Initiated by Dr. Xin Wei, University of Michigan
Ongoing development by the community

TerraMosaic Daily Digest: April 27, 2026

April 27, 2026
TerraMosaic Daily Digest

Daily Summary

Today's papers place landslide science inside a broader geohazard shift from detection toward constrained explanation. The strongest studies connect hidden processes to observable fields: soil-piping experiments show when subsurface conduits drain slopes and when partial blockage triggers shallow failure, while multi-geometry InSAR, CR-InSAR, explainable learning, and finite-element constraints turn landslide and highway-slope deformation maps into mechanically interpretable evidence. Laboratory strain imaging, coupled FDM-DEM surrogate reliability, and landslide-tsunami time-series emulation extend the same principle across scales, replacing inaccessible failure processes with testable deformation, strain, and wave-boundary representations.

The wider set emphasizes infrastructures and hazards whose risk depends on records as much as algorithms. Earthquake photographs, urban flood labels, reservoir-slope deformation, coastal road settlement, volcanic electrical structure, mine pressure waves, freeze-thaw rock damage, and tunnel or pipeline response all show a common movement toward curated evidence, uncertainty-aware modelling, and physical attribution. The most useful remote-sensing and AI papers are not generic accuracy gains; they are methods that preserve process signals, expose failure controls, or supply reusable observations for hazards that remain sparse, transient, or hard to instrument.

Key Trends

Across landslides, floods, earthquakes, volcanic systems, cold-region rock failure, and engineered ground, the technical center of gravity is shifting toward process-constrained observation.

  • Hidden failure mechanisms are being made observable: soil piping, full-field slope strain, freeze-thaw rock fracture, water-sand collapse, rock-concrete shear, rockfill breakage, and coal-pillar dilatancy all use experiments or particle-scale models to expose pre-failure processes that standard monitoring cannot see directly.
  • InSAR is moving from displacement mapping to stability reasoning: reservoir-region landslides, incomplete SAR deformation fields, highway slopes, coastal roads, railway subgrades, and mining landscapes use multi-geometry observations, corner reflectors, interpretable learning, or forecasting to connect deformation patterns to hazard mechanisms.
  • Surrogate models are becoming acceptable only when physically anchored: FDM-DEM slope reliability, landslide-generated tsunami boundaries, subsea-pipeline collapse, THM fatigue, tunnel response, and structural seismic damage show that fast models are most valuable when their inputs, uncertainty, and governing processes remain explicit.
  • Hazard archives and hand-labelled datasets are turning into infrastructure: the INGV earthquake-photo archive, UFO urban-flood labels, Armenian PSHA model, forest-fire event reconstruction, polder history, and flood digital-twin review all strengthen the empirical basis needed for reproducible hazard assessment.
  • Foundation and open-vocabulary remote sensing remain conditional tools: UGDMoE, zero-shot Segment Anything, water-quality downscaling, gap-free SST reconstruction, PolInSAR decomposition, impervious-surface extraction, and ecological-resilience mapping are most relevant where they add transferable structure rather than detached classification scores.

Selected Papers

This digest features 70 selected papers from 2,135 papers analyzed. The sequence opens with soil-piping controls on shallow landslide initiation, then moves through multi-geometry and explainable InSAR for reservoir and highway slopes, full-field strain imaging, coupled FDM-DEM reliability, landslide-generated tsunami surrogates, earthquake and flood datasets, volcanic and mine-hazard monitoring, cold-region rock failure, coastal and railway settlement, tunnel and pipeline response, and process-aware remote-sensing methods.

1. Impact of soil piping on shallow landslide initiation: Experimental observations

Source: Engineering Geology Type: Concepts & Mechanisms Geohazard Type: Shallow landslides Relevance: 9/10

Core Problem: Shallow-landslide initiation models rarely resolve how soil piping changes pore-pressure routing, internal collapse, and the transition from subsurface erosion to slope failure.

Key Innovation: Controlled slope experiments compare no-pipe, closed-pipe, open-pipe, and partly blocked pipe conditions across slope angle and compaction, showing when piping drains the slope and when blockage focuses collapse.

2. Integrating multi-geometry InSAR and explainable machine learning for mapping potential landslide areas in reservoir regions

Source: Remote Sensing of Env. Type: Detection and Monitoring Geohazard Type: Reservoir-region landslides Relevance: 8/10

Core Problem: Reservoir-bank landslide screening is limited when single-geometry InSAR misses directional deformation and machine-learning outputs are difficult to interpret mechanically.

Key Innovation: The study integrates multi-geometry InSAR with explainable machine learning to map potential landslide areas and expose the deformation and terrain factors controlling reservoir-slope instability.

3. Explainable Artificial Intelligence for Estimating Surface Deformation in Landslide Areas with Incomplete SAR Data

Source: Remote Sensing (MDPI) Type: AI / Remote Sensing Geohazard Type: Landslide deformation Relevance: 8/10

Core Problem: Incomplete SAR coverage leaves landslide deformation fields spatially fragmented, weakening displacement interpretation in areas where direct interferometric observations are sparse.

Key Innovation: SBAS-InSAR, ensemble machine learning, and SHAP attribution are combined to estimate missing surface deformation, with XGBoost producing the strongest GNSS-validated performance.

4. Stability Evaluation of Vegetation-Covered Highway Slopes Employing Integrated CR-InSAR and Finite Element Simulation

Source: Remote Sensing (MDPI) Type: Hazard Modelling Geohazard Type: Vegetation-covered highway slopes Relevance: 8/10

Core Problem: Vegetation and sparse coherent targets make conventional InSAR poorly suited to highway slopes where small deformation can still indicate changing stability.

Key Innovation: Artificial corner reflectors, CR-InSAR deformation rates, and finite-element simulation are integrated to evaluate the stability of a vegetation-covered highway slope from observation to mechanics.

5. Development of an image-based method of visualizing maximum principal strain for slope laboratory modeling

Source: Engineering Geology Type: Experimental Monitoring Geohazard Type: Slope physical models Relevance: 8/10

Core Problem: Embedded sensors disturb laboratory slope models and provide only point measurements, limiting observation of strain localization before failure.

Key Innovation: An image-based visualized maximum principal strain method uses non-contact DIC to recover full-field strain magnitude and direction in slope physical modelling.

6. An efficient reliability assessment of coupled FDM-DEM slope stability using physics-informed feature-based surrogate machine learning techniques

Source: Computers and Geotechnics Type: Reliability Assessment Geohazard Type: Slope stability Relevance: 8/10

Core Problem: Coupled FDM-DEM slope simulations can represent failure mechanisms, but their computational cost prevents routine reliability assessment under uncertain parameters.

Key Innovation: A physics-informed feature-based surrogate machine-learning framework emulates coupled FLAC3D-PFC3D slope stability simulations for efficient probabilistic reliability analysis.

7. Multi-Temporal InSAR and Machine Learning for Geohazard Monitoring: A Systematic Review with Emphasis on Noise Mitigation and Model Transferability

Source: Remote Sensing (MDPI) Type: Review Geohazard Type: InSAR/ML geohazard monitoring Relevance: 7/10

Core Problem: Operational geohazard monitoring with multi-temporal InSAR remains constrained by atmospheric noise, decorrelation, incomplete transferability, and weak model generalization.

Key Innovation: A systematic review of 135 studies organizes machine-learning and deep-learning uses in MT-InSAR geohazard monitoring, emphasizing noise mitigation and cross-site transferability.

8. Deep neural network–based prediction of near-field time series for landslide-generated tsunamis: A surrogate framework for boundary input modeling

Source: Ocean Engineering Type: Hazard Modelling Geohazard Type: Landslide-generated tsunamis Relevance: 7/10

Core Problem: Near-field landslide-tsunami simulations are too expensive for repeated boundary-condition generation in rapid or probabilistic hazard workflows.

Key Innovation: A deep neural surrogate predicts water-level and velocity time series near landslide sources, providing boundary inputs for downstream tsunami modelling at reduced cost.

9. The INGV macroseismic photographic database (DFM): a structured photographic collection of earthquake effects in Italy

Source: ESSD Type: Open Dataset Geohazard Type: Earthquake effects Relevance: 7/10

Core Problem: Photographic evidence of earthquake damage is often scattered, inconsistently described, and difficult to reuse for macroseismic interpretation or model validation.

Key Innovation: The INGV DFM database structures Italian earthquake-effect photographs with standardized metadata and EMS-98 damage descriptors under a FAIR data framework.

10. Probabilistic Seismic Hazard Assessment of Armenia Using an Integrated Seismotectonic Framework

Source: GeoHazards (MDPI) Type: Seismic Hazard Assessment Geohazard Type: Earthquakes Relevance: 7/10

Core Problem: Armenia's seismic hazard requires an integrated treatment of active faults, paleoseismic evidence, historical earthquakes, and instrumental seismicity.

Key Innovation: An OpenQuake probabilistic seismic-hazard model combines fault-based and distributed sources through a logic-tree framework to produce updated national ground-motion maps.

11. Urban Flood Observations (UFO): A hand-labeled training and validation dataset of post-flood inundation

Source: ArXiv (Geo/RS/AI) Type: Open Dataset Geohazard Type: Urban floods Relevance: 7/10

Core Problem: Urban flood mapping lacks small, high-quality labelled datasets for validating inundation models under post-event satellite conditions.

Key Innovation: Urban Flood Observations supplies 215 hand-labelled 3 m PlanetScope image chips from 14 flood events and benchmarks segmentation performance for post-flood inundation.

12. Critical Assessment on Digital Twins Implementation for Flood Risk Management and Governance: A Systematic Literature Review

Source: IJDRR Type: Review Geohazard Type: Flood risk management Relevance: 7/10

Core Problem: Digital twins are increasingly proposed for flood-risk management, but implementation pathways and governance requirements remain fragmented.

Key Innovation: The review evaluates flood digital-twin literature and identifies the technical, institutional, and human-nature governance components needed for operational adoption.

13. Unveiling coastal urban road settlement issues from InSAR monitoring and time-series forecasting: insights from Zhuhai, China

Source: Geomatics, Nat. Haz. & Risk Type: Detection and Monitoring Geohazard Type: Coastal urban road settlement Relevance: 7/10

Core Problem: Coastal road systems can undergo spatially uneven settlement controlled by soft ground, hydrology, road type, and extreme meteorological events, but sparse surveys rarely resolve the full network.

Key Innovation: SBAS-InSAR maps Zhuhai road-system settlement from 2016 to 2019, identifies a three-ring deformation structure and prominent segments reaching -150 to -100 mm, and uses time-series forecasting to project short-term settlement of -50 to -30 mm.

14. Three-dimensional electrical structure beneath the Changbaishan volcano (NE China): Implications for magmatic plumbing systems and diverse eruption styles

Source: Geoscience Frontiers Type: Detection and Monitoring Geohazard Type: Volcanic plumbing systems Relevance: 7/10

Core Problem: Changbaishan's diverse eruption styles require a three-dimensional view of the magmatic plumbing system beneath the volcano.

Key Innovation: Dense magnetotelluric imaging resolves the subsurface electrical structure and links conductive pathways to magma storage, ascent, and eruption diversity.

15. Experimental Study on Pressure Wave Propagation in Mine Ventilation Disasters

Source: GeoHazards (MDPI) Type: Early Warning Geohazard Type: Mine ventilation disasters Relevance: 7/10

Core Problem: Coal-and-gas outburst disasters generate pressure waves whose propagation and attenuation in mine ventilation networks remain difficult to quantify.

Key Innovation: A 1:30 scaled mine-ventilation model separates static and dynamic pressure-wave behavior, improving understanding of outburst warning signals and ventilation safety.

16. Competitive interaction between fluids and solids during water–sand mixture collapse beneath a water body

Source: TUST Type: Concepts & Mechanisms Geohazard Type: Mining-related ground collapse Relevance: 7/10

Core Problem: Ground collapse beneath water bodies depends on coupled fluid-solid competition, but the stages from infiltration to visible collapse are hard to observe directly.

Key Innovation: Transparent-soil visualization resolves infiltration activation, collapse, lateral expansion, slope erosion, and self-stabilization during water-sand mixture failure.

17. The role of effective pressure on fracture and failure mode mechanisms in freeze-thaw damaged rocks

Source: Cold Regions Sci. & Tech. Type: Concepts & Mechanisms Geohazard Type: Freeze-thaw rock failure Relevance: 7/10

Core Problem: Freeze-thaw damage changes rock fracture mode in cold mountain environments, yet the role of effective pressure in damaged rock failure is poorly constrained.

Key Innovation: An improved freeze-thaw grain-based DEM model and acoustic-emission analysis link confining pressure, microcrack evolution, and macroscopic failure mode.

18. The end stage of barrier estuary infill: Insights from Jack Smith Lake, Southeastern Australia

Source: Earth Surf. Proc. & Landforms Type: Geomorphic Analysis Geohazard Type: Barrier estuary evolution Relevance: 6/10

Core Problem: Late-stage barrier-estuary infill controls coastal accommodation, wetland evolution, and future exposure, but mature infill examples are rarely resolved in detail.

Key Innovation: Geomorphic mapping, sediment cores, and chronological evidence reconstruct the end-stage infill of Jack Smith Lake and clarify how barrier estuaries approach closure.

19. Decoupling water erosion dynamics in the Jinsha River Basin, China: From historical to future scenarios

Source: Catena Type: Hazard Modelling Geohazard Type: Water erosion Relevance: 6/10

Core Problem: Large-basin erosion hazards reflect coupled climate and land-cover dynamics, making it difficult to separate historical drivers from future scenarios.

Key Innovation: RUSLE, CMIP6 climate projections, PLUS land-use simulation, and machine learning are combined to decouple water-erosion controls in the Jinsha River Basin.

20. Identifying Priority Vegetation Zones for Floodplain Wetland Management Based on Hydrological and Meteorological Drivers

Source: Water Resources Research Type: Ecohydrological Analysis Geohazard Type: Floodplain wetlands Relevance: 6/10

Core Problem: Floodplain wetland vegetation responds to hydrological regulation and meteorological variability, but management zones are often selected without driver attribution.

Key Innovation: Multi-decadal hydrological and meteorological analysis identifies priority vegetation zones for Poyang Lake floodplain-wetland management.

21. Spatiotemporal dynamics of polders in the Chaohu Lake Basin, lower Yangtze River region of China, over the last 1000 years

Source: Catena Type: Historical Geomorphology Geohazard Type: Flood-control polders Relevance: 6/10

Core Problem: Millennial polder construction has reshaped floodplain storage, land reclamation, and flood-control capacity in the lower Yangtze region.

Key Innovation: Historical maps, gazetteers, and remote-sensing evidence reconstruct the 1000-year spatiotemporal evolution of Chaohu Lake Basin polders.

22. Fluctuations of Caspian Sea water due to climate change and the relict evidence in coastal soils

Source: Catena Type: Coastal Hazard Analysis Geohazard Type: Endorheic coastal plains Relevance: 6/10

Core Problem: Caspian Sea-level fluctuations reorganize inundation, salinity, groundwater depth, and soil development along vulnerable coastal plains.

Key Innovation: Coastal soil records are used to infer past water-level changes and the hydrosalinity cycles that accompany climate-driven Caspian shoreline migration.

23. Advantages of integrated geophysical, sedimentary, and dating approaches for reconstructing Holocene tidal inlet evolution: A case study from offshore the Belgian coastal plain

Source: Geomorphology Type: Coastal Geomorphology Geohazard Type: Tidal inlet evolution Relevance: 6/10

Core Problem: Relict tidal inlets are difficult to reconstruct where modern seabed morphology, sediments, and chronology record different parts of the system.

Key Innovation: Integrated geophysical profiles, sedimentological evidence, radiocarbon dating, and OSL dating reconstruct Holocene tidal-inlet evolution offshore Belgium.

24. Spatiotemporal Dynamics and Driving Patterns of Forest Fires in Yunnan Province, China: An Empirical Study Based on Event-Level Reconstruction from Multi-Source Remote Sensing (2012–2024)

Source: Remote Sensing (MDPI) Type: Detection and Monitoring Geohazard Type: Forest fires Relevance: 6/10

Core Problem: Active-fire points fragment real wildfire events and can obscure event-level drivers, recurrence, and spatial clustering.

Key Innovation: Optical-thermal cross-validation and spatiotemporal clustering reconstruct Yunnan forest-fire events from 2012 to 2024 for driver and pattern analysis.

25. Performance assessment of WRF-ELEC and lightning potential index for improved forecasting in Odisha, India

Source: Natural Hazards Type: Early Warning Geohazard Type: Lightning hazards Relevance: 6/10

Core Problem: Operational lightning forecasting remains difficult in monsoon-affected regions where convective electrical structure changes quickly.

Key Innovation: WRF-ELEC and lightning-potential-index forecasts are evaluated for Odisha, India, clarifying how physics-based and proxy approaches support warning skill.

26. Assessing future heat wave patterns in India: insights from a high-resolution regional climate model

Source: Natural Hazards Type: Hazard Projection Geohazard Type: Heatwaves Relevance: 6/10

Core Problem: India's heatwave risk depends on regional timing, duration, and frequency changes that coarse climate products can smooth out.

Key Innovation: A high-resolution regional climate model projects altered heatwave onset, cessation, and frequency patterns across India.

27. Numerical Assessment of Environmental Hazard Mitigation Through River Mouth Structures in Enclosed Aquatic Systems: A Case of Hypoxia Reduction

Source: GeoHazards (MDPI) Type: Mitigation Geohazard Type: Hypoxia and water-quality hazards Relevance: 6/10

Core Problem: Enclosed aquatic systems can develop hypoxia when river-mouth circulation and nutrient transport reinforce stagnant conditions.

Key Innovation: A three-dimensional hydrodynamic-ecosystem model tests river-mouth structures as a mitigation measure for hypoxia reduction.

28. Water salinity in global salt lakes: Monitoring technologies, spatiotemporal dynamics, and socio-environmental consequences

Source: Earth-Science Reviews Type: Review Geohazard Type: Salt lake water stress Relevance: 6/10

Core Problem: Salt-lake salinity governs ecosystem stress, water-resource use, and regional environmental risk, but monitoring approaches and impacts are dispersed across fields.

Key Innovation: The review synthesizes salinity monitoring technologies, spatiotemporal dynamics, and socio-environmental consequences for global salt lakes.

29. Palaeochannels as hidden pathways of water flow in agricultural alluvial landscapes: A review of concepts and suggested future directions for modelling

Source: Journal of Hydrology Type: Review Geohazard Type: Palaeochannel hydrology Relevance: 6/10

Core Problem: Buried palaeochannels can act as hidden pathways for water and solute movement but are often absent from agricultural alluvial-landscape models.

Key Innovation: The review defines palaeochannel structure and function and proposes modelling directions for incorporating hidden pathways into hydrological assessment.

30. Spatiotemporal dynamics of blue-green water resources in the Min-Tuo River Basin: A coupled SWAT-FLUS-CMIP6 modeling approach

Source: Journal of Hydrology Type: Hydroclimate Modelling Geohazard Type: Blue-green water resources Relevance: 6/10

Core Problem: Climate change and land-use change jointly alter blue and green water availability, complicating basin-scale water-security planning.

Key Innovation: A coupled SWAT-FLUS-CMIP6 framework projects spatiotemporal blue-green water dynamics in the Min-Tuo River Basin.

31. Reconstructing High-Resolution Coastal Water Quality Data Based on a Deep Learning Multivariate Downscaling Approach

Source: Remote Sensing (MDPI) Type: AI / Remote Sensing Geohazard Type: Coastal water-quality monitoring Relevance: 6/10

Core Problem: High-resolution coastal water-quality products remain limited by cloud gaps, sparse observations, and scale mismatch among satellite sensors.

Key Innovation: A multivariate deep-learning downscaling framework reconstructs high-resolution coastal water-quality data from physically and statistically linked inputs.

32. Physically-aware deep learning for reconstructing gap-free sea surface temperature in the South China Sea

Source: ISPRS J. Photogrammetry Type: AI / Remote Sensing Geohazard Type: Gap-free geophysical fields Relevance: 6/10

Core Problem: Cloud cover and orbital sampling gaps limit the continuity of sea-surface-temperature fields needed for ocean and coastal hazard applications.

Key Innovation: A physically aware deep-learning model reconstructs gap-free sea-surface temperature in the South China Sea while preserving geophysical consistency.

33. UGDMoE: An Uncertainty-Guided Mixture-of-Experts Decoder for Open-Vocabulary Remote Sensing Segmentation

Source: Remote Sensing (MDPI) Type: AI / Remote Sensing Geohazard Type: Open-vocabulary segmentation Relevance: 6/10

Core Problem: Open-vocabulary remote-sensing segmentation struggles with dense textures, scale variation, and uncertain text-image correspondence.

Key Innovation: UGDMoE uses an uncertainty-guided mixture-of-experts decoder to adapt vision-language representations to remote-sensing segmentation scenes.

34. Zero-shot inference strategies for smallholder (<0.1 ha) agriculture field delineation with the Segment Anything foundation model

Source: Science of Remote Sensing Type: AI / Remote Sensing Geohazard Type: Foundation-model field delineation Relevance: 6/10

Core Problem: Smallholder field boundaries are difficult to delineate because labels are scarce and field sizes are often below 0.1 ha.

Key Innovation: Zero-shot Segment Anything inference strategies are evaluated for very small agricultural fields, clarifying when foundation models can transfer without local training labels.

35. Effect of Real-Time High-Temperature on T-H-M Coupled Properties of Rock–Concrete Interfaces under Shear Condition

Source: Transportation Geotechnics Type: Concepts & Mechanisms Geohazard Type: High-geothermal tunnel interfaces Relevance: 6/10

Core Problem: Rock-concrete interfaces in high-geothermal tunnels experience coupled thermal, hydraulic, and mechanical damage that conventional shear tests underrepresent.

Key Innovation: Real-time high-temperature THM shear experiments quantify how temperature, seepage, and normal stress change interface strength and damage evolution.

36. A cohesive zone model for fatigue under cyclic thermo-hydro-mechanical loading

Source: Computers and Geotechnics Type: Hazard Modelling Geohazard Type: Cyclic THM rock fatigue Relevance: 6/10

Core Problem: Underground structures can accumulate fatigue damage under variable thermo-hydro-mechanical cycles, but models rarely degrade strength and stiffness consistently.

Key Innovation: A cohesive-zone formulation represents tensile-strength and stiffness degradation under cyclic THM loading for progressive fracture analysis.

37. Discrete element investigation of particle breakage and macro–micro behaviour of rockfill materials under flexible boundary triaxial compression

Source: Computers and Geotechnics Type: Concepts & Mechanisms Geohazard Type: Rockfill dam materials Relevance: 6/10

Core Problem: Rockfill deformation and dam stability depend on particle breakage, shape, and fabric evolution that continuum tests can mask.

Key Innovation: Discrete-element triaxial simulations with realistic particle shapes and flexible boundaries link macro-scale rockfill response to micro-scale breakage behavior.

38. A long-term fracture conductivity model for shale considering non-linear creep damage, proppant elastic-elastoplastic-plastic embedment, and hydration swelling effects

Source: Computers and Geotechnics Type: Hazard Modelling Geohazard Type: Shale fracture conductivity Relevance: 6/10

Core Problem: Fracture conductivity in shale decays through creep damage, proppant embedment, and hydration swelling, complicating long-term subsurface flow prediction.

Key Innovation: An analytical model integrates nonlinear creep damage, elastic-elastoplastic-plastic proppant embedment, and hydration swelling to estimate conductivity decline.

39. Mechanical behavior of rocks repaired and reinforced by modified EICP biocementation method

Source: Transportation Geotechnics Type: Mitigation Geohazard Type: Weathered rock reinforcement Relevance: 6/10

Core Problem: Weathered and cracked rock masses lose strength through microstructural damage that conventional repair methods may not penetrate effectively.

Key Innovation: A modified EICP biocementation method repairs and reinforces damaged rock by precipitating cementing products within cracks and pores.

40. Enhancing strain transfer efficiency of fibre optic sensors for geo-structural health monitoring through 3D-printed packaging

Source: JRMGE Type: Detection and Monitoring Geohazard Type: Geo-structural health monitoring Relevance: 6/10

Core Problem: Fiber-optic monitoring of geo-structures is limited when sensor packaging weakens strain transfer from host material to fiber.

Key Innovation: 3D-printed packaging is designed to improve strain-transfer efficiency for distributed fiber-optic sensing in geo-structural health monitoring.

41. Biaxial gradient loading approach with unilateral constraint for testing dilatancy behavior of narrow coal pillar in gob-side entry driving

Source: JRMGE Type: Concepts & Mechanisms Geohazard Type: Coal-pillar subsidence and instability Relevance: 6/10

Core Problem: Narrow coal pillars in gob-side entry driving experience gradient loads and unilateral constraints linked to roof rotation and subsidence.

Key Innovation: A biaxial gradient-loading test with unilateral constraint captures dilatancy behavior and failure evolution in narrow coal pillars.

42. Railway subgrade settlement interval prediction based on collaborative enhancement denoising and bidirectional gated recurrent unit with Monte Carlo Dropout model

Source: RESS Type: Hazard Modelling Geohazard Type: Railway subgrade settlement Relevance: 6/10

Core Problem: Railway subgrade settlement prediction needs uncertainty intervals rather than only point estimates for maintenance and safety decisions.

Key Innovation: Collaborative enhancement denoising and a bidirectional GRU with Monte Carlo dropout produce interval forecasts for subgrade settlement.

43. Seismic risk assessment framework for buildings in seasonally frozen zones considering multiscale seismic intensity measures

Source: Cold Regions Sci. & Tech. Type: Risk Assessment Geohazard Type: Seasonally frozen-zone seismic risk Relevance: 5/10

Core Problem: Building seismic risk in seasonally frozen zones depends on frozen-soil effects that are rarely represented in portfolio-scale loss frameworks.

Key Innovation: A multiscale seismic-intensity framework incorporates seasonally frozen conditions into building seismic-risk assessment.

44. Seismic collapse resistance and fragility assessment of ancient timber structure retrofitted with displacement-amplified rotational friction dampers

Source: Bull. Earthquake Eng. Type: Mitigation Geohazard Type: Seismic collapse of timber structures Relevance: 5/10

Core Problem: Ancient timber structures need retrofits that improve collapse resistance without erasing structural and heritage behavior.

Key Innovation: Displacement-amplified rotational friction dampers are evaluated through collapse-resistance and fragility analysis for retrofitted timber structures.

45. A framework for analyzing seismic resilience and chain-reaction economic losses in substations based on sample optimization strategies

Source: RESS Type: Resilience Assessment Geohazard Type: Seismic substations Relevance: 5/10

Core Problem: Substation earthquake losses can propagate through power systems, but full chain-reaction resilience analysis is computationally intensive.

Key Innovation: A sample-optimization strategy supports seismic resilience and economic-loss analysis for substation systems under cascading effects.

46. Research on the invulnerability of interdependent networks in energy internet under seismic disasters: A case study of Beijing

Source: RESS Type: Resilience Assessment Geohazard Type: Seismic energy internet networks Relevance: 5/10

Core Problem: Interdependent power and information networks can suffer cascading failures during seismic disasters, reducing energy-system robustness.

Key Innovation: A Beijing energy-internet case study evaluates invulnerability in coupled networks under earthquake-induced disruptions.

47. Coupled defect interaction analysis and interpretable collapse pressure prediction of subsea pipelines with uncertainty quantification

Source: Ocean Engineering Type: Risk Assessment Geohazard Type: Subsea pipeline collapse Relevance: 5/10

Core Problem: Subsea pipelines with interacting defects require collapse-pressure predictions that represent uncertainty and expose controlling features.

Key Innovation: Nonlinear analysis, Bayesian neural prediction, and SHAP interpretation quantify coupled-defect effects and uncertainty in collapse pressure.

48. Individual Tree Species Classification in a Mining Area of the Yellow River Basin Using UAV-Based LiDAR, Hyperspectral, and RGB Data

Source: Remote Sensing (MDPI) Type: Detection and Monitoring Geohazard Type: Mining ecological restoration Relevance: 5/10

Core Problem: Mining-area restoration monitoring needs species-level vegetation information that standard optical surveys cannot reliably recover.

Key Innovation: UAV LiDAR, hyperspectral, and RGB data are fused for individual tree-species classification in a restored Yellow River Basin mining area.

49. Stage-dependent topographic controls on vegetation recovery and landscape-pattern evolution under contrasting restoration treatments in an abandoned mine from UAV multispectral time series

Source: Catena Type: Detection and Monitoring Geohazard Type: Abandoned mine restoration Relevance: 5/10

Core Problem: Abandoned-mine restoration trajectories depend on topography and treatment stage, but temporal landscape patterns are often measured too sparsely.

Key Innovation: Eleven UAV multispectral time points track vegetation recovery and landscape-pattern evolution under contrasting restoration treatments.

50. Groundwater flow estimation from energy pile thermal responses: A case study from laboratory experiments to field testing

Source: Engineering Geology Type: Hydrogeological Analysis Geohazard Type: Groundwater flow and energy piles Relevance: 5/10

Core Problem: Groundwater flow alters energy-pile thermal response, which can bias design and also offer an indirect hydrogeological signal.

Key Innovation: Laboratory experiments and field testing estimate groundwater flow from thermal-response behavior of energy piles.

51. Behaviors of Sediment Particles During Erosion Driven by Turbulent Wave Action

Source: GeoHazards (MDPI) Type: Concepts & Mechanisms Geohazard Type: Wave-driven sediment erosion Relevance: 5/10

Core Problem: Sediment-particle erosion under turbulent wave action affects coastal and riverbed stability but remains difficult to observe at process scale.

Key Innovation: The study analyzes sediment-particle behavior during wave-driven erosion to clarify transport and detachment mechanisms.

52. A Fine-Scale Urban Impervious Surface Extraction Method Based on UAV LiDAR and Visible Imagery

Source: Remote Sensing (MDPI) Type: AI / Remote Sensing Geohazard Type: Urban impervious surfaces Relevance: 5/10

Core Problem: Fine-scale impervious-surface extraction affects runoff and urban-flood modelling, but dense urban scenes mix roofs, roads, vegetation, and shadows.

Key Innovation: UAV LiDAR and visible imagery are combined with machine learning and SHAP interpretation to improve fine-scale impervious-surface mapping.

53. An Improved Multiple-Component Decomposition Method of Polarimetric SAR Interferometry Using Refined Volume Scattering Models

Source: Remote Sensing (MDPI) Type: AI / Remote Sensing Geohazard Type: Polarimetric SAR interferometry Relevance: 5/10

Core Problem: PolInSAR decomposition can misattribute volume scattering in complex terrain and vegetation, weakening downstream geophysical interpretation.

Key Innovation: An improved multiple-component decomposition method refines volume scattering models for polarimetric SAR interferometric analysis.

54. Recurrent Climate-Driven Dieback of Subalpine Grasslands in Central Europe Detected from Multi-Decadal Landsat and Sentinel-2 Time Series

Source: Remote Sensing (MDPI) Type: Detection and Monitoring Geohazard Type: Climate-driven vegetation dieback Relevance: 5/10

Core Problem: Subalpine grassland dieback can signal hydroclimatic stress, but recurrence and spatial extent require long satellite records.

Key Innovation: Multi-decadal Landsat and Sentinel-2 time series detect recurrent climate-driven dieback in Central European subalpine grasslands.

55. How long do trees resist prolonged drought? Insights from semiarid regions

Source: Science of Remote Sensing Type: Ecohydrological Analysis Geohazard Type: Drought stress Relevance: 5/10

Core Problem: Tree survival during prolonged drought is uncertain in semiarid regions where resistance time controls ecosystem degradation risk.

Key Innovation: Remote-sensing and ecological analysis estimate how long trees resist prolonged drought across semiarid environments.

56. Revealing the spatiotemporal evolution of landscape ecological resilience in China's Yellow River Basin using remote sensing and explainable AI

Source: Catena Type: AI / Remote Sensing Geohazard Type: Landscape ecological resilience Relevance: 5/10

Core Problem: Regional ecological resilience assessments need interpretable links between remote-sensing indicators, landscape structure, and environmental stress.

Key Innovation: Remote sensing and explainable AI reveal spatiotemporal evolution of landscape ecological resilience in China's Yellow River Basin.

57. Bayesian system identification and damage estimation using nonlinear shear-flexural model and stochastic filtering

Source: Soil Dyn. & Earthquake Eng. Type: Risk Assessment Geohazard Type: Structural seismic damage Relevance: 5/10

Core Problem: Nonlinear structural damage estimation under earthquake loading must infer hidden states from noisy response records.

Key Innovation: Bayesian system identification with stochastic filtering estimates damage using a nonlinear shear-flexural structural model.

58. Seismic fluid-structure interaction in aqueducts: Interface tracking and unsteady mesh motion

Source: Soil Dyn. & Earthquake Eng. Type: Hazard Modelling Geohazard Type: Seismic aqueduct response Relevance: 5/10

Core Problem: Aqueduct seismic analysis requires fluid-structure interaction methods that track moving interfaces without excessive mesh distortion.

Key Innovation: Interface tracking and unsteady mesh motion are used to model seismic fluid-structure interaction in aqueduct systems.

59. Thermo-mechanical coupled SPH modeling for progressive failure of tunnel surrounding rock based on invariance of the spherical stress tensor

Source: TUST Type: Hazard Modelling Geohazard Type: Tunnel surrounding-rock failure Relevance: 5/10

Core Problem: Progressive tunnel surrounding-rock failure under thermal and mechanical coupling is difficult to capture with mesh-based methods during large deformation.

Key Innovation: A thermo-mechanical coupled SPH model based on spherical-stress-tensor invariance simulates progressive rock failure around tunnels.

60. Blast-induced dynamic response and vibration control of double-arch tunnels without the center wall

Source: TUST Type: Mitigation Geohazard Type: Tunnel blasting vibration Relevance: 5/10

Core Problem: Double-arch tunnels without a center wall are sensitive to blast-induced vibration and dynamic response during construction.

Key Innovation: Dynamic-response analysis and vibration-control evaluation clarify how blasting affects center-wall-free double-arch tunnel stability.

61. Distributed fiber-optic monitoring of internal temperature and strain of rock support system subject to fire

Source: TUST Type: Detection and Monitoring Geohazard Type: Tunnel fire and rock support Relevance: 5/10

Core Problem: Rock-support systems exposed to fire can lose integrity internally before surface inspection detects damage.

Key Innovation: Distributed fiber-optic sensing monitors internal temperature and strain in rock supports under fire loading.

62. The advanced displacement-confinement method for design of mechanized shield tunnels in soft ground

Source: TUST Type: Hazard Modelling Geohazard Type: Shield tunnel deformation Relevance: 5/10

Core Problem: Mechanized shield tunnels in soft ground require design methods that connect displacement confinement to ground response and lining demand.

Key Innovation: An advanced displacement-confinement method improves shield-tunnel design by representing soft-ground deformation constraints.

63. Seismic vertical uplift capacity of horizontal plate anchors accounting for excess pore water pressure generation during an earthquake

Source: Computers and Geotechnics Type: Hazard Modelling Geohazard Type: Seismic anchor uplift Relevance: 5/10

Core Problem: Plate-anchor uplift capacity during earthquakes changes as excess pore water pressure develops in saturated ground.

Key Innovation: The study accounts for earthquake-induced excess pore pressure when estimating seismic vertical uplift capacity of horizontal plate anchors.

64. Investigation on pipeline-backfill-trench lateral interaction in clay seabed

Source: Soil Dyn. & Earthquake Eng. Type: Risk Assessment Geohazard Type: Subsea pipeline geohazards Relevance: 5/10

Core Problem: Subsea pipelines can be threatened by lateral interaction among pipeline, backfill, trench, and clay seabed during ground movement.

Key Innovation: Experimental and analytical investigation quantifies pipeline-backfill-trench lateral interaction in clay seabed conditions.

65. Seismic assessment of a restored historical masonry minaret after the Kahramanmaraş, Türkiye earthquakes (MW 7.8; MW 7.6) and numerical investigation of future collapse potential

Source: Soil Dyn. & Earthquake Eng. Type: Risk Assessment Geohazard Type: Masonry seismic collapse Relevance: 4/10

Core Problem: Restored masonry minarets damaged by major earthquakes need post-event assessment of residual capacity and future collapse potential.

Key Innovation: Numerical seismic assessment of a restored historical minaret after the Kahramanmaras earthquakes evaluates likely future collapse mechanisms.

66. Women's lived experiences in post-disaster temporary housing: A comparative study from Türkiye

Source: IJDRR Type: Disaster Recovery Geohazard Type: Post-disaster temporary housing Relevance: 4/10

Core Problem: Temporary housing after earthquakes and other disasters can shape long-term recovery outcomes but is often evaluated without lived-experience evidence.

Key Innovation: Comparative qualitative evidence from Turkiye documents women's experiences in post-disaster temporary housing and highlights recovery constraints.

67. A Narrative Review of Food, Energy, and Water Systems for Disaster Resilient Communities in Puerto Rico

Source: IJDRR Type: Review Geohazard Type: Disaster-resilient communities Relevance: 4/10

Core Problem: Food, energy, and water systems are interdependent during disasters, yet community resilience planning often treats them separately.

Key Innovation: A Puerto Rico-focused narrative review synthesizes food-energy-water system vulnerabilities and resilience options for disaster-prone communities.

68. Effect of intermediate cyclic stress on dynamic characteristics of landfill waste using cyclic true triaxial test

Source: Soil Dyn. & Earthquake Eng. Type: Concepts & Mechanisms Geohazard Type: Landfill seismic response Relevance: 4/10

Core Problem: Landfill waste undergoes complex dynamic property changes under cyclic stress, affecting seismic deformation and stability assessment.

Key Innovation: Cyclic true-triaxial testing quantifies how intermediate cyclic stress changes the dynamic characteristics of landfill waste.

69. 1:1 and 2:1 clay mixtures under controlled hydration: Experiment and molecular dynamics analysis

Source: JRMGE Type: Concepts & Mechanisms Geohazard Type: Hydrated clay behavior Relevance: 4/10

Core Problem: Clay hydration controls swelling, weakening, and interface behavior relevant to engineered ground and barrier materials.

Key Innovation: Experiments and molecular dynamics simulations compare 1:1 and 2:1 clay mixtures under controlled hydration conditions.

70. Experimental and theoretical research on the impact resistance performance of thin-walled viscoelastic vibration isolation devices

Source: Soil Dyn. & Earthquake Eng. Type: Mitigation Geohazard Type: Impact isolation Relevance: 4/10

Core Problem: Thin-walled viscoelastic isolation devices need quantified impact resistance before they can be used for protective civil infrastructure.

Key Innovation: Experimental and theoretical analysis evaluates impact-resistance performance for thin-walled viscoelastic vibration-isolation devices.