TerraMosaic Daily Digest: June 2, 2026
Daily Summary
The June 2 papers treat landslide hazard as a process chain rather than a mapping endpoint. The strongest contributions move beyond event detection toward complete hazard chains, probabilistic time-of-failure estimation, reactivation mechanics, dynamic early warning, and loss accounting. A Cyclone Idai study shows why landslide inventories must include the downstream debris-rich flood continuum; the Landslides papers quantify uncertainty in failure timing and explain rainfall-excavation reactivation of an old landslide; and the slip-zone soil work links residual weakening to a normal-stress threshold that is directly relevant to deep-seated reactivation.
The broader geohazard set is organized by coupling rather than by hazard labels. Large-earthquake scenarios in Türkiye and a virtual co-seismic ground-instability framework connect shaking to landslides, liquefaction, and flood impacts. Climate forcing appears through slush-flow emergence in the Italian Alps and thermomechanical weakening of the Marmolada glacier collapse. Infrastructure papers extend the same logic to railway subsidence, tunnel settlement and spalling, mining overburden deformation, rockburst monitoring, hydraulic-fracturing fault activation, and urban flood hydraulics. The methodological center is observability: optical imagery, PS-InSAR, DS-InSAR, electrical resistivity tomography, distributed fiber sensing, microseismic tomography, and physics-guided machine learning are used to infer evolving terrain or infrastructure state.
Key Trends
Five movements define the issue: consequence-aware landslide inventories, probabilistic early warning, triggering-chain multi-hazard models, cryosphere mass-movement mechanics, and dense observation of infrastructure state.
- Landslide products are being tied to consequences: The Idai continuum paper and the inventory-to-loss reporting study both argue that source-area landslide maps are incomplete unless downstream sediment transfer, debris-rich flooding, exposure, and loss records are linked in one workflow.
- Early warning is becoming explicitly probabilistic: PS-InSAR cross-correlation with Bayesian failure-time prediction, frequency-enhanced displacement warning, and monitoring-informed case studies all treat warning as an uncertainty problem rather than a fixed threshold problem.
- Multi-hazard modelling is converging around triggering chains: The Türkiye earthquake assessment, co-seismic ground-instability tool-chains, arid earthquake-triggered landslide mapping, and landslide-risk time series frame landslides, liquefaction, flooding, and exposure as interacting scenarios.
- Cryosphere hazards are entering mechanical stability models: The slush-flow and Marmolada glacier-collapse studies show that warming, liquid-water content, fracture networks, basal water pressure, and ice strength now need to be modeled as direct controls on rapid mass movement.
- Infrastructure geohazards are adopting dense state observation: DS-InSAR railway subsidence, tunnel settlement prediction, fiber-optic overburden monitoring, mine microseismic tomography, and rockburst acoustic-emission learning all convert distributed sensing into deformation or failure-state estimates.
Selected Papers
This issue contains 44 selected papers from 1,787 papers analyzed. The selected papers are ordered by relevance score and focus on landslide process chains, probabilistic warning, earthquake-triggered ground instability, cryosphere mass movements, debris-flow monitoring, infrastructure deformation, and observation methods that make terrain-state change measurable.
1. Capturing the complete landslide–debris-rich flood continuum for accurate inventory, susceptibility and exposure mapping – lessons from Cyclone Idai
Core Problem: Rainfall-triggered landslide disasters often cause their largest impacts downstream, where sediment-rich flood processes are missed by source-area inventories and susceptibility models.
Key Innovation: Uses Cyclone Idai to argue for an integrated landslide-to-debris-rich-flood inventory, susceptibility, and exposure framework that follows sediment transfer from hillslopes to communities.
2. Probabilistic landslide time-of-failure prediction using Persistent Scatterer Interferometric Synthetic Aperture Radar (PS-InSAR), time series cross-correlation, and Bayesian inference
Core Problem: Progressive landslide rupture is difficult to forecast from noisy satellite displacement records, and conventional time-of-failure methods rarely provide operational confidence intervals.
Key Innovation: Combines PS-InSAR reliability screening, time-series cross-correlation, and Bayesian inference to estimate failure time with uncertainty bounds for early warning.
3. Multi-hazards assessment in Türkiye following large earthquakes
Core Problem: The 2023 Kahramanmaraş earthquakes triggered landslides, liquefaction, infrastructure damage, and flooding, but regional scenario tools still tend to evaluate these hazards separately.
Key Innovation: Builds a 50-year scenario-based framework that couples deterministic earthquake ruptures with landslide, liquefaction, flood, and exposure modelling for southeastern Türkiye.
4. Toward an Automatic Pixel-Based Detection of Earthquake-Triggered Landslides in Arid Environments Using Optical Imagery
Core Problem: Automatic landslide mapping is less reliable in arid terrain because failed and stable slopes can have weak spectral contrast in optical imagery.
Key Innovation: Tests pixel-based optical-image detection strategies for rapid earthquake-triggered landslide mapping in arid environments where conventional vegetation-based cues are limited.
5. Deformation mechanism of landslide reactivation triggered by combined effects of rainfall and excavation: a case study of the Dawafang landslide in Chongqing
Core Problem: Old landslide zones can reactivate under combined rainfall infiltration and excavation, but separating shallow seasonal motion from deeper deformation remains difficult.
Key Innovation: Integrates boreholes, displacement monitoring, InSAR, and numerical simulation to explain rainfall-excavation reactivation and deformation partitioning in the Dawafang landslide.
6. Dynamic landslide early warning based on frequency-domain enhanced deep learning
Core Problem: Warning models often miss nonlinear interactions between displacement trends, seasonal cycles, rainfall, reservoir level, and geological controls.
Key Innovation: Introduces a frequency-enhanced deep learning model that fuses environmental forcing, geological attributes, Fourier-domain periodicity, and MCMC displacement sampling for dynamic warning.
7. Geotechnical hazard to power facilities from the Hongyanzi landslide in Hanyuan County, China: an integrated InSAR monitoring and seepage simulation study
Core Problem: Steep, vegetated reservoir slopes near power facilities are hard to monitor with InSAR alone, and the hydrological controls on motion remain uncertain.
Key Innovation: Combines InSAR deformation interpretation with seepage simulation to diagnose landslide activity and infrastructure hazard at the Hongyanzi slope.
8. Normal stress-dependent frictional weakening mechanism of landslide slip-zone soils revealed by the meso-morphology of shear surface
Core Problem: Residual strength in ancient landslide slip zones changes with normal stress, but the meso-scale shear-surface mechanism is poorly constrained.
Key Innovation: Uses ring shear tests, 3D laser scanning, and GIS roughness analysis to identify a normal-stress threshold beyond which residual friction stabilizes at low values.
9. Climate Warming and Ice Weakening Trigger Alpine Glacier Collapses: The Marmolada Case
Core Problem: Rapid retreat of temperate alpine glaciers increases collapse risk, but linking warming, fracture damage, water pressure, and ice strength remains challenging.
Key Innovation: Develops a three-dimensional thermomechanical stability model for the Marmolada collapse, showing synergistic weakening by englacial warming, fractures, and basal water pressure.
10. A multi-hazard framework to assess co-seismic ground instabilities scenarios: application to a virtual test bed
Core Problem: Regional co-seismic landslide and liquefaction scenarios require physical realism without the data and computational burden of full process models everywhere.
Key Innovation: Creates modular tool-chains that sequence susceptibility, preparatory processes, and transient triggering to generate dynamic ground-instability scenarios.
11. The Occurrence of Widespread Slush Flow Events as an Indicator of Accelerating Climate Change in the Northwestern Italian Alps
Core Problem: Slush flows are rapid water-saturated snow and debris movements whose recent emergence in the northwestern Italian Alps suggests changing cryospheric hazard conditions.
Key Innovation: Documents widespread slush-flow occurrence and interprets it as a climate-sensitive mass-movement signal linked to liquid-water thresholds in snowpacks.
12. Spatiotemporal assessment of landslide risk over large areas: a case study of the Valencian Community (1950–2021)
Core Problem: Landslide risk evolves with land-use decisions, residential expansion, and changing hazard recurrence, yet large-area risk histories remain sparse.
Key Innovation: Reconstructs landslide risk over seven decades in the Valencian Community to connect susceptibility, exposed buildings, and urban expansion through time.
13. How to better link Landslide Inventory Mapping with Loss and Damage Reporting
Core Problem: Landslides are underrepresented in global disaster databases because frequent, locally damaging events often fall below reporting thresholds.
Key Innovation: Proposes tighter linkage between landslide inventory mapping and loss-damage reporting to support multi-hazard early warning and national risk knowledge systems.
14. Redistribution of Basal Forces and Shielding Model for Seismic Signals of Debris Flows Over Loose Thin-Layer Sediments
Core Problem: Debris-flow seismic monitoring can underestimate flow dynamics when loose sediment layers shield basal force fluctuations from bedrock sensors.
Key Innovation: Combines flume experiments, discrete-element simulations, and field validation to model how thin sediment layers redistribute basal forces and attenuate debris-flow seismic signals.
15. System reliability analysis of jointed bedding rock slopes under seismic loading: Integrating multiple failure modes
Core Problem: Jointed bedding rock slopes may fail through multiple coupled modes under seismic loading, but single-mode assessments can underestimate instability.
Key Innovation: Builds a Monte Carlo system-reliability framework that integrates translational, rotational, and mixed failure modes for seismic slope assessment.
16. Surface subsidence monitoring and prediction along high-speed railways using DS-InSAR and machine learning: A case study of the Jining section of Lunan High-Speed Railway
Core Problem: Surface deformation along high-speed rail corridors requires millimetric monitoring and forward prediction to protect operational safety.
Key Innovation: Uses DS-InSAR on Sentinel-1A imagery and Random Forest regression to map and forecast subsidence along the Jining section of the Lunan railway.
17. Seasonal ground deformation at subglacial Katla Volcano, Iceland: observations and models
Core Problem: Subglacial volcanic systems deform seasonally under interacting ice load, meltwater, and magmatic or hydrothermal processes.
Key Innovation: Uses observations and modelling to separate seasonal ground-deformation signals at Katla Volcano, improving interpretation of ice-covered volcanic hazards.
18. Can reservoir ductility reduce the likelihood of fault activation during hydraulic fracturing?
Core Problem: Hydraulic fracturing can reactivate faults, but the role of reservoir ductility in buffering pressure and stress transfer remains unresolved.
Key Innovation: Uses artificial-specimen experiments and time-resolved stress mapping to show how ductile reservoirs suppress abrupt Coulomb stress changes near faults.
19. Evolution of multiscale pore structure in saturated silty clay under non-coaxial F-T cycles
Core Problem: Freeze-thaw cycles alter pore structure, cracking, and moisture migration in saturated silty clay, affecting cold-region slope and foundation stability.
Key Innovation: Combines custom freeze-thaw experiments, CT scanning, mercury intrusion porosimetry, and fractal pore metrics to quantify multiscale structural evolution.
20. A gully catchment-scale model for predicting loess cave density: a case study from the Loess Plateau
Core Problem: Loess cave prediction is limited by a scale gap between local high-resolution mapping and broad regional assessments.
Key Innovation: Uses the gully catchment as an intermediate modelling unit to predict loess cave density from hillslope-channel controls on the Loess Plateau.
21. MVMD-TimeXer: a time series prediction model integrating MVMD and endogenous-exogenous attention for highway photovoltaic slope displacement and strain forecasting
Core Problem: Highway photovoltaic slopes produce noisy, nonlinear displacement and strain time series that are difficult to forecast with conventional sequence models.
Key Innovation: Adds multivariate variational mode decomposition to an attention-based forecasting model to denoise and fuse endogenous and exogenous slope-monitoring variables.
22. Prediction of surface settlement of shield-driven twin tunnels using AutoML-regressor chain model
Core Problem: Twin-tunnel excavation can induce coupled surface settlement above each tunnel, requiring multi-output prediction for construction safety.
Key Innovation: Uses AutoML with a regressor-chain structure to predict paired settlement targets from soil, tunnel geometry, and construction parameters.
23. Dynamic Fracture and Mechanical Behaviour of Rock in Tunnels Adjacent to Existing Pipelines
Core Problem: Tunnel construction near operating pipelines can transmit dynamic loads and fracture surrounding rock, threatening existing infrastructure.
Key Innovation: Combines static and dynamic mechanical tests with fracture-toughness analysis to characterize granite and tuff behavior beneath pipeline-adjacent tunnel excavation.
24. A Numerical Study of the Spalling Failure Mechanism of Tunnel Surrounding Rock Under Excavation and Impact Conditions
Core Problem: Excavation unloading and impact loading can jointly drive progressive spalling damage around mine tunnels.
Key Innovation: Applies a continuum-discontinuum method with node separation and fictitious cracks to simulate tunnel-rock cracking under excavation and impact conditions.
25. Distributed optical fiber characterization of mining-induced overburden deformation characteristics: a simulation experimental study
Core Problem: Mining-induced overburden deformation can lead to roof collapse, surface subsidence, and gas anomalies, but large-scale deformation is hard to observe continuously.
Key Innovation: Uses BOTDA distributed optical fiber sensing in a similarity model to identify rock-mass deformation thresholds and overburden failure patterns.
26. Study on the Prediction and Variation Characteristics of Acoustic Emission Signals in the Rockburst Stage Based on an Improved Fish Swarm Optimization Algorithm
Core Problem: Rockburst early warning depends on acoustic-emission signals whose stage-specific patterns are difficult to classify reliably.
Key Innovation: Combines multidimensional acoustic-emission features, improved fish swarm optimization, XGBoost, and SHAP analysis to predict and interpret rockburst stages.
27. Three-Dimensional Microseismic Tomography of Mines Based on Fast Marching Method and Convolutional Neural Networks
Core Problem: Mine tomography needs efficient velocity imaging despite sparse and uneven ray coverage in complex underground geology.
Key Innovation: Combines fast marching travel-time modelling with a U-Net trained on geology-constrained synthetic data for three-dimensional microseismic velocity imaging.
28. CFD-DEM investigation of high-frequency cyclic-strain-driven suffusion and post-erosion cyclic degradation in gap-graded sand
Core Problem: Train vibrations can aggravate seepage-induced suffusion around metro tunnels, weakening gap-graded sands after fine-particle loss.
Key Innovation: Uses coupled CFD-DEM simulations and cyclic triaxial tests to link vibration frequency, strain amplitude, fine loss, and post-erosion cyclic degradation.
29. An analytical model for permeability coefficient of unsaturated loess considering internal pore collapse
Core Problem: Unsaturated loess permeability can change during seepage-induced pore collapse, affecting analysis of loess-region ground disasters.
Key Innovation: Derives an analytical permeability model that incorporates pore-collapse effects through modified soil-water characteristic and pore-volume functions.
30. Influence of Voids Behind Tunnel Lining on the Mechanical Behaviour and Failure of Large Section Small Clearance Tunnels
Core Problem: Voids behind tunnel linings create discontinuities in ground-lining interaction and can concentrate stress or trigger crack propagation.
Key Innovation: Builds a three-dimensional XFEM ground-structure model to quantify how void size, position, and bilateral coupling affect tunnel-lining damage.
31. Analytical solution and initial instability tendency assessment of deep elliptical tunnels in layered rock under arbitrary 3D far-field stresses
Core Problem: Deep elliptical tunnels in layered rock are governed by misaligned bedding, tunnel geometry, and three-dimensional far-field stresses.
Key Innovation: Develops a closed-form analytical solution and local safety-factor assessment for initial instability around deep elliptical tunnel perimeters.
32. Dynamic response and long-term effects of tunnels in expansive mudstone under high-speed train loading
Core Problem: Expansive mudstone can swell and concentrate stress in tunnel linings, while repeated high-speed train loading amplifies long-term deformation risk.
Key Innovation: Constructs a vehicle-track-tunnel dynamic model to evaluate lining response and local surrounding-rock expansion under train-induced cyclic loading.
33. Physics-Informed Machine Learning for Short-Term Flood Prediction
Core Problem: Purely data-driven flood models can violate basic hydrological behavior and perform poorly under data scarcity or extremes.
Key Innovation: Adds hydrological trend-alignment constraints to an LSTM loss function, improving physically plausible short-term hydrograph prediction.
34. An approximate surface-sewer coupling framework for urban flood modeling under data scarcity
Core Problem: Urban flood models need surface-sewer coupling, but detailed sewer network data are often unavailable.
Key Innovation: Represents sewer interaction through road-derived flow routes, stormwater inlet distributions, flood-prone units, and drainage influence zones.
35. Three-Dimensional Characterization of Lateral Subsurface Flow in an Alpine Forested Hillslope on the Tibetan Plateau Using Time-Lapse Electrical Resistivity Tomography
Core Problem: Rainfall-runoff generation on alpine slopes depends on lateral subsurface flow through soil and weathered bedrock, but three-dimensional monitoring remains limited.
Key Innovation: Uses time-lapse 3D electrical resistivity tomography and a morphology-based interpretation framework to quantify lateral subsurface flow pathways.
36. How does the roots architecture affect the runoff-infiltration-erosion processes under varying rainfall events?
Core Problem: Vegetation roots alter rainfall partitioning, infiltration, and erosion, but architecture-specific controls remain difficult to parameterize.
Key Innovation: Evaluates how root-system architecture modifies runoff, infiltration, and erosion responses across rainfall events, informing shallow-slope hydrology.
37. Performance-based assessment of tropical cyclone parametric precipitation models for flood simulation applications
Core Problem: Flood simulations driven by tropical cyclones depend on precipitation models whose performance varies by storm structure and application context.
Key Innovation: Assesses parametric cyclone precipitation models for flood-simulation use, clarifying where simplified rainfall forcing is reliable.
38. Variable contributions of vertical land motion to sea-level change inferred at tide gauges
Core Problem: Relative sea-level change at tide gauges reflects both ocean signals and vertical land motion, complicating coastal hazard attribution.
Key Innovation: Infers spatially variable vertical land motion contributions to sea-level change, improving interpretation of coastal exposure trends.
39. Uncovering Insights of Compound Flooding with Data-Driven AI
Core Problem: Compound flooding arises from interacting hydrological and coastal drivers that are difficult to interpret from observations alone.
Key Innovation: Applies data-driven AI to extract patterns in compound-flooding behavior, offering transferable tools for multi-driver flood-risk analysis.
40. LaVIDE: Language-Prompted Satellite Change Detection via Map-Image Alignment
Core Problem: Satellite change detection needs stronger semantic control when the relevant change class is specified by task language rather than fixed labels.
Key Innovation: Aligns maps, imagery, and language prompts for satellite change detection, a transferable capability for event-specific hazard mapping.
41. Zeolite amendments in silty loam soils: Impacts on infiltration, runoff and soil erosion
Core Problem: Runoff and sediment yield in semi-arid soils depend on amendment type and application method, but nano- and bulk-zeolite effects are not directly comparable.
Key Innovation: Uses rainfall-simulator experiments to compare zeolite and nanozeolite effects on infiltration, runoff, sediment concentration, and sediment yield.
42. Effect of vegetation stem cover on overland flow velocity in the Loess Plateau
Core Problem: Overland-flow velocity controls erosion and sediment transport, but the hydraulic effect of vegetation stem cover is hard to generalize in loess landscapes.
Key Innovation: Quantifies how vegetation stem cover modifies overland-flow velocity, supporting erosion and runoff parameterization for the Loess Plateau.
43. Integration of GIS and MCDM for temporary shelter site selection in earthquake-risk zone: a case study with sensitivity analysis
Core Problem: Post-earthquake shelter siting must account for exposure, accessibility, and suitability under spatial uncertainty.
Key Innovation: Combines GIS and multi-criteria decision-making with sensitivity analysis to identify temporary shelter sites in an earthquake-risk zone.
44. Multi-level assessment of flood risk perception and flood behaviour
Core Problem: Flood risk reduction depends on how individual perception and community context translate into protective behaviour.
Key Innovation: Uses a multi-level assessment to connect flood risk perception with behavioural response across social and spatial scales.