodplain patches.
Key Innovation: Each patch represents a relatively homogeneous, contiguous area of floodplain that is distinct from adjacent patches based on vegetation and soil salinity characteristics at a resolution of 10 m. We find that floodplains along channels with snowmelt-dominated flow regimes have a significantly higher aggregation than those dominated by rainfall flow regimes and floodplains of braided rivers have significantly less diversity.
31. Analyses of Glacial Sediments From Princess Elizabeth Land, East Antarctica Using a Standardised Computational Approach: Grain Size Analysis Tool (GSAT)
Core Problem: Statistical parameters derived from grain size data offer insights into grain size distribution (GSD), facilitating the interpretation of depositional environments and sediment transportation history.
Key Innovation: GSD of glacial sediments is complex, often exhibiting bimodal, trimodal and polymodal distributions, complicating the calculation of statistical parameters. The variance in the results ranged from 0.01 to 1.95, and in some instances, reached up to 4.93 and 49.34 for the skewness and kurtosis values, respectively.
32. A Satellite-Based Estimate of the Contribution of Filamentary Structures to Lateral Carbon Transport in the Pacific and Atlantic Upwelling Systems
Core Problem: Cross-shelf interactions and the associated lateral transport of organic carbon remain poorly quantified and constrained by observations, despite their importance for the ocean carbon cycle.
Key Innovation: Our results provide a scalable observational framework for monitoring lateral carbon transport. We find an offshore transport of 239 ± 25 to 272 ± 28 TgC yr −1, with filaments contributing 8 ± 2 to 10 ± 3 TgC yr −1 (less than 5%).
33. Twenty Years of Satellite Observations Reveal a Shift in the Diurnal Trend of Dust Optical Depth Over West Africa
Core Problem: West Africa is the largest global source of mineral dust, with significant impacts on climate, air quality, and biogeochemical cycles.
Key Innovation: While temporal variations in dust loading from daily to interannual scales are well studied, the long-term evolution of the diurnal cycle itself is often overlooked. We find a striking summertime shift in the diurnal contrast of dust loading: the afternoon-minus-morning DOD difference (DOD) increases significantly over the record, flipping from negative near 2003 to positive by 2022.
34. A Threshold-Free Clustering Framework Disentangles Policy-Linked Greening From Urbanization-Driven Browning in the Yangtze River Delta of China
Core Problem: Accurately characterizing grid-level vegetation dynamics in urban agglomerations remains challenging due to high spatial heterogeneity and complex temporal responses to urbanization.
Key Innovation: We developed a threshold-free clustering framework that integrates the LinCoIndex and structural similarity (SSIM) index to address this issue.
35. SWOT Observations Reveal Basin-Scale Reservoir Operating Patterns
Core Problem: However, complex seasonal operating patterns remain opaque due to data scarcity, limiting the accuracy of hydrological modeling and water management.
Key Innovation: Applying a two-step framework to 305 reservoirs (more than 0.5 km²) in the Ganjiang River Basin (Yangtze), we found that single-cycle regulation (33.4%) and dual-cycle regulation (40.3%) patterns are dominant. Here we show that the Surface Water and Ocean Topography satellite overcomes these challenges by providing full-coverage water surface elevation measurements for a basin-wide network of small reservoirs.
36. Changes in the Global Pattern and Magnitude of Summertime Heatwave Metrics Are Largely Explained by a Mean Shift
Core Problem: Heatwave severity has increased over time, and the degree of change varies over the globe.
Key Innovation: Our results suggest that the use of threshold-based metrics can result in apparent differences in heat trends that can be unified through the proposed shift framework. We demonstrate that the observed spatial pattern of changes in common heatwave metrics can largely be explained by local mean warming.
37. Experimental investigation of scour evolution around a vertical-axis tidal turbine under combined wave-current action
Core Problem: Scour around vertical-axis tidal turbine (VATT) foundations can threaten the long-term stability of tidal-stream energy systems.
Key Innovation: This study experimentally investigates scour evolution around a VATT foundation under combined wave-current action, focusing on the effects of flow intensity, wave height, tip clearance, tip speed ratio ( TSR), and water depth. These results show that VATT scour assessment should consider not only maximum scour depth but also turbine-specific spatial bed deformation.
38. Scour reduction performance of multi-layer collars around a cylindrical monopile
Core Problem: Local scour around offshore wind monopiles can reduce foundation embedment and threaten long-term structural performance.
Key Innovation: This study experimentally investigates single-, double-, and triple-layer collar systems under steady clear-water currents, with emphasis on the effects of collar width, elevation, vertical arrangement, layer number, and flow intensity. The results identify the lower-collar elevation and width as the primary design parameters, while upper collars provide supplementary control of residual scour.
39. Wave breaking characteristics and short-term morphodynamic responses under energetic wave conditions in the Yellow River Estuary
Core Problem: This study investigates the relationships between wave-breaking types and short-term nearshore morphodynamic change in the Yellow River Estuary using an integrated numerical framework.
Key Innovation: This study investigates the relationships between wave-breaking types and short-term nearshore morphodynamic change in the Yellow River Estuary using an integrated numerical framework. Results suggest that depth-induced wave breaking acts as dominant energy sink, forming a stable, shore-parallel high-energy dissipation band at the kilometer scale in the nearshore.
40. Load transfer mechanism of deep braced excavation considering one-strut failure in marine and terrestrial deposit soft soil
Core Problem: Extensive coastal infrastructure construction has led to excavation failures causing social and economic losses.
Key Innovation: Hence, this study investigates the mechanism of one-strut failure in deep braced excavation in marine and terrestrial deposit soft soil, based on a real project in coastal areas. The results indicate that one-strut failure has a significant impact on marine and terrestrial deposit soft soil, with both load transfer ratio and load increment ratio exceeding 20% and an influence radius up to 15 m.
41. Comparison of EDGAR, ODIAC, and MEIC Grid CO₂ Emission Inventories: Historical and Current Versions
Core Problem: High-resolution gridded CO₂ emission inventories underpin carbon cycle research and facilitate bridging global and national climate mitigation targets with regional emission assessments and site-specific mitigation actions.
Key Innovation: Major datasets including ODIAC, EDGAR, MEIC-global and MEIC-China adopt inconsistent spatial proxies and accounting frameworks, causing notable discrepancies. The results reveal global total emission deviations of only 7.97% and 2.03%, yet pronounced regional heterogeneity.
42. Multi-Instrumental Evidence of the 2025 Absorbing Aerosol Perturbation at the RADO-Bucharest Observatory
Core Problem: This paper presents a multi-parameter characterisation of the atmospheric composition at the RADO-Bucharest observatory, a regional WMO-GAW and ACTRIS facility in southeastern Europe, by anchoring recent observations within a multi-annual baseline (2015-2024).
Key Innovation: The application of unified inversion frameworks is demonstrated through case studies of smoke and mineral dust, applying the Generalized Retrieval of Atmosphere and Surface Properties (GRASP) algorithm to retrieve vertically distributed aerosol microphysics through lidar-photometer integration.
43. Satellite Mapping of Annual Center Pivot Irrigation Expansion in Africa’s Hyper-Arid Regions from 1972 to 2025
Core Problem: Currently, these systems are being increasingly deployed across Africa’s hyper-arid regions, where scarce surface water and minimal rainfall force heavy reliance on groundwater extraction.
Key Innovation: In this study, we mapped CPISs from both Landsat and Sentinel-2 archived satellite imagery in Africa’s hyper-arid regions annually from 1972 to 2025. Using a Cascade Mask R-CNN instance segmentation model with a Swin Transformer backbone, we achieved an average precision of 83.62%.
44. Detecting Early Successional Stages in a Glacier Forefield Through a Hierarchical Sentinel-2 Classification Framework
Core Problem: Early ecological succession on newly exposed glacier terrain is difficult to separate consistently from rock, debris, snow and water in satellite imagery.
Key Innovation: Uses a hierarchical Sentinel-2 classification and a natural chronosequence to map moisture-sensitive successional stages across a glacier forefield.
45. A Magnetic Anomaly Detection Method Based on Multi-Feature Classification-Fusion Neural Network in Colored Noise Background
Core Problem: Magnetic anomaly detection (MAD) is a core technology for the detection of ferromagnetic targets, yet traditional methods such as the Orthogonal basis function (OBF) detector suffer from severe performance degradation in low signal-to-noise ratio (SNR) and Gaussian colored noise environments.
Key Innovation: To address this issue, this paper proposes a novel neural network architecture based on manual feature extraction, namely the Partitioned Classification Network with 42 features (PCN-42). Simulation results demonstrate that compared with the OBF detector, the detection probability of the PCN-42 detector is improved by 45-75 percentage points, reaching over 80% in different magnetic moment directions and approximately 90% at.
46. Interpretable Multi-Year Winter Wheat Mapping with Sentinel-1/2 Time Series: SHAP-Based Feature Selection and Bayesian-Optimized Machine Learning
Core Problem: Accurate winter wheat mapping is important for interannual planting-area monitoring, yet models developed independently for each year require repeated sample collection, feature selection, and parameter tuning.
Key Innovation: This study developed an interpretable cross-year transfer approach for winter wheat mapping in the Tailan River Irrigation District, Xinjiang, China, using Sentinel-1/2 time-series imagery. The Bayesian-optimized Random Forest model achieved overall accuracies of 93.82% in the 2025 source-year validation and 92.13% and 91.01% in the 2023 and 2024 target-year tests, with corresponding Kappa coefficients of 0.92, 0.90, and 0.89.
47. Evidence-Based Reliability Assessment of Spatial Transfer Learning for Satellite-Derived Ground Deformation Monitoring
Core Problem: Satellite-derived ground deformation monitoring has become an essential tool for infrastructure management; however, the spatial heterogeneity of Interferometric Synthetic Aperture Radar (InSAR) observations limits reliable assessment in regions with sparse measurement coverage.
Key Innovation: This study presents an evidence-based reliability assessment framework for spatial transfer learning using European Ground Motion Service (EGMS) observations. The proposed local-transfer fusion achieved a mean RMSE of 1.720 mm yr⁻¹, outperforming multi-source transfer strategies.
48. Polar Summer Snow and Ice Albedo Feedbacks Assessed by Satellite Observations and Radiative Kernels
Core Problem: Snow and ice albedo feedback (SIAF) is a key process linking cryospheric changes to global warming.
Key Innovation: Satellite radiative kernels provide another choice for estimating radiative feedbacks, bypassing the complexity of climate model radiative transfer processes. An overall positive SIAF is observed in the Arctic, while Antarctic SIAF shows pronounced spatial heterogeneity, with negative feedback mainly over the East Antarctic coastal sea-ice regions and positive feedback over the West Antarctic marginal seas.
49. Failure Criterion of Rock-Shotcrete Composites Under Hydro-Mechanical Coupling: Incorporating Interface Inclination and Seepage Pressure
Core Problem: The evolution of mechanical properties, failure modes, permeability, and crack propagation of composites was investigated.
Key Innovation: In this study, triaxial compression hydro-mechanical coupling experiments were conducted on rock-shotcrete composites with various interface inclinations (0°, 15°, 30°, and 45°), accompanied by flow volume and acoustic emission monitoring. The results indicate that the 45° interface exerts the most pronounced weakening effect on the mechanical performance of composites.
50. Prestressed Rock Bolt Support Considering the Influence of Blocky Rock Mass Quality
Core Problem: Blocky surrounding rock separated by discontinuities is widely encountered in metro tunnel engineering, and variations in block geometry can substantially affect the deformation and load-transfer behavior of the surrounding rock and the response of prestressed rock-bolt support.
Key Innovation: Taking a Qingdao metro tunnel as the engineering background, a generalized notched rectangular block model was established from the statistical characteristics of field discontinuities. These favorable cases also showed relatively continuous stress-transfer paths and limited stress-reduction zones.
51. Drought-driven spatially heterogeneous crop-value decline in the United Kingdom’s food valley
Core Problem: Agricultural drought is an increasingly important risk-management challenge because it threatens crop productivity, farm profitability, and food-system resilience.
Key Innovation: While drought impacts are widely documented, most assessments focus on average yield losses and provide limited insight into the spatial distribution, variability, and uncertainty of economic impacts, masking local vulnerabilities. Results show substantial spatial heterogeneity in crop value, with potato and winter wheat forming a high-value production corridor in the southeast.
52. RoofSeg: An edge-aware transformer-based network for end-to-end roof plane segmentation
Core Problem: Point-cloud segmentation must preserve boundaries and planar geometry under noise, occlusion and variable sampling density.
Key Innovation: Combines edge-aware attention with adaptive mask and plane-geometry losses in an end-to-end transformer, offering transferable tools for structural and terrain feature extraction.
53. Real-time image absolute orientation framework with Optical-GNSS data fusion for lightweight UAVs
Core Problem: However, achieving absolute accuracy under inertial-inaccessible constraints remains challenging due to the scale drift of visual odometry and the non-stationary noise of global positioning measurements.
Key Innovation: To address these operational challenges, this paper presents a practical hierarchical visual-GNSS fusion framework designed for real-time onboard deployment. However, achieving absolute accuracy under inertial-inaccessible constraints remains challenging due to the scale drift of visual odometry and the non-stationary noise of global positioning measurements.
54. SAAD-SR: A multi-altitude aerial object detection benchmark evaluated using an altitude-conditioned distillation detector
Core Problem: However, detection reliability is often degraded by flight height variation, environmental variability, and reduced object pixel support.
Key Innovation: To address these challenges, the Scale-Aware Aerial Dataset for Scale Robustness (SAAD-SR) is introduced as a large-scale altitude-stratified benchmark comprising 154K UAV frames, of which 104,255 images are fully annotated for object detection, yielding approximately 1.6M labeled object instances across eight classes under diverse illumination, weather, and scene conditions.
55. DecAttNet: A decoder-centric multi-task attention network for rapid, large-scale land use change detection and 2 M resolution products generation
Core Problem: To address the issues of insufficient resolution and a lack of fine-grained categories in traditional land use and land cover (LULC) products, this study integrates lightweight deep learning, a decoder-centric multi-task learning architecture, and a significant change attention mechanism to propose a deep learning framework (Decoder-Centric Attention Network, DecAttNet) for large-scale, high-precision land use change mapping.
Key Innovation: To address the issues of insufficient resolution and a lack of fine-grained categories in traditional land use and land cover (LULC) products, this study integrates lightweight deep learning, a decoder-centric multi-task learning architecture, and a significant change attention mechanism to propose a deep learning framework (Decoder-Centric Attention Network, DecAttNet) for large-scale, high-precision land use change mapping.
56. On the role of observational representativeness in satellite-constrained calibration of lake temperature models
Core Problem: Satellite-derived surface temperature is increasingly used to constrain lake hydrodynamic models where in situ profile observations are sparse.
Key Innovation: This study compared full-profile calibration, direct transfer of a reference-site parameter set, and satellite-constrained transfer calibration within an off-river storage reservoir in southeastern Australia. Satellite-constrained recalibration did not outperform direct transfer, increasing weighted RMSE by 3.2%, 29.2%, and 199.5% across the three sites.
57. Geological constraints on the spatial distribution and morphology of Erosion gullies in the black soil region of Northeast China: a case study of the Heihe area
Core Problem: Gully erosion threatens soil resources and agricultural production in the black soil region of Northeast China, but geological effects are difficult to separate from terrain and land use.
Key Innovation: We combined 5407 erosion gullies from the First National Water Resources Census with 1:250,000 lithological and soil-forming parent material maps, a 12.5 m digital elevation model, and land-use data for the Heihe area. These results support using substrate information to refine terrain-based gully erosion risk zoning.
58. A hybrid CNN-Mamba network with interpretable mixture-of-experts routing for tunnel lining surface defects segmentation
Core Problem: Accurate segmentation of tunnel lining surface defects is fundamental to service condition assessment, yet existing networks suffer from limited modeling efficiency, weak cross-defect generalization, and poor interpretability.
Key Innovation: An interpretable geometric-structure, region large-kernel, and photometric-frequency mixture of experts (GRPMoE) with spatial routing is further introduced, forming the proposed GRPMoE-CNN-Mamba network. Experimental results show that GRPMoE-CNN-Mamba improves mAP@0.5, F1-score, recall and mIoU by 5.48%, 4.76%, 7.25% and 2.41%, respectively.
59. Quantitative effects of artificial lighting parameters on rock discontinuity detection using structure-from-motion photogrammetry in tunnel environments
Core Problem: This study quantitatively investigates how artificial lighting and image acquisition parameters control the reliability of Structure-from-Motion (SfM)-based rock discontinuity detection in tunnel environments.
Key Innovation: This study quantitatively investigates how artificial lighting and image acquisition parameters control the reliability of Structure-from-Motion (SfM)-based rock discontinuity detection in tunnel environments. The most stable results were achieved using a 1.0 m capture distance, 7-8 m lighting distance, 5500 K CCT, and 1000 lx illuminance.
60. Physics-guided symbolic regression of nonlocal transport memory from sparse observations in aquatic systems
Core Problem: Predicting non-Fickian transport in aquatic systems from sparse observations is challenging because historical effects and spatial heterogeneity often control present behavior.
Key Innovation: We introduce a physics-guided, sequential hybrid framework that couples nonlocal transport theory with deep-learning-assisted symbolic regression to infer transport memory directly from one-dimensional observations, such as breakthrough curves (BTCs) or spatial concentration snapshots.
61. Spectral-enhanced cross-domain network: a novel daily runoff prediction model based on energy amplification and cross-domain multi-scale collaboration
Core Problem: However, existing deep learning models exhibit a spectral bias-a tendency to prioritize low-frequency trends over high-frequency signals-resulting in extreme- value smoothing that masks critical information about peak flows.
Key Innovation: To address these challenges, we propose the Spectral-Enhanced Cross-domain Network (SEC), a novel daily runoff forecasting model that integrates a spectral energy modulator with cross-domain multi-scale collaboration.
62. High-spatiotemporal-resolution GRACE(-FO)-constrained terrestrial water storage reconstruction captures the hydrological footprints of successive typhoons in southern mainland China
Core Problem: The southern coastal region of mainland China faces persistent typhoon-related hydrological hazards, highlighting the need to monitor rapid variations in terrestrial water storage during extreme precipitation events.
Key Innovation: This study proposes a cascading framework that combines empirical-statistical reconstruction with machine-learning downscaling of reconstruction parameters to generate a climate-driven terrestrial water storage anomaly (TWSA) reconstruction at high spatiotemporal resolution. Cross-product comparisons with GRACE(-FO)-based products and estimates from land surface models showed overall consistency.
63. Integrated remote sensing retrieval of surface and root-zone soil moisture through physical mechanisms-guided machine learning
Core Problem: However, the limited penetration capability of remote sensing signals into the soil presents significant challenges for the retrieval of RZSM.
Key Innovation: To address these challenges, this study proposed an integrated retrieval framework for SSM and RZSM that combines physical mechanisms with ML models. Physical and machine learning (ML) retrieval models require extensive labeled data to achieve high-accuracy retrieval, which poses difficulties in regions with limited in-situ data.
64. A hybrid peridynamics-finite element method for thermo-hydro-mechanical modeling of frost cracking in frozen soils
Core Problem: Freeze-induced soil damage requires simultaneous representation of heat transfer, water migration, phase change and unconstrained crack growth.
Key Innovation: Couples peridynamics for crack evolution with finite elements for thermo-hydro-mechanical transport, reproducing damage initiation and propagation without prescribed crack paths.
65. Soil-structure interaction on resilient systems, A state-of-the-art review of modeling approaches, design provisions, mitigation strategies and emerging computational tools
Core Problem: Seismic resilience assessments often fragment soil-structure interaction, mitigation and post-event functionality across separate modelling traditions.
Key Innovation: Integrates coupled ground-structure mechanics, code practice, mitigation, fragility and emerging physics-informed tools into a unified resilience-focused review.
66. Multiscale physics-informed neural network for road effective roughness identification based on vehicle-road coupled system
Core Problem: However, traditional roughness identification methods are often limited by stationary assumptions and exhibit insufficient robustness under complex, non-stationary road conditions, particularly in the presence of localized surface defects.
Key Innovation: To address these limitations, this paper proposes a multiscale physics-informed neural network (MS-PINN) framework for ReR reconstruction based on vehicle-road coupled dynamics, embedding vehicle-road dynamic equilibrium, tire-road contact feasibility, multiscale frequency characteristics, and power spectral density (PSD) consistency into a unified learning architecture.
67. Significant Imprints of Vertical Resolution on Scale Interactions in the Global Model for Prediction Across Scales
Core Problem: Global kilometer-scale model simulations have demonstrated clear benefits from refining horizontal resolution in representing mesoscale weather phenomena, yet the role of vertical resolution remains comparatively underexplored.
Key Innovation: To address this gap, 40-day experiments with the model for prediction across scales systematically varied both horizontal and vertical grid spacing. The results show that vertical resolution impinges on the mesoscale KE spectrum, substantially influencing the kinematics of the mid-latitude upper-troposphere, and underscoring the importance of carefully prescribing vertical grid spacing in high-resolution global models.
68. Lakes Modify the Magnitude and Timing of the Northern Hemisphere Terrestrial Cryosphere Radiative Effect
Core Problem: Vast areas of Earth's Northern Hemisphere are covered in lakes that form seasonal ice cover, yet their influence on the terrestrial Cryosphere Radiative Effect (CrRE t) has not previously been quantified.
Key Innovation: Here we use 22 years of satellite data to constrain the contribution of lakes to CrRE t. We find that, per unit area, lakes have a significantly higher CrRE t than land (−14.4 vs. −8.2 W/m²).
69. Comparative Field Evaluation of Active and Passive Transient-Based Techniques for Pipeline Condition Assessment
Core Problem: Buried pipeline condition must be inferred from pressure transients whose range and diagnostic value differ between active and passive measurements.
Key Innovation: Compares controlled and ambient transient methods in a field network, quantifying detection range, accuracy and complementary sensitivity to localized features.
70. Experimental Study on Mechanical Characteristics and Energy Evolution of Sandstone Under True Triaxial Cyclic Loading and Unloading Conditions
Core Problem: Deep rock is subjected to true triaxial cyclic loading in a complex stress environment during deep mining.
Key Innovation: The effect of intermediate principal stress on the mechanical behavior and energy evolution of rock corresponding to principal stress directions exhibits significant differences, which are essential for disaster prevention in deep mining. Differences in the effect of intermediate principal stress on these characteristics corresponding to different principal stress directions were revealed.
71. Space-time dynamics of rainfed wheat yield in Ethiopia: Insights from remote sensing
Core Problem: Ethiopia is the largest wheat producer in sub-Saharan Africa, yet national food security efforts are constrained by coarse administrative level yield data that lack spatial and temporal detail.
Key Innovation: To bridge this gap, we present a spatio-temporally explicit assessment of rainfed wheat productivity, stability, and trends across Ethiopia from 2016 to 2023 using a multi-scale, data-driven framework. Model performance reached an R² of 0.56 (RMSE = 0.85 t ha⁻1) under random cross-validation and declined to 0.29 under spatial block cross-validation, with approximately 80% of the rainfed wheat area falling within the model’s.
72. The role of poroelastic residual stresses in hydraulic fracture deflection
Core Problem: While material heterogeneity is known to alter local stress fields, the mechanism by which a mismatch in poroelastic properties controls fracture trajectories remains poorly understood.
Key Innovation: While material heterogeneity is known to alter local stress fields, the mechanism by which a mismatch in poroelastic properties controls fracture trajectories remains poorly understood.
73. Mechanism of Soft Soil Stabilization Using a CS-AR-Activated FASP System
Core Problem: To address the poor engineering performance of sludge-like soft soils, the high carbon emissions associated with conventional cement stabilization, and the unstable alkalinity supply of single-alkali activation systems, this study develops a solid waste-based FA-GGBFS-PS (FASP) stabilization system, in which a precursor consisting of fly ash (FA), ground granulated blast furnace slag (GGBFS), and phosphorus slag (PS) is.
Key Innovation: To address the poor engineering performance of sludge-like soft soils, the high carbon emissions associated with conventional cement stabilization, and the unstable alkalinity supply of single-alkali activation systems, this study develops a solid waste-based FA-GGBFS-PS (FASP) stabilization system, in which a precursor consisting of fly ash (FA), ground granulated blast furnace slag (GGBFS), and phosphorus slag (PS) is.
74. A multiscale framework for predicting swelling pressure in compacted bentonite incorporating molecular dynamics and modified Poisson-Boltzmann theory
Core Problem: Existing swelling-pressure models are largely empirical and lack general applicability, while classical diffuse double layer (DDL) theory relies on simplifying assumptions that limit its validity under highly compacted conditions.
Key Innovation: This study proposes a multiscale swelling-pressure prediction framework that couples molecular dynamics (MD) simulations, modified Poisson-Boltzmann (MPB) theory, and fractal geometry. The simulations reveal that, under highly compacted conditions, ion hydration leads to heterogeneous ion distributions characterized by ion crowding near montmorillonite surfaces and redistribution toward the interlayer center.
75. Mechanical behavior of persistently jointed rock masses under uniaxial compression
Core Problem: It plays a decisive role in controlling the strength and failure modes of rock specimens containing a single persistent inclined joint.
Key Innovation: It plays a decisive role in controlling the strength and failure modes of rock specimens containing a single persistent inclined joint. The results indicate that peak strength increases nonlinearly with JRC, with different trends before and after the turning point at JRC = 10, depending on the rock strength group.
76. Effects of biochar amendment and root density on water and gas permeability of unsaturated soils under wetting and drying cycles
Core Problem: Biochar and plant roots can modify soil WAGP, yet their combined effects in unsaturated soils are poorly understood.
Key Innovation: This study examined soils amended with 0%, 5%, and 10% biochar and vetiver grass roots at three densities (0, 3, and 6 clumps), compacted to 90% relative density. Results showed that biochar and roots significantly reduced WAGP.
77. Damage degradation characteristics and constitutive modeling of hot dry granite subjected to cyclic thermal shock: An acoustic emission perspective
Core Problem: Liquid nitrogen (LN2), as a waterless fracturing medium, is promising for enhancing the permeability of hot dry rock (HDR) geothermal reservoirs owing to its cryogenic temperature and environmental compatibility.
Key Innovation: To clarify the damage degradation mechanism of hot dry granite under cyclic thermal shock, uniaxial compression tests were performed on granite specimens subjected to different heating temperatures, ranging from 250 °C to 850 °C, followed by 1-9 LN2 cooling cycles. The results show that both uniaxial compressive strength (UCS) and elastic modulus decrease significantly with increasing temperature and LN2 cooling cycles, with.