TerraMosaic Daily Digest: June 24, 2026
Daily Summary
The June 24 literature is led by landslide papers that are unusually explicit about geometry, kinematics, and near-surface process. UAV-LiDAR and SIMWE modeling separate erosion behavior on vegetated and non-vegetated landslide slopes; the Joffre Peak study reconstructs how a multi-stage rockslide changed its kinematic state; SBAS-InSAR and XGBoost-SHAP reveal clustered active landslides along the upper Yellow River; and field, borehole, inclinometer, and FLAC3D evidence explain the excavation-driven reactivation of the Hongyan landslide. Urban tropical slope diagnosis and unsaturated residual-soil mechanics extend the same message: landslide assessment is strongest when it resolves the physical boundary conditions that make instability observable.
The multi-hazard layer is methodologically broad but not diffuse. A Natural Hazards and Earth System Sciences review synthesizes data-driven climate multi-hazard and multi-risk assessment, while an arXiv flood-detection paper uses topological constraints to improve optical-SAR mapping under cloud and sensor ambiguity. Ground fissure segmentation, fault and lineament mapping near a flash-flood mitigation dam, freeze-thaw expansive-soil testing, and CPT-informed vibro-compactibility classification show how remote sensing, geophysics, and laboratory mechanics are being converted into hazard-relevant decision variables.
A third group widens the context from event detection to system response. Himalayan agricultural losses under compound climate extremes, Winter Storm Uri household-burden data, complex-terrain groundwater and ecosystem service zoning, and mountain weathering studies connect terrain, water, infrastructure, and vulnerability. Several lower-score but useful AI and remote-sensing papers then supply transferable methods: multimodal change detection, foundation-model embeddings in complex terrain, methane-plume segmentation, decision-aligned uncertainty evaluation, and zero-shot LiDAR scene flow. The day therefore reads as a practical inventory of how hazard science moves from maps to mechanisms, and from mechanisms to accountable monitoring.
Key Trends
Five movements define this issue: process-resolved landslide analysis, data-governed multi-hazard modeling, geotechnical constraints on mapped risk, operational remote sensing, and AI methods designed for transfer and decisions.
- Landslide papers are converging on measurable process states: Slope erosion, multi-stage rockslide kinematics, active deformation clusters, reactivation mechanisms, and GPR-derived geometry all focus on observables that can constrain failure behavior rather than only susceptibility labels.
- Multi-hazard modeling is becoming more data-governed: The NHESS review, topological flood detection, ground-fissure segmentation, and climate-loss analysis emphasize data structure, model interpretability, and hazard-specific validation over generic prediction accuracy.
- Geotechnical evidence is being linked back to landscape-scale risk: Unsaturated residual-soil shear strength, freeze-thaw expansive soil damage, vibro-compactible soil classification, swelling pressure mitigation, and seismic soft-soil response provide material constraints for mapped hazards.
- Remote sensing is most useful when it returns a physical or operational variable: SBAS-InSAR landslide clusters, UAV-LiDAR erosion inputs, GPR slope geometry, fault mapping near a dam, and optical-SAR flood detection all produce deformation, geometry, lineament, or inundation information that can be audited.
- Transferable AI methods are shifting toward robustness and decision alignment: Change detection, foundation embeddings, plume segmentation, uncertainty evaluation, and zero-shot LiDAR motion estimation are relevant because they address cross-domain transfer, weak labels, and decision-specific reliability.
Selected Papers
The selected papers span landslide-slope erosion, multi-stage rockslide kinematics, active-landslide clustering, ancient-landslide reactivation, urban GPR slope geometry, unsaturated residual-soil mechanics, climate multi-hazard risk, topology-aware flood detection, ground fissure segmentation, dam-site structural controls, freeze-thaw expansive soils, vibro-compactible soil classification, mountain morphotectonics, storm vulnerability, groundwater mapping, remote-sensing foundation models, uncertainty evaluation, LiDAR scene flow, and unsaturated seismic-wave propagation. This issue contains 33 selected papers from 2297 papers analyzed.
1. Comparing erosion processes in vegetated and non-vegetated landslide slopes using the physically based SIMWE model and UAV-LiDAR
Core Problem: Vegetation modifies runoff and erosion on landslide-prone slopes, but its effect is difficult to quantify at the scale of individual failure surfaces.
Key Innovation: Combines UAV-LiDAR elevation data, measured infiltration, laboratory soil properties, and SIMWE simulations to compare erosion on vegetated and non-vegetated landslide slopes.
2. Analysis of the May 2019 Joffre Peak landslides, British Columbia, Canada: The importance of considering changes in slope kinematics
Core Problem: Multi-stage rockslides can change geometry, boundary conditions, and failure mechanisms as successive events unfold, making static back-analysis incomplete.
Key Innovation: Reconstructs the Joffre Peak landslides with remote sensing, structural analysis, discontinuous numerical modeling, and high-resolution point clouds to track evolving kinematics.
3. Spatial Clustering and Controlling Factors of Active Landslides in the Upper Yellow River: Insights From SBAS-InSAR and XGBoost-SHAP Modeling
Core Problem: The spatial organization and controlling factors of active landslides remain uncertain in the upper Yellow River corridor, where infrastructure exposure is high.
Key Innovation: Uses SBAS-InSAR deformation fields, improved Faster R-CNN detection, spatial clustering, XGBoost, and SHAP interpretation to identify 963 active landslides and their controls.
4. Reactivation characteristics and failure mechanism of the Hongyan landslide in Guiyang city, China
Core Problem: Ancient landslides in mountainous cities can reactivate under staged excavation, but the mechanism depends on lithology, weak layers, and anthropogenic disturbance.
Key Innovation: Integrates field surveys, boreholes, inclinometer monitoring, and FLAC3D simulation to explain the deformation and failure mechanism of the Hongyan landslide.
5. Landslide geometry in urban tropical zones: integrating analyses of ground-penetrating radar and traditional geophysical assessment
Core Problem: Dense tropical cities often occupy steep, rainfall-sensitive terrain, but urban infrastructure and steep access make landslide geometry difficult to resolve.
Key Innovation: Combines GPR and standard penetration testing to delineate subsurface landslide geometry and instability conditions in a hazard-prone area of Belo Horizonte, Brazil.
6. Review article: Harnessing data-driven methods for climate multi-hazard and multi-risk assessment
Core Problem: Climate hazards increasingly compound across space and time, yet multi-risk assessments often lack coherent data pipelines, explainability, and validation.
Key Innovation: Synthesizes machine learning, multivariate statistics, Earth observation, ensemble modeling, explainable AI, and future-scenario methods for climate multi-hazard assessment.
7. Shear strength and dilatancy behavior of unsaturated diabase and granite residual soils
Core Problem: Residual soils on slopes can fail through suction-controlled strength changes, but intact-weathering effects are often simplified in stability analysis.
Key Innovation: Measures suction, shear strength, and dilatancy in granite and diabase residual soils to constrain unsaturated behavior relevant to slope instability and collapse.
8. Topology-Informed Neural Networks for Flood Detection in Optical and Synthetic Aperture Radar Imagery
Core Problem: Flood detection from satellite imagery remains difficult when clouds, sensor differences, and fragmented inundation patterns break conventional pixelwise classifiers.
Key Innovation: Introduces topology-informed neural constraints for optical and SAR flood detection, targeting spatial connectivity and shape consistency in rapid hazard mapping.
9. PGD-UNet++: A Combining Direction-Awareness and Skeletal Connectivity Loss Segmentation Network for Ground Fissures Detection
Core Problem: Ground fissures are thin, discontinuous, and visually confusable in UAV imagery, yet they can damage infrastructure and mining landscapes.
Key Innovation: Adds position priors, global-local feature decoupling, and skeleton connectivity loss to improve automated fissure segmentation in coal-mining UAV imagery.
10. Integrated geophysical and remote sensing approach for understanding structural controls near Al Harjah Dam, Asir Terrane, Saudi Arabia
Core Problem: Small dams in structurally complex terrains require fault and lineament characterization because these controls affect water storage, seepage, and flood-mitigation reliability.
Key Innovation: Integrates remote sensing, aeromagnetic data, and field observations to map fault activity and structural controls around Al Harjah Dam.
11. Cyclic Freezing-Thawing Induced Stresses on the Mechanical Response, Cracking Behavior, and Microstructural Evolution of Expansive Soil in Cold Region: A Mechanistic-based Study
Core Problem: Seasonal freeze-thaw cycles alter expansive-soil strength and cracking, creating durability and deformation risks for cold-region infrastructure and slopes.
Key Innovation: Combines strength testing, void-ratio evolution, ultrasonic velocity, crack-image analysis, CT scanning, and FESEM to resolve macro- and microstructural damage.
12. Re-evaluating the classification of vibro-compactible soils using CPT and an engineering-geology-informed classification and regression tree
Core Problem: Modern ground-improvement practice needs a better boundary for identifying soils that can be improved by vibratory densification.
Key Innovation: Links CPT measurements, dynamic soil response, field densification tests, and an engineering-geology-informed classification tree to update vibro-compactibility criteria.
13. Reassessing fluvial landscape evolution along the passive margin of the Western Ghats, India: Evidence from tectonic reactivation and knickpoint dynamics
Core Problem: Passive-margin terrain can retain active tectonic and knickpoint signals that influence incision, relief, and long-term slope conditioning.
Key Innovation: Uses basin morphometry, river profiles, chi analysis, channel steepness, and knickpoint distributions to reassess reactivated fault controls in the Western Ghats.
14. Geomorphic evidence for present-day uplift and sediment bypass in the northern South American Foreland Basins
Core Problem: Foreland basins are usually treated as sediment sinks, but present-day uplift and sediment bypass can reshape erosion, floodplain stability, and mountain-front coupling.
Key Innovation: Combines regional slope maps, surface geology, stream profiles, channel steepness, and hypsometry to identify youthful uplift zones across northern South America.
15. Morphotectonic evolution of the Kas-Demre Karst Plateau: Cascade-type poljes, submerged poljes and Demre Stream Canyon, Western Taurus, Turkiye
Core Problem: Karst plateaus can encode coupled tectonic, erosional, and hydrological evolution that conditions flood routing, groundwater movement, and terrain instability.
Key Innovation: Maps faults, lithology, poljes, canyon morphology, and erosional surfaces to reconstruct Quaternary extension and karst-landscape evolution in the Western Taurus.
16. Diverse Connections Between Climate Extremes and Agricultural Loss Across the Southern Slopes of the Himalayas
Core Problem: Drought, heat stress, and extreme precipitation affect agricultural losses differently across Himalayan cropping systems, complicating regional vulnerability assessment.
Key Innovation: Quantifies probability- and magnitude-dominated loss patterns across the southern Himalayan slopes and separates the contribution of multiple climate extremes.
17. Investigating the determinants of household capabilities burden during power outages: the case of Winter Storm Uri
Core Problem: Conventional vulnerability indices rarely measure how utility outages disrupt household capabilities during compound winter-storm disasters.
Key Innovation: Uses household survey data from Winter Storm Uri to quantify burden across cooking, heating, refrigeration, and other disrupted capabilities.
18. Groundwater potential zone mapping using geographic information system and remote sensing in the case of upper Jemma sub-basin, North Shewa zone, Amhara Regional State, Ethiopia
Core Problem: Groundwater access and recharge potential depend on lithology, slope, drainage, land cover, and lineament structure, but many basins lack integrated maps.
Key Innovation: Combines GIS, remote sensing, AHP, and thematic hydrogeological factors to delineate groundwater potential zones in the upper Jemma sub-basin.
19. Ecosystem service trade-offs and ecological functional zoning in the Shiyang River Basin of northwest China
Core Problem: Arid inland basins require management zones that balance water provision, soil conservation, and carbon storage under strong mountain-oasis-desert gradients.
Key Innovation: Uses multi-source datasets and spatial ecological models to quantify water yield, soil conservation, carbon storage, and ecological trade-offs from 2000 to 2022.
20. Reduced West Antarctic Ice Sheet Size During Prominent Quaternary Interglacials Constrained by Iceberg-Rafted Debris Provenance in the Amundsen Sea
Core Problem: Past West Antarctic Ice Sheet retreat constrains future sea-level risk, but offshore provenance records remain sparse in the Amundsen Sea sector.
Key Innovation: Uses argon ages of iceberg-rafted debris minerals to infer reduced ice-sheet extent during prominent Quaternary interglacials.
21. High degree of soil development and large carbon stocks in soils above the last glacial maximum trimline in the North-Western Italian Alps
Core Problem: Mountain soils above glacial trimlines preserve older geomorphic surfaces that affect carbon storage, periglacial features, and terrain evolution.
Key Innovation: Compares soil development and carbon stocks above and below the Last Glacial Maximum trimline using a modified profile-development index and soil analyses.
22. Supply-limited weathering across global water-limited mountains governed by water availability and denudation
Core Problem: Water-limited mountain landscapes can experience weathering constraints that influence sediment supply, slope material properties, and long-term hazard conditioning.
Key Innovation: Analyzes global water-limited mountains to show how water availability and denudation govern supply-limited weathering regimes.
23. CSD-GAN: Content and Style Decoupled Generative Adversarial Network for Unsupervised Multimodal Remote Sensing Image Change Detection
Core Problem: Disaster mapping often requires comparing heterogeneous sensors, but modality differences can cause semantic drift in unsupervised change detection.
Key Innovation: Separates content and style representations in a GAN framework to improve multimodal remote-sensing change detection without dense paired labels.
24. Evaluating the Efficacy of AlphaEarth Embeddings for Tree Species Mapping in Complex Terrain
Core Problem: Complex terrain introduces spectral confusion and topographic effects that challenge remote-sensing foundation models outside benchmark settings.
Key Innovation: Compares AlphaEarth embeddings with traditional multisource features for tree-species mapping in topographically fragmented Sichuan landscapes.
25. Methane-Plume Segmentation From Hyperspectral Satellite Imagery Via Multimodal Deep Learning
Core Problem: Operational environmental monitoring needs accurate segmentation of weak methane plume signals in hyperspectral satellite imagery.
Key Innovation: Combines physically meaningful methane cues with transformer-based RGB representations to improve plume segmentation accuracy with lower computational cost.
26. Decision-Aligned Evaluation of Uncertainty Quantification
Core Problem: Generic uncertainty metrics can look well calibrated while being poorly aligned with the decisions that hazard models are meant to support.
Key Innovation: Defines decision-alignment and prior-weighted utility metrics to evaluate whether uncertainty quantification improves downstream decision utility.
27. UniFlow: Zero-Shot LiDAR Scene Flow for Autonomous Vehicles
Core Problem: Cross-sensor LiDAR transfer remains difficult, limiting motion estimation for mobile mapping and rapid post-event terrain or infrastructure surveys.
Key Innovation: Shows that low-level scene-flow models can benefit from cross-dataset training and transfer to unseen LiDAR sensors without architecture changes.
28. Reducing Lateral Swelling Pressure on Sheet Pile Walls Using EPS Geofoam and Sand Layers
Core Problem: Water-absorbing expansive soils can impose large lateral pressures on retaining systems, increasing deformation and serviceability risk.
Key Innovation: Uses large-scale physical model tests to compare EPS geofoam and sand layers as buffers for swelling pressure and sheet-pile wall displacement.
29. Dynamic performance evaluation of wind turbines on coastal soft-soil sites subjected to oblique-incident seismic waves
Core Problem: Coastal wind turbines on deep soft soils experience coupled site amplification and multidirectional seismic loading that are often simplified in design.
Key Innovation: Evaluates wind-turbine dynamic response under oblique incident waves to quantify how soft-soil stratigraphy and wave direction affect seismic demand.
30. Investigation of stress-strain behavior of calcareous sand sites using a novel integrated inversion analysis framework
Core Problem: Calcareous sand sites can deform differently under earthquake loading, complicating site-response and foundation performance assessment.
Key Innovation: Combines shaking-table data and inversion analysis to estimate stress-strain behavior of calcareous sand sites under increasing seismic acceleration.
31. Performance of base-isolated structures with eddy current damper tuned tandem mass dampers-inerters considering soil-structure interaction
Core Problem: Base-isolation performance can change when soil-structure interaction is included, especially for advanced damper-inerter systems.
Key Innovation: Develops an equivalent linearized state-space model for base-isolated structures with eddy-current damper tuned tandem mass dampers-inerters under SSI.
32. Analytical modeling of P-wave scattering by a spherical cavity in an unsaturated porous half-space
Core Problem: Cavities in unsaturated porous ground can alter surface displacement and stress concentration during seismic-wave propagation.
Key Innovation: Derives an analytical P-wave scattering solution for a spherical cavity in an unsaturated half-space and quantifies saturation-dependent wave effects.
33. Reflection characteristics of plane S waves at the free surface of unsaturated porous thermo-viscoelastic media
Core Problem: Thermo-hydro-mechanical coupling in unsaturated soils can affect reflected seismic energy and near-surface ground response.
Key Innovation: Derives reflected-wave amplitude and energy ratios for thermo-viscoelastic unsaturated media and evaluates the role of saturation, temperature, and incidence angle.