TerraMosaic Daily Digest: June 11, 2026
Daily Summary
The June 11 issue is dominated by slope-process papers that tighten the link between geomorphic controls, hydromechanical triggering, and operational prediction. A rainfall-induced shallow-landslide study in volcanic terrain shows that aspect-dependent initiation can arise from tectonic fabric and microtopography, not only from vegetation or soil-hydraulic contrasts. The loess-paleosol study resolves rainfall failure as a coupled problem of preferential flow in loess and low-permeability paleosol layers, while the anti-dip rock-slope work combines shaking-table tests and DEM to explain dynamic toppling under seismic loading. Landslide inventory and risk papers extend this process view into mapping: the Three Gorges Reservoir Area inventory identifies more than 10,000 landslide relics with field validation, and the Doti District study turns a 1997-2023 Himalayan inventory into susceptibility and vulnerability estimates with machine learning.
The monitoring and forecasting papers move from deterministic maps toward uncertainty-aware warning systems. Reservoir landslide displacement is treated through prediction intervals rather than point forecasts; open-pit slope monitoring links hybrid deep learning to probabilistic time-to-failure; and particle filtering is tested for rainfall-infiltration slope deformation with explicit assimilation-window design. Cryospheric hazards appear as both immediate and landscape-scale problems: weekly activation windows for retrogressive thaw slumps provide an early-warning clock, while plateau-scale explainable machine learning maps future thaw-slump susceptibility. Flood and erosion papers broaden the hazard chain from post-fire debris-flow modeling and levee piping to dam overtopping, flood-scene extraction, flood-depth estimation, bedload shape effects, gully erosion, scour, engineered logjams, and rip-channel evolution. The issue also includes geotechnical and seismic contributions on tunnel-slope coupling, hydrate-reservoir particle migration, coal-mine roof-fracture modeling, hydrochemical earthquake precursors, and earthquake early-warning ground-motion prediction.
Key Trends
Five movements define the issue: coupled slope-failure physics, uncertainty-aware landslide prediction, operational cryosphere and wildfire hazards, flood-system diagnostics, and infrastructure-scale geotechnical coupling.
- Slope failure is being framed as coupled structure-flow-mechanics, not isolated triggering: The shallow-landslide aspect study, loess-paleosol seepage model, anti-dip seismic-slope experiments, soil-rock mixture analysis, and high-steep slope-tunnel shaking tests all resolve instability through interacting stratigraphy, permeability, discontinuities, and dynamic loading.
- Landslide prediction is shifting from classification to uncertainty-aware time evolution: The reservoir-landslide prediction-interval framework, open-pit time-to-failure model, Transformer-LSTM slope-stability emulator, particle-filter assimilation tests, and CETransUNet inventory mapping move the field toward warnings that report both state and confidence.
- Cryospheric and wildfire-conditioned hazards are becoming operational early-warning problems: Post-fire debris-flow modeling tests burn heterogeneity and hyperspectral inputs, while retrogressive thaw-slump studies define weekly activation windows and geographically explainable susceptibility shifts under warming.
- Flood science is linking rapid extraction, hydraulic failure, sediment motion, and infrastructure response: Deep-learning flood-scene extraction, RS-FloodXDepth, levee-piping indicators, coupled SPH dam-breach simulation, bedload shape correction, monopile scour, and engineered-logjam hydraulics all target actionable variables rather than flood extent alone.
- Geotechnical hazard modeling is becoming more explicitly coupled across infrastructure systems: Tunnel-slope seismic-rainfall coupling, slurry shield optimization, coal-roadway fracture reconstruction, offshore wind reliability under wind-wave-earthquake loading, and hydrate-reservoir fine-particle migration extend hazard analysis into underground, coastal, and energy infrastructure.
Selected Papers
This issue centers on rainfall-induced and cryospheric slope processes, post-fire debris-flow modeling, landslide risk and inventory construction, uncertainty-aware displacement forecasting, levee and dam failure, flood-depth mapping, erosion and scour, earthquake early warning, tunnel-slope coupling, and geotechnical infrastructure response. This issue contains 38 selected papers from 1564 papers analyzed.
1. Tectonic and microtopographic controls on aspect-dependent distribution of rainfall-induced shallow landslides
Core Problem: Aspect-dependent shallow-landslide initiation is often attributed to vegetation or soil-hydraulic contrasts, but those mechanisms do not explain the bimodal aspect pattern observed after a rainstorm in volcanic terrain.
Key Innovation: Analyzes more than 700 shallow landslides in Zhuji City and links their southeast-northwest aspect dependence to regional tectonic fabric and microtopographic controls.
2. Assessing wildfire spatial variability and hyperspectral data in debris-flow modeling
Core Problem: Post-fire debris-flow models can miss burn heterogeneity when basin-averaged dNBR is used as the main predictor.
Key Innovation: Tests alternative dNBR metrics and compares multispectral satellite and hyperspectral aircraft inputs for improving USGS M1 post-fire debris-flow likelihood modeling.
3. Machine learning-based landslide risk assessment in Doti district, western Nepal
Core Problem: Himalayan districts need risk assessments that combine susceptibility mapping with exposure and vulnerability rather than stopping at landslide occurrence probability.
Key Innovation: Builds a 1,147-landslide inventory for 1997-2023 and compares LR, SVM, RF, and XGBoost, with XGBoost reaching high discrimination for landslide risk assessment.
4. Comprehensive inventory of landslide relics in the three gorges reservoir area: spatial distribution and controlling factors
Core Problem: Fragmented landslide databases have limited systematic analysis of relic-landslide distribution in the Three Gorges Reservoir Area.
Key Innovation: Constructs a multi-source, field-validated inventory of 10,645 landslide relics covering 1,339.68 square kilometers and quantifies spatial controls with statistical analysis.
5. Rainfall-induced landslide mechanisms in loess-paleosol slopes: combined effects of preferential flow pathways and low-permeability layers
Core Problem: Interbedded loess-paleosol slopes fail through coupled seepage contrasts that standard single-permeability analyses can oversimplify.
Key Innovation: Develops a hydro-mechanical framework combining dual-permeability moisture migration in loess with single-permeability paleosol behavior to resolve rainfall failure mechanisms.
6. Weekly thermal-hydrologic activation windows of retrogressive thaw slumps: Implications for early warning of slope instability in degrading permafrost regions
Core Problem: Early warning for thaw-slump activation requires weekly-scale thermal and hydrologic indicators rather than seasonal permafrost status alone.
Key Innovation: Combines ground-surface temperature, active-layer soil moisture, a snow-cover proxy, and a 30-day snow-memory index to define weekly activation windows on the Qinghai-Tibetan Plateau.
7. Seismic response and critical acceleration analysis of anti-dip stratified rock slopes: Insights from shaking table tests and numerical simulations
Core Problem: Anti-dip stratified rock slopes can develop toppling deformation under earthquakes, but their acceleration response and critical instability thresholds remain hard to quantify.
Key Innovation: Integrates large-scale shaking-table experiments, white-noise dynamic-property tracking, and DEM simulations using the TH14 deformation mass as prototype.
8. Ensemble prediction intervals for reservoir landslide displacement with a novel cost function
Core Problem: Most machine-learning displacement forecasts provide point estimates, limiting their value for landslide early warning under uncertainty.
Key Innovation: Develops a Bootstrap-LUBE ensemble with a prediction-interval cost function and tests it on Shuping landslide GPS monitoring data.
9. GeoHazards, Vol. 7, Pages 75: Prediction of Rainfall-Induced Slope Stability Spatiotemporal Evolution Based on a Hybrid Transformer-LSTM Deep Learning Framework
Core Problem: Numerical simulation of rainfall-affected slope factor of safety is computationally expensive, while single machine-learning models often miss coupled spatiotemporal behavior.
Key Innovation: Builds a COMSOL fluid-structure coupling dataset and trains a Transformer-LSTM framework to emulate the spatiotemporal evolution of slope stability.
10. Remote Sensing, Vol. 18, Pages 1974: CETransUNet: An Intelligent Landslide Identification Method Based on Collaborative Optimization of Global Context and Dual Attention Mechanisms
Core Problem: Semantic segmentation of landslides remains sensitive to scale variation and spectral similarity with surrounding terrain.
Key Innovation: Constructs a Nyingchi earthquake co-seismic landslide dataset and proposes CETransUNet, combining ResNet, Transformer context, coordinate attention, and edge-guided attention.
11. Prediction of open-pit slope displacement and time-to-failure based on hybrid deep learning
Core Problem: Mine-slope early warning needs both continuous displacement prediction and probabilistic remaining-time-to-failure estimates.
Key Innovation: Combines a TCN-BiLSTM-Attention displacement model with Bayesian-regression time-to-failure estimation for monitored open-pit rock slopes.
12. Dynamic response of high-steep rock slope-tunnel system with cataclastic rock belt under seismic-rainfall coupling: Shaking table test
Core Problem: High-steep slope-tunnel systems with cataclastic rock belts can respond nonlinearly when rainfall preconditioning and seismic loading occur together.
Key Innovation: Uses large shaking-table tests and time-frequency analysis to identify acceleration amplification, cataclastic-rock effects, and failure mechanisms under coupled seismic-rainfall forcing.
13. Rapid flood scene extraction using deep learning for disaster emergency response
Core Problem: Typhoon emergency response needs inundation estimates before landfall, when real-time imagery and hydrological measurements may be unavailable.
Key Innovation: Develops a deep-learning flood-scene extraction strategy designed for fast pre-landfall emergency response where conventional hydrodynamic simulation is too slow.
14. RS-FloodXDepth: Enhancing Remote Sensing-Derived Flood Extent and Estimating Flood Depth Using a Hydrologically Guided Region-Growing Method and High-Resolution DEMs
Core Problem: Remote sensing often maps flood extent but misses obscured inundation and rarely estimates flood depth needed for damage assessment.
Key Innovation: Introduces a hydrologically guided region-growing method that uses high-resolution DEMs to enhance flood maps and infer flood depths.
15. Climate-driven thaw slump susceptibility on the Qinghai-Tibet plateau using geographically explainable machine learning
Core Problem: Static permafrost-hazard assessments cannot explain how climate change shifts retrogressive thaw-slump drivers across space.
Key Innovation: Uses an inventory of 3,693 retrogressive thaw slumps, ensemble susceptibility modeling, SHAP, and GeoShapley to map present and future driver changes.
16. Multi-scale time-series framework for piping-induced levee failure using a piping potential index
Core Problem: River levees may fail through progressive piping before overtopping, but early-warning indicators must integrate river stage, groundwater, rainfall, and antecedent soil moisture.
Key Innovation: Defines a Piping Potential Index from reconstructed hourly hydro-environmental variables and evaluates lead time, duration, and signal prominence for six South Korean levee failures.
17. Mechanisms of seepage and overtopping-induced failure in earthen dams: A coupled two-layer SPH investigation
Core Problem: Dam-breach modeling often starts overtopping simulations from simplified seepage states, obscuring coupled internal saturation and surface erosion.
Key Innovation: Builds a two-stage two-layer SPH framework that first establishes a stable phreatic line and then simulates overtopping-induced breach development.
18. Influence of topsoil depth and rock block content on the stability of soil-rock mixture slopes
Core Problem: The stability of soil-rock mixture slopes depends on internal heterogeneity and surface-layer geometry, but their combined effect is difficult to generalize.
Key Innovation: Uses ABAQUS finite-element analysis with adaptive stabilization and energy-based failure criteria to vary topsoil depth, rock-block content, and interface strength.
19. Spatiotemporal scaling of gully erosion susceptibility: Integrating seasonal dynamics across regional and watershed contexts
Core Problem: Static or annual-mean predictors can obscure seasonal processes that trigger gully initiation.
Key Innovation: Develops a multi-scale susceptibility framework that decomposes precipitation and NDVI into seasonal periods across regional and watershed scales.
20. Widespread ancient bedrock landslide deposits facilitate deep weathering for storage and access of organic carbon
Core Problem: Mountain soil-carbon estimates often miss thick, weathered landslide deposits that store organic carbon below the conventional 30 cm sampling depth.
Key Innovation: Combines a Western Oregon landslide chronosequence with an inventory of nearly 10,000 dated landslides to quantify deep soil organic carbon associated with relict landslide deposits.
21. Two-phase flow in clayey silt hydrate reservoirs drives migration and clogging of fine clay particles
Core Problem: Fine-particle migration during hydrate production can alter permeability and contribute to engineering-induced marine geohazards.
Key Innovation: Tests illite migration under gas-brine two-phase flow and identifies flow-controlled clogging mechanisms in clayey silt hydrate reservoirs.
22. Stochastic dynamic reliability of monopole-supported offshore wind turbines under coupled wind-wave-earthquake loading considering soil spatial variability
Core Problem: Offshore wind turbines face coupled wind, wave, seismic, and spatially variable soil loading, but reliability models often simplify the soil field.
Key Innovation: Combines spectral representation of soil heterogeneity with probability density evolution to evaluate stochastic dynamic reliability under coupled hazards.
23. Particle filtering for infiltration-induced slope deformation: assimilation window design and observation-error covariance matrix R strategies
Core Problem: Slope early warning from pore pressure and displacement observations is sensitive to when assimilation starts and how observation errors are represented.
Key Innovation: Couples a three-phase finite-element slope model with particle filtering and tests assimilation-window and error-covariance strategies against centrifuge experiments.
24. Stereology-based three-dimensional fracture modeling for coal-mine roadway roof strata and bidirectional validation
Core Problem: Roof-fracture geometry controls underground roadway support, but borehole observations are usually two-dimensional.
Key Innovation: Combines optical televiewer fracture extraction, stereology-based size inversion, DFN reconstruction, and bidirectional validation for three-dimensional roof-strata modeling.
25. Improving peak ground acceleration prediction and seismic intensity estimation using ensemble learning for earthquake early warning in the Taiwan region
Core Problem: On-site earthquake early warning needs rapid estimates of PGA and intensity from the initial P-wave segment under complex regional path effects.
Key Innovation: Uses ensemble learning to improve PGA prediction and seismic intensity estimation for Taiwan early-warning applications.
26. GeoHazards, Vol. 7, Pages 73: Experimental Insight on Hydraulic Performance of Surface Roughness in Eco-Engineered Flood Defenses
Core Problem: Hybrid flood defenses need quantitative evidence on how surface roughness and vegetation alter floodwave energy and timing.
Key Innovation: Uses controlled experiments on a dike-moat-vegetation system to show that increased roughness can reduce flow energy and delay flood arrival.
27. GeoHazards, Vol. 7, Pages 74: Investigation of Atmospheric Circulation Regimes for Wildfire, Flood and Rainfall Extremes in Greece
Core Problem: Wildfire, flood, and extreme-rainfall events in Greece are usually analyzed separately despite shared atmospheric circulation controls.
Key Innovation: Combines ERA5-driven HYSPLIT back trajectories and k-means clustering to compare air-mass regimes across wildfire, flood, and rainfall extremes.
28. Decoding multicomponent hydrochemical anomalies: a synergistic detection model for earthquake forecasting
Core Problem: Thermal-spring hydrochemistry can respond to tectonic activity, but single-parameter anomalies are noisy for real-time earthquake forecasting.
Key Innovation: Uses Bayesian change-point analysis and multicomponent anomaly detection for Qujiang and Wana spring records near the Xiaojiang-Red River fault intersection.
29. A novel rigid-plastic finite element method for slope stability analysis incorporating tensile failure
Core Problem: Classical slope-stability analyses often neglect tensile failure, potentially overestimating stability in steep slopes.
Key Innovation: Introduces a lower-bound rigid-plastic finite-element formulation that adds tensile strength constraints while preserving convexity through Gershgorin-circle inequalities.
30. A novel physical temperature triggering (PTT) method for simulating slope deformation in laboratory testing: Preliminary testing
Core Problem: Slope model tests need repeatable triggering methods that simulate geomaterial strength reduction without imposing unrealistic mechanical disturbance.
Key Innovation: Uses temperature-sensitive jelly-wax bonding in a ball-skeleton material to trigger deformation and compare concave and convex slope responses.
31. A Stress-Partition Constitutive Model for Glacial Till-Ice Mixtures considering Temperature, Confining pressure, and Ice Content Effects
Core Problem: The mechanical behavior of glacial till-ice mixtures changes with temperature, confining pressure, and ice content, affecting rock-glacier hazard assessment.
Key Innovation: Develops an SPF-GIM constitutive model that partitions stress between a till skeleton and pore ice as the load-bearing structure changes.
32. Do backcountry skiers react to the avalanche forecast? Measuring avalanche terrain exposure using a continuous score
Core Problem: Avalanche forecast effectiveness is difficult to evaluate because terrain exposure is usually measured coarsely from trip choices.
Key Innovation: Introduces a continuous GPS-track exposure score and applies Bayesian analysis to 26,703 backcountry tracks to test behavioral response to avalanche-danger ratings.
33. Adapting the effective flow work method for monopile scour estimation under combined wave-current loading
Core Problem: Monopile scour under combined wave-current loading remains uncertain because process-faithful methods often require unavailable velocity time series.
Key Innovation: Adapts the effective flow work method and introduces synthetic effective flow work to estimate scour without direct flow-velocity measurements.
34. Hydrodynamic and ecological effects of engineered logjams: Insights from field observations
Core Problem: Engineered logjams are used to modify floodplain hydraulics, but full-scale field measurements of flow structure and geomorphic effects remain limited.
Key Innovation: Uses large-scale particle image velocimetry in a restored Emme River reach to quantify jets, wakes, and habitat-relevant hydraulic effects.
35. Evolution of rip channel systems under morphodynamic state constraints
Core Problem: Beach-state classifications used for rip-current hazard assessment do not directly resolve the hydrodynamic-sedimentary evolution of rip channels.
Key Innovation: Combines numerical simulations, field observations, and controlled experiments from 54 headland beaches to quantify rip-channel three-dimensionality under shoreline curvature and wave-direction constraints.
36. A machine learning-based approach for predicting slurry properties and optimizing formulations in slurry shield tunneling
Core Problem: Slurry shield stability depends on slurry properties, but laboratory-based formulation optimization is slow.
Key Innovation: Combines Bayesian optimization, GPR, SVR, XGBoost, LightGBM, SHAP, and NSGA-II to predict slurry performance and invert for balanced formulations.
37. Influence of biochar addition to clayey sand on desiccation cracks formation and water infiltration under wetting and drying
Core Problem: Alternating wetting and drying can produce desiccation cracks that increase infiltration and pore-pressure buildup on roadside slopes.
Key Innovation: Uses column tests with rice-husk and wood-chip biochar amendments to quantify crack intensity, infiltration, and pore-pressure evolution through repeated wetting-drying cycles.
38. Revealing irregular sediment shape effects in bedload transport rate based on state probability correction
Core Problem: Flood-routing and sediment-transport models often assume spherical grains, limiting their ability to represent natural bedload motion.
Key Innovation: Introduces Corey-shape-factor corrections into a state-probability bedload framework using Markov-chain transitions among sediment-motion states.