TerraMosaic Daily Digest: Jan 29, 2026
Daily Summary
This digest synthesizes 186 selected papers and focuses on landslide process mechanics and slope evolution, seismic source-to-ground response pathways, coastal and submarine hydro-geomechanics. Top-ranked studies examine flood generation and hydroclimatic forcing, satellite and LiDAR-based deformation monitoring, and coastal and submarine hazard coupling.
Across the full set, evidence converges on mechanism-constrained analysis with operational relevance, especially for infrastructure-focused hazard performance and flood generation, routing, and hydroclimatic forcing. The strongest contributions pair interpretable process evidence with monitoring or forecasting workflows that support warning design and risk prioritization.
Key Trends
- Landslide studies increasingly resolve process chains: Contributions connect triggering conditions, slope deformation, and mobility outcomes, improving the basis for warning thresholds and scenario testing.
- Seismic hazard research links source behavior to ground response: Recurring topics connect rupture or loading conditions with geotechnical performance and consequence assessment.
- Coastal and submarine hazards are treated as coupled systems: Wave, mass-transport, and shoreline processes are analyzed together with engineering implications.
- Infrastructure-facing outputs are increasingly decision-ready: Asset performance is evaluated with uncertainty-aware frameworks to support mitigation and maintenance prioritization.
- Flood analyses are becoming event-specific and process-based: Papers emphasize precipitation structure, antecedent wetness, and catchment controls rather than static hazard descriptors.
Selected Papers
This digest features 186 selected papers from 1141 papers analyzed (out of 2633 RSS items) across multiple journals. Each paper has been evaluated for its relevance to landslide and broader geohazard research and includes links to the original publications.
1. Comprehensive analysis of rainfall-deformation characteristics for railway slopes - a case study of the Dazhou-Chengdu railway in China
Core Problem: Railway slope risk management lacks integrated multi-scale deformation diagnostics and quantitative rainfall-deformation linkage analysis.
Key Innovation: The study combines UAV and field mapping with InSAR-based time-series deformation analysis and a comprehensive rainfall-correlation metric, showing that cumulative multi-day rainfall better explains slope deformation than intensity alone.
2. Enhancing surface cracking and erosion resistances in lateritic soil with eco-friendly surface spraying agents
Core Problem: Lateritic slopes are highly susceptible to desiccation cracking and rainfall erosion during early vegetation establishment, undermining long-term stability.
Key Innovation: Through comparative laboratory testing of eco-friendly spray treatments, the work identifies polyacrylamide as the most effective agent for simultaneously improving tensile strength, suppressing cracking and erosion, and preserving plant growth.
3. SMFCs-driven lead remediation: morphological transformation and targeted migration in contaminated soils
Core Problem: Optimizing lead remediation with soil microbial fuel cells requires understanding how contamination level and electrode spacing jointly control electrochemical performance and Pb migration.
Key Innovation: Dual-chamber SMFC experiments across Pb concentrations and cathode-anode spacings quantify power-remediation tradeoffs, document cathodic Pb deposition and migration, and identify operating conditions that maximize remediation efficiency.
4. Water resources modeling in Wadi Numan, Western Saudi Arabia using remote sensing and GIS-based multi-criteria
Core Problem: Water-resource prospecting in hyper-arid Wadi Numan demands robust integration of geomorphic, structural, hydrologic, and climatic controls into predictive mapping.
Key Innovation: The study fuses 14 remote-sensing/GIS layers using frequency-ratio and evidential-belief-function models, validates outputs with wells and AUC metrics, and demonstrates stronger performance of the frequency-ratio approach for groundwater potential zoning.
5. Tsunamigenic potential of unstable masses in the Gulf of Pozzuoli, Campi Flegrei, Italy
Core Problem: The potential hazard posed by gravitational instabilities (unstable masses) in the Campi Flegrei area and their ability to trigger tsunamis has received little consideration.
Key Innovation: Reconstruction and numerical simulation of four landslide-tsunami scenarios to assess tsunami energy distribution, identify affected coastal stretches, and explore dispersion and resonance effects, providing insights into the tsunamigenic potential of unstable masses.
6. AI in remote sensing and satellite image processing-a review
Core Problem: Rapid expansion of AI-based remote sensing workflows requires a critical synthesis of real-world applications and unresolved technical bottlenecks.
Key Innovation: This review consolidates AI applications in land-cover mapping, object detection, climate monitoring, and disaster management, and distills key priorities in generalizability, data fusion, and computational efficiency for next-generation systems.
7. Development of a Novel Expansion-Type Energy-Dissipating Device for Flexible Rockfall Barriers: Static and Dynamic Performance Analysis
Core Problem: Existing energy dissipating devices (brakes) in flexible rockfall barriers often display inadequate performance stability, inefficient energy dissipation capacity, susceptibility to dynamic loading, and corrosion vulnerability, leading to unexpected activation force or total energy dissipation.
Key Innovation: Introduced a novel aluminum alloy expansion-type brake with mixed friction and plastic deformation energy dissipation mechanisms. A sample was produced and tested to verify manufacturability and stable performance. Quasi-static and drop hammer impact tests characterized static and dynamic performance, leading to recommendations for dynamic coefficients for activation and working loads.
8. Effect of Seepage Behavior on Frost-Heaving Pressure of Fractured Rocks: Numerical Approach
Core Problem: In cold regions, naturally fractured rocks undergo freezing cracks when frozen, potentially causing reduced stability with risk of rock falls, and the influence of seepage behavior on this process is not fully understood.
Key Innovation: Developed a multi-field coupled numerical simulation method integrating ice–water phase change, rock fracture propagation, seepage, and heat transfer, proposing an equivalent water expansion method for frost heaving. The study demonstrated that freezing direction significantly influences frost heave pressure and that permeability thresholds (10^-16 m^2 and 10^-18 m^2) determine the significance of frost-heaving pressure and seepage, respectively, providing insights for cold-region rock engineering.
9. Direct cause of the prehistoric catastrophe revealed by the sedimentary provenance source of the overburden layer in the Lajia ruins in western China
Core Problem: Identifying the direct cause of a prehistoric catastrophe that buried the Lajia ruins in western China.
Key Innovation: Uses sedimentary provenance analysis of the overburden layer to reveal the direct cause of the prehistoric catastrophe, likely a major geohazard event.
10. Effects of cylindrical-obstacle spacing on granular shock wave interactions in gravity-driven flows
Core Problem: Understanding the complex dynamics of granular flows, specifically how obstacles affect shock wave interactions in gravity-driven flows, which is critical for modeling debris flows and rock avalanches.
Key Innovation: Investigation of the effects of cylindrical-obstacle spacing on granular shock wave interactions in gravity-driven flows, providing fundamental insights into landslide-related phenomena like debris flows.
11. Experimental study on variable-mass seepage model of water-rich altered granite
Core Problem: Investigating the catastrophic mechanism of water inrush and mud outburst when tunnels traverse water-rich altered granite strata, which exhibits significant 'variable mass' seepage failure.
Key Innovation: Established a three-stage cyclic mechanism ('seepage channel connection - rock particle migration - local skeleton collapse') for variable mass seepage failure, identifying water pressure as the dominant factor and initial porosity's complex non-linear impact, providing a theoretical basis for disaster prevention.
12. Fatigue damage evolution mechanism of sandstone under Freeze–Thaw and loading rate coupling based on NMR and AE techniques
Core Problem: Understanding the fatigue mechanical response and damage mechanisms of sandstone under the coupled effects of freeze-thaw cycles and varying loading rates, which contributes to rock mass degradation and slope instability in cold climates.
Key Innovation: Demonstrated that freeze-thaw damage significantly weakens sandstone by increasing macropores and microcracks, and that while higher loading rates can increase cyclic strength, freeze-thaw damage remains the primary cause of strength degradation, providing insights for rock engineering stability.
13. Experimental study on failure mechanism of clay seabed induced by gas migration
Core Problem: Shallow gas in marine clay sediments poses significant geohazards to offshore infrastructure, but predictive frameworks linking sediment properties to seabed failure patterns remain inadequate.
Key Innovation: Through quasi-2D gas injection experiments, identifies three distinct gas migration modes (discrete bubble, stable cavity, conduit-dominated pipe flow) and establishes a phase diagram highlighting shallow-water, high-strength sediment zones as having the highest geohazard potential.
14. Fuzzy decision-driven multi-dimensional seismic fragility analysis of shield tunnels in liquefiable strata
Core Problem: Accurate seismic fragility analysis for shield tunnels in liquefiable strata is essential, but current methodologies overlook uncertainties from sampling insufficiency and inhomogeneity, and rely on empirically selected IMs and single DMs.
Key Innovation: Presents an advanced framework for fuzzy decision-driven multi-dimensional seismic fragility analysis, developing fuzzy probabilistic seismic demand models, determining optimal IMs using fuzzy AHP and TOPSIS, and conducting multi-dimensional fragility analysis with dual DMs for improved reliability.
15. Accounting for nonuniform correlation structure: Bayesian evidence-based selection of 3D auto-correlation functions for loess properties
Core Problem: Accurately capturing the site-specific spatial variability of geotechnical properties in loess deposits is essential for reliable geotechnical design and risk assessment, but conventional 3D random field modeling assumptions may not hold.
Key Innovation: Proposes a Bayesian selection framework to select appropriate 3D auto-correlation functions (ACFs) for loess properties, demonstrating that conventional assumptions of shared ACF structures may not hold and providing evidence-based guidance for property-specific random field models.
16. Influence of Lateral Load on the Seismic Behavior of Pin-Piled Jacket Structures for Offshore Wind Turbines Founded in Liquefiable Soils
Core Problem: The influence of lateral load on the seismic behavior of pin-piled jacket structures for offshore wind turbines founded in liquefiable soils needs to be understood for reliable design.
Key Innovation: Investigates the influence of lateral load on the seismic behavior of pin-piled jacket structures for offshore wind turbines in liquefiable soils.
17. Bearing behavior and damage mechanisms of GFRP micro uplift piles in coastal areas: Field tests and numerical simulations
Core Problem: Understanding the bearing capacity and damage mechanisms of GFRP micro uplift piles used in coastal environments.
Key Innovation: Combination of field tests and numerical simulations to assess the performance and resilience of piles against coastal forces.
18. Turbulence characteristics and energy cascade in wave-current flow over an adverse sloping bed
Core Problem: Understanding the fundamental turbulence characteristics and energy transfer mechanisms in combined wave-current flow over an adverse sloping bed.
Key Innovation: Detailed analysis of flow dynamics and energy cascade, providing critical insights for modeling coastal erosion and submarine slope stability.
19. Preserving Native Spatial Resolution in Long-Term Satellite Datasets Through Improved Projection Algorithms
Core Problem: Temporal gravity field solutions from GRACE-FO are constrained by aliasing effects from imperfect background models, and the combined impact of updated ocean tide (FES2022) and nontidal dealiasing (AOD1B RL07) products on KBR- and LRI-based estimations is insufficiently quantified.
Key Innovation: Assesses the influence of FES2022 and AOD1B RL07 on GRACE-FO monthly gravity field solutions, demonstrating that these updated models effectively reduce noise levels and enhance temporal consistency of mass variation signals, with LRI-based solutions showing more pronounced noise reduction, thus refining background models is crucial for LRI potential.
20. DTWSTSR: Dual-Tree Complex Wavelet and Swin Transformer Based Remote Sensing Images Super-Resolution Network
Core Problem: Critical soil moisture (SM) thresholds for large wildfires remain poorly characterized across different data sources, and volumetric SM measurements vary in magnitude and dynamic range across satellite products, making direct comparisons challenging.
Key Innovation: Calculates the Fraction of Available Water (FAW) from SMOS, AMSR2, SMAP, and CCI satellite observations to explore antecedent conditions for major wildfires, revealing low FAW (plant stress < 0.50, extreme drought < 0.20) throughout the preceding month, demonstrating that FAW thresholds provide a robust framework for identifying SM levels predisposing areas to wildfire danger.
21. Water retention behaviour of 3D-printed double-structured soil
Core Problem: The hydro-mechanical and water-retention behavior of 3D-printed double-structured soils remains poorly characterized despite growing engineering interest.
Key Innovation: The paper reconstructs drying-wetting water-retention paths using multi-scale volumetric monitoring (including 3D laser scanning), links fabric evolution to retention behavior through MIP/SEM analyses, and develops a model that explicitly accounts for dual-structure pore networks.
22. Enhancing two-week live fuel moisture content forecasts through biophysical modelling and remote sensing data assimilation
Core Problem: Improving the accuracy and lead time of live fuel moisture content forecasts, which are critical for wildfire prediction.
Key Innovation: Enhancing two-week live fuel moisture content forecasts by integrating biophysical modeling with remote sensing data assimilation techniques.
23. Peridynamics modeling of multi-fissure occurrence in the subsiding North China plain
Core Problem: Modeling the occurrence of multiple fissures in subsiding plains.
Key Innovation: Application of peridynamics modeling to simulate multi-fissure occurrence in the subsiding North China Plain.
24. Global compound drought–hot events: insights from a 3D-event based framework, intercontinental synchronization, and the evolving influence of climatic drivers
Core Problem: Understanding global compound drought-hot events, their intercontinental synchronization, and evolving climatic drivers.
Key Innovation: Providing insights into global compound drought-hot events using a 3D-event based framework, analyzing intercontinental synchronization and evolving climatic drivers.
25. A locally resonant seismic metamaterial with a low-frequency broadband bandgap
Core Problem: Developing effective strategies to attenuate low-frequency broadband seismic waves to protect structures and infrastructure from earthquake damage.
Key Innovation: Design and analysis of a locally resonant seismic metamaterial capable of creating a low-frequency broadband bandgap for vibration attenuation, offering a novel approach to earthquake mitigation.
26. Flood risks to the financial stability of residential mortgage borrowers: an integrated modeling approach
Core Problem: Uninsured households struggle to meet funding needs for property damage from flooding, leading to financial distress and credit constraints, with little known about their prevalence and drivers.
Key Innovation: A simulation-based approach to estimate the impact of uninsured flood damage on residential mortgage borrowers' financial conditions, identifying those likely to have unmet funding needs due to flood-related credit constraints.
27. Coseismic surface rupture probabilities from earthquake cycle simulations: influence of fault geometry
Core Problem: Existing Probabilistic Fault Displacement Hazard Analysis (PFDHA) models are limited by incomplete empirical datasets and do not adequately capture the influence of physical fault parameters like geometry on surface rupture occurrence and variability.
Key Innovation: Using the RSQSim earthquake simulator to systematically evaluate the influence of various fault geometries (connectivity, dip, sinuosity) on the probability and spatial variability of primary surface rupture, demonstrating its key role in seismic hazard assessment.
28. BuRNN (v1.0): a data-driven fire model
Core Problem: Fires are complex phenomena that are challenging to model numerically, and existing process-based models have limitations in accurately simulating burned area.
Key Innovation: Presenting BuRNN (v1.0), a data-driven model using Long Short-Term Memory networks to simulate burned area globally, outperforming process-based models and providing insights into regional fire drivers.
29. High-resolution geodetic velocities reveal role of weak faults in deformation of Tibetan Plateau
Core Problem: Understanding the mechanisms of deformation in the Tibetan Plateau and the specific role of weak faults.
Key Innovation: High-resolution geodetic velocities revealing the significant role of weak faults in the deformation of the Tibetan Plateau.
30. Extreme plate boundary localization promotes shallow earthquake slip at the Japan Trench
Core Problem: Understanding the factors that promote shallow earthquake slip at subduction zones like the Japan Trench.
Key Innovation: Discovery that extreme plate boundary localization is a key factor promoting shallow earthquake slip at the Japan Trench.
31. The analysis of primary circulation flow patterns and environmental parameter characteristics of short-duration heavy precipitation during the warm seasons in the Western Tianshan Mountains, Xinjiang
Core Problem: Understanding the atmospheric circulation patterns and environmental conditions that lead to short-duration heavy precipitation in the Western Tianshan Mountains, a trigger for floods and landslides.
Key Innovation: Identified three dominant circulation patterns (Low-Trough, Low-Vortex, Eastward-moving Waves) and their distinct characteristics, showing how they influence the intensity, location, and daily cycle of heavy rainfall, providing a basis for precipitation forecasting.
32. Optimizing methods of excavation dewatering for deformation control in strata with leaky aquifers: a case study in Tianjin, China
Core Problem: Excavation dewatering in strata with leaky aquifers often leads to excessive deformation and building settlement, with a lack of research and countermeasures for this specific problem.
Key Innovation: Proposes and verifies a countermeasure combining deep-shallow-wells with groundwater recharge, demonstrating its effectiveness in significantly reducing confined aquifer drawdown and building settlement compared to conventional mixed-well schemes.
33. Probabilistic analysis of soil nails placed in random soil fields with rotated anisotropic strength
Core Problem: The influence of rotated anisotropic random fields of soil strength on the probabilistic margins of safety for soil nail arrangements in excavations has not been examined.
Key Innovation: Examines for the first time the influence of rotated anisotropic random fields of soil strength on probabilistic safety margins for soil nails, identifying a detectable worst bedding direction for failure probability, and providing lessons for random finite difference method application.
34. Probabilistic analysis for large strain radial consolidation of soft soils considering creep
Core Problem: Effective design of prefabricated vertical drains (PVDs) requires accurate prediction of soil consolidation behavior, incorporating creep and spatial variability in soil properties.
Key Innovation: Develops a probabilistic analysis framework integrating random field theory, piecewise-linear method, and Monte Carlo simulation to evaluate long-term consolidation of soft soils with PVDs, accounting for various factors and validating against field measurements.
35. Novel erosion law based on CFD–DEM simulations and its application in hydromechanical modeling of gap-graded soils
Core Problem: Existing erosion laws for suffusion often do not incorporate the influence of stress state into the mass exchange between liquid and solid phases, limiting hydromechanical modeling accuracy.
Key Innovation: Develops a novel erosion law incorporating stress state into mass exchange for suffusion using CFD–DEM simulations, integrates it into a four-constituent hydromechanical framework, and assesses its ability to capture soil behavior before and after suffusion.
36. Hydrodynamics-based geometric optimisation and parametric analysis of a wing-type floating breakwater
Core Problem: Improving the design and performance of wing-type floating breakwaters for coastal protection.
Key Innovation: Application of hydrodynamics-based geometric optimization and parametric analysis to enhance breakwater efficiency.
37. An analytical solution and its data-driven applicability domain for oscillating submerged horizontal plate breakwaters
Core Problem: Developing a robust analytical solution for oscillating submerged horizontal plate breakwaters and defining its practical applicability.
Key Innovation: Provision of an analytical solution complemented by a data-driven applicability domain for effective coastal protection design.
38. Dynamic Strength and Energy Dissipation Characteristics of Coal Specimens Under Multiaxial Dynamic-Static Loading
Core Problem: Investigating the multi-axial stress states of coal under deep mining conditions and the disaster-inducing mechanisms associated with coupled static-dynamic disturbances is crucial for preventing catastrophic incidents (dynamic hazards) in underground coal mines.
Key Innovation: Performed systematic experiments using a true triaxial Hopkinson bar system, revealing significant pre-applied static load dependence of coal's dynamic strength, loading-configuration-dependent correlations between energy dissipation density and dynamic strength, and identifying mesoscopic mechanisms, providing a theoretical foundation for preventing dynamic hazards in underground mining.
39. Characterizing and Modeling the Dynamic Compressive Behavior of Rock-Concrete Combined Body Considering Rock Type and Joint Roughness Characteristics
Core Problem: Lack of understanding of the influence mechanisms of strain rate variations, lithological differences, and joint roughness coefficients (JRC) on the dynamic compressive behavior and failure characteristics of rock-concrete composites, which are used for rock-burst hazard prevention.
Key Innovation: Systematically investigated these influences through controlled impact tests, revealing significant strain rate effects and greater sensitivity of hard rock composites to JRC variations (4-20). Established a dynamic constitutive model based on elastoplasticity theory, incorporating JRC, strain rate, and rock type, which accurately predicts compressive properties across various strain rate levels.
40. From Local Spalling to Blasting Jet: Lithology-Dependent Failure Mechanisms of Composite Rocks Under Varied Loading Paths
Core Problem: Understanding the destabilization and fracturing mechanisms of composite rocks under varied loading paths and lithologies is vital for evaluating coal mine strata stability and providing early warnings for dynamic hazard risks.
Key Innovation: Investigated fracturing mechanisms using acoustic emission (AE), digital image correlation (DIC), and high-speed camera technologies, identifying elastic energy accumulation as the key factor influencing failure modes (local spalling to blasting jet). The study characterized distinct failure processes and proposed monitoring elastic energy accumulation for early warning of dynamic hazard risks.
41. Gouge Epidotization and Lateral Heterogeneity Control Granitoid Fault Friction in Geothermal Reservoirs
Core Problem: Understanding the impacts of rock epidotization and its heterogeneity on the frictional strength and stability of earthquake-capable faults in granitoid rocks within active geothermal reservoirs.
Key Innovation: Completed fault reactivation experiments on laboratory faults containing homogeneous and lateral heterogeneous epidote-quartz mixed gouges. Results showed that elevating epidote proportion or patch size enhances instability, with a transition from stable to unstable frictional sliding occurring at >60 vol.% epidote content for homogeneous faults, aiding in understanding enhanced instability potential in EGS reservoirs.
42. A New Sea-Bolt for Time-Delayed Rockbursts: Investigation of Anchorage Force and Energy-Absorption Characteristics
Core Problem: Time-delayed rockbursts in tunnels pose severe risks, and existing reinforcement solutions may not adequately address their mechanical and deformation characteristics.
Key Innovation: Proposed and evaluated a novel self-expanding, energy-absorbing rock bolt (SEA-bolt) by integrating a self-expanding cement roll with an energy-absorbing compression-sleeve structure. Tests showed it increases anchorage force by 79.9% and achieves superior static (119.4 kJ) and dynamic (79.7 kJ) energy-absorption capacities, confirming its strong potential for rockburst mitigation and control.
43. Millisecond Delay Time Optimization and Its Effect on Blasting Vibration Reduction in an Open Pit Mine
Core Problem: The negative effects of blasting vibrations on open slopes and buildings in open-pit mines, requiring optimization of blasting parameters for vibration reduction.
Key Innovation: Proposed a reasonable delay time calculation method based on interval estimation and staggered-phase superposition. The optimal delay time of 22 ms (for a dominant principal frequency of 23 Hz) was determined and verified by numerical simulation and field tests, achieving significant average vibration reduction rates (29.51% horizontal radial, 10.14% horizontal tangential, 23.68% vertical) compared to original delay times.
44. Probabilistic estimation of shear-wave velocity profiles and site classes from surface wave surveys for accelerometer stations in the southern Korean Peninsula
Core Problem: The need for accurate site-correction of seismic recordings and reliable shear-wave velocity (VS) profiles for accelerometer stations in seismically active regions to improve seismic hazard analysis, while addressing non-uniqueness in VS profile and site parameter estimations.
Key Innovation: Developed a probabilistic inversion framework using an ensemble-based stochastic approach for joint inversion of MASW and MAM data. This approach provides VS profiles and site parameters (VS30, soil layer, bedrock depth) with quantified uncertainties for 21 accelerometer stations, enabling more reliable ground motion predictions for probabilistic seismic hazard analyses.
45. Cumulative effects due to seismic sequences in RC frames designed with force- and energy-based approaches
Core Problem: Current seismic standards predominantly focus on a single reference earthquake, neglecting the cumulative impact of earthquake sequences, which can lead to severe structural damage or collapse even when not exceeding design-level peak ground accelerations.
Key Innovation: Investigated the seismic response of a 6-story reinforced concrete bare frame, designed with both force- and energy-based approaches, when subjected to near- and far-field earthquake sequences. Findings indicate that dissipated energy is a more adequate parameter than maximum deformation for evaluating the evolution of cumulative damage during a seismic sequence.
46. Investigating the catastrophe mechanism and evolution of anti-frost subgrade in high-speed railways under extreme climatic events
Core Problem: N/A (Abstract not provided)
Key Innovation: N/A (Abstract not provided)
47. Complex Network-Based Detection and Forecasting of High-Intensity Tropical Cyclones
Core Problem: N/A (Abstract not provided)
Key Innovation: N/A (Abstract not provided)
48. Random modeling of dynamic wind loads in tropical cyclones and its application to transmission tower fragility and reliability analyses
Core Problem: N/A (Abstract not provided)
Key Innovation: N/A (Abstract not provided)
49. InSAR analysis using both co- and cross-polarized data at Death Valley, California from 2017–2025
Core Problem: Monitoring and analyzing ground deformation in Death Valley, California, using advanced remote sensing techniques.
Key Innovation: Application of InSAR analysis, utilizing both co- and cross-polarized data, for ground deformation monitoring in Death Valley over an extended period.
50. Mapping land uses following tropical deforestation with location-aware deep learning
Core Problem: Accurately mapping complex land use changes that occur after tropical deforestation, which is a key factor in landslide susceptibility.
Key Innovation: A location-aware deep learning approach for mapping land uses in areas affected by tropical deforestation.
51. Environmental resistance parameters generated by soil and semi-arid crops under inter-rill erosion and overland flow with low Reynolds numbers
Core Problem: Quantifying environmental resistance parameters of soil and semi-arid crops under inter-rill erosion and overland flow conditions, which are precursors to shallow landslides.
Key Innovation: Generates and analyzes environmental resistance parameters for soil and crops under specific erosion conditions, contributing to understanding landslide initiation mechanisms.
52. Development of a predictive, risk-based model to assess the effects of maintenance decisions on vertical mine shaft structures
Core Problem: Assessing the impact of maintenance decisions on the structural integrity and safety of vertical mine shaft structures, which are susceptible to ground conditions and potential ground deformation.
Key Innovation: Development of a predictive, risk-based model to evaluate these effects, enabling better maintenance planning and hazard management for mine infrastructure.
53. Investigation into the stress corrosion behavior of cable bolts under different tensile stresses
Core Problem: Understanding the stress corrosion behavior of cable bolts under various tensile stresses to ensure their long-term reliability as ground support elements.
Key Innovation: An investigation into the mechanical and chemical degradation of cable bolts, providing crucial insights into their durability and effectiveness as a mitigation measure in challenging ground conditions.
54. An integrated analytical approach for predicting structural performance and cracking behavior in composite linings of deep hydraulic tunnels
Core Problem: Predicting the structural performance and cracking behavior of composite linings in deep hydraulic tunnels, which are subject to high ground pressures and potential deformation.
Key Innovation: An integrated analytical approach developed to accurately predict the structural responses and cracking patterns of deep hydraulic tunnel linings, enhancing design and safety.
55. Regional mechanisms of hillslope erosion and soil and water conservation pathways on the Loess Plateau
Core Problem: Understanding regional mechanisms of hillslope erosion and identifying effective conservation pathways.
Key Innovation: Identifying regional mechanisms of hillslope erosion and proposing soil and water conservation pathways on the Loess Plateau.
56. How much historical data do we need? The role of data recency and training period length in LSTM-based rainfall-runoff modeling
Core Problem: Determining optimal historical data requirements for LSTM-based rainfall-runoff modeling.
Key Innovation: Investigating the role of data recency and training period length in LSTM-based rainfall-runoff modeling.
57. Stress integration algorithm of the small-strain Unified Hardening model for soils based on multistage Homotopy continuation method
Core Problem: Developing an efficient and robust stress integration algorithm for a complex soil constitutive model.
Key Innovation: A stress integration algorithm for the small-strain Unified Hardening model for soils based on the multistage Homotopy continuation method.
58. Cyclic response analysis of air-injected calcareous sand: An experimental study on excess pore water pressure development and volumetric strain changes
Core Problem: Understanding the cyclic behavior of calcareous sand, particularly the development of excess pore water pressure and volumetric strain, which are critical for assessing liquefaction potential during earthquakes.
Key Innovation: Experimental study on the cyclic response of air-injected calcareous sand, providing insights into its behavior under dynamic loading relevant to earthquake-induced liquefaction.
59. Brief communication: Towards disability inclusive risk management
Core Problem: People with disabilities face heightened vulnerability during disasters and are underrepresented in risk management planning and response.
Key Innovation: A pilot study in Tyrol, Austria, assessing flood exposure and disaster preparedness in facilities serving people with disabilities, identifying critical gaps and emphasizing the need for inclusive disaster risk management.
60. Coastal-Cosmo-Model (CCMv1): a cosmogenic nuclide model for rocky coastlines
Core Problem: Understanding the long-term evolution of rocky coasts requires models that can account for complex interactions between exposure, erosion, and sea level, constrained by empirical observations.
Key Innovation: Introducing Coastal-Cosmo-Model version 1 (CCMv1), a modular forward modeling framework to reconstruct coastal histories from in situ cosmogenic nuclide concentrations, allowing flexible inversion of platform histories and testing of erosion hypotheses.
61. Earthquake sensors buried in the quietest spot on Earth
Core Problem: There is a need for ultra-low-noise seismic observations beneath the South Pole to probe Earth’s interior and monitor Antarctic ice movement.
Key Innovation: Two seismometers deployed beneath the South Pole leverage an exceptionally low-noise setting to improve sensitivity for probing planetary interior structure and Antarctic ice motion.
62. Automated Marine Biofouling Assessment: Benchmarking Computer Vision and Multimodal LLMs on the Level of Fouling Scale
Core Problem: Traditional marine biofouling assessment relies on hazardous and unscalable diver inspections, necessitating automated classification of biofouling severity on the Level of Fouling (LoF) scale.
Key Innovation: Investigates automated classification of biofouling severity using both custom computer vision models and large multimodal language models (LLMs) on an expert-labelled dataset. Computer vision models showed high accuracy at extreme LoF categories, while LLMs achieved competitive performance without training and provided interpretable outputs, suggesting hybrid methods for scalable and interpretable assessment.
63. ACFormer: Mitigating Non-linearity with Auto Convolutional Encoder for Time Series Forecasting
Core Problem: Time series forecasting (TSF) faces challenges in modeling complex intra-channel temporal dependencies and inter-channel correlations, with linear architectures struggling to capture non-linear signals despite their efficiency in global trends.
Key Innovation: ACFormer, an architecture designed to reconcile the efficiency of linear projections with the non-linear feature-extraction power of convolutions, captures fine-grained information via a shared compression module, preserves temporal locality via gated attention, and reconstructs variable-specific temporal patterns, achieving state-of-the-art performance.
64. Adapting the Behavior of Reinforcement Learning Agents to Changing Action Spaces and Reward Functions
Core Problem: Reinforcement Learning (RL) agents often struggle in real-world non-stationary environments, particularly when reward functions shift or the available action space expands, requiring on-the-fly adaptation without full retraining.
Key Innovation: MORPHIN, a self-adaptive Q-learning framework, enables on-the-fly adaptation by integrating concept drift detection with dynamic adjustments to learning and exploration hyperparameters, adapting to changes in reward functions and action spaces while preserving prior policy knowledge, achieving superior convergence and continuous adaptation.
65. COMET-SG1: Lightweight Autoregressive Regressor for Edge and Embedded AI
Core Problem: Time-series prediction on edge and embedded AI systems requires lightweight, stability-oriented autoregressive regressors that prioritize bounded long-horizon behavior, as prediction errors accumulate over time.
Key Innovation: COMET-SG1, a lightweight, stability-oriented autoregressive regression model, operates through linear behavior-space encoding, memory-anchored transition estimation, and deterministic state updates, achieving competitive short-horizon accuracy with significantly reduced long-horizon drift and a compact parameter footprint for edge deployment.
66. A generalized two-step model for predicting thermal conductivity of bentonite-based mixtures in nuclear waste repositories
Core Problem: Accurate prediction of thermal conductivity (TC) of buffer/backfill materials is critical for high-level radioactive waste repository performance, and existing models have limited accuracy.
Key Innovation: Proposes a generalized "two-step" TC prediction model for bentonite-based mixtures, incorporating solid-phase correction and water-air balance factors, achieving significantly superior predictive accuracy (R2 0.868) compared to existing models.
67. Monitoring the Performance of a Steel-Reinforced Mixed MSE Abutment Wall during and after Construction
Core Problem: The performance of steel-reinforced mixed MSE abutment walls during and after construction needs to be monitored to ensure stability and identify potential issues.
Key Innovation: Monitors and analyzes the performance of a steel-reinforced mixed MSE abutment wall during and after construction.
68. Surface spalling segmentation algorithm of underwater concrete structures based on sonar images: Auxiliary loss and dynamic training
Core Problem: Accurately detecting and segmenting surface spalling damage in underwater concrete structures using sonar images.
Key Innovation: Development of a novel segmentation algorithm incorporating auxiliary loss and dynamic training for improved damage detection.
69. Validation of high-resolution surface soil moisture time series retrieved by means of SAR interferometry
Core Problem: Validating the accuracy and reliability of high-resolution surface soil moisture data derived from SAR interferometry.
Key Innovation: Validation of high-resolution surface soil moisture time series retrieved using SAR interferometry, a crucial input for various geohazard models.
70. Behavior of coarse-grained soils for railway subgrade layers with different saturation and drainage conditions under cyclic loading
Core Problem: Understanding the behavior of coarse-grained soils under cyclic loading, especially with varying saturation and drainage, which is crucial for assessing ground deformation and liquefaction potential.
Key Innovation: Experimental study on the cyclic behavior of coarse-grained soils for railway subgrade layers, highlighting the influence of saturation and drainage conditions.
71. Stability of near-fault surrounding rocks using microseismic monitoring subjected to mining disturbance
Core Problem: Assessing the stability of rock masses near faults under mining-induced disturbances, which can lead to ground deformation or rockfalls.
Key Innovation: Using microseismic monitoring to evaluate and understand the stability of near-fault surrounding rocks subjected to mining disturbance.
72. SEEPS4ALL: an open dataset for the verification of daily precipitation forecasts using station climate statistics
Core Problem: The computation of precipitation forecast verification scores (like SEEPS) is not straightforward due to the need for precipitation climatology information, hindering the assessment of forecast performance.
Key Innovation: Introducing SEEPS4ALL, an open dataset and tools that democratize the use of climate statistics for verifying daily precipitation forecasts, showcasing its application for both deterministic and probabilistic forecasts, which is crucial for early warning systems.
73. Minimum-Cost Network Flow with Dual Predictions
Core Problem: Improving the performance of classic algorithms like minimum-cost network flow using machine-learned predictions.
Key Innovation: Proposes the first minimum-cost network flow algorithm augmented with a dual prediction, based on the $\varepsilon$-relaxation algorithm. It provides time complexity bounds in terms of the infinity norm prediction error and proves sample complexity bounds for PAC-learning the prediction, demonstrating significant speedups (e.g., 12.74x on traffic networks) through empirical validation.
74. ScatterFusion: A Hierarchical Scattering Transform Framework for Enhanced Time Series Forecasting
Core Problem: Time series forecasting faces significant challenges due to complex temporal dependencies at multiple time scales.
Key Innovation: Introducing ScatterFusion, a novel framework that integrates a Hierarchical Scattering Transform Module (HSTM) for multi-scale invariant features, a Scale-Adaptive Feature Enhancement (SAFE) module, a Multi-Resolution Temporal Attention (MRTA) mechanism, and a Trend-Seasonal-Residual (TSR) decomposition-guided loss function for robust forecasting.
75. AWGformer: Adaptive Wavelet-Guided Transformer for Multi-Resolution Time Series Forecasting
Core Problem: Time series forecasting requires capturing patterns across multiple temporal scales while maintaining computational efficiency, especially for multi-variate and non-stationary series.
Key Innovation: Introducing AWGformer, a novel architecture that integrates an Adaptive Wavelet Decomposition Module (AWDM), a Cross-Scale Feature Fusion (CSFF) mechanism, a Frequency-Aware Multi-Head Attention (FAMA) module, and a Hierarchical Prediction Network (HPN) for enhanced multi-resolution time series prediction.
76. TimeCatcher: A Variational Framework for Volatility-Aware Forecasting of Non-Stationary Time Series
Core Problem: Existing lightweight MLP-based models for time series forecasting struggle with long-term prediction of highly non-stationary series, especially during abrupt fluctuations, due to an implicit local stationarity assumption.
Key Innovation: Proposing TimeCatcher, a Volatility-Aware Variational Forecasting framework that extends linear architectures with a variational encoder to capture latent dynamic patterns and a volatility-aware enhancement mechanism to detect and amplify significant local variations, outperforming baselines in high-volatility scenarios.
77. Learning Contextual Runtime Monitors for Safe AI-Based Autonomy
Core Problem: AI-based controllers in cyber-physical systems can degrade sharply in unfamiliar environments, and traditional ensemble methods dilute specialized strengths; a monitoring framework is needed to identify and exploit contextual strengths for safety.
Key Innovation: A novel framework for learning context-aware runtime monitors for AI-based control ensembles, reformulating safe AI control as a contextual monitoring problem, which continuously observes context to select the best controller, providing theoretical safety guarantees and improved utilization of controller diversity.
78. Li-ViP3D++: Query-Gated Deformable Camera-LiDAR Fusion for End-to-End Perception and Trajectory Prediction
Core Problem: End-to-end perception and trajectory prediction for autonomous driving from raw sensor data is challenging, as modular pipelines restrict information flow and existing fusion schemes for cameras and LiDAR often rely on heuristics, leading to suboptimal utilization of information.
Key Innovation: Li-ViP3D++, a query-based multimodal PnP framework, introduces Query-Gated Deformable Fusion (QGDF) to integrate multi-view RGB and LiDAR in query space, aggregating image evidence, extracting LiDAR context, and applying query-conditioned gating to adaptively weight cues, jointly optimizing detection, tracking, and trajectory forecasting.
79. GraphAllocBench: A Flexible Benchmark for Preference-Conditioned Multi-Objective Policy Learning
Core Problem: Existing benchmarks for Preference-Conditioned Policy Learning (PCPL) in Multi-Objective Reinforcement Learning (MORL) are restricted to toy tasks and fixed environments, limiting realism and scalability for approximating diverse Pareto-optimal solutions.
Key Innovation: GraphAllocBench, a flexible benchmark built on CityPlannerEnv (a novel graph-based resource allocation sandbox environment), provides a rich suite of problems with diverse objectives and high-dimensional scalability, along with new evaluation metrics (PNDS, OS) to capture preference consistency, exposing limitations of existing MORL approaches.
80. Supervised Guidance Training for Infinite-Dimensional Diffusion Models
Core Problem: While diffusion models provide expressive priors in function space for Bayesian inverse problems, the theory of conditioning them to sample from the posterior remains open, and the guidance term for conditional score decomposition is intractable.
Key Innovation: The paper proves that infinite-dimensional diffusion models can be conditioned using an infinite-dimensional Doob's h-transform and proposes Supervised Guidance Training, a simulation-free score matching objective, enabling efficient and stable posterior sampling for Bayesian inverse problems in function spaces.
81. Active Learning for Decision Trees with Provable Guarantees
Core Problem: The theoretical understanding of active learning label complexity for decision trees, particularly the disagreement coefficient, is limited, and general active learning algorithms with strong guarantees for decision trees are lacking.
Key Innovation: The paper provides the first analysis of the disagreement coefficient for decision trees under natural assumptions and presents the first general active learning algorithm for binary classification that achieves a multiplicative error guarantee and polylogarithmic label queries, with near-optimal dependence on error tolerance.
82. GNN Explanations that do not Explain and How to find Them
Core Problem: Explanations provided by Self-explainable Graph Neural Networks (SE-GNNs) can be suboptimal, misleading, and even unambiguously unrelated to how SE-GNNs infer labels, yet most faithfulness metrics fail to identify these degenerate explanations.
Key Innovation: The paper identifies a critical failure mode where SE-GNN explanations do not explain, showing that degenerate explanations can be maliciously planted or emerge naturally, and introduces a novel faithfulness metric that reliably marks these unfaithful explanations, highlighting the need for reliable auditing.
83. Microwave dry-back of compacted unbound granular materials: an experimental and numerical study
Core Problem: Current dry-back methods for compacted unbound granular pavement layers are time-consuming and weather-dependent, delaying construction and affecting optimal strength/stiffness.
Key Innovation: Investigates microwave drying as an alternative, demonstrating substantial reduction in drying time through experimental and numerical analyses, and providing recommendations for optimizing microwave drying to enhance energy efficiency in field implementation.
84. Hydraulic gradient based low-gravity simulation system for geomaterials
Core Problem: Existing low-gravity experimental simulation techniques for extraterrestrial geotechnical research face limitations in cost, duration, and scalability.
Key Innovation: Presents a novel ground-based low-gravity system using the Hydraulic Gradient Similitude Method, developing a small-scale apparatus that provides stable low-gravity environments for conventional granular materials, validated through CPT responses.
85. Effect of side shear zones on T-bar penetration resistance in clay—upper bound analysis
Core Problem: The contribution of side shear zones to the resistance of the T-bar penetrometer in undrained clay is not fully understood, impacting the accuracy of undrained strength measurements.
Key Innovation: Provides insight into the contribution of side shear zones, deriving a quantitative relation between T-bar resistance and undrained shear strength using the upper bound theorem, introducing a resistance factor Ns, and validating results against finite element analyses.
86. Drained and undrained weakening and post-reconsolidation strength recovery
Core Problem: The effect of clay fraction on drained and undrained strain-weakening and subsequent strength recovery through reconsolidation in different soil types is not fully understood.
Key Innovation: Investigates strain-weakening and strength recovery using DSS and DRS tests on three materials, highlighting potential for undrained strength recovery for pre-sheared specimens and the dependence of undrained frictional-weakening on clay fraction.
87. Membrane Behavior of Exhumed Geosynthetic Clay Liners
Core Problem: Understanding the long-term membrane behavior of exhumed geosynthetic clay liners is important for their performance in geotechnical applications.
Key Innovation: Investigates the membrane behavior of exhumed geosynthetic clay liners.
88. Influence of Particle Shape on the Shear Behavior of Biopolymer-Treated Soil under Monotonic and Cyclic Loading
Core Problem: The influence of particle shape on the shear behavior of biopolymer-treated soil under monotonic and cyclic loading is not fully characterized.
Key Innovation: Investigates the influence of particle shape on the shear behavior of biopolymer-treated soil under various loading conditions.
89. SAMYOLO: Integrating Super-Resolution and Vision Mamba for Enhanced Small Object Detection
Core Problem: Satellite datasets are growing, creating challenges in efficiently projecting measurements onto geographical grids, with current processing steps propagating errors and increasing computational time, leading to loss of native spatial resolution.
Key Innovation: Presents a new interpolation algorithm for satellite missions (demonstrated with SMOS SSS processor) that keeps measurements in the instrument coordinate system until final product generation, and a novel algorithm to project measurements by weighting them based on actual acquisition area, numerically optimized for efficiency and parallelization, preserving high resolution.
90. Seasonal Dependence of Topographic Effect on Mountainous Land Surface Temperature Retrieval: Insights From Landsat-8 Data
Core Problem: Existing multiscale encoders for change detection in remote sensing images often result in blurry change boundaries due to not directly enhancing or enriching high-frequency features, especially for targets with various scales.
Key Innovation: Proposes MEDS-CD, a multiscale enhanced change detection method based on detail supplementation, which designs a multiscale edge-guided adaptive filter to extract and compensate high-frequency information, and introduces a multiscale convolution modulation module and fuzzy weighting strategy to fully exploit restored high-frequency information for change extraction and edge enhancement.
91. Dual-Branch Prototype Enhancement Network for Few-Shot Hyperspectral Image Classification
Core Problem: Previous studies have incorporated topographic effects in mountainous LST (MLST) retrieval, but a systematic quantification of the seasonal magnitude of these effects remains lacking.
Key Innovation: Employs Landsat-8 data to quantify seasonal topographic effects (ΔLST) on MLST retrieval across the Tibet Plateau, revealing that ΔLST is primarily controlled by topographic factors (SSP, SVF, LSE), is most pronounced in summer, and exhibits clear dependence on slope and aspect, improving understanding of topographic influences on MLST.
92. GAFF: Global Attention Feature Flow Network for Optical and SAR Image Registration Under Geometric Transformations
Core Problem: Vegetation segmentation in remote sensing imagery faces challenges like extreme scale variance, spectral ambiguity, and complex boundary characteristics, with CNNs limited in long-range dependency modeling and Transformers impractical for high-resolution images due to quadratic complexity.
Key Innovation: Proposes BRSMamba, a boundary-aware network integrating two novel modules with noncausal state-space duality (NC-SSD) to enhance boundary preservation and global context modeling, using a boundary-subject fusion perception module and a boundary-body resolution module to inject boundary awareness into the NC-SSD state-transition matrix, achieving high accuracy and efficiency.
93. A Deep Neural Network for Subpixel Target Detection Using Dimensionality Reduction in Hyperspectral Imaging
Core Problem: Registration of optical and SAR images under geometric distortions is challenging due to inherent radiometric and geometric differences, diverse terrain, and distinct noise patterns, compromising accuracy, robustness, and generalization of existing methods.
Key Innovation: Introduces GAFF, a global attention feature flow network combining model-driven and data-driven methodologies, leveraging a hybrid CNN-transformer architecture with modality independent region descriptors (MIRD) for feature extraction, employing a global attention mechanism for robust correspondences, and a hierarchical feature flow refinement module (HFRM) with a combined weight loss function for optimization, achieving high accuracy and generalization.
94. MEETNet: Morphology-Edge Enhanced Triple-Cascaded Network for Infrared Small Target Detection
Core Problem: Super-resolution reconstructions of remote sensing images often suffer from over-smoothed textures and structural distortions, failing to accurately recover intricate details of ground objects.
Key Innovation: Proposes DTWSTSR, a remote sensing image super-resolution network combining Dual-Tree Complex Wavelet Transform and Swin Transformer, which enhances texture detail reconstruction by fusing frequency-domain and spatial-domain features, includes a Multiscale Efficient Channel Attention mechanism, and a Kolmogorov–Arnold Network based on branch attention, achieving superior performance and generalization.
95. Mechanism of the influence of high-temperature and high-pressure storage conditions on coal pore structure and methane adsorption kinetics
Core Problem: Deep coal seams operate under high-temperature and high-pressure conditions that alter pore architecture and methane adsorption behavior, but the controlling mechanisms are not fully resolved.
Key Innovation: By integrating pore-structure characterization, isothermal adsorption experiments, and molecular-dynamics simulations, the study quantifies temperature-pressure effects on adsorption capacity and identifies distinct micropore/mesopore controls on methane adsorption kinetics.
96. Simulation of anisotropic seepage in 3D heterogeneous coal fractured by liquid nitrogen
Core Problem: Designing liquid-nitrogen fracturing for coal permeability enhancement requires quantitative understanding of post-fracture 3D pore heterogeneity and anisotropic seepage.
Key Innovation: Micro-CT-based 3D reconstruction coupled with COMSOL seepage simulations reveals LN2-induced increases in pore-throat connectivity and pronounced directional permeability, clarifying pressure-gradient-dependent anisotropic flow behavior in heterogeneous coal.
97. Experimental Investigation of Dynamic Shear Fatigue Properties of Granite Under Constant and Variable Amplitude Cyclic Loading
Core Problem: Rock engineering often suffers shear fatigue failure due to cyclic dynamic disturbances, but the fatigue behaviors of rock subjected to cyclic dynamic shear, especially under constant and variable amplitude loading, have not been fully clarified.
Key Innovation: Performed cyclic dynamic shear experiments on granite via the split Hopkinson pressure bar, revealing that peak shear stress and stiffness decrease with increasing impacts. The study determined a fatigue threshold (0.68 times critical impact stress), showed that higher amplitude loads earlier lead to quicker damage accumulation, and identified intergranular cracks and grain fragmentation as primary fatigue mechanisms.
98. A Study of a Method for Division of the Comprehensive Homogeneous Domain and Domain Design of Bolt–Mesh Support Parameters
Core Problem: Roadway support in complex fractured rock masses presents a challenge due to variability in composition and structure, requiring targeted support designs for homogeneous domains to enhance safety and stability.
Key Innovation: Proposed a method for delineating structural and rock mass quality homogeneous domains based on the spatial distribution characteristics of blockiness (MBi). A theoretical framework for comprehensive homogeneous domain division was established and applied to develop a domain-specific bolt-mesh support design in the Tongkeng mine, which effectively controls roof deformation and enhances surrounding rock stability.
99. Exploring multiscale videogrammetry techniques for analyzing rock mass discontinuities in geological formations
Core Problem: N/A (Abstract not provided)
Key Innovation: N/A (Abstract not provided)
100. Modeling the spatial structural network of layered rock masses using an innovative hierarchical method
Core Problem: N/A (Abstract not provided)
Key Innovation: N/A (Abstract not provided)
101. Assessing the building façade vulnerability to multiple hazards
Core Problem: N/A (Abstract not provided)
Key Innovation: N/A (Abstract not provided)
102. Probabilistic modeling of the risk and resilience of transportation systems for community resilience analysis
Core Problem: N/A (Abstract not provided)
Key Innovation: N/A (Abstract not provided)
103. Impact of heterogeneous behaviors of subarea managers on the recovery of urban water distribution systems after disaster
Core Problem: N/A (Abstract not provided)
Key Innovation: N/A (Abstract not provided)
104. Policy modeling for urban energy resilience to extreme heat: A multi-agent simulation framework in Chongqing, China
Core Problem: N/A (Abstract not provided)
Key Innovation: N/A (Abstract not provided)
105. Sensitivity estimation of stochastic output with respect to distribution parameters of stochastic inputs
Core Problem: Accurately estimating the sensitivity of stochastic outputs to the distribution parameters of stochastic inputs.
Key Innovation: A method for sensitivity estimation of stochastic output with respect to distribution parameters of stochastic inputs.
106. D2R: A distance metric for exploring network structural robustness enhancement potential
Core Problem: Quantifying and exploring the potential for enhancing the structural robustness of networks.
Key Innovation: D2R, a novel distance metric for exploring network structural robustness enhancement potential.
107. A general multivariate gamma process with moment matching estimation for degradation modeling
Core Problem: Developing a general and robust multivariate model for degradation processes.
Key Innovation: A general multivariate gamma process with moment matching estimation for degradation modeling.
108. A spectral index using generic global endmembers from Landsat multispectral data for mapping urban areas
Core Problem: Developing an effective and generalizable method for mapping urban areas using Landsat multispectral data.
Key Innovation: A novel spectral index utilizing generic global endmembers from Landsat multispectral data for robust urban area mapping.
109. Energy balance effects of extreme snow events on shallow frozen and thawed surfaces in highland pastoral areas
Core Problem: Understanding the energy balance effects of extreme snow events on shallow frozen and thawed surfaces in highland pastoral areas, which can influence ground stability.
Key Innovation: Investigates how extreme snow events impact energy balance on shallow frozen/thawed surfaces, providing insights into processes relevant to ground deformation.
110. Unified nonlinear modelling of longitudinal equivalent bending stiffness for circular and non-circular shield tunnels
Core Problem: Developing a unified nonlinear model to accurately represent the longitudinal equivalent bending stiffness of both circular and non-circular shield tunnels.
Key Innovation: A unified nonlinear modelling approach for tunnel bending stiffness, improving structural design and analysis for various tunnel geometries.
111. 3D reality and deep zoom image framework for inspection of an undersea multi-chamber tunnel
Core Problem: Efficiently inspecting the structural integrity and condition of undersea multi-chamber tunnels.
Key Innovation: A 3D reality and deep zoom image framework for detailed and comprehensive inspection of undersea tunnel infrastructure.
112. Towards low-carbon construction of metro station foundation pit: A probabilistic digital twin framework with self-supervised learning capability
Core Problem: Achieving low-carbon construction for metro station foundation pits while ensuring safety and efficiency.
Key Innovation: A probabilistic digital twin framework with self-supervised learning for optimizing low-carbon construction processes in metro station foundation pits.
113. Mechanical performance and damage characteristics of segmental joints in semi-rigid element immersed tunnel under bending deformation: A combined experimental and numerical study
Core Problem: Understanding the mechanical performance and damage characteristics of segmental joints in semi-rigid element immersed tunnels under bending deformation.
Key Innovation: A combined experimental and numerical study providing insights into the structural behavior and damage mechanisms of immersed tunnel joints.
114. Predicting river turbidity in Pine Island Bayou using machine learning techniques coupled with variational mode decomposition
Core Problem: Accurately predicting river turbidity.
Key Innovation: Using machine learning techniques coupled with variational mode decomposition for river turbidity prediction.
115. The performance of quadratic finite-discrete element method (qFDEM) and its potential advantages
Core Problem: Improving the performance and understanding the advantages of numerical methods for complex geotechnical and rock mechanics problems.
Key Innovation: Development and performance analysis of the quadratic finite-discrete element method (qFDEM), a numerical technique with potential for simulating rock fracture and granular flows.
116. Influence of particle size scaling on the static and cyclic behaviour of railway slab track-bed materials: Laboratory and DEM study
Core Problem: Understanding how particle size scaling affects the static and cyclic behavior of granular materials, which is fundamental to geotechnical engineering and can inform geohazard analysis.
Key Innovation: Laboratory and Discrete Element Method (DEM) study investigating the influence of particle size scaling on the static and cyclic behavior of railway track-bed materials.
117. Incorporation of a hypoplastic material model for sandy soils into a dynamic ALE formulation suitable for structures subjected to moving loads
Core Problem: Accurately modeling the dynamic behavior of sandy soils under moving loads, which is foundational for understanding dynamic ground response during geohazards.
Key Innovation: Incorporation of a hypoplastic material model for sandy soils into a dynamic Arbitrary Lagrangian-Eulerian (ALE) formulation for structures subjected to moving loads.
118. A time-dependent hydration-driven bonded-particle model for simulating strength evolution in cement-stabilised granular soils: Experimental and DEM insights
Core Problem: Accurately simulating the time-dependent strength evolution in cement-stabilized granular soils, which is crucial for ground improvement and mitigation strategies.
Key Innovation: Development of a time-dependent hydration-driven bonded-particle model for simulating strength evolution in cement-stabilized granular soils, supported by experimental and DEM insights.
119. Multi-objective optimization design based on surrogate modelling for concrete column-supported embankment on soft ground
Core Problem: Optimizing the design of concrete column-supported embankments on soft ground to ensure stability and prevent ground deformation.
Key Innovation: Multi-objective optimization design based on surrogate modeling for concrete column-supported embankments on soft ground, aiming to improve stability.
120. Drainage and imbibition along main and scanning curves: A pore-scale morphology approach
Core Problem: Characterizing and understanding pore-scale fluid flow (drainage and imbibition) in soils, which is fundamental to hydrological processes influencing rainfall-induced landslides.
Key Innovation: A pore-scale morphology approach to study drainage and imbibition along main and scanning curves, providing insights into unsaturated soil behavior.
121. Water glass-enhanced biocementation: Crystal precipitation evidence and environmental dependence analysis
Core Problem: Understanding the mechanisms and environmental factors influencing water glass-enhanced biocementation for soil/rock improvement.
Key Innovation: Providing crystal precipitation evidence and analyzing the environmental dependence of water glass-enhanced biocementation, a technique potentially relevant for ground stability.
122. Exploring the influence of mixing energy on strength of sand treated by deep soil mixing
Core Problem: Understanding how mixing energy affects the strength of sand treated by deep soil mixing, a ground improvement technique.
Key Innovation: Exploring the influence of mixing energy on the strength of sand treated by deep soil mixing, which can enhance ground stability.
123. Mechanical properties and mechanisms of soft clay treated by all-industrial by-product binder in alkali-sulfate activating framework
Core Problem: Improving the mechanical properties of soft clay using sustainable binders derived from industrial by-products.
Key Innovation: Investigating the mechanical properties and mechanisms of soft clay treated by an all-industrial by-product binder in an alkali-sulfate activating framework, relevant for ground improvement.
124. QUADICA v2: extending the large-sample data set for water QUAlity, DIscharge and Catchment Attributes in Germany
Core Problem: The need for a more comprehensive and updated dataset combining water quality, discharge, and catchment attributes for large-sample studies in Germany.
Key Innovation: QUADICA v2, an extended dataset incorporating more recent data, additional water quality and driver variables, and more stations with concurrent water quantity data, facilitating a comprehensive understanding of ecological impacts and water quality dynamics, relevant for flood modeling.
125. Regime-Adaptive Bayesian Optimization via Dirichlet Process Mixtures of Gaussian Processes
Core Problem: Standard Bayesian Optimization (BO) assumes uniform smoothness across the search space, an assumption violated in multi-regime problems (e.g., molecular conformation search), leading to miscalibrated uncertainty.
Key Innovation: Proposes RAMBO, a Dirichlet Process Mixture of Gaussian Processes that automatically discovers latent regimes during optimization, each modeled by an independent GP with locally-optimized hyperparameters. It derives collapsed Gibbs sampling for efficient inference and introduces adaptive concentration parameter scheduling, demonstrating consistent improvements over state-of-the-art baselines on multi-regime objectives.
126. DiSa: Saliency-Aware Foreground-Background Disentangled Framework for Open-Vocabulary Semantic Segmentation
Core Problem: Open-vocabulary semantic segmentation using Vision-Language Models (VLMs) suffers from foreground bias (ignoring background regions) and limited spatial localization (blurred object boundaries) due to VLMs being pre-trained on salient, object-centric image-text pairs.
Key Innovation: Introduces DiSa, a novel saliency-aware foreground-background disentangled framework. It explicitly incorporates saliency cues in a Saliency-aware Disentanglement Module (SDM) to separately model foreground and background ensemble features, and proposes a Hierarchical Refinement Module (HRM) for pixel-wise spatial contexts and channel-wise feature refinement, consistently outperforming state-of-the-art methods.
127. Domain Expansion: A Latent Space Construction Framework for Multi-Task Learning
Core Problem: Training a single network with multiple objectives often leads to conflicting gradients that degrade shared representations, forcing them into a compromised state suboptimal for any single task, a problem termed latent representation collapse.
Key Innovation: Introduces Domain Expansion, a framework that prevents latent representation collapse by restructuring the latent space itself. It uses a novel orthogonal pooling mechanism to construct a latent space where each objective is assigned to a mutually orthogonal subspace, yielding an explicit, interpretable, and compositional latent space where concepts can be directly manipulated.
128. Semi-Supervised Masked Autoencoders: Unlocking Vision Transformer Potential with Limited Data
Core Problem: Training Vision Transformers (ViTs) is challenging when labeled data is scarce, despite the abundance of unlabeled data.
Key Innovation: Proposes Semi-Supervised Masked Autoencoder (SSMAE), a framework that jointly optimizes masked image reconstruction and classification using both unlabeled and labeled samples with dynamically selected pseudo-labels. SSMAE introduces a validation-driven gating mechanism that activates pseudo-labeling only after the model achieves reliable, high-confidence predictions, reducing confirmation bias and consistently outperforming supervised ViT and fine-tuned MAE, especially in low-label regimes.
129. Going NUTS with ADVI: Exploring various Bayesian Inference techniques with Facebook Prophet
Core Problem: Facebook Prophet's built-in inference methods (L-BFGS-B, NUTS MCMC) limit the application of alternative Bayesian inference techniques, and its fluent API design lacks flexibility for implementing custom modeling ideas for time-series forecasting.
Key Innovation: Developed a complete reimplementation of the Prophet model in PyMC, which enables extending the base model and evaluating/comparing multiple Bayesian inference methods (full MCMC, MAP estimation, Variational inference). This provides a more flexible framework for custom modeling and detailed analysis of sampling approaches, convergence diagnostics, and computational efficiency.
130. PASS: Ambiguity Guided Subsets for Scalable Classical and Quantum Constrained Clustering
Core Problem: Pairwise-constrained clustering, which augments unsupervised partitioning with must-link (ML) and cannot-link (CL) constraints, adds complexity and struggles with data scalability, especially in niche applications like quantum or quantum-hybrid clustering.
Key Innovation: Proposes PASS, a pairwise-constraints and ambiguity-driven subset selection framework that preserves ML and CL constraint satisfaction while allowing scalable, high-quality clustering solutions. PASS collapses ML constraints into pseudo-points and offers two selectors (constraint-aware margin rule, information-geometric rule) to collect high-information subsets, achieving competitive SSE at substantially lower cost and remaining effective where prior approaches fail.
131. MAPLE: Self-supervised Learning-Enhanced Nonlinear Dimensionality Reduction for Visual Analysis
Core Problem: Existing nonlinear dimensionality reduction methods like UMAP can struggle with complex manifolds, high intra-cluster variance, and curved manifold structures in high-dimensional data, leading to less clear visual cluster separations.
Key Innovation: Presents MAPLE, a new nonlinear dimensionality reduction method that enhances UMAP by employing a self-supervised learning approach to more efficiently encode low-dimensional manifold geometry. It uses maximum manifold capacity representations (MMCRs) to untangle complex manifolds, producing clearer visual cluster separations and finer subcluster resolution than UMAP while maintaining comparable computational cost.
132. Causal-Driven Feature Evaluation for Cross-Domain Image Classification
Core Problem: Out-of-distribution (OOD) generalization remains a fundamental challenge in real-world classification, where test distributions often differ substantially from training data, and domain-invariant representations are not necessarily causally effective for prediction.
Key Innovation: Revisits OOD classification from a causal perspective, proposing to evaluate learned representations based on their necessity and sufficiency under distribution shift. Introduces an explicit segment-level framework that directly measures causal effectiveness across domains, providing a more faithful criterion than invariance alone and demonstrating consistent improvements in OOD performance.
133. Quartet of Diffusions: Structure-Aware Point Cloud Generation through Part and Symmetry Guidance
Core Problem: Prior methods for shape generation treat it holistically or only support part composition, lacking explicit modeling of both part composition and symmetry for structure-aware point clouds.
Key Innovation: Introducing the Quartet of Diffusions, a framework using four coordinated diffusion models to learn distributions of global shape latents, symmetries, semantic parts, and their assembly, ensuring guaranteed symmetry, coherent part placement, and diverse, high-quality outputs with fine-grained control.
134. Nonlinear Dimensionality Reduction with Diffusion Maps in Practice
Core Problem: The practical application of Diffusion Maps for nonlinear dimensionality reduction is significantly influenced by data preprocessing, parameter settings, and component selection, which are not comprehensively discussed in literature.
Key Innovation: Providing a practice-oriented review of Diffusion Map, illustrating pitfalls, and showcasing a technique for identifying the most relevant components, demonstrating that the first components are not necessarily the most relevant ones.
135. CCMamba: Selective State-Space Models for Higher-Order Graph Learning on Combinatorial Complexes
Core Problem: Most topological deep learning methods for higher-order relational structures rely on local message passing via attention, incurring quadratic complexity and limiting scalability and rank-aware information aggregation in combinatorial complexes.
Key Innovation: Proposing Combinatorial Complex Mamba (CCMamba), the first mamba-based neural framework that reformulates message passing as selective state-space modeling, enabling adaptive, directional, and long-range information propagation in linear time without self-attention, with improved scalability and robustness.
136. Advancing Open-source World Models
Core Problem: The need for advanced open-source world simulators that offer high fidelity, robust dynamics, long-term contextual consistency, and real-time interactivity across diverse environments.
Key Innovation: Presenting LingBot-World, an open-sourced world simulator stemming from video generation, which offers high fidelity and robust dynamics in broad environments, minute-level horizon with contextual consistency, and real-time interactivity, aiming to empower content creation, gaming, and robot learning.
137. Unsupervised Ensemble Learning Through Deep Energy-based Models
Core Problem: Combining predictions from multiple learners without access to ground truth labels or additional data is challenging, especially in data-scarce or privacy-sensitive environments where evaluating individual classifier performance is difficult.
Key Innovation: A novel deep energy-based method for constructing an accurate meta-learner using only individual learner predictions, capable of capturing complex dependence structures between them, with theoretical guarantees for conditionally independent learners.
138. Robust Distributed Learning under Resource Constraints: Decentralized Quantile Estimation via (Asynchronous) ADMM
Core Problem: Decentralized learning on resource-constrained edge devices requires communication-efficient, robust, and memory-light algorithms, but existing asynchronous decentralized ADMM-based methods for robust estimation (e.g., median) often require memory that scales with node degree.
Key Innovation: AsylADMM, a novel gossip algorithm for decentralized median and quantile estimation, designed for asynchronous updates and requiring only two variables per node, offering theoretical guarantees, fast convergence, and enabling quantile-based trimming for robustness.
139. Ranking-aware Reinforcement Learning for Ordinal Ranking
Core Problem: Ordinal regression and ranking are challenging due to inherent ordinal dependencies that conventional methods struggle to model.
Key Innovation: Ranking-Aware Reinforcement Learning (RARL), a novel RL framework, explicitly learns ordinal relationships through a unified objective integrating regression and Learning-to-Rank, driven by a ranking-aware verifiable reward and enhanced by Response Mutation Operations for improved exploration.
140. Regularized Gradient Temporal-Difference Learning
Core Problem: Existing convergence analyses for Gradient Temporal-Difference (GTD) learning algorithms, used for off-policy policy evaluation, rely on the restrictive assumption that the feature interaction matrix (FIM) is nonsingular, leading to instability or degraded performance when it is singular.
Key Innovation: R-GTD, a regularized GTD algorithm, is proposed by reformulating the mean-square projected Bellman error (MSPBE) minimization, guaranteeing convergence to a unique solution even when the FIM is singular, with established theoretical convergence guarantees and explicit error bounds.
141. DIVERSE: Disagreement-Inducing Vector Evolution for Rashomon Set Exploration
Core Problem: Systematically exploring the Rashomon set of deep neural networks (models matching accuracy but differing in predictive behavior) is challenging, as existing methods often require retraining or gradient access.
Key Innovation: DIVERSE, a framework that augments a pretrained model with Feature-wise Linear Modulation (FiLM) layers and uses Covariance Matrix Adaptation Evolution Strategy (CMA-ES) to search a latent modulation space, generating diverse, high-performing yet functionally distinct model variants efficiently without retraining.
142. Optimal Transport Group Counterfactual Explanations
Core Problem: Existing methods for group counterfactual explanations struggle with generalizing to new group members, rely on strong model assumptions (e.g., linearity), or poorly control the counterfactual group geometry distortion.
Key Innovation: An explicit optimal transport map is learned that sends any group instance to its counterfactual without re-optimization, minimizing the group's total transport cost, enabling generalization with fewer parameters, preserving group geometry, and outperforming baselines, especially for non-linear models.
143. Is Pure Exploitation Sufficient in Exogenous MDPs with Linear Function Approximation?
Core Problem: Despite empirical evidence that greedy, exploitation-only methods work well in Exogenous MDPs (Exo-MDPs), existing theoretical regret guarantees rely on explicit exploration or tabular assumptions, leaving a gap in understanding why pure exploitation is sufficient.
Key Innovation: Pure Exploitation Learning (PEL) and LSVI-PE are proposed, providing the first general finite-sample regret bounds for exploitation-only algorithms in Exo-MDPs, demonstrating that exploration is unnecessary and achieving strong performance without optimism, using novel tools like counterfactual trajectories and Bellman-closed feature transport.
144. SA-PEF: Step-Ahead Partial Error Feedback for Efficient Federated Learning
Core Problem: In federated learning, biased gradient compression with error feedback (EF) can lead to slow residual error decay and gradient mismatch under non-IID data, stalling progress in early training rounds.
Key Innovation: SA-PEF (step-ahead partial error feedback), which integrates step-ahead correction with partial error feedback, guarantees convergence to stationarity under heterogeneous data and partial client participation, achieving faster target accuracy by controlling residual contraction and accelerating early training.
145. Less is More: Clustered Cross-Covariance Control for Offline RL
Core Problem: Offline reinforcement learning faces a fundamental challenge with distributional shift, where scarce data or datasets dominated by out-of-distribution (OOD) areas exacerbate harmful TD cross covariance, biasing optimization and degrading policy learning.
Key Innovation: Clustered Cross-Covariance Control for TD (C^4) is developed, using partitioned buffer sampling to restrict updates to localized replay partitions and an explicit gradient-based corrective penalty to cancel covariance-induced bias, mitigating excessive conservatism in OOD areas and improving returns by up to 30%.
146. Smoothing the Black-Box: Signed-Distance Supervision for Black-Box Model Copying
Core Problem: Black-box model copying from hard-label outputs is a discontinuous surface reconstruction problem, severely limiting the efficient recovery of boundary geometry when original training data or model internals are unavailable.
Key Innovation: A distance-based copying (distillation) framework is proposed that replaces hard-label supervision with signed distances to the teacher's decision boundary, converting copying into a smooth regression problem that exploits local geometry, improving fidelity and generalization accuracy over hard-label baselines.
147. Conditional PED-ANOVA: Hyperparameter Importance in Hierarchical & Dynamic Search Spaces
Core Problem: Estimating hyperparameter importance (HPI) in conditional search spaces, where hyperparameter presence or domain depends on others, is challenging because existing PED-ANOVA assumes fixed, unconditional spaces, leading to misleading estimates.
Key Innovation: Conditional PED-ANOVA (condPED-ANOVA), a principled framework for estimating HPI in conditional search spaces, introduces a conditional HPI for top-performing regions and derives a closed-form estimator that accurately reflects conditional activation and domain changes, providing meaningful importances.
148. Incorporating data drift to perform survival analysis on credit risk
Core Problem: Survival analysis for credit risk implicitly assumes stationary data, but mortgage portfolios are exposed to various forms of data drift, impacting model robustness.
Key Innovation: Proposed a dynamic joint modeling framework that integrates a longitudinal behavioral marker with a discrete-time hazard formulation, combined with landmark one-hot encoding and isotonic calibration, consistently outperforming classical survival models across simulated data drift scenarios.
149. X-SAM: From Segment Anything to Any Segmentation
Core Problem: Large Language Models (LLMs) lack pixel-level perceptual understanding, and the Segment Anything Model (SAM) has limitations in multi-mask and category-specific segmentation, failing to integrate all segmentation tasks into a unified architecture.
Key Innovation: X-SAM, a streamlined Multimodal Large Language Model (MLLM) framework that extends segmentation to "any segmentation" by introducing a novel unified framework for advanced pixel-level perceptual comprehension, proposing Visual GrounDed (VGD) segmentation, and using a unified training strategy, achieving state-of-the-art performance on various benchmarks.
150. Estimating Stem Water Content From Tower-Based L-Band Tomographic Radar Using a Single-Scattering Model of a Uniform Layer: A First in Situ Experiment
Core Problem: Understanding the underlying scattering processes and their link to forest vegetation water content (VWC) for spaceborne synthetic aperture radar missions, and accurately estimating stem water content.
Key Innovation: Used tower-based L-band tomographic radar observations to measure canopy-only backscatter, which was anticorrelated with stem water content, and developed a single-scattering model that effectively captures diurnal and long-term variations in stem water content with an RMSE of 4%.
151. From physics to machine learning and back: Part I - Learning with inductive biases in prognostics and health management (PHM)
Core Problem: Integrating physics-based knowledge with machine learning for improved prognostics and health management.
Key Innovation: Exploring the role of inductive biases in learning for prognostics and health management, bridging physics and machine learning.
152. Variational Bayesian data assimilation with time-varying multi-physics-informed neural network for solving dimension-reduced probability density evolution equation
Core Problem: Efficiently solving dimension-reduced probability density evolution equations, especially with time-varying and multi-physics aspects.
Key Innovation: A variational Bayesian data assimilation approach combined with a time-varying multi-physics-informed neural network for solving complex probability density evolution equations.
153. Inferring failure processes via causality analysis: from event logs to predictive fault trees
Core Problem: Inferring complex failure processes and constructing predictive fault trees from event logs.
Key Innovation: A causality analysis method to infer failure processes and generate predictive fault trees from event logs.
154. A stage-dependent Markov-switching fractional Brownian motion model for reliability analysis considering random effects and long-term memory
Core Problem: Developing a robust model for reliability analysis that accounts for stage-dependent behavior, random effects, and long-term memory.
Key Innovation: A stage-dependent Markov-switching fractional Brownian motion model for reliability analysis.
155. TARD: Test-time domain adaptation for robust fault detection under evolving operating conditions
Core Problem: Achieving robust fault detection in systems operating under evolving conditions where training and test data distributions differ.
Key Innovation: TARD (Test-time domain adaptation) for robust fault detection under evolving operating conditions.
156. Reliability estimation for the multicomponent stress-strength model based on objective Bayesian method
Core Problem: Estimating reliability for multicomponent stress-strength models using an objective Bayesian approach.
Key Innovation: An objective Bayesian method for reliability estimation in multicomponent stress-strength models.
157. Uncertainty-guided alignment for unsupervised domain adaptation in regression
Core Problem: Performing unsupervised domain adaptation for regression tasks, especially when dealing with uncertainty between source and target domains.
Key Innovation: An uncertainty-guided alignment approach for unsupervised domain adaptation in regression.
158. An active learning multi-fidelity Kriging model for predicting the expected lifetime of time- and space-dependent structural systems with multi-fidelity data
Core Problem: Efficiently predicting the expected lifetime of time- and space-dependent structural systems using multi-fidelity data.
Key Innovation: An active learning multi-fidelity Kriging model for lifetime prediction of structural systems.
159. Modelling complexity in system safety: Generalizing the D<sup>2</sup>T<sup>2</sup> methodology
Core Problem: Generalizing the D2T2 methodology to better model complexity in system safety.
Key Innovation: A generalization of the D2T2 methodology for modeling complexity in system safety.
160. Anomaly detection via Gaussian-adaptive reset observer: An entropy approach for predictive maintenance
Core Problem: Detecting anomalies for predictive maintenance in a robust and adaptive manner.
Key Innovation: Anomaly detection via a Gaussian-adaptive reset observer using an entropy approach for predictive maintenance.
161. Health state assessment method for complex systems based on belief rule base with dual interpretability
Core Problem: Assessing the health state of complex systems with a focus on interpretability.
Key Innovation: A health state assessment method for complex systems based on a belief rule base with dual interpretability.
162. PANet: A multi-scale temporal decoupling network and its high-resolution benchmark dataset for detecting pseudo changes in cropland non-agriculturalization
Core Problem: Accurately detecting subtle or 'pseudo' changes related to cropland non-agriculturalization using remote sensing.
Key Innovation: PANet, a multi-scale temporal decoupling network, and a high-resolution benchmark dataset for detecting pseudo changes in cropland non-agriculturalization.
163. Identifying green leaf and leaf phenology of large trees and forests by time series PlanetScope and Sentinel-2 images and the chlorophyll and green leaf indicator (CGLI)
Core Problem: Accurately identifying green leaf and leaf phenology in large trees and forests using satellite imagery.
Key Innovation: A method utilizing time series PlanetScope and Sentinel-2 images and the chlorophyll and green leaf indicator (CGLI) to identify green leaf and leaf phenology.
164. Empowering tree-scale monitoring over large areas: Individual tree delineation from high-resolution imagery
Core Problem: Achieving accurate individual tree delineation from high-resolution imagery for large-area tree-scale monitoring.
Key Innovation: A method for individual tree delineation from high-resolution imagery, enabling tree-scale monitoring over large areas.
165. Unveiling spatiotemporal forest cover patterns breaking the cloud barrier: Annual 30 m mapping in cloud-prone southern China from 2000 to 2020
Core Problem: Accurately mapping annual forest cover patterns in persistently cloud-prone regions over long time periods.
Key Innovation: A method for annual 30m forest cover mapping in cloud-prone southern China from 2000 to 2020, overcoming cloud obstruction.
166. SmartQSM: a novel quantitative structure model using sparse-convolution-based point cloud contraction for reconstruction and analysis of individual tree architecture
Core Problem: Accurately reconstructing and analyzing the complex 3D architecture of individual trees from point cloud data.
Key Innovation: SmartQSM, a novel quantitative structure model employing sparse-convolution-based point cloud contraction for individual tree architecture reconstruction and analysis.
167. Mapping melliferous tree species in Kenya via one-class classification with hyperspectral unsupervised domain adaptation
Core Problem: Accurately mapping specific melliferous tree species in Kenya using hyperspectral data, especially with limited labeled data.
Key Innovation: A method for mapping melliferous tree species in Kenya using one-class classification combined with hyperspectral unsupervised domain adaptation.
168. Improvement of the consistency among long-term global land surface phenology products derived from AVHRR, MODIS, and VIIRS observations
Core Problem: Addressing inconsistencies and improving the comparability of long-term global land surface phenology products from different satellite sensors.
Key Innovation: Methods to improve the consistency among long-term global land surface phenology products derived from AVHRR, MODIS, and VIIRS observations.
169. Optimized CNN-BiLSTM-Attention with hybrid signal denoising: a novel interpretable framework for prediction of shield tunneling advance speed
Core Problem: Accurately predicting the advance speed of shield tunneling machines, which is influenced by complex ground conditions and operational parameters.
Key Innovation: A novel interpretable framework combining CNN-BiLSTM-Attention with hybrid signal denoising for enhanced prediction of tunneling advance speed.
170. A new rock cutting and splitting method for hard-rock excavation: methodology, scaled model test and numerical modelling, and field validation
Core Problem: Developing an efficient and effective method for excavating hard rock formations.
Key Innovation: A new rock cutting and splitting method validated through scaled model tests, numerical modeling, and field application for hard-rock excavation.
171. Muck particle transport behavior and Cutterhead-Agitator synergistic optimization for EPB shield tunneling in sandy pebble strata
Core Problem: Optimizing muck particle transport and cutterhead-agitator synergy during EPB shield tunneling in challenging sandy pebble strata.
Key Innovation: Investigation and optimization of muck transport behavior and cutterhead-agitator synergy to improve efficiency and stability in EPB tunneling through difficult ground.
172. Predicting disc cutter forces for hard rock TBM cutterhead modeling: a comparative analysis of modified CSM semi-theoretical model and hybrid deep learning approach
Core Problem: Accurately predicting disc cutter forces for hard rock TBM cutterhead modeling to optimize excavation performance.
Key Innovation: A comparative analysis of a modified semi-theoretical model and a hybrid deep learning approach for predicting TBM disc cutter forces in hard rock.
173. Influence of inclined groove on rock-crushing behavior of TBM cutter
Core Problem: Investigating how the design of an inclined groove affects the rock-crushing behavior of a TBM cutter.
Key Innovation: Analyzing the impact of inclined groove geometry on TBM cutter performance and rock-crushing efficiency.
174. Experimental study of highly mineralized groundwater-based microbial polysaccharose slurry for slurry shield tunneling
Core Problem: Developing an effective and environmentally friendly slurry for shield tunneling, particularly in the presence of highly mineralized groundwater.
Key Innovation: An experimental study on a microbial polysaccharose slurry based on highly mineralized groundwater for improved slurry shield tunneling.
175. Time-Feature Fused Transformer model: A study on TBM performance prediction and attention evolution patterns
Core Problem: Accurately predicting TBM performance and understanding the evolution patterns of attention in the prediction model.
Key Innovation: A Time-Feature Fused Transformer model for TBM performance prediction, offering insights into attention evolution patterns.
176. Multiphysics modelling of millimetre-wave ablation of geological materials
Core Problem: Modeling the ablation process of geological materials using millimetre-waves.
Key Innovation: Multiphysics modeling of millimetre-wave ablation of geological materials.
177. A numerical manifold method based on numerical integration: Eliminating explicit manifold element generation
Core Problem: Simplifying the implementation and improving the efficiency of the Numerical Manifold Method (NMM) by eliminating explicit manifold element generation.
Key Innovation: Development of a Numerical Manifold Method based on numerical integration, which removes the need for explicit manifold element generation.
178. Influence of consolidation history on the transitional behavior of silty clay
Core Problem: Understanding how the consolidation history affects the transitional behavior of silty clay, a fundamental aspect of soil mechanics relevant to ground stability.
Key Innovation: Investigation of the influence of consolidation history on the transitional behavior of silty clay.
179. The effect of selecting different stress states in laboratory tests on the predicted permanent deformation of soils
Core Problem: Understanding how the choice of stress states in laboratory tests influences the prediction of permanent soil deformation, a fundamental aspect of soil mechanics.
Key Innovation: Investigation of the effect of selecting different stress states in laboratory tests on the predicted permanent deformation of soils.
180. Effects of multi-source solid waste-based stabilizing materials on the improvement of high-salinity silt
Core Problem: Improving the engineering properties of high-salinity silt using sustainable waste-based stabilizing materials.
Key Innovation: Investigation of the effects of multi-source solid waste-based stabilizing materials on the improvement of high-salinity silt, potentially applicable to ground stabilization.
181. BuildingSense: a new multimodal building function classification dataset
Core Problem: Existing building function classification methodologies suffer from poor model interpretability and inadequate multimodal feature fusion, and there's a lack of fine-grained multimodal datasets for this purpose.
Key Innovation: Constructing BuildingSense, a novel multimodal and fine-grained dataset for building function classification, and evaluating state-of-the-art large models on it, demonstrating their effectiveness in comprehending multimodal spatial data, which can inform exposure assessments.
182. How well can we quantify when 1.5 °C of global warming has been exceeded?
Core Problem: Quantifying when the 1.5 °C global warming limit has been exceeded is challenging due to definitional ambiguities, observational limitations, and the non-stationary nature of the climate.
Key Innovation: A comprehensive community methodological overview proposing clear and reasoned ways to quantify the exceedance of global warming levels, combining lines of evidence and methodologies to estimate the present long-term warming level robustly.
183. Questioning the Endorheic Paradigm: water balance dynamics in the Salar del Huasco basin, Chile
Core Problem: Understanding water balance dynamics in arid endorheic basins is crucial for sustainable water resources management, but the endorheic assumption itself might be flawed for some basins.
Key Innovation: Examining rainfall and evaporation drivers of groundwater recharge and water balance in the Salar del Huasco basin using a modified semi-distributed rainfall-runoff model and satellite data, challenging the endorheic assumption and highlighting groundwater-evaporation trade-offs.
184. Increased surface water evaporation loss induced by reservoir development on the Loess Plateau
Core Problem: Evaporation losses from surface water bodies due to global-scale reservoir construction are poorly understood, especially in the context of climate change, with most studies overlooking aquatic evaporation dynamics.
Key Innovation: Investigating water body evaporation on the Loess Plateau using a modified Penman model and remote sensing data, revealing a significant upward trend in total evaporation volume primarily driven by the expansion of small- and medium-sized reservoirs and check dams, impacting water resources.
185. Co-simulation of continuous and categorical variables: application in the Shuiyindong gold deposit modeling
Core Problem: Accurately characterizing concealed orebodies in complex geological environments, specifically addressing the spatial interdependence between continuous (grade) and categorical (lithology) variables for 3D modeling and uncertainty assessment.
Key Innovation: Implemented a unified co-simulation workflow integrating Plurigaussian simulation and Turning Bands algorithm to co-simulate ore grade and rock type, effectively representing spatial anisotropy and variability for enhanced understanding of ore distribution and identifying zones for further exploration.
186. Accurate intelligent modeling of mud loss while drilling wells via soft computing methods
Core Problem: Accurately predicting mud loss volume in drilling operations to mitigate economic losses, operational delays, and environmental concerns.
Key Innovation: Developed and evaluated advanced ensemble machine learning models (RF, AdaBoost, DT, and a custom stacking framework) for mud loss prediction, demonstrating AdaBoost's superior performance (R2 of 0.828) and identifying key influencing factors (mud viscosity, solid content, hole size, differential pressure).