TerraMosaic Daily Digest: July 7, 2026
Daily Summary
The July 7 literature is unusually concentrated in landslide science. Several papers move beyond mapping or single-factor susceptibility and instead model full process chains: a rock-ice avalanche in the Yanzigou Basin is reconstructed as a transition from frozen-slope failure to debris flow and flooding; Pleistocene landslide dams in the Jinsha River are tied to valley evolution and outburst-flood deposits; and cascading flood-landslide scenarios are used to reframe infrastructure hazard assessment. The landslide modeling papers also push toward operational prediction: slope-unit runout modeling, rainfall-threshold forecasting in data-scarce mountains, physics-anchored numerical forecasting, Bayesian geostatistical fusion of machine learning and slope stability, and differentiable or particle-based large-deformation simulation all target the same bottleneck: how to make landslide forecasts physically interpretable without losing regional scalability.
A second theme is failure mechanics under coupled disturbance. Prior vibration followed by seepage changes slip-zone response, desiccation cracks reshape drought-rainfall slope failure, progressive melting weakens soil-ice landslide materials, reservoir fluctuation drives softening-hardening transitions in slip zones, and wet-dry or freeze-thaw cycles degrade expansive and colorful clays. These studies provide useful mechanics for cold-region, reservoir-bank, and climate-stressed slopes, where hazard evolves through repeated hydrologic or thermal loading rather than a single trigger.
Remote sensing and sensing-driven AI are also more directly tied to geohazards than on many days. Object-based residual graph attention is used for Himalayan landslide detection, XGBoost and multi-source geospatial data support susceptibility mapping, InSAR enters posterior probability updating and subsidence monitoring, fiber-optic sensing and satellite radar are combined for ground-subsidence warning, and seismic/acoustic signals are used for debris-flow identification, shear stage classification, natural debris-flow velocity profiling, and bedding-rock slope damage quantification. The broader hazard set includes GLOF monitoring gaps, surge-induced GLOF reconstruction, flood regime shifts, dryland flooding, fault-crossing tunnels, liquefaction, seismic site classification, and infrastructure reliability under landslide-prone loading.
Key Trends
The issue is driven by five linked directions: landslide process-chain modeling, coupled disturbance mechanics, uncertainty-aware AI, sensing-constrained warning, and multi-hazard terrain systems.
- Landslide research is shifting from static maps to process-chain forecasting: Runout, rainfall thresholds, physics-anchored forecasting, cascading flood-landslide scenarios, and landslide-dam reconstructions all model how failures propagate through terrain, infrastructure, and hydrologic systems.
- Coupled disturbance history is becoming central to slope mechanics: Seepage after vibration, drought-rainfall cracking, melting soil-ice mixtures, reservoir fluctuation, freeze-thaw, and wet-dry cycling show that prior loading controls present-day failure response.
- Uncertainty-aware AI is entering landslide prediction at multiple scales: Probabilistic deep learning, Bayesian geostatistics, GA-optimized superlearners, posterior updating with deformation, and ensemble data assimilation all move landslide models toward quantified uncertainty rather than only classification accuracy.
- Sensing is becoming a physical constraint rather than just a data source: InSAR, GNSS, distributed fiber optics, optical fiber sensing, seismic signals, acoustic emission, conductivity probes, and UAV/remote-sensing observations are being used to infer slip surfaces, damage stages, deformation, and flow structure.
- The surrounding hazard context broadens from landslides to multi-hazard terrain systems: GLOFs, dryland floods, flood-frequency nonstationarity, fault-crossing tunnels, liquefiable slopes, subsidence, tsunami-like retaining-wall loading, and coastal cliffs extend the issue from slope failure to connected geohazard systems.
Selected Papers
The selected papers cover rock-ice avalanche cascades, seepage-triggered landslides, slope-unit runout modeling, co-seismic landslide uncertainty, drought-rainfall cracked slopes, soil-ice landslide materials, soil-rock mixture SPH, Himalayan landslide detection, landslide-dammed river history, seismic bedding-rock slopes, rainfall-threshold warning, Bayesian slope-stability fusion, landslide numerical forecasting, debris-flow detection and mechanics, InSAR-updated susceptibility, reservoir-bank landslides, GLOFs, ground subsidence, liquefaction, fault-crossing tunnels, and multi-hazard flood-landslide infrastructure assessment. This issue contains 87 selected papers from 2531 papers analyzed.
1. Multi-frequency seismic evidence for phase transitions in a rock-ice avalanche-triggered cascading Hazard in the Yanzigou Basin, Southeastern Tibetan Plateau
Core Problem: With ongoing global climate warming, rock-ice avalanches in cold, high-mountain regions are becoming increasingly frequent, posing severe threats to the cryospheric environment and human society.
Key Innovation: This study examines a large rock-ice avalanche that occurred on 25 July 2024 at Yanzigou Glacier on Mount Gongga and reconstructs the cascade hazard evolution from rock-ice avalanche to debris flow and subsequent flooding based on multi-source observations.
2. Effects of prior vibration and seepage on sliding responses in slip zones: implications for seepage-triggered landslides after seismic and hydrological disturbances
Core Problem: Seepage-triggered landslides are widespread in mountainous regions, where slopes commonly experience seismic and hydrological disturbances prior to failure.
Key Innovation: In this study, a newly developed dynamic shear-seepage coupling (DSSC) test system was used to investigate the sliding responses of natural slip zone materials under disturbances with different vibration intensities (amplitude, duration, frequency) and seepage pressures.
3. Heteroscedastic probabilistic deep learning reveals environmental controls on co-seismic landslide area density variability
Core Problem: Earthquake-induced landslides are a major secondary hazard in mountainous regions, and their size plays a key role in determining overall impact.
Key Innovation: In this study, we propose a probabilistic framework to model landslide area density using a Feature Tokenizer Transformer with a heteroscedastic lognormal head.
4. LARMSU: A semi-empirical landslide runout model based on slope units
Core Problem: Landslide runout models are essential tools for delineating potential inundation areas, assessing risks, and designing mitigation structures.
Key Innovation: In this study, we propose a semi-empirical LAndslide Runout Model based on Slope Units (LARMSU) to simulate runout and assess susceptibility in potentially unstable regions.
5. Roles of desiccation crack depth and surface intensity in affecting clay slope failures during extreme drought-rainfall events
Core Problem: Desiccation cracks significantly alter slope stability during extreme rainfall, yet the relative influence of crack depth versus crack intensity factor (CIF) across soil types remains poorly understood.
Key Innovation: In this study, four centrifuge tests were conducted on slopes subjected to various antecedent temperature loading followed by an extreme rainfall with 100-year return period.
6. A global dataset of fluid-based pre-seismic anomalies in crustal environments
Core Problem: Fluid-related pre-seismic anomalies are widely reported, but they remain difficult to compare because studies use inconsistent parameters, baselines, time lags, and site metadata.
Key Innovation: The paper compiles a global, structured dataset of more than 900 fluid-related anomalies associated with over 500 earthquakes, standardizing geochemical, hydrological, and physical descriptors for reproducible seismic-hazard analysis.
7. An enhanced SPH framework to simulate dynamic motion of soil-rock mixtures considering discontinuity contact
Core Problem: Conventional SPH simulations of soil-rock mixture slope motion often underrepresent soil-rock contact, inter-block interaction, and discontinuity behavior.
Key Innovation: The paper adds explicit soil-rock, rock-rock, and soil-block contact treatments to an SPH framework and validates the approach on collapse and SRM-landslide simulations.
8. Strength degradation of soil-ice landslide materials under progressive melting
Core Problem: Distinct from conventional earth-rock dams, ice within soil-ice landslide dams undergoes phase transitions due to temperature fluctuations.
Key Innovation: Therefore, this study investigates the mechanical behavior and structural degradation of these uncemented mixtures across varying initial ice contents and progressive melting stages.
9. Scalable landslide detection in complex Himalayan topography via a novel object-based residual graph attention network
Core Problem: High-mountain environments are some of the most fragile and difficult environments to manage.
Key Innovation: Here, we propose an open-source object-based Landslide Residual Graph Attention Network (LANDS-ResGAT) for landslide segmentation in rapidly transforming mountains.
10. Catastrophic processes in bedding rock slopes under multiphase seismic actions: Insights from shaking table tests and optical fiber sensing
Core Problem: In seismically active regions, landslides often result from the cumulative damage caused by multiple seismic events.
Key Innovation: In this study, shaking table tests and optical fiber sensing technologies, including optical frequency domain reflectometry (OFDR) and distributed acoustic sensing (DAS), are employed to investigate the dynamic response and evolution of bedding rock slopes under multiphase seismic activity.
11. Late Pleistocene landslide-dammed events in the Panzhihua-Wudongde reach of the Jinsha River in southeastern Tibetan: Implications for the origin of the Longjie Formation
Core Problem: The origin of the Longjie Formation has remained debated in a reach of the Jinsha River shaped by landslide damming, lake formation, and outburst-flood processes.
Key Innovation: New sedimentological and luminescence evidence identifies two Late Pleistocene landslide-dammed lake events and links their deposits to valley evolution and downstream sediment transfer.
12. Developing a Novel GIS-Based Index for Physical Coastal Vulnerability
Core Problem: Coastal zones globally face escalating risks from climate-induced hazards, such as shoreline erosion, saltwater intrusion, and flooding resulting from rising sea levels.
Key Innovation: The results indicate that coastal vulnerability is not uniform across the UK, with England identified as the most vulnerable hotspot.
13. Dual Dynamics of Erosion and Clogging in Granite Residual Soil: Pore-Scale Evolution and Seepage Response by in-situ CT Scanning
Core Problem: Background: Granite residual soil (GRS) is highly susceptible to rainfall-induced landslides, and its seepage behavior and slope stability are fundamentally governed by the dynamic evolution of pore structure under hydraulic erosion.
Key Innovation: Results: The results show that hydraulic erosion triggers a systematic coarsening and integration of the pore network, with the average equivalent pore diameter increasing by 6.81-9.93%, the connected porosity increasing by 4.94-8.23%, and the flow path tortuosity decreasing by 21.92-24.26%.
14. Integrating machine learning and physics-based slope stability through Bayesian geostatistics: An uncertainty-aware framework for landslide prediction
Core Problem: Effective landslide susceptibility assessment requires a modelling framework that integrates heterogeneous geospatial information while explicitly accounting for predictive uncertainty.
Key Innovation: This study developed a Bayesian geostatistical approach that combined susceptibility prediction from a data-driven random forest (RF) classifier with factor of safety (Fs) output from a physics-based model − transient rainfall infiltration and grid-based regional slope-stability (TRIGRS) − using integrated nested Laplace approximation (INLA).
15. Landslide hazard prediction in data-scarce mountainous region using satellite-based frequentist rainfall thresholds
Core Problem: Landslides triggered by intense rainfall pose persistent threats to infrastructure and communities, particularly in data-scarce mountainous regions.
Key Innovation: This study presents an integrated spatiotemporal framework for landslide hazard prediction in the Muzaffarabad region, NW Himalayas, Pakistan.
16. Landslide numerical forecasting mode: A physics-anchored and data-refined paradigm shift
Core Problem: Accurate and real-time landslide numerical forecasting is critical for geohazard mitigation but remains challenged by the lag in kinematic methods and the parameter uncertainty in physical models.
Key Innovation: To overcome this bottleneck, this study pioneers the physics-anchored and data-refined (PADR) landslide numerical forecasting mode.
17. Softening-hardening transition in landslide slip zones under cyclic reservoir fluctuations: Insights into the deformation evolution process
Core Problem: Reservoir landslides exhibit pronounced sensitivity to cyclic water-level fluctuations, yet current models often fail to capture the effect of cyclic loading induced by long-term reservoir water fluctuations, remaining constrained by static assumptions.
Key Innovation: The results indicate that the strain evolution mechanism under cyclic loading is fundamentally different from that under static creep.
18. The macro-meso-micro mechanical properties and numerical simulation analysis of expansive soil under wet-dry and freeze-thaw cycles
Core Problem: In cold and arid regions, expansive soil slopes in open channel projects are highly susceptible to landslides and frost heaving damage under the influence of wet-dry and freeze-thaw (WDFT) cycles.
Key Innovation: The results indicate the following: (1) The numerical model, based on the “soil particle expansion and shrinkage-water particle expansion method” effectively captures the “water expansion and shrinkage” behavior of expansive soil.
19. A deep learning framework for stage classification and short-term prediction of deep shear evolution of landslides based on acoustic emission signals
Core Problem: The shear-zone-controlled deep shear evolution in landslides exhibits pronounced stage dependence.
Key Innovation: On this basis, this study conducted a five-stage physical simulation under laboratory conditions and proposed AE-StageNet, a dual-task deep learning framework for stage classification and short-term prediction.
20. A deformation-informed posterior probability updating framework for landslide susceptibility assessment: application to the Motuo area, China
Core Problem: Landslide susceptibility assessment is essential for hazard prevention and risk management in mountainous regions.
Key Innovation: In this study, surface deformation refers to measurable ground motion detected by InSAR, represented by line-of-sight (LOS) cumulative displacement and deformation velocity, and treated as dynamic evidence of slope activity.
21. An optimized random forest model for debris flow identification based on seismic signals
Core Problem: Accurate identification of debris-flow events is essential for reliable seismic-based monitoring and early-warning systems in mountainous regions.
Key Innovation: This study proposes an optimized Random Forest (RF) model for debris-flow signal identification, with an emphasis on robustness, interpretability, and applicability across diverse geomorphological settings.
22. Deformation and failure mechanisms of multi-sliding zone landslide under water level fluctuation: Insights from laboratory experiments
Core Problem: The presence of multi-sliding zones has been recognized in many large-scale landslides in the Three Gorges Reservoir area (TGRA), as reflected by their distinctive deformation behaviors.
Key Innovation: In this study, to investigate the deformation and failure mechanisms of the multi-sliding zone landslide under reservoir water level fluctuation, a simplified physical model is developed from the Jiuxianping landslide with two sliding zones.
23. Cascading flood-landslide scenarios reframe the hazard assessment of exposed infrastructure
Core Problem: Conventional multi-hazard assessments in catchment-infrastructure configurations based on spatial overlay do not always capture physically interacting hazard chains, because they lack systematic protocols for identifying possible cascading activation.
Key Innovation: Results show that landslide-induced channel obstruction generates backwater effects that reduce considerably the flood return period required to inundate the most vulnerable WWTP components.
24. Channel position-dependent erosivity of steep debris flows combining gravitational and erosive processes: Towards predictive modeling
Core Problem: Debris flow erosivity is the most sensitive parameter for run-out length and impact, at the same time it is a poorly understood process, thus inhibiting predictive modeling for mitigation measures.
Key Innovation: In this paper, we have evaluated the erosivity of debris flows in Bavaria from 5 deeply eroding channels with boulder-, gravel-, or regolith-dominated beds.
25. Modeling large-deformation dynamic behavior of saturated soils using the Hybrid Element Particle Method (HEPM)
Core Problem: Modeling the dynamic behavior of saturated porous media is crucial for predicting geotechnical hazards.
Key Innovation: To address these issues, this study extends the Hybrid Element Particle Method (HEPM) to dynamic coupled problems in saturated porous media.
26. Reliability analysis of soil-rock mixture slopes considering soil particle gradation
Core Problem: Soil-rock mixture (SRM) slopes are widely encountered in colluvial, residual, and landslide deposits, and their reliability can be influenced by both rock and soil properties.
Key Innovation: This paper thus proposes a new reliability analysis method to consider the randomness in both rock and soil distributions, with particular focus on soil particle gradation (SPG).
27. How well can we resolve? Derivation of vertical velocity profiles from paired conductivity measurements in natural debris flows
Core Problem: Debris flows are characterized as highly concentrated sediment-water mixtures transiting steep headwater channels.
Key Innovation: This study investigates the derivation of velocity estimates based on cross-correlation of paired conductivity sensors.
28. Mechanical degradation mechanisms of colorful clay subjected to dry-wet cycling
Core Problem: Yunnan Jianshui colorful clay is a silty clay with certain swelling-shrinkage behavior, and it undergoes significant mechanical property degradation under frequent dry-wet cycles, which may induce geological disasters such as landslides.
Key Innovation: To reveal its degradation mechanism, this study conducted 0 to 10 dry-wet cycles and employed direct shear tests, unconfined compressive strength tests, CT scanning, mercury injection test (MIP), and scanning electron microscopy with energy-dispersive spectroscopy (SEM/EDS) for comprehensive investigation.
29. Physics-informed dynamic reliability evaluation of landslide-prone unequal-wall-thickness girth welds based on far-field magnetic signals
Core Problem: Oil and gas transmission pipelines in landslide-prone areas are exposed to time-varying internal pressure, lateral soil loading, and geometric/material discontinuities at unequal-wall-thickness girth welds, making dynamic reliability assessment difficult.
Key Innovation: This study proposes a physics-informed dynamic reliability framework for landslide-prone girth welds.
30. A novel unified distance potential-based contact framework for NMM-MPM simulation of landslide-structure interaction
Core Problem: A novel numerical manifold method and material point method (NMM-MPM) coupling framework is proposed based on a unified distance-potential formulation for contact modeling, enabling a unified treatment of NMM-NMM and NMM-MPM contacts.
Key Innovation: To account for the finite size of material points, triangular manifold elements are geometrically offset to construct virtual rounded manifold elements, transforming the contact problem into a point to boundary interaction.
31. An automated GA-optimized superlearner framework for landslide susceptibility mapping in complex mountainous regions
Core Problem: Landslide susceptibility mapping in complex mountains is sensitive to inventory noise, model choice, and hyperparameter tuning.
Key Innovation: The study builds a genetic-algorithm-optimized automated SuperLearner and couples it with spatial regression to improve prediction, interpretability, and disaster-zoning transferability.
32. Landslide Susceptibility Mapping Using Multi-Source Geospatial Data and XGBoost
Core Problem: Regional landslide susceptibility mapping depends on heterogeneous terrain, hydrologic, geologic, and land-surface predictors that must be integrated without losing spatial interpretability.
Key Innovation: The study applies XGBoost to multi-source geospatial predictors for landslide susceptibility mapping, providing a tree-ensemble benchmark for data-driven zoning and terrain-risk screening.
33. A new probabilistic rockfall trajectory model incorporating fragmentation, rock mass quality and block shape variability
Core Problem: Rockfall fragmentation significantly influences the dynamics and hazard footprint of rockfall events, yet it remains poorly represented in most trajectory models.
Key Innovation: This study presents RockFRAG, a novel probabilistic lumped-mass trajectory model that explicitly incorporates fragmentation processes, block shape variability and rock-mass condition within an experimentally calibrated framework.
34. Centrifuge modeling of seismic mechanisms and liquefaction-induced instability of gently submerged slopes
Core Problem: Coastal infrastructure located on gently submerged slopes is highly vulnerable to earthquake-induced ground deformation.
Key Innovation: This study investigates the seismic response of loose sandy slopes ( D r = 30%) with inclinations of 10° and 20° using geotechnical centrifuge modeling under fully saturated conditions and controlled one-way and two-way shaking.
35. Assessment of Bearing Response and Failure Mode Transition in Unsaturated Soil Slope-Footing Systems under Varying Hydraulic Regimes and Geometric Conditions
Core Problem: Rapid infrastructure development in hilly and mountainous regions has intensified the need for constructing geostructures on or near natural slopes, where variable soil profiles, fluctuating groundwater, and pronounced climatic variations present significant geotechnical challenges.
Key Innovation: This study investigates the response of an unsaturated soil slope-footing system under varying hydraulic conditions, including no-flow, steady-state (infiltration and evaporation), and transient fluxes governed by climatic inputs such as precipitation, temperature, and relative humidity.
36. Effects of ground motion types and rubber-sand-concrete damping layer on the seismic fragility of tunnel linings
Core Problem: Tunnels are critical infrastructure components that are vulnerable to seismic hazards, particularly near-fault ground motions.
Key Innovation: This study uses the conventional lining under near-fault pulse-like motions as a reference case and evaluates the RSC constrained damping lining under near-fault pulse-like, near-fault non-pulse-like, and far-field motions.
37. Built to last: Long-term stability of steep deglaciated slopes of alpine karst plateaus (Northern Calcareous Alps, Austria)
Core Problem: Steep limestone plateaus of the Northern Calcareous Alps rise abruptly above deeply incised valleys, yet despite their sharp relief, they rarely experience large-scale slope failures.
Key Innovation: To understand why these massifs remain remarkably stable, we investigated five alpine karst plateaus: Göll, Untersberg, Hagengebirge, Tennengebirge, and Totes Gebirge, using a combination of speleothem geochronology, structural mapping, limit equilibrium analysis and elastic stress modelling.
38. Rapid seismic screening of non-engineered URM buildings in liquefaction-prone areas
Core Problem: Rapid visual screening is essential for prioritising interventions in densely populated, data-scarce areas.
Key Innovation: This study proposes an engineering-driven framework for rapid susceptibility screening that combines k-means clustering with parameter weighting derived from the analytic hierarchy process (AHP).
39. Three-dimensional dynamic interaction between shallow-founded building and shield tunnel in liquefiable sand deposits
Core Problem: A large number of buildings are constructed near the subway shield tunnel with the rapid development of urbanization in China.
Key Innovation: This study established a refined three-dimensional numerical model to elucidate the dynamic response mechanism of shallow-founded building and shield tunnel in liquefiable sand.
40. Automatic P-wave Arrival Picking for Microseismic Signals of Mines Using a Hybrid Neural Network with Constrained Attention and CNN-BiLSTM
Core Problem: Microseismic (MS) monitoring is a critical technology for ground pressure control, water inrush warning, and slope stability management in mining, as outlined by national standards and industry regulations.
Key Innovation: This study introduces a hybrid model, the Dual-channel Seismology-prior Constrained Attention Convolutional Neural Network-Bidirectional Long Short-Term Memory (Dual-ch Sp-CA-CNN-BiLSTM), designed for automatic P-wave arrival time picking.
41. Towards early warning of catastrophic ground subsidence through integrated fibre-optic sensing and satellite radar imaging
Core Problem: Catastrophic ground collapses, such as sinkholes, pose severe risks to infrastructure and public safety.
Key Innovation: This paper consolidates experimental and numerical investigations of DFOS alongside a separate case study demonstrating satellite imaging for early detection of ground subsidence.
42. An integrated ERT-MASW probabilistic framework for liquefaction susceptibility assessment: A case study in Pohang
Core Problem: Liquefaction assessment based on sparse boreholes or a single geophysical method can miss local heterogeneity and produce discontinuous risk maps.
Key Innovation: The framework fuses ERT, borehole lithology, MASW-derived shear-wave velocity, and liquefaction potential indices to produce probabilistic liquefaction maps for the Pohang earthquake site.
43. Experimental investigation of seismic response of waste dump slopes under pre- and post-rainfall conditions
Core Problem: Waste dump slopes are a common type of artificial slope, and the potential geological hazards that may be triggered by waste dumps pose a non-negligible risk to regional road traffic safety.
Key Innovation: A shaking table physical model test was conducted to investigate the seismic response characteristics of waste dump slopes under pre- and post-rainfall conditions in this study.
44. Systemic Seismic Risk Analysis of Process Plants Considering Physical-chemical and Domino Consequences
Core Problem: Seismic risk assessments for process plants often omit cascading physical-chemical failures, production disruption, deterioration, and downtime.
Key Innovation: The proposed probabilistic framework couples seismic hazard, component damage, domino consequences, process-network reliability, repair cost, production loss, and time-dependent deterioration.
45. Virtual testbed for multi-risk assessment: application to the Inland RETURNVILLE
Core Problem: Urban areas are characterized by high level of exposure in terms of population, assets, and highly interconnected systems, which can lead to multiple cascading effects under single and multi-hazards scenarios.
Key Innovation: This study presents an application for the Inland RETURNVILLE, the VTB representing an inland urban settlement exposed to earthquake and volcanic hazards, with the aim of illustrating multi-risk interactions modelling both at the building level - where ashfall increases seismic vulnerability - and the system level - where post-earthquake building collapses.
46. First-order estimates of slip rates and earthquake recurrence intervals of offshore seismogenic structures from the Taiwan Earthquake Model (TEM) project
Core Problem: To accurately assess future seismic hazard, it is necessary to obtain reliable estimates of slip rates and earthquake recurrence intervals for seismogenic structures.
Key Innovation: The results show that faults in the Ryukyu subduction upper plate and eastern offshore Taiwan are the most active, with slip rates >15 mm/yr and earthquake recurrence intervals 6 kyr.
47. Full-3D Electrical Resistivity Tomography (ERT) for mining activity planning and sinkhole assessment: A case study in the Murisengo gypsum quarry
Core Problem: Mining excavation design requires preliminary investigation surveys to identify the location of the exploitable resource.
Key Innovation: Also, monitoring activities should be planned since the degradation of the mechanical properties of rock volumes may trigger instability.
48. Impact of topography and near-surface layers on ground motion: a finite difference study of the 2011 Tohoku earthquake
Core Problem: The 2011 Tohoku earthquake (M w 9.0) took place in the Northeast Japan (Tohoku) forearc area in a megathrust zone and struck off the Pacific coast of Tohoku.
Key Innovation: This study investigates the influence of topography and near-surface velocity layers on seismic ground motion, using the 2011 Tohoku earthquake (Mw 9.0) as a case study.
49. Integrating multi-source remote sensing and numerical simulation approaches for enhanced flood hazard assessment
Core Problem: Extreme flooding is expected to intensify under a warming climate, which emphasizes the need for rapid and accurate flood mapping to support effective risk assessment.
Key Innovation: In this study, we integrate multi-source RS data, including optical and synthetic aperture radar (SAR) imagery, with NS to improve the characterization of extreme floods.
50. Mapping subsidence, uplift, and seasonal deformation from InSAR: Responses to groundwater decline and recharge in China's Qingxu-Jiaocheng region
Core Problem: Land subsidence caused by long-term groundwater overextraction has been a persistent hazard in the Qingxu-Jiaocheng area of the Taiyuan Basin, China, while recent management policies have led to substantial aquifer recovery and emerging uplift.
Key Innovation: To quantify this transition, we integrated L-band ALOS-1 (2007-2010) and C-band Sentinel-1 (2015-2023) data using the SBAS-InSAR technique to construct the spatiotemporal evolution of surface deformation.
51. Nonparametric fragility analysis of asphalt concrete core embankment dams based on bidirectional ground-motion composite intensity measures and explainable machine learning
Core Problem: The dynamic analysis of asphalt concrete core embankment dams (ACCEDs) is highly nonlinear, and their postearthquake responses are affected by both horizontal and vertical seismic motions.
Key Innovation: To address these challenges, a data-driven nonparametric fragility analysis framework is developed in this study.
52. Stochastic Assessment of Static Liquefaction and Slope Stability Using NORSAND and Random Fields
Core Problem: Static liquefaction and slope stability analyses require uncertainty-aware constitutive modeling of contractive sandy soils.
Key Innovation: The study embeds NORSAND in a random finite element framework to quantify stochastic pore-pressure buildup, post-peak softening, and bearing-capacity variability in loose sandy slopes.
53. An additive framework for developing a hybrid VS30 model: Incorporating geological information into the existing SCK Model for an updated VS30 map of Chinese mainland
Core Problem: V S30, the time-averaged shear-wave velocity to 30 m depth, is a key parameter in seismic site classification and hazard analysis.
Key Innovation: This study presents an additive modeling framework using a Bayesian updating approach for V S30 estimation and mapping.
54. Enhancing rainfall-induced slope stability prediction via the ensemble Kalman filter data assimilation
Core Problem: In the transient analysis of rainfall-induced slope response, data assimilation using the ensemble Kalman filter (EnKF) provides an effective framework for integrating pore-water pressure monitoring data into numerical models.
Key Innovation: This study investigates how spatial variability in porosity and monitoring schemes jointly affect parameter uncertainty reduction and slope stability prediction.
55. Physical model tests on deformation and failure mechanisms of expansive soil slopes with inherent fissures under groundwater fluctuations
Core Problem: Deeply excavated expansive soil slopes in large water diversion projects are subjected to repeated wetting-drying cycles, rainfall infiltration, groundwater fluctuations, and inherent fissures, yet their coupled effects remain insufficiently understood from physical model observations.
Key Innovation: This study developed a large-scale physical model to investigate the deformation and failure behavior of expansive soil slopes under controlled groundwater fluctuations, cyclic wetting-drying, prolonged rainfall, and fissure analogues.
56. Seismic vulnerability analysis of deeply buried water diversion tunnels
Core Problem: Unlike shallow-buried tunnel projects such as subway tunnels, deep-buried water diversion tunnels are characterized by complex stress conditions, significant disturbances during construction, complex mechanisms of action of various components, and severe consequences of failure.
Key Innovation: This paper takes a deeply buried water diversion tunnel section of a water system interconnection project as the research object.
57. Slope stability analysis of Mount St. Helens slope using Scoops3D and machine learning
Core Problem: The current study investigates a hybrid intelligent modeling framework to examine the three-dimensional (3D) stability of the Mount St. Helens (MSH) slope when seismic loading and groundwater effects are combined.
Key Innovation: Stability analyses were performed for five seismic coefficients k e = 0.0 - 0.20 and eleven pore-pressure ratio scenarios r u = 0.00 t o 0.50 , generating a total dataset of 2,750 samples (50 material property samples × 5 seismic sets × 11 pore-pressure cases).
58. Soil carbon and nitrogen dynamics during the thermokarst lake succession under climate change on the Tibetan Plateau
Core Problem: Thermokarst lakes (TLs) induced by permafrost thaw can disturb soil carbon and nitrogen cycling, creating positive feedbacks to climate warming.
Key Innovation: For the first time, we investigated the expanding, stable, drying, and drained TLs on the TP, each with distinct microfeatures.
59. Study on post-earthquake driving performance evaluation and driving recovery strategy of high-speed railway bridges under different damping ratios
Core Problem: As a core design parameter of high-speed railway (HSR) track-bridge system (TBS), the structural damping ratio (SDR) directly affects the system's energy dissipation efficiency and structural damage under earthquakes, thereby restricting train operational safety and comfort.
Key Innovation: To accurately quantify its influence mechanism and support efficient post-earthquake operational decision-making for bridges, this study focuses on the CRTS III TBS, investigating post-earthquake driving performance and evaluating traffic capacity.
60. A particle-scale investigation of reliquefaction behavior in saturated sand considering stress history effects
Core Problem: Saturated sand subjected to multiple earthquake events may undergo reliquefaction, which differs fundamentally from initial liquefaction and is strongly governed by stress history.
Key Innovation: In this study, a Computational Fluid Dynamics-Discrete Element Method (CFD-DEM) coupled numerical framework incorporating centrifuge dynamic similarity principles is developed to continuously simulate the full sequence of liquefaction, reconsolidation, and reliquefaction.
61. Evaluating gas-driven fracture and ejecta dynamics in permafrost blowout craters via model-based generalized interpolation material point framework
Core Problem: Gas blowout craters in permafrost regions form through gas overpressure, cap failure, and dynamic ejecta transport.
Key Innovation: In this work, we develop a unified process-based modeling framework with the use of the Generalized Interpolation Material Point (GIMP) method to evaluate gas-driven fracturing and ejecta dynamics in frozen soils.
62. Evidence-optimized physics-informed digital twin for probabilistic prediction and risk assessment of wall deflection in braced excavations
Core Problem: Accurate prediction and risk assessment of time-varying wall deflection during staged braced excavations remain challenging due to evolving system behavior and multiple sources of uncertainty.
Key Innovation: Results show that sequential data assimilation drives progressive posterior contraction of key parameters and improves the agreement between predicted and measured wall-deflection profiles.
63. Integrated use of CPT and VS for seismic site classification and liquefaction potential evaluation considering soil aging effects
Core Problem: Soil aging can alter seismic site classification and liquefaction resistance, but it is often neglected in routine CPT- and VS-based assessments.
Key Innovation: The paper derives an aging factor from CPT-VS relations and incorporates it into a Bayesian liquefaction probability model for risk-informed seismic site characterization.
64. Investigations of a vertical dual-control bearing with variable-stiffness and variable-friction for control of metro- and earthquake-induced vibrations
Core Problem: Buildings along metro lines in high-intensity seismic zones face dual demands for vibration comfort and seismic resilience.
Key Innovation: The results indicate that the VSVFS bearing exhibits multi-stage variable-stiffness and variable-friction characteristics.
65. Physics-informed deep learning framework for prediction of cone resistance from seismic reflection data
Core Problem: Accurate characterization of seabed soil properties is essential for foundation design and wind farm layout optimization in offshore renewable energy projects.
Key Innovation: This study proposes a physics-informed deep learning framework for predicting CPT cone resistance (q c) from time-domain seismic reflection data, with P-wave velocity (v p) used as the central physical intermediary linking geophysical observations and geotechnical properties.
66. Probabilistic prediction of maximum scour depth downstream of sluices and culverts using a tabular foundation model
Core Problem: Accurate prediction of maximum scour depth downstream of hydraulic structures is essential for safe design and risk-informed assessment.
Key Innovation: This study investigates the application of TabPFN (Tabular Prior-Data Fitted Network), a pre-trained foundation model for tabular data, for probabilistic prediction of scour depth downstream of culverts and sluice gates.
67. Regional process-based analysis of trends in flood peaks and volumes using hourly data
Core Problem: Most studies on flood trends rely on daily discharge data at the station scale, limiting their ability to disentangle contrasting trends in different flood-generation mechanisms.
Key Innovation: Here, we propose a process-based regional framework to investigate trends across different flood types using hourly data.
68. Seasonal frost impact on liquefaction-induced ground lateral spreading
Core Problem: Liquefaction-induced lateral spreading poses a significant hazard to the built transportation infrastructure in cold regions, where seasonal freezing and thawing alter the ground crust's hydraulic and mechanical behavior.
Key Innovation: This study investigates the impact of seasonally frozen ground crust on liquefaction-induced lateral spreading displacement.
69. Seismic responses of a fault-crossing tunnel with a variable-stiffness composite lining structure: Theoretical analysis and shaking table test verification
Core Problem: Fault-crossing tunnels are exposed to concentrated seismic deformation that single mitigation measures often cannot control.
Key Innovation: The study develops and tests a variable-stiffness composite lining, combining theoretical analysis and shaking-table validation for tunnels crossing active faults.
70. Time-dependent seismic resilience assessment of pile-supported wharves under climate-modified chloride-induced corrosion
Core Problem: Pile-supported wharves operate in aggressive marine environments where aging, chloride-induced corrosion, and climate-modified temperature and humidity alter seismic vulnerability and post-earthquake recovery.
Key Innovation: This study proposes a hazard-consistent, time-dependent framework to quantify the seismic resilience of deteriorating pile-supported wharves over a 50-year service life by integrating site-specific probabilistic seismic hazard analysis, time-dependent fragility curves, and damage-state-dependent recovery functions.
71. Influence of multi-dimensional ground motion intensity measures on seismic resilience of embankments on liquefiable soils
Core Problem: Scalar ground-motion intensity measures do not fully capture amplitude, duration, and frequency effects on embankment resilience over liquefiable soils.
Key Innovation: The study evaluates multi-dimensional intensity measures to improve prediction of permanent deformation and functional degradation in liquefiable embankment systems.
72. Seismic fragility of nuclear power plant structure considering raft-pile-soil-structure interaction in improved soil
Core Problem: The seismic response of Nuclear Power Plant (NPP) structures is a critical factor in ensuring their safety and operational integrity during seismic events.
Key Innovation: This study comprehensively investigates the effects of nonlinearity and raft-pile-soil-structure interaction (RSSI) on the seismic response of NPP structures, focusing on different foundation types: raft foundations, pile group foundations, and combined piled raft foundations.
73. A novel framework for Identifying and predicting stratified land deformation: a case study of the Beijing Plain
Core Problem: Regional land subsidence research has advanced rapidly in recent decades, and integrating InSAR with machine learning has become an important research direction.
Key Innovation: To address this limitation, we propose a framework that integrates SAR data, stratified deformation ratios from extensometers, and hydrogeological information.
74. A spatial recursion framework for nonstationary flood frequency analysis considering river network dependence and reservoir routing
Core Problem: Flood frequency analysis (FFA) is critical for hydraulic design and flood risk management, yet the traditional stationarity assumption is increasingly challenged by climate change and intensive human activities.
Key Innovation: Results show that, except for two upstream stations, flood peaks at all other regulated nodes exhibit significant nonstationary behavior.
75. Emerging dryland flooding
Core Problem: Drylands are typically considered as low flood risk regions due to their aridity and predominance of dry river channels.
Key Innovation: To adapt to the expected increase in hazard, exposure and risk, we propose six transferable solutions that span from engineering strategies to nature-based solutions to enhance flood resilience and minimize flood damage in these often-marginalized drylands.
76. Multi-station rainfall-runoff-operation model for assessing flood characteristics in a cascading dam system under the impacts of climate change and sediment dynamics
Core Problem: Flood hazard in cascading dam systems depends jointly on rainfall-runoff response, dam operation, climate change, and sediment-driven storage loss.
Key Innovation: The study trains a multi-station deep-learning rainfall-runoff-operation model and uses large-ensemble climate projections to quantify future flood-frequency and spillway-capacity pressures.
77. SAR2LiDAR: A deep learning-based cross-modal matching framework between SAR images and LiDAR point clouds
Core Problem: Synthetic aperture radar (SAR) and light detection and ranging (LiDAR) are two fundamental active remote sensing technologies.
Key Innovation: In this paper, a deep learning-based SAR-to-LiDAR (SAR2LiDAR) matching framework is proposed, which enables direct matching between SAR images and LiDAR point clouds.
78. Small lakes, catastrophic impacts: Precursor signals and monitoring gaps revealed by the 2025 Gyirong GLOF
Core Problem: On 8 July 2025, an uncatalogued glacial lake (0.83 km2) in Gyirong, Tibet, burst and destroyed the China-Nepal Friendship Bridge, causing at least 9 deaths and about 30 persons missing (per official reports as of August 2025).
Key Innovation: The lake was absent from inventories not because a 1 km2 threshold excluded it, but because it had only recently formed, and because several regional inventories have incomplete coverage below roughly 10 ha and lag behind the rapid formation of supraglacial and newly formed moraine lakes between update cycles.
79. Do CMIP6 models reduce the uncertainty in sub-daily extreme precipitation projections?
Core Problem: Sub-daily extreme precipitation projections remain uncertain, even though they are central to flood and rainfall-triggered hazard assessment.
Key Innovation: The paper compares CMIP6 and CMIP5 projections to test whether newer climate models reduce uncertainty in future sub-daily precipitation extremes.
80. Effect of rainfall infiltration on the stability of geosynthetic reinforced walls in unsaturated soil
Core Problem: Rainfall is a primary trigger of instability in geosynthetic reinforced soil (GRS) walls, as infiltration alters the hydromechanical behavior of both the unsaturated backfill and the soil-reinforcement interface, particularly for walls backfilled with marginal soils.
Key Innovation: This study establishes an analytical framework that integrates a rainfall-dependent effective stress model under transient infiltration into the limit equilibrium solution, namely the top-down procedure, to evaluate the stability of GRS walls during rainfall events.
81. Effects of non-persistent discontinuities on dynamic responses of rock slopes
Core Problem: Local nonlinear amplification of seismic motion caused by discontinuities is a critical factor contributing to the dynamic instability of rock slopes.
Key Innovation: This study first investigates the dynamic response of a slope with a single controlling non-persistent joint at the top using shaking table tests, and then validates the reliability of the numerical model using experimental results.
82. Efficient three-dimensional finite element modeling of bridge seismic response in liquefiable river valleys: validation and case studies
Core Problem: Bridge response in liquefiable river valleys is controlled by coupled site liquefaction, lateral spreading, pile-superstructure interaction, and valley geometry.
Key Innovation: The paper validates an efficient 3D coupled finite-element workflow against laboratory tests and applies it to quantify bridge deformation under liquefiable and non-liquefiable site conditions.
83. Evaluation of ground-motion intensity measures for seismic response of fault-crossing tunnels
Core Problem: Seismic damage mechanisms of fault-crossing tunnels are complex and involve the coupled effects of fault offset and seismic ground-motion.
Key Innovation: In this study, a series of shaking table tests is conducted to investigate the applicability of different ground-motion intensity measures (IMs) for evaluating the seismic response of fault-crossing tunnels.
84. How uncertain are multi-source precipitation products in characterizing extreme precipitation?
Core Problem: Given the increasing risks of extreme precipitation and the growing number of available precipitation products, it is necessary to comprehensively quantify uncertainties in how these products characterize extreme precipitation.
Key Innovation: This study uses 24 precipitation products across six categories over seven climatic zones in China to quantify uncertainty in the characterization of extreme precipitation and to identify consistent patterns in daily variability, annual trends, total amounts, event detection, and spatiotemporal displacement errors.
85. Improved understanding of California coastal cliff morphodynamics using airborne lidar and vegetation filtering (2016-2023)
Core Problem: Coastal cliffs are widespread along California coastlines, yet their erosion remains difficult to quantify at statewide scales because of (1) limited high-resolution data, (2) intensive processing requirements, and (3) challenges in mapping ground elevations beneath vegetation.
Key Innovation: This study collected airborne lidar data across the California coast in 2023 and compared it with 2016 lidar to evaluate coastal cliff changes.
86. Integrating GNSS-Derived Zenith Wet Delay into a Weather Foundation Model Improves Precipitation Forecasting
Core Problem: Global Navigation Satellite Systems (GNSS), best known for positioning, also serve weather science, as atmospheric water vapour delays their signals.
Key Innovation: Here we present the first integration of GNSS-derived ZWD into Aurora, a state-of-the-art weather foundation model.
87. Seismic response of subway station structures in soil deposits with liquefiable interlayer under near-fault pulse-like ground motions
Core Problem: Near-fault ground motions with forward-directivity velocity pulses can impose concentrated deformation demands on underground structures in liquefiable ground, but their influence on subway station structures embedded in liquefiable interlayered deposits remains insufficiently understood.
Key Innovation: In this study, a two-dimensional nonlinear effective-stress soil-structure interaction model was developed for a representative shallow-buried two-story, three-span subway station.