TerraMosaic Daily Digest: May 13, 2026
Daily Summary
The May 13 papers are dominated by slope-hazard studies that turn remote or indirect observations into mechanical state variables. The strongest contribution is a Remote Sensing of Environment framework that couples TS-InSAR displacement decomposition with spatial kinematic learning, making landslide displacement prediction less dependent on sparse ground gauges and local triggering records. The same observation-to-mechanics logic appears across the landslide and rock-slope cluster: Suoertou deformation is reconstructed from rainfall, seismic forcing, and fault stress; the Afsarabad landslide is interpreted through geomorphology, mineralogy, geotechnics, and geophysics; GBInSAR is paired with thermo-mechanical modelling to explain thermally driven rock-slope motion; acoustic-emission networks separate deep shear-zone signals; and UAV-derived geometry feeds dynamic fragmentation simulations for rockfall source blocks.
Earthquake, tsunami, flood, cryosphere, and fire papers extend that pattern from slope processes to broader geohazard systems. A California stress model built from more than 810,000 focal mechanisms provides a fault-scale stress and instability reference for a complex plate boundary, while the updated CEDIT catalogue turns more than nine centuries of earthquake-induced ground failures into a bias-aware inventory for multi-hazard analysis. Tsunami work combines event-scale Kamchatka modelling, open GPU-accelerated SWEpy simulation, and vegetation-based run-up attenuation experiments; flood work adds metropolitan shallow-water surrogates, street-scale drainage vulnerability, scene-adaptive water mapping, and rainstorm disruption propagation through transport networks. The climate-linked papers sharpen the physical context for future hazards: Nature connects cumulative CO2 to compound extremes, Nature Geoscience links deglaciation to hydrate dissolution through subglacial groundwater flushing, PolarLakes and Svalbard studies expand cryosphere monitoring, and permafrost, drought, precipitation, wildfire, and hydrological-extreme papers describe regime shifts rather than isolated events.
Key Trends
The day’s methodological center is the conversion of measurements into hazard-state variables: displacement components, stress fields, ground-effect inventories, failure geometries, tsunami wave periods, flood hydraulics, and climate-driven regime shifts.
- Landslide papers now forecast motion rather than only mapping susceptibility. TS-InSAR kinematics, rainfall-seismic-fault coupling, AE shear-zone separation, and GBInSAR thermo-mechanics all estimate evolving displacement or precursor state.
- Rock-slope hazard is being resolved through 3D geometry and fragmentation. UAV structural models, dynamic earthquake simulations, block-fragmentation DEM, and rainfall-driven rolling mechanics show why simplified slope geometries miss failure controls.
- Seismic and tsunami studies build reusable hazard evidence. California stress fields, CEDIT ground-effect inventories, Kamchatka waveforms, SWEpy simulations, and coastal forest experiments provide constraints that can be reused across scenarios.
- Flood diagnostics are becoming faster and more operational. Metropolitan flood surrogates, urban street-risk maps, scene-adaptive water indices, precipitation nowcasting, and transport-network waterlogging models shorten the path from rainfall to impact.
- Climate-linked hazards are treated as shifting regimes. Compound extremes, deglaciation-driven hydrate dissolution, Antarctic supraglacial lakes, Svalbard debris-cover ablation, permafrost hydrology, drought nonstationarity, and wildfire risk all point to moving baselines.
Selected Papers
This issue contains 49 selected papers from 1,841 papers analyzed. The leading papers convert observations into hazard-state variables: InSAR-derived landslide kinematics, California fault-stress fields, earthquake-induced ground-effect inventories, 3D rock-slope failure geometry, thermo-mechanical GBInSAR displacement, acoustic-emission shear-zone signals, and coastal subsidence rates that alter future flood exposure. The broader set adds Kamchatka tsunami modelling, vegetation-based tsunami attenuation, open tsunami solvers, flood digital twins, street-scale urban flood risk, cover-collapse sinkholes, hydrate-bearing submarine slope stability, volcanic degassing from ambient noise, compound climate extremes, deglaciation-driven hydrate dissolution, Antarctic lake monitoring, permafrost hydrologic shifts, wildfire risk prediction, and deformation-focused tunnel and CO2-storage geomechanics.
1. A unified deep learning framework coupling InSAR-derived spatiotemporal kinematic features for robust landslide displacement prediction
Core Problem: Remote slow landslides often lack ground displacement and triggering-factor records, limiting stability assessment and operational forecasting.
Key Innovation: A Remote Sensing of Environment framework decomposes Sentinel-1 TS-InSAR displacement into monotonic and seasonal components, learns spatial kinematic consistency among monitoring points, and predicts displacement without requiring dense in-situ trigger observations.
2. Stress Model of California: Fault‐Stress Interactions Across a Complex Plate Boundary System From Focal Mechanisms of Small Earthquakes
Core Problem: California's seismic hazard depends on heterogeneous stress orientation, faulting regime, and fault instability across a complex plate-boundary fault system.
Key Innovation: Machine-learning phase picking and REFOC focal-mechanism inversion produce 810,562 high-quality focal mechanisms and 2D/3D stress models for 350 major faults, resolving stress transfer, instability, and fault-regime contrasts across California.
3. Evaluating completeness and consistency in earthquake-induced ground effects inventorying: insights from the last release of the Italian CEDIT catalogue
Core Problem: Multi-hazard seismic risk analysis requires complete and internally consistent inventories of earthquake-induced ground failures, but historical catalogues are uneven and biased.
Key Innovation: The updated Italian CEDIT catalogue documents 4,256 ground effects from 215 earthquakes since 1117 AD and evaluates completeness, spatial coverage, multi-event overlap, and GIS-based survey gaps.
4. Identifying Key Factors for the Collapse Range of Cover-Collapse Sinkholes
Core Problem: Cover-collapse sinkholes can produce abrupt ground failure, but collapse range prediction remains sensitive to subsurface and cover-material controls.
Key Innovation: A GeoHazards study identifies key factors governing cover-collapse sinkhole extent, supporting more explicit hazard zoning for sinkhole-prone terrain.
5. Multi-factor coupling effects on deformation mechanisms of the suoertou landslide: insights from 3D numerical modeling and field analysis
Core Problem: Creep-type landslides can evolve under coupled hydrological, seismic, and tectonic forcing, but single-driver analyses miss progressive failure pathways.
Key Innovation: Field data, boreholes, geotechnical testing, and 3D FLAC3D modelling quantify how rainfall, seismic shaking, and Pingding-Huama fault stress jointly control the Suoertou landslide's deformation and safety factor.
6. Numerical analysis of the influence of 3D geometry and seismic input on rock-slope failures triggered by moderate to strong earthquakes: An example from the 2016 Central Italy seismic sequence
Core Problem: Earthquake-triggered rock-slope failures depend on 3D geometry, discontinuity orientation, and directional seismic amplification that simplified analyses cannot represent.
Key Innovation: UAV, geological-structural, and geomechanical data drive static and dynamic simulations of the Valle Orteccia rock slide, reconstructing near-critical conditions during the 2016 Central Italy sequence.
7. Multidisciplinary insights into landslide stability: geomorphology, mineralogy, and geotechnical analysis of the Afsarabad landslide
Core Problem: The Afsarabad landslide remains hazardous because material controls and slope structure have not been jointly characterized.
Key Innovation: Geomorphology, stratigraphy, XRD mineralogy, shear testing, geophysics, and finite-element stability analysis identify expansive clay minerals, low safety factors, and layered controls on failure.
8. Integrated GBInSAR and thermo-mechanical modeling for rock slope displacement analysis across geometrical complexities
Core Problem: Thermal cycles can drive precursor-scale rock-slope deformation, but geometry and material controls complicate interpretation of GBInSAR displacement.
Key Innovation: GBInSAR observations from two Italian Alps sites are coupled with thermo-mechanical FEM simulations to show how curvature, inclination, aspect, heat capacity, and elastic parameters concentrate thermal stress and displacement.
9. AE-CSNet: A deep learning framework for acoustic emission signal classification and separation in deep landslide monitoring
Core Problem: Acoustic-emission monitoring of deep landslides is confounded when multiple shear zones generate mixed signals, reducing displacement-inversion reliability.
Key Innovation: AE-CSNet combines layered active-waveguide sensor design with deep signal classification and separation to distinguish acoustic emissions from different shear-zone depths.
10. Realistic geometric modeling and dynamic fragmentation simulation of potential unstable blocks in high and steep rock slopes
Core Problem: Rockfall hazard simulations often simplify source blocks and ignore fragmentation, reducing reliability for high and steep slopes.
Key Innovation: A UAV-photogrammetry, 3D kinematic, and DEM framework identifies 186 potential unstable blocks at Shuangjiangkou Hydropower Station and simulates realistic block geometry and fragmentation dynamics.
11. Assessment of land subsidence and future flood hazards for coastal cities of the Gulf of Mexico using time series InSAR
Core Problem: Sea-level rise risk in coastal cities is amplified by local subsidence that may remain hidden without deformation monitoring.
Key Innovation: Sentinel-1 time-series InSAR maps 2019-2024 deformation in Houston, New Orleans, and Tampa, identifies previously unrecognized subsidence zones, and links them to future flooding hazard.
12. Primarily tsunami modeling of the Mw 8.8 Kamchatka Peninsula earthquake on July 29, 2025
Core Problem: Megathrust tsunami hazard requires rapid validation of earthquake source models against distributed ocean observations.
Key Innovation: Numerical tsunami simulations and DART waveform analysis reproduce the July 29, 2025 Kamchatka tsunami and identify coherent 30-70 minute long-period energy across Pacific gauges.
13. Harnessing nature’s defenses against tsunami run-up: strengthening coastal forests and strategic planning for planting sites
Core Problem: Green tsunami defenses must be designed around vegetation density, planting arrangement, and gaps rather than treated as generic roughness.
Key Innovation: Dam-break bore experiments show how high-density and staggered coastal forest configurations reduce wave heights, while open gaps weaken attenuation and guide practical planting-site planning.
14. Post Foundations for Rockfall Protection Systems: Experimental Investigations and New Developments
Core Problem: Rockfall protection systems have standardized nets and support components, but foundation design under service and maximum impact loads remains under-specified.
Key Innovation: More than 175 real-scale tests evaluate micropile, grout, pipe reinforcement, and concrete foundation configurations under SEL and MEL impact loading.
15. Rolling Susceptibility of a Partially Buried Spherical Block During Extreme Rainfall
Core Problem: Spherical granite blocks on residual-soil slopes can roll during extreme rainfall, but their mechanism differs from rectangular block toppling.
Key Innovation: Theoretical and laboratory analysis derives a rolling-stability formulation controlled by pore-water pressure and block weight for partially buried spherical blocks.
16. Kinematic solution of hydrate-bearing submarine slope stability during CH4-CO2 replacement
Core Problem: Gas hydrate exploitation can alter submarine slope stability through coupled hydrate dissociation, CO2 replacement, and reservoir restoration processes.
Key Innovation: A kinematic stability solution evaluates hydrate-bearing submarine slope response throughout CH4-CO2 replacement, linking gas production and CO2 sequestration to geotechnical safety.
17. Monitoring diffuse volcanic degassing with seismic ambient noise
Core Problem: Diffuse volcanic gas release is difficult to monitor continuously, especially when gas pathways evolve without obvious eruptive signals.
Key Innovation: NHESS demonstrates that seismic ambient noise can detect and track diffuse volcanic degassing, expanding non-invasive monitoring of volcanic unrest and gas transport.
18. Feature-based framework for evaluating deep learning models in post-disaster building damage assessment under diverse natural hazards
Core Problem: Remote-sensing damage models can appear accurate while failing under hazard, geography, or feature-domain shifts.
Key Innovation: A feature-based evaluation framework tests deep-learning post-disaster building-damage models across diverse natural hazards, exposing model behavior beyond aggregate accuracy.
19. Study on the spatial distribution of urban flooding and street flood risk in drainage-vulnerable city under extreme precipitation
Core Problem: Drainage-vulnerable cities need spatially explicit flood-risk diagnosis at street scale during extreme rainfall.
Key Innovation: A Natural Hazards study maps urban flooding distribution and street flood risk under drainage vulnerability, supporting urban flood-preparedness targeting.
20. Toward AI-Driven Digital Twins for Metropolitan Floods: A Conditional Latent Dynamics Network Surrogate of the Shallow Water Equations
Core Problem: Metropolitan flood digital twins require fast surrogates that retain shallow-water dynamics without running full hydrodynamic models for every scenario.
Key Innovation: A conditional latent dynamics network emulates shallow-water equations for AI-driven metropolitan flood forecasting and scenario exploration.
21. SWEpy: an open-source GPU-accelerated solver for near-field inundation and far-field tsunami modeling
Core Problem: Tsunami modelling needs open, efficient tools that can handle both near-field inundation and far-field propagation.
Key Innovation: SWEpy provides a GPU-accelerated shallow-water-equation solver for tsunami and inundation modelling, lowering barriers for reproducible hazard simulation.
22. Enhanced response of extreme compound events to cumulative CO2 emissions
Core Problem: Compound extremes can respond nonlinearly to cumulative emissions, affecting long-term risk under climate change.
Key Innovation: A Nature study quantifies enhanced responses of extreme compound events to cumulative CO2 emissions, linking climate forcing directly to multi-hazard extremes.
23. Gas hydrate dissolution triggered by subglacial groundwater flushing during deglaciation
Core Problem: Deglaciation can reroute groundwater and destabilize gas hydrate systems, but the hydrological trigger remains poorly constrained.
Key Innovation: Nature Geoscience links subglacial groundwater flushing to gas hydrate dissolution during deglaciation, connecting cryosphere change with subsurface carbon and geohazard processes.
24. PolarLakes: A Bi-Weekly Dataset of Supraglacial Lakes on Antarctic Ice Shelves from Multi-Sensor Satellite Observations (2015–2024)
Core Problem: Supraglacial lake evolution on Antarctic ice shelves is a key hydrofracture-relevant process but lacks consistent multi-sensor temporal records.
Key Innovation: PolarLakes provides a 2015-2024 bi-weekly dataset of Antarctic ice-shelf supraglacial lakes from multi-sensor satellite observations.
25. FireScope: Wildfire Risk Raster Prediction with a Chain-of-Thought Oracle
Core Problem: Wildfire risk prediction must represent spatial heterogeneity and rare-event behavior across raster landscapes.
Key Innovation: FireScope introduces a wildfire risk raster prediction framework using chain-of-thought oracle guidance to improve spatial risk estimation.
26. Combined change detection analysis of the impacts resulting from heavy rainfall events in mountainous areas
Core Problem: Heavy rainfall can leave spatially complex impacts in mountain terrain that are difficult to separate from background land-cover variability.
Key Innovation: Combined change-detection analysis maps impacts of heavy rainfall events in mountainous areas, improving post-event assessment from remote sensing.
27. Mechanisms driving hydrologic regime shifts in small catchments underlain by continuous permafrost during prolonged warming
Core Problem: Continuous permafrost catchments can shift hydrologic regime as warming changes storage, flow paths, and thaw processes.
Key Innovation: Catena identifies mechanisms driving hydrologic regime shifts in small catchments underlain by continuous permafrost during prolonged warming.
28. Continuous change monitoring of supraglacial lakes during melt and non-melt seasons with multi-source satellite imagery and deep learning
Core Problem: Supraglacial lakes evolve differently during melt and non-melt seasons, requiring continuous mapping across sensors.
Key Innovation: Multi-source satellite imagery and deep learning are used to monitor supraglacial lake changes through melt and non-melt seasons.
29. Assessing Glacier Dynamics and Debris-Cover-Controlled Ablation in Svalbard: Insights from Gåsbreen (2003−2023)
Core Problem: Debris cover can modify glacier ablation and motion, complicating interpretation of long-term glacier change.
Key Innovation: A Gåsbreen case study reconstructs 2003-2023 glacier dynamics and debris-cover-controlled ablation in Svalbard.
30. Multi-Quantile Regression for Extreme Precipitation Downscaling
Core Problem: Extreme precipitation downscaling must preserve tail behavior rather than optimize only mean error.
Key Innovation: A multi-quantile regression approach targets the distributional tails needed for flood and landslide-triggering rainfall applications.
31. Sensitivities of mean and extreme streamflow to climate variability across Europe
Core Problem: Hydrological resilience planning requires knowing how mean, maximum, and minimum streamflow respond to precipitation and temperature variability.
Key Innovation: More than 7,000 European catchments are used to quantify regional elasticities of mean and extreme streamflow to annual and seasonal climate drivers.
32. ST-DualNet: A Spatiotemporal Dual-Branch Neural Network Model for Short-Term Precipitation Forecasting
Core Problem: Short-term precipitation forecasts often underperform for heavy rainfall cores and unstable temporal dependencies.
Key Innovation: ST-DualNet combines enhanced ConvLSTM, dilated convolution, Transformer components, CBAM attention, and 3D convolution to improve radar-echo precipitation forecasting.
33. A physics-guided and scene-adaptive water index for consistent surface water mapping across regions and seasons
Core Problem: Water indices can fail in turbid, bloom-affected, icy, shadowed, or high-albedo scenes, weakening flood and reservoir mapping.
Key Innovation: A physics-guided Robust Water Index combines green, NIR, SWIR, and red-SWIR terms in a scene-adaptive form for cross-region and cross-season surface-water mapping.
34. Evolution of nonstationary hydrological drought characteristics in the UK under warming
Core Problem: Hydrological drought risk is becoming nonstationary as warming alters drought duration, severity, and propagation.
Key Innovation: HESS analyzes the evolution of nonstationary hydrological drought characteristics in the UK under warming.
35. Oscillation-induced synchronization hubs in global hydrological extremes
Core Problem: Global hydrological extremes can synchronize across distant regions through climate oscillations, but these hubs are poorly mapped.
Key Innovation: GRACE terrestrial-water-storage anomalies are used to build event-synchronization networks and identify climate-index controls on global hydrological extreme hubs.
36. Quantifying hydrological acceleration: A process-based framework and its application in arid regions
Core Problem: Static water-balance metrics miss the acceleration of hydrological fluxes under warming, especially in arid inland basins.
Key Innovation: A rate-of-hydrological-change index integrates observations, models, and reanalysis data to quantify atmosphere-to-runoff acceleration and glacier-loss contributions in Northwest China.
37. Soil Crack Width Controls Preferential Flow Velocity Through Drag Partitioning
Core Problem: Preferential flow through cracks controls infiltration and runoff response, but crack-scale velocities are hard to measure directly.
Key Innovation: Optical frequency-domain reflectometry and a drag-partitioning framework link crack width to preferential-flow velocity in laboratory and field slope monitoring.
38. Mapping water erosion susceptibility in Northeastern Algeria using two machine learning models
Core Problem: Water erosion susceptibility varies spatially with terrain and land-surface controls, requiring field inventories and interpretable predictors.
Key Innovation: Random Forest and Gradient Tree Boosting with Boruta feature selection map water-erosion susceptibility in northeastern Algeria and identify slope-length controls.
39. Calibrating Hydraulic-Geomechanical Models for CO₂ Storage Using Fibre Optic Strain Sensing: Insights from an Otway Basin Field Trial
Core Problem: CO2 storage safety depends on coupled geomechanical response, but reservoir models are difficult to calibrate from surface measurements alone.
Key Innovation: Distributed fibre-optic strain sensing from the Otway Basin field trial is used for history matching hydraulic-geomechanical models during CO2 injection.
40. Seepage characteristics and transformation mechanisms of point–linear–planar infiltration in loess tunnels
Core Problem: Water infiltration can trigger water inrush, collapse, crown failure, and surface cracking during loess tunnel construction.
Key Innovation: Field observations and physical model tests classify point, linear, and planar infiltration modes and link seepage pathways to axial force redistribution and surrounding-rock deformation.
41. Resilience analysis of multi-modal public transport networks under rainstorm waterlogging disasters: A case study in Chengdu, China
Core Problem: Rainstorm waterlogging can propagate disruptions through bus and rail systems, but resilience analysis often misses dynamic network propagation.
Key Innovation: A Chengdu case study models dynamic propagation of waterlogging-induced disruptions and identifies drainage capacity as the dominant resilience control.
42. Comparative Analysis of Near-Storm Environmental Characteristics of Tornadoes in Northern and Southern China Based on Himawari-8 Satellite and ERA5 Data
Core Problem: Tornado hazard assessment needs regional comparisons of near-storm environments using consistent remote-sensing and reanalysis data.
Key Innovation: Himawari-8 and ERA5 are combined to compare near-storm environments of northern and southern China tornadoes.
43. Observation of urban karst flows inform planning, design and construction of green stormwater infrastructure
Core Problem: Urban stormwater design in karst terrain can fail if subsurface conduit and fissure flows are not observed directly.
Key Innovation: Journal of Hydrology shows how observations of urban karst flows can inform planning, design, and construction of green stormwater infrastructure.
44. Advancing aquifer recharge forecasting through hybrid explainable AI and hydrological modeling
Core Problem: Aquifer recharge forecasting is uncertain because recharge is rarely measured directly at basin scale and subsurface heterogeneity is strong.
Key Innovation: A serial hybrid XAI framework learns from hydrological-model recharge estimates, identifies missed recharge events, and explains drivers using SHAP in the Edwards aquifer system.
45. The funnel effect of maize exacerbates rill erosion at the plant base in karst sloping farmlands
Core Problem: Vegetation can protect soil through roots but also concentrate rainfall through canopy stemflow, complicating erosion prediction on karst slopes.
Key Innovation: Simulated rainfall experiments with artificial maize isolate aboveground funnel effects and show large increases in runoff and sediment yield at plant bases.
46. Seismic analysis of a monopile-supported offshore wind turbine subjected to long-period ground motions considering soil-structure interaction
Core Problem: Offshore wind turbines in seismic regions require soil-structure interaction assessment under long-period ground motion.
Key Innovation: Ocean Engineering analyzes seismic response of a monopile-supported offshore wind turbine considering soil-structure interaction.
47. Dynamic risk-informed path planning for fire evacuation: An incremental reliability assessment approach
Core Problem: Evacuation routes can become unsafe as fire risk evolves, requiring real-time reliability-aware replanning.
Key Innovation: RTT-D* Lite uses dual risk thresholds, time-bounded reachability, and risk-averse cost reconstruction to maintain evacuation reliability under late-stage fire scenarios.
48. Can Greece Solve Its Wildfire Problem?
Core Problem: Mediterranean wildfire risk is shaped by governance, land management, and emergency-response structure as well as fuel and weather.
Key Innovation: A GeoHazards article evaluates whether Greece can solve its wildfire problem, framing wildfire mitigation as a coupled environmental and institutional challenge.
49. An investigation into the shear mechanical properties and failure features at the interface of soil-rock strata
Core Problem: Excavation stability in bi-material ground depends on soil-rock interface roughness, burial depth, and shear failure mechanisms.
Key Innovation: Field morphology, direct shear tests, and PFC mesoscale simulations quantify shear phases and failure patterns at silty clay-weathered sandstone interfaces.