TerraMosaic Daily Digest: Feb 27, 2026
Daily Summary
This February 27, 2026 digest is defined by a clear shift from hazard description to mechanism-resolving, decision-ready analysis. The highest-impact studies explicitly connect source uncertainty to consequence: earthquake-parameter uncertainty to millisecond tsunami ensemble forecasts, extreme monsoon forcing to coupled landslide-flash-flood failure chains, and fault-zone degradation to spatially explicit landslide susceptibility across tectonically active terrain.
A second signal is engineering translation under real constraints. Several papers move beyond post hoc interpretation to deployable risk controls: probabilistic wedge-failure design for underground drifts, theory-grounded rockburst warning thresholds, MT-InSAR-enhanced corridor susceptibility zoning, and CNN-accelerated random-field embankment reliability. In parallel, cryosphere and hydroclimate studies quantify operationally relevant change in sea ice, snow processes, and permafrost-coupled foundation response, tightening the link between climate diagnostics and infrastructure safety.
Key Trends
The strongest contributions combine physical interpretability, uncertainty quantification, and operational relevance.
- Hazard forecasting is becoming uncertainty-explicit: tsunami, rainfall-triggered slope failure, and seasonal hydroclimate studies increasingly model uncertainty as a first-class output rather than a residual error term.
- Landslide science is integrating structural geology with deformation sensing: new work fuses fault-damage characterization, MT-InSAR time series, and geomorphometric indicators to separate where slopes are weak from when they are likely to fail.
- Geotechnical risk methods are moving toward probabilistic design workflows: wedge stability, embankment failure, and tunnel vibration analyses now emphasize probability, parameter sensitivity, and design-level thresholds instead of deterministic safety factors alone.
- Cryosphere papers are shifting from climate diagnostics to hazard implications: sea-ice, ice-shelf, and frozen-ground studies increasingly report quantities that feed directly into flood, slope, and infrastructure risk decisions.
- General AI papers are being retained only when transfer value is explicit: algorithmic advances are prioritized when they improve geohazard monitoring, inversion, or risk governance under sparse or biased field data.
Selected Papers
This digest features 78 selected papers from 524 papers analyzed 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. Machine learning-based ensemble tsunami forecasting: A dual-branch SP-MLP approach for the South China Sea
Core Problem: Addressing uncertainties in preliminary earthquake information for rapid and accurate tsunami hazard assessment and forecasting, especially under strict time constraints.
Key Innovation: Developed a novel segmented physically coupled multilayer perceptron (SP-MLP) with a dual-branch architecture to directly predict full tsunami wave-height time series from earthquake source parameters, enabling rapid, high-resolution probabilistic forecasts within milliseconds per realization.
2. Catastrophic rainfall-induced landslides and flash floods in the Darjeeling Himalaya, Eastern India: Insights from a post-monsoonal hydro-geomorphic extreme event
Core Problem: Understanding the mechanisms and triggers of catastrophic rainfall-induced landslides and flash floods in the Darjeeling Himalaya, particularly following extreme hydro-geomorphic events.
Key Innovation: Providing insights into the characteristics and causes of a specific catastrophic rainfall-induced landslide and flash flood event in the Darjeeling Himalaya, Eastern India.
3. Preliminary inferences on the August 5, 2025, catastrophic debris flow in Dharali, India: insights from the satellite imageries and geomorphological analysis of the Kheer Gad watershed
Core Problem: Understanding the key factors governing the evolution and transformation of catastrophic debris flow events in mountainous watersheds, particularly the interplay of hydroclimatic forcing and geomorphological attributes, and the need for comprehensive assessments.
Key Innovation: Integrating satellite imagery and detailed morphometric assessments of the Kheer Gad watershed to infer how watershed characteristics (erosion potential, valley floor width-to-height ratios, stream length-gradient indices, knickpoints) amplified a catastrophic debris flow event, and suggesting mitigation measures for similar settings.
4. Field investigation and numerical simulation of transmission tower landslides under rainfall
Core Problem: Managing geohazards that threaten critical infrastructure, specifically understanding the mechanisms of rainfall-induced seepage, slope stability, and seismic response leading to transmission tower landslides, and developing effective countermeasures.
Key Innovation: Combining field investigation, laboratory testing, and transient ABAQUS-based numerical simulation to achieve a mechanistic understanding of time-dependent coupled hydro-mechanical and seismic effects on a transmission tower landslide, evaluating the effectiveness of countermeasures like drainage ditches and micro steel pipe retaining walls.
5. Integrated risk-based design approach for wedge hazard in underground mining drifts
Core Problem: Existing stability analysis methods for underground excavations cannot fully determine the probability of occurrence and size of wedges, which is critical for quantifying the hazard and risk associated with wedge failure.
Key Innovation: Presents a comprehensive DFN-based risk analysis approach that considers complex structural regimes and 3D excavation profiles to estimate the location and size of critical wedges, quantify risk, and optimize ground support systems.
6. Anomalies of land-atmospheric parameters in the 2023 Turkey earthquake (Mw7.8) using GNSS and ERA5
Core Problem: Detecting and understanding pre-seismic or co-seismic anomalies in land surface and atmospheric parameters associated with major earthquakes to potentially aid in monitoring and analysis.
Key Innovation: Observed and analyzed anomalies of land surface and atmospheric multi-parameters using GNSS and ERA5 data in relation to the 2023 Turkey earthquake (Mw7.8), indicating a potential for identifying earthquake-related precursors or effects.
7. Optimizing landslide susceptibility zoning for clean energy transmission corridors in loess regions using MT-InSAR and boosting ensemble framework
Core Problem: Conventional landslide susceptibility assessment (LSA) methods fail to incorporate the variability of ground deformation characteristics, leading to risk overestimation in dynamically active areas like clean energy transmission corridors.
Key Innovation: Proposes an optimized LSA framework integrating Multi-Temporal Interferometric Synthetic Aperture Radar (MT-InSAR) for long-term surface deformation patterns with a Boosting ensemble framework (CatBoost model) to refine initial susceptibility assessments, significantly reducing overestimation by incorporating varied surface deformation trends.
8. Spatial heterogeneity of rock mass degradation and landslide susceptibility in the fault damage zone, Northwestern Himalayas, Pakistan
Core Problem: Fault Damage Zones (FDZs) experience significant mechanical deterioration of rock masses due to neo-tectonic stresses, which critically affects slope stability and landslide development, requiring a better understanding of this relationship.
Key Innovation: Provides the first macro-scale FDZ characterization in the Himalayas, demonstrating heterogeneous rock mass degradation and its clear relationship with fault-induced deformation, rock mass quality, and increased landslide susceptibility near Intense Deformation Zones (IDZs), offering a predictive framework for landslide hazards.
9. Numerical modeling of the out-of-plane dynamic response of masonry gable walls via a high-fidelity block-based finite element modeling approach – part I: Blind prediction
Core Problem: Reliably predicting the dynamic out-of-plane (OOP) response and collapse mechanisms of unreinforced masonry gable walls under earthquake loading, especially with limited experimental data.
Key Innovation: A novel 3D block-based finite element model, extended for dynamic OOP bending, accurately predicted key experimental observations (collapse onset, failure mechanisms, displacement demands, dynamic responses) of masonry gables under seismic loading, demonstrating its potential for complementing physical testing and extending understanding of seismic performance.
10. Intelligent detection method for tunnel water inrush disasters based on deep learning
Core Problem: Existing tunnel monitoring methods for water inrush disasters have limitations in monitoring coverage, real-time performance, data intuitiveness, and accurate identification.
Key Innovation: This paper proposed YOLOv8-WI, an intelligent recognition algorithm for tunnel water inrush disasters based on deep learning, which integrates a lightweight backbone, multi-fusion attention mechanism, and MPDIoU loss, achieving significant improvements in detection accuracy, computational efficiency, and deployment complexity on a newly constructed TWI dataset.
11. Determination of critical stress for rockburst warning: A model incorporating borehole-sensor interaction and its validation
Core Problem: Rockburst real-time monitoring and early warning systems lack a fundamental theoretical method for establishing monitoring thresholds for stress sensors installed in boreholes, based on the quantitative interaction between the sensor and borehole.
Key Innovation: Developed a theoretical damage model for the interaction between stress sensors and boreholes in rockburst-prone roadways, characterizing driving forces and deriving characteristic formulae. Proposed a theoretical criterion and method to determine the warning threshold of stress sensors, directly related to rock strength, brittleness, mining stress, and borehole size, validated with a low error range (4.2%–6.5%) in typical rockburst mines.
12. Integrating 3D Point Cloud analysis for potentially unstable rock blocks characterization: a method for assessing size and shape distribution
Core Problem: Traditional methods for characterizing potentially unstable rock blocks in rockfall risk analysis have limitations in accurately assessing size and shape distribution, hindering comprehensive geoengineering assessments.
Key Innovation: Developed a novel workflow integrating high-resolution 3D point cloud analysis for characterizing potentially unstable rock blocks. The method includes block segmentation, 3D modeling, quantitative extraction of volume, surface area, and shape, and introduces a new parameter for block flattening to refine shape classification, providing detailed insights for rockfall risk management.
13. Enhancing random field-based limit equilibrium analysis of road embankment failures on soft soils using convolutional neural networks
Core Problem: Conventional Monte Carlo simulations for probabilistic geotechnical analysis of road embankment stability are computationally expensive, especially when considering spatial variability of soil properties and non-stationary conditions.
Key Innovation: Developed a probabilistic framework integrating random field theory with convolutional neural networks (CNNs) to efficiently evaluate the stability of road embankments, demonstrating that CNNs can accurately reproduce probabilistic stability outcomes while substantially reducing computational time for real-time risk assessment.
14. Active protothrusts and fluid highways: Seismic noise reveals hidden subduction dynamics in Cascadia
Core Problem: Understanding the complex interactions of strain accumulation, fault slip, and fluid migration in subduction zones and their influence on earthquake behavior and dynamics, particularly in the Cascadia region.
Key Innovation: Using a decade of ambient seismic data to reveal distinct regional variations in Cascadia's subduction dynamics, showing differences in megathrust locking and fluid migration, and suggesting fluid transport modulates earthquake behavior and promotes rupture arrest.
15. Erosion and compressibility characteristics of dispersive soil amended with calcinated coal gangue and xanthan gum
Core Problem: Dispersive soils rapidly disintegrate in water, compromising supported structures due to monovalent sodium ions, necessitating effective stabilization methods to control dispersivity, erosion, and compressibility.
Key Innovation: Examining the performance of calcinated coal gangue (CCG) and xanthan gum (XG) in stabilizing dispersive soil, demonstrating significant reduction in dispersion ratio and erosion depth, and improved compressibility characteristics, attributed to strengthened soil aggregation and void filling, recommending the modified soil for short-term applications.
16. GIS-based mapping of bearing capacity and liquefaction hazard for the Srinagar metropolitan region of Kashmir
Core Problem: The Srinagar Metropolitan Region, with its variable sedimentary deposits, requires a comprehensive evaluation and mapping of bearing capacity and liquefaction hazards to support safe infrastructure development and land-use planning.
Key Innovation: Develops two novel GIS-based indices, the Bearing Capacity Hazard Index (BCHI) and the Liquefaction Hazard Index (LHI), by integrating multiple geotechnical parameters through an Analytical Hierarchy Approach to create detailed hazard maps for the region.
17. Analysis of horizontal dynamic response of end-bearing piles in saturated frozen soil under hydro-thermo-mechanical-seismic coupling effects
Core Problem: The horizontal vibration response of piles in saturated frozen soil under seismic SH-waves is significantly controlled by the superstructure and temperature, requiring a comprehensive dynamic coupling model.
Key Innovation: This research constructed a dynamic coupling model of saturated frozen soil-temperature field-pile-superstructure under vertically incident SH-waves, deriving analytical solutions and revealing the influence of parameters like superstructure load and pile slenderness ratio on horizontal vibration response, providing theoretical support for seismic design in permafrost regions.
18. Advancing annual-scale significant wave height prediction: A comprehensive skill assessment of M5P decision tree approach
Core Problem: Accurate and computationally efficient annual-scale prediction of significant wind-wave height is crucial for maritime safety and coastal management, but existing ML/numerical models can be complex and computationally demanding.
Key Innovation: Demonstrates that the M5P Decision Tree, with monthly data segmentation and wind speed as the most influential predictor, offers a practical, interpretable, and computationally efficient approach for significant wave height forecasting, outperforming traditional models in accuracy (RMSE as low as 0.059).
19. Solid-liquid transition behavior and shear strength degradation mechanisms of deep-sea sediments induced by water content variation and external disturbance
Core Problem: Understanding the solid-liquid transition behavior and shear strength degradation mechanisms of deep-sea sediments under external disturbance and varying water content.
Key Innovation: Developed a unified constitutive model coupling water content and shear rate effects on soil structure, accurately describing post-disturbance shear strength degradation, based on systematic disturbance mechanics tests.
20. Reassessing boreal wildfire drivers enables high-resolution mapping of emissions for climate adaptation
Core Problem: Current regional-scale methods for estimating boreal wildfire emissions lack the spatial precision needed to understand flux drivers and inform carbon storage strategies.
Key Innovation: Developed a highly constrained procedure for estimating wildfire emissions at local and regional scales by reassessing heat development and expanding satellite-derived fire radiative power applications, emphasizing fine-scale variation in climate-sensitive fuel loading.
21. Consistent tangent operator for elastoplastic model of frozen-thawed rock incorporating crack closure
Core Problem: Accurately modeling the nonlinear elastoplastic behavior of frozen-thawed rock, particularly due to crack closure and cumulative plastic strain under mechanical loading, while ensuring computational efficiency and convergence in numerical simulations.
Key Innovation: Develops a consistent tangent operator (CTO) based on a fully implicit Newton method and a modified DP yield function with a non-associated flow rule, which improves computational efficiency and accuracy in capturing crack closure evolution and accumulative plastic response of frozen-thawed rock.
22. A finite detonation equivalent load for numerical analysis of tunnel blasting vibrations
Core Problem: Conventional numerical simulations of tunnel blasting-induced vibrations, relying on instantaneous detonation, inaccurately predict vibration responses and lack explicit incorporation of key finite detonation characteristics.
Key Innovation: Proposes a Finite Detonation Equivalent Load (FDEL) method that explicitly incorporates initiation position, detonation direction, and velocity of detonation, accurately reproducing stress wave superposition and reducing computational cost for predicting tunnel blasting vibrations.
23. Quasi‐Periodic Behavior of Tropical Cyclone Precursors and Its Phase Matching With the Insolation Diurnal Cycle
Core Problem: The formation of tropical cyclones (TCs) remains a significant scientific challenge, particularly understanding the role of quasi-periodic behavior (QPB) in TC genesis.
Key Innovation: Demonstrated robust quasi-periodic behavior (QPB) during TC genesis in idealized numerical simulations, identifying it as a convectively coupled inertia-gravity oscillation. Showed that phase matching between the QPB and the solar diurnal cycle accelerates TC genesis due to state-dependent responses of precursors to diurnal radiation.
24. Assessment of lateral capacity of semi-rigid monopile for offshore wind turbines in marine clay
Core Problem: Current upper-bound methods for assessing the lateral capacity of offshore monopiles are inadequate for semi-rigid piles in marine clay, which exhibit a unique triple failure mode (shallow wedge, middle full flow, deep rotational failure).
Key Innovation: A novel upper-bound solution that rigorously incorporates the combined translational-rotational deflection and the unique triple failure mechanism of semi-rigid piles in clay, providing a theoretically robust tool for predicting their ultimate lateral capacity.
25. Dynamic response analysis of a 15 MW semi-submersible FOWT under the freak wave and extreme operation gust
Core Problem: The operational survivability of ultra-large floating offshore wind turbines (FOWTs) is severely challenged by complex extreme marine environments, particularly the combined impact of freak waves and extreme operational gusts, leading to unknown dynamic responses and failure modes.
Key Innovation: A fully coupled aero-hydro-structure-mooring model framework revealing a novel aerodynamic unloading-rebound-slamming coupling mechanism under combined extreme events, which amplifies buoy motions and tower-base bending moments, and identifies distinct mooring system failure modes.
26. A generalised method for spudcan penetration analysis in multi-layered soils
Core Problem: Accurately predicting spudcan penetration behavior in multi-layered soils is challenging because existing industry standards use simplified 'wished-in-place' approaches, and mechanism-based models lack generalizability to complex stratigraphy.
Key Innovation: A unified framework synthesizing two decades of research into spudcan penetration mechanisms, incorporating dynamic evolution of soil failure, soil plug dynamics, and cross-layer interactions into a top-down predictive algorithm, offering improved accuracy and broader applicability for multi-layered seabeds.
27. Research on traveling characteristics of propelled deep-sea mining vehicle based on experiment and MBD-DEM simulation
Core Problem: Traditional tracked deep-sea mining vehicles cause significant sediment disturbance, conflicting with stringent environmental standards.
Key Innovation: Proposed a hybrid propeller-powered system with a skid grounding mechanism, and used MBD-DEM simulations and experiments to optimize travel parameters, significantly reducing travel resistance (74.01%) and sinkage depth (62.69%) compared to tracked mechanisms, thus lowering environmental disturbance.
28. Accelerated identification of hardening soil model parameters employing tree-based machine learning algorithms
Core Problem: The time-consuming and inconsistent nature of identifying constitutive model parameters for elasto-plastic double hardening soil models, which limits the practical application of advanced models in geotechnical engineering.
Key Innovation: Proposed a machine learning approach using tree-based algorithms to rapidly and consistently identify hardening soil model parameters from laboratory test data. This method significantly reduces determination time (by orders of magnitude) and improves accuracy compared to manual and automated optimization, promoting the use of advanced constitutive models in practice.
29. Test mining sites at the moon: A review of resource extraction opportunities and challenges at de Gerlache, Shackleton, and Sverdrup-Henson Regions
Core Problem: Identifying and assessing suitable lunar south polar regions for future resource extraction, considering uncertainties in volatile resource distribution, accessibility, and various engineering and geological constraints, including hazards.
Key Innovation: A site feasibility assessment of nine lunar south polar locations, integrating geological characteristics, environmental conditions, and an optimization-based scoring framework to compare sites based on resource potential, engineering constraints, and "geologic hazards," providing a tool for lunar mining planning and site prioritization.
30. Peatland dynamics and sensitivities under hyperhumid climates: Paleoecological evidence from glacier-valley peatlands in the southeastern Tibetan Plateau
Core Problem: Understanding the long-term development, carbon accumulation histories, and sensitivity of peatlands in hyperhumid glacier valleys to climate change, particularly their vulnerability to degradation.
Key Innovation: Reveals that peatland development in the southeastern Tibetan Plateau was triggered by decreased precipitation, underwent fen-to-bog transitions under shifting climate, and that carbon accumulation rates were controlled by localized hydrological disturbances, making them vulnerable to natural degradation under a wetter future.
31. Flow resistance continuity from overland to rill flows: evidence from experimental data and theoretical implications
Core Problem: Lack of a continuous flow resistance law applicable from overland to rill flows, which is required for accurate modeling of soil erosion and transport processes.
Key Innovation: Used literature flume measurements to demonstrate a continuity in flow resistance from overland to rill flows, theoretically deducing and calibrating a new flow resistance law applicable for rectangular cross-sections varying from very wide to narrow, with high accuracy.
32. A frictional contact approach for saturated soil-structure dynamic interaction considering fluid-solid coupling and contact coupling effects
Core Problem: Existing models for dynamic soil-structure interaction in saturated soils struggle to accurately characterize nonlinear frictional contact behaviors (sticking-sliding-separation) and variable boundary conditions at the interface, especially considering both fluid-solid and contact coupling effects.
Key Innovation: Proposes a dynamic frictional contact model based on the u-p formulation and a two-pass penalty method, incorporating Mohr-Coulomb criterion and effective stress, to accurately characterize "sticking-sliding-separation" states and variable constraints at saturated soil-structure interfaces.
33. Adaptive stable node-based smoothed upper bound finite element limit analysis considering discontinuous velocity fields for soil-structure interaction problems
Core Problem: Need for accurate, robust, and efficient numerical methods for soil-structure interaction problems that can effectively consider discontinuous velocity fields along interfaces to predict plastic regions and limit loads.
Key Innovation: Proposed an adaptive stable node-based smoothed upper bound finite element limit analysis (adaptive SNS-UBFELA) method, incorporating discontinuous velocity fields and an adaptive mesh refinement scheme, demonstrating higher accuracy, numerical robustness, and computational efficiency for soil-structure interaction analysis.
34. Drilling parameter acquisition and rock strength determination on rotary-percussive drilling for drill-and-blast excavation
Core Problem: Lack of reliable methods for determining in situ rock strength, leading to reliance on limited laboratory test results for drill-and-blast excavation design and rock mass characterization.
Key Innovation: Proposed a novel method for estimating in situ rock compressive and tensile strengths using drilling monitoring data from a developed rotary-percussive drilling test system, demonstrating a nonlinear correlation between drilling work done and rock strength under in situ stress conditions.
35. Major Melting Event on the Ross Ice Shelf, Antarctica, Connected With Enhanced Atmospheric Turbulence
Core Problem: Understanding the atmospheric conditions and mechanisms driving extensive surface melting events on the Ross Ice Shelf, particularly the role of atmospheric water vapor and turbulent mixing, which are challenging to measure in remote polar regions.
Key Innovation: Used atmospheric estimates from a network of 13 GNSS receivers on the Ross Ice Shelf to quantify the relative strength of atmospheric turbulence, finding it was four times greater during the January 2016 melting event. This demonstrates a novel application of GNSS for remotely observing unusual atmospheric conditions connected to ice-shelf melting.
36. ANBI-UAV: A Framework for Automated Noise Barrier Inspection Using UAV Imagery in High-Speed Railway
Core Problem: Traditional noise barrier inspection methods for high-speed railways are inefficient, costly, and impractical for large-scale assessments, leading to undetected structural defects that compromise noise insulation and safety.
Key Innovation: Proposed ANBI-UAV, a novel framework for automated noise barrier inspection using Uncrewed Aerial Vehicle (UAV) imagery and a specialized YOLO-based network (ANBINet) with simplified C3 and multiple attention modules, achieving high performance in defect detection and demonstrating practical application for large-scale infrastructure monitoring.
37. Refining the Lagrangian approach for moisture source identification through sensitivity testing of assumptions using BTrIMS1.1
Core Problem: Lagrangian models for moisture source identification rely on untested assumptions, leading to ambiguous or inaccurate results in understanding precipitation patterns, which are crucial for extreme weather events.
Key Innovation: Comprehensive sensitivity testing of Lagrangian model assumptions (parcel number, release height, vertical movement, identification methods) leading to an improved BTrIMS1.1 model, providing critical information for more accurate moisture source identification relevant to precipitation patterns.
38. NorESM2–DIAM: a coupled model for investigating global and regional climate-economy interactions
Core Problem: Existing climate-economy models are limited by coarse spatial aggregation, simplified climate representation, or lack of interaction, hindering accurate assessment of climate change impacts and associated risks.
Key Innovation: Development of NorESM2–DIAM, the first framework to fully couple an Earth System Model with a high-resolution cost-benefit Integrated Assessment Model, enabling spatially disaggregated analysis of climate-economy interactions and potential incorporation of extreme events.
39. Implementation of a multi-layer snow scheme in the GloSea6 seasonal forecast system: impacts on land–atmosphere interactions and climatological biases
Core Problem: Traditional single-layer snow schemes in seasonal forecast systems inadequately represent snowpack thermal processes, leading to climatological biases in land-atmosphere interactions, soil moisture, temperature, and precipitation, which are critical for hydrological hazard prediction.
Key Innovation: Implementation of a multi-layer snow scheme in the GloSea6 seasonal forecast system, which delays snowmelt, improves soil moisture memory, enhances evaporative cooling, and mitigates near-surface warming biases, leading to improved simulation of temperature and precipitation relevant for hydrological hazard forecasting.
40. Incorporating natural variability in master recession curves
Core Problem: Traditional Master Recession Curves (MRCs) often represent an average, failing to capture the natural variability (aleatory uncertainty) in streamflow recession due to diverse antecedent hydroclimatic and storage conditions, which are important for hydrological assessments.
Key Innovation: Hypothesis and demonstration of a continuum of recession curves, represented by a family of percentile curves, which captures the natural variability in streamflow recessions, supported by a two-store qualitative model.
41. Optical logic convolutional neural network
Core Problem: Optical analog computing is susceptible to perturbations and relies on digital-to-analog/analog-to-digital converters, while optical digital computing lacks flexibility for broader AI applications like inference.
Key Innovation: Proposal and demonstration of an optical logic convolutional neural network (OLCNN) using optical logic convolutional operators for high-speed pattern generation, image edge extraction, and classification, pioneering a logic-driven paradigm for energy-efficient optical AI hardware.
42. From ecological sensemaking to sensegiving: a way to develop a scientific project that reaches out to society
Core Problem: A persistent gap exists between geosciences and society, leading to natural heritage being primarily viewed as biotic elements and hindering collective actions for conservation in a rapidly changing planet.
Key Innovation: Analyzed an interdisciplinary exchange project that uses organizational science concepts to bridge the gap between geosciences and society, fostering a holistic view of natural heritage and enabling knowledge-sharing activities for conservation and protection of natural sites.
43. Shear Behavior of Recycled Coarse Aggregate Concrete–Soil Interfaces Subjected to Sulfuric Acid Corrosion
Core Problem: Ensuring the practical application and durability of recycled coarse aggregate concrete piles in chemically aggressive (sulfuric acid) soil environments, requiring an understanding of their interfacial shear behavior.
Key Innovation: Large-scale direct shear tests demonstrated that both prolonged sulfuric acid corrosion and increased recycled coarse aggregate replacement ratio significantly enhance interfacial shear strength and displacement. A time-dependent modification to the conventional hyperbolic model was proposed, improving predictive accuracy for corroded interfaces.
44. Quantitative reconstruction of the paleoclimate over the past 200 years in the northern margin of the east Asian summer monsoon, China
Core Problem: Lack of quantitative climate data to understand long-term climatic and environmental changes in the northern margin of the East Asian summer monsoon (NMEASM), a region vulnerable to climate change.
Key Innovation: Provides a high-resolution quantitative reconstruction of January temperature and rainy season precipitation over the past ~200 years in the NMEASM, revealing significant warming and increased precipitation, vegetation shifts, and climatic cycles linked to ENSO and PDO.
45. Vegetation increases carbon and nitrogen content in riparian soils along high-Arctic headwater streams
Core Problem: Understanding how changes in riparian vegetation cover in the High-Arctic influence carbon and nitrogen content in riparian soils and their potential runoff to streams and downstream ecosystems under a warmer climate, given the sensitivity of permafrost regions.
Key Innovation: Demonstrates that higher vegetation cover in High-Arctic riparian soils leads to increased organic C, total N, and dissolved organic carbon (DOC) content, suggesting that increased vegetated areas under a warmer climate will increase C and N available for runoff, potentially influencing stream ecosystems and contributing to permafrost thaw and erosion.
46. Elucidating the influence of grain boundaries and microstructure characteristics on the mechanical behavior of laminated shale: Insights from EBSD, nano-indentation characterization, and Mori-Tanaka upscaling method
Core Problem: Accurate characterization of the mechanical properties of laminated shale is critical for optimizing unconventional hydrocarbon extraction, but grain-scale influences on its multi-scale mechanical behavior remain insufficiently understood.
Key Innovation: Integrated EBSD, nano-indentation, and the Mori-Tanaka homogenization method to investigate the influence of grain boundary states and microstructure characteristics on shale's multi-scale mechanical behavior. Revealed significant variability in micro-mechanical properties based on grain position, morphology, and boundary types, and quantitatively related grain-scale features to macroscopic properties, providing insights for refining hydraulic fracturing strategies.
47. Origin-dependent creep resistance in shale chlorite revealed by nanoindentation
Core Problem: The long-term mechanical properties of clay minerals govern the stability of hydraulic fractures in shale, but the mechanisms underlying the differential creep behavior between authigenic and detrital chlorite, which have distinct properties, remain poorly understood.
Key Innovation: Systematically investigated the origin-dependent creep properties of detrital and authigenic chlorite in shale through SEM and nanoindentation tests. Revealed that authigenic chlorite exhibits superior mechanical properties (higher elastic modulus, hardness, and creep resistance) and clarified the mechanisms responsible for these differences, providing critical insights for predicting long-term deformation in shale reservoirs.
48. Hydrological effects of vegetation greenness change in China: a first-order assessment of the role of moisture recycling
Core Problem: Previous studies evaluating hydrological impacts of vegetation greening often neglect the role of moisture recycling, leading to potential biases in assessing water availability and runoff.
Key Innovation: Conducted a first-order, offline diagnostic assessment integrating a remotely sensed ET model, moisture-tracking dataset, and Budyko framework to quantify the effects of vegetation greenness change on evapotranspiration, precipitation, and water availability, highlighting the significant contribution of moisture recycling to precipitation and runoff.
49. Divergent hydrological responses of multi-depth soil moisture variability to multiple factors on the Loess Plateau, China
Core Problem: Limited understanding of the internal mechanisms, synergistic, and interactive influences of multiple factors regulating multi-depth soil moisture (SM) variability.
Key Innovation: Examined spatiotemporal patterns and dominant seasonal differences of surface and root zone SM on the Loess Plateau, clarifying divergent mechanisms of independent and synergistic effects of multiple factors (precipitation, vegetation, circulation factors), and highlighting significant interactions between vegetation and climatic factors.
50. Enhancement of Glover’s equation for non-steady state drainage
Core Problem: The classical Glover's equation for non-steady state drainage, used to determine drain spacing, exhibits significant prediction errors for short times because it neglects higher-order terms of the exponential series.
Key Innovation: Proposed a methodology to incorporate additional terms of the exponential series into Glover's equation, enabling more accurate and reliable determination of drain spacing independently of time, especially for shorter time periods (e.g., t = 3 days) with relative errors below 3.4%.
51. Advanced hypoplastic contact modelling of field tests for tension piles in layered soils
Core Problem: Conventional Coulomb friction models fail to accurately reproduce the complete non-linear load-displacement response of soil-pile interfaces, especially during unloading-reloading, leading to less reliable foundation design.
Key Innovation: Presents the first 3D back-analysis of a hypoplastic contact model, benchmarked against full-scale tensile load tests, using a physically consistent calibration framework derived from torsional interface shear tests, significantly improving accuracy in modeling soil-pile interaction in complex layered soils.
52. North Pacific Model Biases Influence Kuroshio Extension Atmospheric Circulation Patterns
Core Problem: Coupled climate models struggle to accurately reproduce Kuroshio Extension (KE) atmospheric circulation patterns and their underlying mechanisms, often exhibiting cold sea surface temperature (SST) biases.
Key Innovation: Evaluated coupled models using satellite-derived and reanalysis products, showing that a stronger latent heat flux response is linked to better simulation of KE atmospheric circulation patterns. Identified that models with cold SST biases over the Kuroshio-Oyashio Extension region systematically underperform, and increasing resolution does not consistently alleviate these biases, suggesting structural model issues.
53. Estimating Ice Cover on the Great Lakes Using Seismic Ambient Noise
Core Problem: Quantifying the relationship between lake-microseism noise and ice cover to enable seismic monitoring of ice changes, especially in remote regions where direct measurements are challenging.
Key Innovation: Quantified the covariance between seismic data from a station near Lake Superior and satellite-derived lake ice cover (2009-2024). Developed a linear model that explains ~87% of the variance in 0.5-2Hz seismic noise during coldest months, demonstrating that seismic data can reliably track lake ice and wave activity, offering new monitoring possibilities.
54. Anisotropic Melt Inclusions as a Confounding Signal for Ice‐Penetrating Radar Observations
Core Problem: Conventional radar sounding analyses either ignore melt or assume spherical melt inclusions, obscuring the anisotropic contributions of melt and potentially leading to significant errors in estimates of subsurface hydrology, water content, and ice fabric.
Key Innovation: Used geometric mixing models to calculate reflectivity, attenuation, and birefringence of temperate ice containing anisotropic melt. Found that anisotropic melt can introduce significant deviations in radar measurements (e.g., up to 30-43% error in water content estimates) and that even small melt fractions can reproduce birefringence signals previously attributed to ice fabric, calling for a critical reassessment of radar analyses.
55. Improving Heterogeneous Aquifer Characterization Using a Deep Learning‐Based Ensemble Smoother With a Hybrid Prior Strategy
Core Problem: Accurate characterization of aquifer hydraulic properties is crucial, but most traditional inverse methods rely on unbiased prior assumptions, which are frequently not met in real-world applications, significantly compromising characterization accuracy.
Key Innovation: Introduced a deep learning-based ensemble smoother (ESDL) with a novel hybrid prior strategy that combines multiple plausible prior assumptions, enabling robust estimation of unknown, heterogeneous aquifer hydraulic properties without relying on unbiased prior assumptions, outperforming traditional methods in accuracy and stability.
56. A multi-parameter model for predicting propeller-jet scour depth based on maximum axial flow velocity
Core Problem: Existing models for propeller-jet scour depth prediction often omit or inadequately represent propeller geometry, boundary conditions, and sediment properties, and fail to reflect the dominant role of flow velocity.
Key Innovation: Proposed a multi-parameter prediction model for propeller-jet scour depth based on the maximum axial flow velocity, derived from flume experiments, to synergistically represent flow velocity and other parameters, enhancing physical interpretability and predictive accuracy.
57. The performance of shared suction caisson anchors in sand for floating offshore wind arrays with taut and catenary mooring
Core Problem: Uncertainties regarding the ability of shared suction caisson anchors to maintain station keeping for floating offshore wind arrays under multidirectional storm loading, potentially leading to over-conservative design.
Key Innovation: Developed a modeling approach combining mooring load simulations from floating offshore wind turbines with nonlinear dynamic finite element analysis using an advanced constitutive model (Sanisand-MS) to investigate the performance of shared suction caisson anchors in sand under extreme sea states.
58. Towards a multi-robot architecture for bearing-only tracking of underwater acoustic sources using vector sensors
Core Problem: Real-time estimation of underwater acoustic source positions using multiple AUVs with bearing-only observations, especially under communication-constrained conditions and asynchronous data arrivals.
Key Innovation: A distributed multi-vehicle architecture combining onboard Acoustic Vector Sensor (AVS) data processing with a pseudo-linear Bearing Only Tracking (BOT) algorithm, validated through scaled sea experiments and simulations, demonstrating practical viability for real-time underwater target tracking.
59. A novel Contact-Based relative positioning methodology for underwater inspection robots: Conceptualization and numerical validation
Core Problem: Underwater inspection robots struggle to maintain stable relative positions to targets in challenging environments (shallow water, strong currents, high turbidity) where traditional ranging methods fail.
Key Innovation: A novel contact-based relative localization system that measures robot position and attitude relative to an inspection target using two contacting arms, coupled with a control framework for auto-regulation of distance and heading, enabling effective surface-following inspection under disturbances.
60. Environment90m – globally standardized environmental variables for freshwater science at high spatial resolution
Core Problem: Establishing a global baseline for freshwater habitats and biodiversity is difficult due to the lack of standardized, high-resolution environmental information to characterize freshwater habitats worldwide.
Key Innovation: Creation of the Environment90m dataset, which aggregates 45 topographic/hydrographic, 19 climate, 22 land cover, 15 soil, aridity, and streamflow variables into 726 million sub-catchments at 90m resolution, along with tools for data access and processing, supporting spatial freshwater biodiversity science.
61. Mapping key soil micronutrients across the Tibetan Plateau
Core Problem: The distribution patterns of soil micronutrients in high-altitude ecosystems like the Tibetan Plateau are poorly quantified, hindering the understanding of critical ecological functions and sustainable ecosystem management.
Key Innovation: Assembly of a plateau-wide dataset from 526 sites, measurement of four micronutrients (Fe, Mn, Zn, V) using XRF/ICP-MS, and application of Random Forest models to quantify controls and generate 1 km high-resolution spatial maps with uncertainty layers, supporting benchmarking of micronutrient cycling models and ecosystem management.
62. Global 30-m annual cropland extent dynamics (2000–2024): A consistent baseline of structural evolution and regional disparities
Core Problem: Current agricultural monitoring relies on fragmented, inconsistent baselines that fail to capture the continuous, annual nature of agricultural management and distinguish active cultivation from permanent structural changes, hindering accurate assessment of food security and planetary boundaries.
Key Innovation: Generation of the first Global 30-m Annual Cropland Extent Dynamics (GACED30) dataset (2000–2024) using a continuous mapping framework, spectral-semantic sample alignment, and rule-based processing, providing a reliable baseline for monitoring global agriculture's structural evolution and regional disparities.
63. GlobalGeoTree: a multi-granular vision-language dataset for global tree species classification
Core Problem: The scarcity of large-scale, labeled datasets has constrained progress in global tree species mapping using remote sensing data for biodiversity monitoring, forest management, and ecological research.
Key Innovation: Introduction of GlobalGeoTree, a comprehensive multi-granular vision-language dataset with 6.3 million geolocated tree occurrences paired with Sentinel-2 image time series and auxiliary environmental variables, along with a baseline model (GeoTreeCLIP) demonstrating improved zero- and few-shot classification.
64. Compact deep neural network models of the visual cortex
Core Problem: The need for predictive yet parsimonious deep neural network (DNN) models of the visual cortex, as existing models are often large and computationally opaque.
Key Innovation: Developed a method to compress large, highly predictive DNN models (60 million parameters) into compact models (5,000 times fewer parameters) with comparable accuracy for predicting visual neuron responses, revealing a general computational principle of the visual cortex.
65. Vectorized instructive signals in cortical dendrites
Core Problem: Testing the theoretical hypothesis that neural circuits implement single-phase vectorized learning at the cellular level by processing feedforward and feedback information in separate dendritic compartments.
Key Innovation: Demonstrated the existence of vectorized instructive signals in cortical dendrites of mice during a neurofeedback task, showing neuron-specific teaching signals that predict learning changes and can be disrupted by optogenetic perturbation, unveiling a potential mechanism for solving credit assignment in the brain.
66. Enable active mobility of untethered miniature robots in high-resistance multiphase medium
Core Problem: Untethered miniature robots exhibit limited mobility in high-resistance multiphase media (e.g., solid-liquid mixtures, granular media) due to insufficient force generation or strategies to reduce medium resistance.
Key Innovation: Development of RoboIMP, an impact-driven magnetic untethered miniature robot capable of high-force outputs and using reciprocal oscillation to fluidize the surrounding medium, enabling effective locomotion in challenging environments.
67. Enabling Fieldwork for All (EFFA) Framework: Supporting physical, social, financial, and psychological safety in the field
Core Problem: Research and recommendations on fieldwork safety are largely siloed by scientific discipline, limiting the comprehensive spread of data and discussion needed to address physical, social, financial, and psychological safety effectively.
Key Innovation: Synthesized cross-disciplinary literature on fieldwork safety, highlighting four interconnected facets (physical, social, financial, psychological), and proposed the Enabling Fieldwork for All (EFFA) framework to provide data-driven resources for principal investigators and foster cross-disciplinary sharing for future research on field safety.
68. Predicting initial accident states in hazardous chemical road transportation: A causal and interpretable machine learning approach
Core Problem: Predicting the initial accident state and identifying key influencing factors in hazardous chemical road transportation is difficult, especially when accident data are limited.
Key Innovation: Develops an accident state prediction framework combining causal inference and machine learning, proposing a Random Forest model with Inverse Probability Weighting (RF-IPW) to account for causal effects and using SHAP for model interpretation and factor contribution assessment.
69. An integrated approach for allocating engineered and procedural safety measures to mitigate fire-induced domino effects based on two-stage optimization
Core Problem: Optimally allocating engineered and procedural safety measures to mitigate fire-induced domino effects in storage tank farms, to enhance risk management and save investment, as existing methods often target only a single type of safety measure.
Key Innovation: Proposes a two-stage optimization model that uses graph theory and event sequence diagrams for the collaborative allocation of engineered and procedural safety measures, demonstrating enhanced effectiveness and cost reduction compared to single-measure approaches.
70. A decade of sea ice concentration retrieved from sentinel-1
Core Problem: Traditional passive microwave sensors for sea ice concentration (SIC) monitoring provide coarse resolution, limiting detailed observation of Arctic sea ice dynamics.
Key Innovation: This study introduces DMI-ASIP, the first decade-long (2014–2024) pan-Arctic SIC record derived from high-resolution (0.5 km) Sentinel-1 SAR imagery, demonstrating superior performance over PMW-based records, particularly in the marginal ice zone.
71. A parameterization-based framework for quantifying the cooling effects of urban blue–green spaces: Comparative evaluation with buffer and Gaussian models
Core Problem: Prevailing methods for evaluating the cooling effects of urban blue–green spaces (BGS) suffer from limited compatibility, generalizability, and interpretability.
Key Innovation: This study proposes the Land Surface Temperature Parameterization Method (LSTPM) to physically characterize BGS cooling effects using local temperature gradients, deriving four indicators (CI, CD, CE, CG), and demonstrating its superior discrimination and precision compared to Buffer and Gaussian models.
72. Mapping snow algae in West Antarctic Peninsula from Sentinel-2 images using a machine learning based approach
Core Problem: The distribution of snow algae in the West Antarctic Peninsula (WAP) remains poorly understood due to their high spatiotemporal variability and complex polar imaging conditions.
Key Innovation: This research develops a robust, large-scale machine learning framework for mapping green and red snow algae from Sentinel-2 images, outperforming traditional methods, and providing a 2019–2024 composite distribution and interannual variability assessment for WAP.
73. Ensemble Machine Learning and Landsat Observations Reveal Seasonal and Spatial Dynamics of Water Quality in a River-Influenced Estuarine System
Core Problem: Conventional empirical algorithms struggle to capture nonlinear bio-optical relationships, limiting long-term and spatially extensive monitoring of coastal water quality in optically complex estuarine environments.
Key Innovation: This study integrates high-resolution field observations with Landsat data to develop ensemble machine learning models (XGB, RF) for estimating chlorophyll-a, CDOM, and turbidity in the western Mississippi Sound, generating a decade-long time series and revealing seasonal and spatial dynamics.
74. Transition of soil microbial abundances and functions from passive response to active adaptation over 15 repeated freeze-thaw cycles
Core Problem: Understanding how soil microbes progressively adapt to repeated freeze-thaw cycles and their impact on soil biogeochemical processes, which can affect soil stability.
Key Innovation: Reveals that soil microbial biomass and functional gene abundances shift from passive response to active adaptation over repeated freeze-thaw cycles, impacting carbon and nitrogen cycling and contributing to soil microbial resilience under changing climate conditions.
75. 13C-traced microbial CO2 fixation in grassland topsoil: More bacterial necromass carbon enrichment than shrub and forest soils
Core Problem: Clarifying the regulatory mechanism of vegetation types on microbial CO2 fixation and its allocation to microbial biomass and necromass, and ultimately soil organic carbon (SOC) accumulation across different vegetation regions, particularly in geohazard-prone areas like the Loess Plateau.
Key Innovation: Shows that grassland soils on the Loess Plateau efficiently accumulate fixed CO2 into SOC via bacterial necromass through microbial CO2 fixation, with vegetation type regulating soil properties and microbial communities to modulate carbon-fixation pathways and SOC accumulation.
76. Remote sensing of photobleaching in fluorescent dissolved organic matter: a coupled non-Markovian kinetics and polarized radiative transfer framework
Core Problem: Insufficient representation of nonlinear decay and cumulative light-dose effects of fDOM photobleaching in remote-sensing models, hindering accurate understanding of energy distribution and spectral responses in natural waters.
Key Innovation: Establishes a physically consistent framework coupling non-Markovian kinetics, spectral energy closure, and polarized radiative transfer, improving the description of fDOM photobleaching and remote-sensing retrievals, demonstrated by field observations and reliable forward simulations.
77. A high-resolution dataset revealing the dynamical variations of the relative humidity in China
Core Problem: Lack of high spatial and temporal resolution atmospheric relative humidity data, which is crucial for studying climate change impacts on agricultural production and understanding climate fluctuations.
Key Innovation: Developed a 1 km x 1 km daily relative humidity raster dataset for China (1951–2020) using a Random Forest spatial interpolation framework, and analyzed its multiscale spatiotemporal variation characteristics across agricultural zones.
78. KL divergence-guided transfer learning for data-driven shield tunneling under distribution shift
Core Problem: Significant distribution shifts between source and target domains often compromise transfer learning performance in data-driven models for engineering applications like shield tunneling clogging prediction, and it remains unclear under what conditions transfer learning offers tangible improvements.
Key Innovation: Introduced multi-dimensional Kullback-Leibler (KL) divergence to quantify domain discrepancy and explain model transferability, developed an adaptive fine-tuning approach incorporating KL divergence as a regularization term, and proposed a transferability criterion, demonstrating effectiveness for clogging prediction in tunnel engineering.