TerraMosaic Daily Digest: May 7, 2026
Daily Summary
May 7's papers are anchored by mechanics, routing, and reliability. An implicit stabilized SPFEM remains stable for large-deformation geotechnical dynamics and reproduces the Shenzhen 12.20 landslide runout; other slope-failure papers move toward rainfall-conditioned stability databases, cracked pile-stabilized slopes under groundwater fluctuation, and long-timescale reconstruction of a Jinsha River landslide-dammed lake and breach flood. Postfire debris-flow prediction, GLOF-oriented lake typing in the Parlung Zangbo Basin, and heavy-rainfall sediment export from deglaciating Alpine catchments extend the same theme: hazard magnitude is being constrained through event volumes, routing geometry, and process-specific hydrometeorological forcing rather than by hazard occurrence alone.
The day also shows a sharp split between fast surrogate modelling and reliability-based engineering. A graph neural network produces probabilistic dike-breach flood maps while screening physical plausibility with mass conservation; a Neural Process densifies noisy station rainfall with radar context; InSAR is evaluated under backfill-mining subsidence conditions; and coastal change detection shifts from wave-height thresholds to longitudinal wave power. Engineering papers treat uncertainty as a design variable across liquefied tailings strength, coir-reinforced liquefaction resistance, offshore monopile pile running, cement-improved monopiles under wind-wave-seismic loading, aging jacket seismic reliability, tunnel lining defects, structural-plane-controlled tunnel instability, and clay-liner contaminant transport.
Key Trends
The central methodological movement is from mapping where hazards may occur to modelling how much deformation, flow, or loss a system can sustain.
- Large deformation and runout modelling are becoming numerically more mature: the implicit SPFEM paper, dense granular-flow rheology, shield-chamber CFD-DEM, muck segmentation, and tunnel surrounding-rock experiments all focus on post-yield motion, fragmentation, and pressure transfer rather than only pre-failure stability.
- Hydrometeorological triggers are being converted into volume and threshold estimates: rainfall-conditioned slope stability, postfire debris-flow volume, glacial-lake classification, Alpine sediment flux, glacierized-catchment discharge downscaling, and California hydroclimate emergence papers quantify forcing intensity, timing, and storage response.
- Fast hazard surrogates are being constrained by physics or uncertainty checks: the dike-breach GNN uses mass conservation to flag plausible simulations, DropsToGrid returns calibrated rainfall uncertainty, and InSAR mining-subsidence monitoring is tested against leveling rather than treated as a black-box displacement product.
- Geotechnical reliability is moving from deterministic safety factors toward spatial variability and degradation: pile running, high-overburden liquefaction, coir-reinforced sands, aging offshore jackets, MSE walls, compacted clay liners, and offshore monopiles all expose how uncertainty, corrosion, soil fabric, or coupled loading changes design margins.
- Hazard histories are being used as quantitative boundary conditions: the Jinsha River paleolake, Glacier Peak tree-ring anomaly, Dutch dwelling mounds, Andaman-Nicobar-Sumatra prism segmentation, Los Angeles retrofit records, and Western Andean thrust evolution convert archival or geologic memory into constraints on future exposure.
Selected Papers
This issue contains 51 selected papers from 1,907 papers analyzed. The leading papers resolve implicit large-deformation landslide simulation, rainfall-conditioned slope instability, Jinsha River landslide-dammed lake evolution, postfire debris-flow volume prediction, cracked pile-stabilized slope stability, probabilistic dike-breach flood mapping, glacial lake type identification for GLOF assessment, InSAR mining-subsidence monitoring, and coastal morphological change detection. The broader set connects liquefaction and tailings strength, offshore foundation reliability, seismic retrofit history, compound cyclone and hurricane losses, heavy-rainfall sediment export, glacierized-catchment discharge, volcanic chronology, tunnel lining defects, tunnel rock-mass instability, underground cavern reliability, TBM muck interpretation, compacted clay liner transport, and hazard-history reconstruction.
1. An implicit stabilized smoothed particle finite element method for dynamic large-deformation analysis in geomechanics
Core Problem: Large-deformation slope failures require stable numerical schemes that can handle quasi-static loading, low-frequency dynamics, and runout without explicit-time-step limitations.
Key Innovation: An implicit stabilized smoothed particle finite element method embeds nodal integration stabilization in a Newmark framework and reproduces elasto-plastic slope failure, column collapse, and the Shenzhen 12.20 landslide runout.
2. Rainfall instability prediction for three types of landslide-prone slopes using various machine learning algorithms
Core Problem: Rapid stability prediction for landslide-prone slopes must incorporate rainfall duration and intensity together with geometry and soil mechanics.
Key Innovation: LHS sampling, SEEP/W, SLOPE/W, and eight machine-learning models generate a rainfall-aware stability database for homogeneous, accumulation-layer, and coal-measure strata slopes, with XGBoost giving the strongest transfer to real cases.
3. Chronology and evolution process of the Zhaizicun ancient landslide-dammed lake in the middle reaches of the Jinsha River, southwest China
Core Problem: The timing, storage, and outburst dynamics of large paleo landslide-dammed lakes remain poorly constrained in tectonically active river corridors.
Key Innovation: OSL dating, DEM-based paleolake reconstruction, DB-IWHR breach modelling, and HEC-RAS hydraulics reconstruct the Zhaizicun dammed lake and a high-energy outburst flood of about 109,702 m3/s.
4. An improved empirical model for predicting postfire debris-flow volume in the western United States
Core Problem: Postfire debris-flow volume models often perform well only in the regions where they were calibrated.
Key Innovation: A 227-event western U.S. database and 36 rainfall, terrain, and fire predictors yield an empirical volume model that matches regional models and improves performance in data-limited areas.
5. Stability analysis of pile-stabilized slopes with cracks under groundwater level fluctuations
Core Problem: Pile-reinforced slopes with vertical cracks respond nonlinearly to groundwater change, surcharge, and suction-dependent shear strength.
Key Innovation: A Fredlund-Xing nonlinear shear-strength model is coupled to a log-spiral failure mechanism and work-energy solution to estimate anti-sliding force for pre-existing and newly formed cracks.
6. Probabilistic flood hazard mapping for dike-breach floods via graph neural networks
Core Problem: Probabilistic dike-breach mapping is computationally expensive and difficult to validate when thousands of scenarios are needed.
Key Innovation: A multi-scale shallow-water-equation graph neural network represents hydraulic structures, runs about 10,000 times faster than Delft3D, and uses mass-conservation error to flag physically plausible simulations.
7. A Novel Approach for Automatic Glacial Lake Type Identification in the Parlung Zangbo Basin via Glacially Fed Lake Subdivision
Core Problem: Large-scale GLOF assessment needs automated lake typing that distinguishes glacier contact, glacier-fed routing, and dam-relevant spatial relationships.
Key Innovation: A rule-based classification scheme combines glacier centerlines, retreat zones, and meltwater flow paths to classify 1,429 lakes in the Parlung Zangbo Basin.
8. Applicability and Feasibility of InSAR-Based Mining Subsidence Monitoring Under Overburden Isolated Grouting Backfill Mining Conditions
Core Problem: Large-gradient caving subsidence degrades InSAR coherence, while backfill mining changes deformation magnitude and detectability.
Key Innovation: Sentinel-1 SBAS-InSAR and leveling compare caving and overburden isolated grouting backfill mines, showing substantially lower RMSE and more reliable deformation monitoring under backfill conditions.
9. Longitudinal wave power as a proxy for coastal change detection
Core Problem: Wave-height thresholds can miss directional forcing that controls coastal morphological response near inlets and river mouths.
Key Innovation: Longitudinal wave power and optimized peak-over-threshold criteria identify erosional and accretional events more consistently than wave height across Delft3D experiments and two Iberian field sites.
10. Liquefied strength from CPT under high overburden stress: a nonlinear critical state approach
Core Problem: Liquefied strength estimates rely on case histories dominated by low confining stresses, but modern tailings facilities can operate under much higher overburden.
Key Innovation: A nonlinear critical-state extension for CPT-based liquefied undrained strength incorporates confining-stress effects and provides risk-informed parameter guidance.
11. Two potential seismic failure mechanisms for 2×2 rock-socketed pile groups in liquefiable soil: Centrifuge shaking table experiments
Core Problem: Pile foundations in liquefiable sites can fail through coupled horizontal and vertical load effects rather than a single dominant mechanism.
Key Innovation: Centrifuge shaking-table tests on 2x2 rock-socketed pile groups reproduce two-hinge and three-hinge seismic failure modes controlled by lateral spreading, inertia, and vertical-load buckling.
12. Mitigation of Liquefaction Using Natural Coir Reinforcement: Insights from Simple Shear and Shaking Table Model Tests
Core Problem: Liquefiable sands need mitigation strategies that improve cyclic resistance and post-liquefaction strength without relying only on synthetic reinforcement.
Key Innovation: Simple shear tests and shaking-table models show that coir fibres increase cyclic resistance by up to 30% and improve pre- and post-liquefaction shear strength through tensile confinement.
13. From Drops to Grid: Noise-Aware Spatio-Temporal Neural Process for Rainfall Estimation
Core Problem: Rainfall-triggered hazards need dense, uncertainty-aware precipitation fields, but sparse stations and radar context are hard to fuse robustly.
Key Innovation: DropsToGrid uses a Neural Process with multi-scale spatial features, temporal attention, radar context, and noisy private weather stations to generate calibrated stochastic rainfall fields across regions.
14. Andaman-Nicobar-Sumatra subduction zone: Morphological and mechanical segmentation from high-resolution multibeam bathymetry
Core Problem: Along-strike forearc structure modulates rupture behaviour, but the Andaman-Nicobar-Sumatra prism lacked a contiguous high-resolution morphotectonic synthesis.
Key Innovation: Multibeam bathymetry and seismic reflection data map four accretionary-prism domains across the 2004 rupture zone, linking wedge geometry, frontal-thrust vergence, and basal coupling to rupture propagation.
15. Dynamic response of monopile offshore wind turbines in cement-improved clay under seismic and environmental loads
Core Problem: Offshore wind monopiles in soft clay face coupled wind-wave-seismic demands that are not captured by single-hazard foundation checks.
Key Innovation: 3-D finite-element simulations evaluate cement-improved clay reinforcement under environmental and seismic loads, quantifying reductions in tower deflection, settlement, and rotation.
16. Thirty years of earthquake retrofit in the city of Los Angeles: Trends, drivers, and implications for seismic resilience
Core Problem: Retrofit uptake for vulnerable building types can diverge from risk geography and social vulnerability, weakening city-scale seismic resilience.
Key Innovation: More than 29,000 permits, census data, spatial clustering, and a Puente Hills earthquake scenario quantify retrofit trends and loss-reduction benefits for cripple-wall and soft-story buildings.
17. Reliability analysis of pile running for offshore large-diameter monopiles considering soil spatial variability
Core Problem: Pile running during large-diameter monopile installation can be underestimated when seabed soil variability is simplified.
Key Innovation: Coupled Eulerian-Lagrangian modelling, random fields, and Monte Carlo simulation link soil spatial variability to penetration velocity, failure probability, and design safety factor.
18. Flood volume allocation method for flood hazard mapping using river model with levee scheme
Core Problem: Global river flood models often omit levee protection and therefore misallocate inundation volume across protected floodplains.
Key Innovation: CaMa-Flood is extended with levee heights and fractions, frequency-analysis flood volumes, and high-resolution downscaling constrained by terrain storage.
19. Combining hazard, exposure and vulnerability data to predict historical United States hurricane losses
Core Problem: Wind-speed-only hurricane scales poorly explain historical U.S. losses.
Key Innovation: A statistical loss model combines optimally weighted hazard, exposure, and vulnerability variables and reduces RMSE against observed losses relative to wind-only predictors.
20. The relationships between extreme winter North Atlantic extratropical cyclone hazards and modes of seasonal climate variability
Core Problem: Winter North Atlantic cyclone hazards co-vary across wind, waves, and precipitation, but their relationship to climate modes is uneven spatially.
Key Innovation: NAO and East Atlantic Pattern modes are linked to compound seasonal maxima in wind gusts, wave heights, daily precipitation, and winter precipitation across Europe.
21. Linking heavy rainfall to suspended sediment fluxes in a deglaciating Alpine catchment
Core Problem: Deglaciating catchments may experience more intense rainfall while losing glacier-driven sediment supply.
Key Innovation: A 21-year hourly precipitation and sediment record separates sub-daily convective events from long-duration storms, showing increasingly flashier sediment transport regimes.
22. Downscaling Daily Discharge to Sub‐Daily Scales for Alpine Glacierized Catchments
Core Problem: Daily discharge obscures diel meltwater cycles that matter for alpine hydropower, ecosystems, and flood risk.
Key Innovation: A maximum-entropy downscaling framework reconstructs sub-daily discharge from daily flows using 45 years of 15-minute observations in glacier-fed Swiss Alpine catchments.
23. Unprecedented Shifts in Hydrology Are Emerging Across California's Critical Basins: An Evaluation From 0.5 to 3.5°C
Core Problem: Water-resource and flood-risk planning needs warming thresholds rather than scenario-only projections.
Key Innovation: A 19-model downscaled ensemble identifies warming levels at which snowfall, peak snowpack, runoff timing, and 5- to 100-year runoff intensities emerge from historical variability.
24. Late 17th-century cooling potentially linked to the most recent eruption of Glacier Peak (Washington, USA) recorded in tree rings
Core Problem: The most recent Glacier Peak eruption remains difficult to date because historical and tephrochronological constraints are sparse.
Key Innovation: Latewood blue-intensity tree-ring reconstruction, eruption-response benchmarking, Indigenous accounts, and tephra evidence identify a localized 1696 cooling anomaly consistent with recent volcanic influence.
25. Non-stationary dynamics of compound climate extremes: a WRF-CMIP6-GAMLSS framework for southeastern China
Core Problem: Compound climate extremes are non-stationary under climate change and require flexible joint modelling.
Key Innovation: A WRF-CMIP6-GAMLSS framework models changing distributions of compound extremes for southeastern China, supporting time-varying hazard assessment.
26. Failure evolution and mechanical response of permeable ribbed double-arch tunnels with insufficient secondary lining thickness
Core Problem: Local lining-thickness defects can trigger load redistribution and progressive instability in permeable ribbed double-arch tunnels.
Key Innovation: Scaled model tests and numerical simulations quantify stiffness degradation, rib-arch failure evolution, rock plastic deformation, and support stress amplification under defect scenarios.
27. Instability damage mechanism of tunnel surrounding rock with different locations and inclination angles of structural plane
Core Problem: Tunnel instability depends on the position and inclination of structural planes, but their failure-stage mechanics remain hard to generalize.
Key Innovation: Orthogonal model tests, biaxial loading, high-speed imaging, and DIC resolve four damage modes and three instability stages for structural-plane-controlled tunnel surrounding rock.
28. Soil-structure interaction of metro tunnel in layered strata: mechanical mechanism, parameter analysis and field monitoring
Core Problem: Metro tunnel behaviour in layered strata depends on coupled soil-structure interaction that simplified design models often miss.
Key Innovation: Mechanical modelling, parameter analysis, and field monitoring are combined to evaluate layered-ground controls on tunnel response.
29. Mechanical behavior and crack evolution mechanism of layered cracked rock samples reinforced by grouting–anchor bolt
Core Problem: Crack defects and bedding interfaces weaken rock masses and complicate reinforcement design.
Key Innovation: Uniaxial compression, acoustic emission, high-speed imaging, and grouting-bolt comparison show how synergistic reinforcement delays crack coalescence and redistributes energy dissipation.
30. Consolidation performance of oil sands tailings: data-driven insights and practical implications for tailings landforms closure and design
Core Problem: Tailings landform closure depends on uncertain compressibility and hydraulic conductivity measured by different laboratory methods.
Key Innovation: Five consolidation testing methods are compared with error propagation and hypothetical design scenarios to quantify how method choice affects long-term geotechnical predictions.
31. Bayesian-updated seismic reliability of aging jacket offshore platforms under corrosion degradation
Core Problem: Offshore jacket platforms lose seismic capacity as corrosion progresses and inspection data update structural state.
Key Innovation: Bayesian updating is used to revise reliability estimates for aging jacket platforms under corrosion degradation and earthquake loading.
32. Reliability analysis of underground water-sealed storage cavern stability with emphasis on support stress distribution
Core Problem: Underground water-sealed caverns require stability assessment that accounts for support stress distribution and uncertainty.
Key Innovation: A reliability-analysis framework evaluates cavern stability with emphasis on support-stress redistribution in underground storage construction.
33. A constitutive rheological model for dense flows of non-spherical particles incorporating frictional and rearrangement effects
Core Problem: Dense flows of non-spherical particles require rheological laws that represent frictional contacts and particle rearrangement.
Key Innovation: A constitutive model incorporates particle shape, frictional effects, and rearrangement mechanisms to improve dense-flow simulation.
34. An innovative simulation approach based on the extended finite element method for modeling the interaction between hydraulically induced and natural fractures in deep underground engineering
Core Problem: Deep rock engineering must predict how hydraulically induced fractures interact with natural discontinuities.
Key Innovation: An extended finite-element simulation approach models fracture interaction and propagation in deep underground settings.
35. Pre-seismic geomagnetic fusion anomaly extraction based on Spatially Weighted Non-negative Tensor Factorization
Core Problem: Geomagnetic precursor studies require separation of weak, spatially variable pre-seismic anomalies from background variability.
Key Innovation: A spatially weighted non-negative tensor factorization framework fuses geomagnetic information to extract pre-seismic anomaly patterns.
36. MuckSeg: A deep learning-based algorithm for automated in-situ TBM muck morphology analysis
Core Problem: Muck morphology on TBM conveyor belts is difficult to measure automatically because chips overlap and mix with water or mud.
Key Innovation: MuckSeg combines semi-smoothed non-binary supervision, dense instance segmentation, and post-processing to estimate muck size and shape for downstream rock-grade interpretation.
37. Multidimensional evaluation and quantification of tunnel shotcrete roughness using point cloud data
Core Problem: Single roughness indices poorly locate shotcrete surface defects relevant to waterproof membrane installation.
Key Innovation: A multidimensional roughness index uses curvature, convexity-concavity, scale, developability, and machine-learning segmentation on tunnel point clouds.
38. Horizontal displacement of segmental linings in small-radius curved shield tunnels under asymmetric thrust
Core Problem: Small-radius curved shield tunnels can experience asymmetric thrust and horizontal lining displacement.
Key Innovation: A mechanical analysis of segmental linings quantifies displacement response under asymmetric thrust conditions.
39. Flame evolution and thermal characteristics of train fires in high-altitude tunnels under longitudinal ventilation
Core Problem: High-altitude tunnel fires couple carriage geometry, ventilation, and heat-release-rate evolution.
Key Innovation: Experiments and revised thermal modelling characterize flame evolution, heat release, and ceiling temperature distribution under longitudinal ventilation.
40. Numerical Analysis of the Impact of Reinforcement Stiffness on the Evolution of Maximum Tensile Load at Incipient Failure in MSE Walls
Core Problem: MSE wall design methods do not consistently capture how reinforcement stiffness changes tensile load at incipient failure.
Key Innovation: Finite-element backanalysis and parametric design charts relate reinforcement stiffness, toe resistance, and lateral earth pressure coefficients.
41. Coupling Influences of Temperature Change and Large-Strain Consolidation on One-Dimensional Transport of Organic Contaminants in a Compacted Clay Liner
Core Problem: Temperature and large-strain consolidation jointly alter contaminant breakthrough in compacted clay liners.
Key Innovation: A coupled one-dimensional transport model evaluates diffusion, hydraulic conductivity, sorption, consolidation, and liner service life under leachate temperature changes.
42. Multi-scale reinforcement mechanisms and fatigue behavior of loess subgrade grouted with permeable polyurethane polymer
Core Problem: Loess subgrades require reinforcement strategies that improve fatigue resistance and multi-scale mechanical performance.
Key Innovation: Permeable polyurethane polymer grouting is evaluated from microstructure to fatigue behaviour for reinforced loess subgrade.
43. CFD-DEM simulations on the flow mechanisms and pressure transfer characteristics of foam-conditioned sands in a shield chamber
Core Problem: Shield tunnelling performance depends on how conditioned sand flows and transfers pressure inside the chamber.
Key Innovation: CFD-DEM simulations resolve flow mechanisms and pressure-transfer characteristics of foam-conditioned sands.
44. Lateral static response analysis of horizontally loaded large-diameter monopiles using an improved T-P model
Core Problem: Large-diameter monopiles require lateral response models that remain reliable across offshore design conditions.
Key Innovation: An improved T-P model is developed for lateral static response analysis of horizontally loaded large-diameter monopiles.
45. Depression development in the Gonghe Basin, northeastern Qinghai-Tibetan Plateau: An approach to estimating dust emission during the last glacial
Core Problem: Dust-emission records in eroding source areas are often poorly preserved.
Key Innovation: Terrace mapping, OSL dating, and deflation-rate estimates reconstruct Gonghe Basin depression development and last-glacial dust flux.
46. Artificial dwelling mounds in Dutch wetlands: Regional archives of human-landscape interactions (500 BCE – Present)
Core Problem: Artificial dwelling mounds are fragile archives of long-term flood adaptation and wetland landscape engineering.
Key Innovation: A national-scale inventory of almost 9,000 Dutch dwelling mounds links mound construction to storm surge, fluvial flooding, shoreline shifts, and subsidence.
47. Erosion depth and seasonality jointly regulate soil nitrogen transformations during early vegetation restoration in dry-hot valleys
Core Problem: Vegetation restoration on eroding dry-hot valley slopes depends on how erosion depth and seasonality regulate nitrogen transformations.
Key Innovation: Field evidence links erosion depth and seasonal dynamics to soil nitrogen transformation during early vegetation restoration.
48. Microbial adaptation to coupled organic carbon transport and hydrodynamics in raindrop-detached versus sheet-eroded sediments
Core Problem: Raindrop detachment and sheet erosion transport organic carbon and microbial communities differently.
Key Innovation: Rainfall flume experiments, SOC fractionation, sequencing, and hydraulic metrics identify selective sorting controls on microbial diversity in transported sediments.
49. Spatial and temporal evolution of ecosystem services and their interrelationships in the alpine zone: A case study of the source regions of the Yangtze River and the Yellow River
Core Problem: Alpine ecosystem services respond differently across permafrost and seasonally frozen ground as warming and humidification progress.
Key Innovation: InVEST, RUSLE, RWEQ, and frozen-ground modelling quantify 1990-2020 ecosystem-service synergies and trade-offs in the Yangtze and Yellow River source regions.
50. Tectonic evolution of the western slope of the Central Andes (36°S) and the building of the Western Andean Thrust System
Core Problem: Active fore-arc fault systems can inherit geometry from older extensional basins, affecting seismic structure and uplift.
Key Innovation: Structural geology and U-Pb geochronology reconstruct inversion, shortening, uplift pulses, and Quaternary deformation along the Western Andean Thrust System.
51. Quantifying bound water conversion in Na-montmorillonite through molecular dynamics simulations
Core Problem: Clay-rich soils in energy and waste systems change water structure and volume under elevated temperature.
Key Innovation: Molecular dynamics simulations quantify d-spacing, interlayer water density, and bound-water conversion in Na-montmorillonite across hydration states and temperatures.