TerraMosaic Daily Digest: May 12, 2026
Daily Summary
The May 12 literature is anchored by a shift from mapping hazard presence toward estimating mechanical state and decision thresholds. The landslide papers are the clearest example: curved basal-surface mechanics explain why deglaciated deep-seated slopes may re-equilibrate or fail; InSAR links Baihetan Reservoir landslide motion to impoundment and rainfall forcing; adaptive K-means and extreme-value thresholds convert creeping displacement into operational warning states; and point-cloud spherical density analysis improves the structural reading of roadcut slopes. Together, these studies treat landslides as evolving systems with basal geometry, hydrologic forcing, discontinuity fabric, and nonstationary kinematics, rather than as static susceptibility polygons.
A parallel signal runs through the earthquake, infrastructure, hydrology, wildfire, and cryosphere papers. The Gonghe enhanced-geothermal catalog shows that weak pre-existing structures and borehole-scale stress rotation, not tensile fracture creation alone, organize induced seismicity; the mainland China surface-rupture database provides standardized rupture evidence for seismic-hazard analysis; and sequential earthquake-tsunami fragility work quantifies how liquefaction pre-damage changes coastal breakwater failure probability. Hydroclimate papers push hazard assessment toward transferable observation systems: foundation-model river cameras reduce water-level calibration demands, stochastic storm-transposition domains receive a statistical test, GNSS/GRACE deformation is used to define compound dry-hot events, and karst hydrology is constrained by geochemical signatures. The day also adds high-value baselines for climate-stressed terrain, including PROMICE Greenland ice velocity maps, a Southern Andes snow-depth dataset, and Nature evidence that Afroalpine fire has entered an elevation band with no analogue in the past 12,000 years.
Key Trends
The common methodological move is to translate heterogeneous observations into variables that can support action: landslide equilibrium states, deformation anomalies, surface-rupture geometry, seismicity controls, flood-relevant water levels, compound-event indices, and infrastructure fragility curves.
- Landslide papers now infer evolving mechanical state. Basal geometry, glacier buttressing, reservoir-rainfall forcing, creeping-displacement anomalies, seismic corridor response, and point-cloud discontinuity fabrics become explicit analytical targets.
- Seismic hazard work is organized around structure, stress, and catalogs. Induced seismicity at Gonghe, mainland China surface ruptures, geothermal permeability scaling, fracture reactivation, and Bay Area site response all turn high-resolution observations into hazard constraints.
- Hydroclimate monitoring is becoming transferable across sites. Foundation-model water-level estimation, statistical storm-transposition domains, GNSS/GRACE dry-hot indices, karst-aware SWAT modelling, and physics-aware runoff learning reduce local calibration burdens.
- Infrastructure risk is evaluated through deformation and sequence effects. Earthquake-tsunami breakwaters, utility tunnels, rock caverns, saline soils, geobag revetments, and boulder-impact tunnel protection are assessed by damage mechanisms rather than hazard intensity alone.
- Cryosphere and fire baselines sharpen climate-risk interpretation. Greenland ice velocity, Southern Andes snow depth, Andean wetlands, glacier-monitoring uncertainty, Amazon fire hydrology, wildfire shift learning, and Afroalpine fire history add observational anchors for emerging hazards.
Selected Papers
This issue contains 53 selected papers from 1,925 papers analyzed. The leading papers translate heterogeneous observations into mechanical or decision-relevant variables: basal stability in deglaciating landslides, reservoir-driven deformation histories, adaptive early-warning thresholds, rock discontinuity sets, induced-seismicity controls, surface-rupture geometry, and sequential earthquake-tsunami fragility. The broader set adds temporal InSAR deformation prediction, deltaic subsidence, post-mining uplift, fault-leakage strain monitoring, storm-transposition domains, foundation-model river water levels, GNSS/GRACE compound dry-hot events, karst-aware hydrological extremes, glacier and snow datasets, wildfire prediction, emergent alpine fire, post-disaster multimodal reasoning, and deformation-focused geotechnical infrastructure studies.
1. Deep‐Seated Landslide Stability on Curved Basal Surfaces With and Without Glacier Buttressing
Core Problem: Deglaciating mountain slopes can accelerate, stabilize, or fail after ice loss, but the mechanical controls on these divergent trajectories remain unclear.
Key Innovation: A two-dimensional Mohr-Coulomb model with curved basal surfaces and explicit ice buttressing shows how basal curvature can self-stabilize deep landslides after glacier retreat while buttress loss narrows the range of stable geometries.
2. Pre‐Existing Structures and Stress Variations Jointly Control the Induced Seismicity in Enhanced Geothermal System of Gonghe Basin, China
Core Problem: Seismic hazard in enhanced geothermal systems depends on whether injection creates new fractures or reactivates inherited structures under heterogeneous stress.
Key Innovation: A machine-learning earthquake catalog from 20 surface stations at Gonghe resolves 7,346 events and links their ring-shaped distribution to chlorite-rich weak faults, borehole-scale stress rotation, and mapped natural structures.
3. The mainland China surface rupture database (MCSRD): development, structure, and descriptive analysis
Core Problem: Earthquake-hazard analysis in mainland China needs a standardized, queryable record of historical and modern surface ruptures.
Key Innovation: The MCSRD compiles and structures mainland China surface rupture information, giving fault-rupture studies a consistent database for rupture geometry, segmentation, displacement, and descriptive analysis.
4. InSAR analysis of reservoir-induced landslides in the Baihetan Reservoir: Evolution across impoundment and response to combined reservoir-rainfall forcing
Core Problem: Large reservoirs can reactivate unstable slopes through coupled water-level and rainfall forcing, but their deformation histories are often spatially fragmented.
Key Innovation: Time-series InSAR reconstructs Baihetan Reservoir landslide evolution across impoundment stages and links displacement changes to combined reservoir-level and rainfall forcing.
5. Landslide early warning system: Single-centroid K-means and extreme value theory-based thresholds
Core Problem: Fixed empirical thresholds are brittle for creeping landslides because baseline displacement rates drift through time and differ among sites.
Key Innovation: A single-centroid K-means and extreme-value-theory framework dynamically updates normal-behavior centroids and Vigilance/Alarm thresholds from displacement, velocity, acceleration, and rainfall features across seven landslide cases.
6. Kinematic-based seismic landslide risk assessment to linear projects in mountainous areas: a case study of the railway project in western China
Core Problem: Railway and transport corridors in mountainous terrain require slope-failure risk assessment that links earthquake motion to kinematic slope response.
Key Innovation: A case study in western China develops kinematic-based seismic landslide risk assessment for linear projects, translating slope response into infrastructure-relevant hazard evaluation.
7. Experimental Verifications and Applications of Discontinuous Deformation Analysis in Block-Type Rock Slope Failure and Movement Mechanisms
Core Problem: Block-type rock slopes fail through discontinuity-controlled movement that continuum approaches may oversimplify.
Key Innovation: Experimental verification and discontinuous deformation analysis are combined to test and apply block-scale failure and movement mechanisms for rock-slope stability assessment.
8. Orientation-based colour mapping and spherical density analysis for rock discontinuity detection in 3D point clouds of roadcut slopes
Core Problem: Stereographic projection can distort subvertical or closely spaced discontinuity sets, weakening slope-stability interpretation from 3D point clouds.
Key Innovation: Axis-rotation optimization, orientation-based colour mapping, and spherical density analysis detect discontinuity sets directly on the unit sphere and are implemented in open-source tools for roadcut slopes.
9. Interaction of flowslides with fixed and movable barriers: centrifuge tests on Baskarp and Vesuvian soils
Core Problem: Debris and flowslide barriers must be designed against moving soil masses whose impact and runout depend on barrier mobility and soil state.
Key Innovation: Centrifuge tests using Baskarp and Vesuvian soils compare fixed and movable barriers, clarifying barrier-soil interaction during flowslide impact.
10. DEM analysis of the dynamic responses of the gabion tunnel lining components subjected to boulder impact
Core Problem: Tunnel portals exposed to rockfall need protection systems that dissipate boulder impact energy without excessive structural deformation.
Key Innovation: Model-test-calibrated DEM simulations quantify how impact energy and gabion netting layers control impact force, depression, rockfill crushing, and buffering efficiency of gabion tunnel lining components.
11. Sequential earthquake–tsunami probabilistic fragility analysis of composite breakwaters on liquefiable soil
Core Problem: Breakwater damage under an earthquake followed by tsunami loading is difficult to evaluate efficiently when liquefaction and hydrodynamic forcing interact.
Key Innovation: A unified finite-element method replaces costly CFD tsunami simulations with equivalent loads and generates 14,400 intensity-demand pairs to quantify displacement fragility after seismic pre-damage.
12. Twenty-first century emergence of alpine fire in Central African mountains
Core Problem: Afroalpine mountains have been considered too cool and moist to burn, leaving little context for recent high-elevation fires.
Key Innovation: Nature reports that a twenty-first-century fire in the Rwenzori Mountains was the first to affect elevations above 3,800 m in 12,000 years, identifying alpine fire as a newly emerging tropical mountain hazard.
13. A Physics-Guided Deep Learning Method for Temporal InSAR Surface Deformation Monitoring and Prediction: A Case Study of Lishi District, Shanxi Province, China
Core Problem: InSAR deformation series are valuable for subsidence and slope monitoring, but prediction can lose physical consistency when treated as a purely statistical sequence.
Key Innovation: A physics-guided deep-learning method predicts temporal InSAR surface deformation in Lishi District, integrating process constraints with deformation time-series learning.
14. Unraveling the Spatial Heterogeneity of Land Subsidence in the Yellow River Delta: A Spatially Adaptive Ensemble Learning Approach
Core Problem: Deltaic subsidence is spatially heterogeneous and difficult to attribute with globally fixed models.
Key Innovation: A spatially adaptive ensemble-learning approach maps and interprets land-subsidence heterogeneity in the Yellow River Delta from remote-sensing and environmental predictors.
15. Monitoring Post-Mining Surface Uplift Induced by Mine Flooding Using EGMS and PSInSAR: A Case Study from the Upper Silesian Coal Basin (Poland)
Core Problem: Closed mines can undergo surface uplift after flooding, but regional monitoring must separate subtle deformation from background noise.
Key Innovation: EGMS and PSInSAR are used to monitor mine-flooding-induced uplift in the Upper Silesian Coal Basin, linking post-mining hydrology to measurable surface motion.
16. Monitoring and prediction of internal strains upon supercritical CO2 leakage along the fault
Core Problem: Geological CO2 storage requires early detection of fault-controlled leakage and associated subsurface strain changes.
Key Innovation: The study monitors and predicts internal strains during supercritical CO2 leakage along a fault, connecting leakage processes with deformation observables.
17. Influence of Temperature on Seismic Moment as a Diagnostic for Crustal Permeability Creation
Core Problem: Subsurface stimulation needs constraints on how permeability increases relate to induced microearthquakes across temperature conditions.
Key Innovation: Concurrent flow-through reactivation experiments and geothermal-reservoir observations establish a temperature-dependent power-law relationship between permeability change and seismic moment.
18. Anatomy of a Reactivated Hydraulic Fracture: An Integrated In Situ Study of Aseismic Slip, Source Mechanisms, and Model Validation
Core Problem: Hydraulic fractures can reactivate aseismically or seismically, but the in-situ mechanics and model validation remain difficult to observe directly.
Key Innovation: An integrated in-situ study documents fracture reactivation, source mechanisms, aseismic slip, and model behavior to improve interpretation of fluid-driven deformation.
19. A data-driven framework for basin delineation and site response characterization using spatial clustering and Bayesian modeling: A case study for the San Francisco Bay area
Core Problem: Urban seismic response can vary sharply across basins, yet delineation of site-response regions is often subjective.
Key Innovation: Spatial clustering and Bayesian modelling are used to delineate basins and characterize site response in the San Francisco Bay Area.
20. Mitigating Site‐Specific Data Dependency in Image‐Based Water Level Estimation Using Vision Foundation Models
Core Problem: Camera-based river gauges usually require long local calibration records, limiting rapid deployment for flood monitoring.
Key Innovation: A one-shot foundation segmentation model extracts water masks and a Static Observer Flooding Index, reducing high-water-level estimation error by about half compared with end-to-end regressors.
21. Hydrologic Whiplash in the Mississippi River Basin: Mechanisms and Projections
Core Problem: Hydrologic whiplash stresses infrastructure and water management, but its mechanisms differ from isolated flood or drought events.
Key Innovation: Observations and CESM2-LENS projections show strong twenty-first-century increases on western Mississippi tributaries and identify precipitation, snowmelt, and soil-moisture mechanisms behind transitions.
22. An L-moments-based hypothesis test to identify homogeneous storm transposition regions
Core Problem: Stochastic storm transposition needs defensible homogeneous rainfall domains before it can support national flood-risk estimation.
Key Innovation: The SLAM procedure uses spatial L-moments of annual maxima to statistically test transposition-domain homogeneity and improve extreme-rainfall frequency estimates.
23. Insights into hydrological compound dry-hot features using GNSS/GRACE-derived terrestrial water storage and land surface temperature
Core Problem: Compound dry-hot events are often defined meteorologically, missing hydrological drought driven by terrestrial-water-storage deficits.
Key Innovation: A GNSS 3D deformation and GRACE joint inversion estimates Amazon water storage and builds a standardized hydrological compound index for dry-hot-event detection.
24. Integrating an improved SWAT model and geochemical constraints to assess climate change impacts on hydrological processes and extremes in mixed karst and non-karst basin
Core Problem: Hydrologic projections in mixed karst and non-karst basins can be unrealistic if aquifer dynamics and geochemical constraints are omitted.
Key Innovation: A SWAT-Duliujiang model adds a karst three-reservoir module and geochemical constraints to project streamflow, sediment yield, and extreme monthly flows under CMIP6 scenarios.
25. Comprehensive Assessment of Grassland Fire Hazards Based on Multi-Source Data in Inner Mongolia
Core Problem: Grassland fire hazard varies with vegetation, weather, and landscape context and needs multi-source mapping rather than single-index screening.
Key Innovation: A Remote Sensing study integrates multi-source data to assess grassland fire hazards across Inner Mongolia.
26. Environment-Adaptive Preference Optimization for Wildfire Prediction
Core Problem: Wildfire events are rare and high-impact, so models trained on historical meteorological distributions often under-detect fires in shifted environments.
Key Innovation: Environment-Adaptive Preference Optimization retrieves locally aligned data and combines supervised fine-tuning with preference optimization to improve wildfire prediction under environmental change.
27. PROMICE: Greenland ice velocity maps v2026
Core Problem: Ice-sheet and outlet-glacier dynamics require consistent, near-operational velocity mosaics with known uncertainty and coverage limits.
Key Innovation: PROMICE v2026 provides 200 m, 12-day-updated Greenland ice velocity maps from Sentinel-1 SAR with adaptive offset tracking, improved mosaicking, validation, and uncertainty reporting.
28. The Southern Andes Daily Snow Depth Dataset (2010–2024): Quality–Controlled Dataset from Chile and Argentina
Core Problem: Southern Andes hydrology is constrained by sparse, inconsistent snow observations, limiting streamflow forecasting and water-risk modelling.
Key Innovation: A daily 2010-2024 snow-depth dataset from 81 Chilean and Argentinian stations is quality-controlled and released for hydrological model forcing, calibration, and reanalysis evaluation.
29. Instruct-ICL: Instruction-Guided In-Context Learning for Post-Disaster Damage Assessment
Core Problem: Disaster-response visual assessment needs fast, reliable reasoning without training a new task-specific model for every event.
Key Innovation: Instruction-guided in-context learning uses one multimodal LLM to generate structured reasoning guidance for another, improving post-disaster VQA accuracy on FloodNet.
30. Long-Term Deformation Prediction of Tunnel Surrounding Rock Fusing Bio-inspired Optimization and Bidirectional Temporal Modeling
Core Problem: Tunnel stability assessment needs reliable long-term deformation prediction under heterogeneous surrounding-rock behavior.
Key Innovation: Bio-inspired optimization and bidirectional temporal modelling are fused to predict surrounding-rock deformation over time.
31. Deformation Characteristics of Surrounding Rock in a Large Underground Opening Under High In Situ Stress and Complex Rock Mass Structures: A Case Study of Yebatan Hydropower Station
Core Problem: Large underground hydropower openings can deform under high in-situ stress and structurally complex rock masses.
Key Innovation: A Yebatan Hydropower Station case study characterizes surrounding-rock deformation mechanisms and stability controls in a large underground opening.
32. Coupling Characteristics of Geo-stress and Stratigraphic Structure and Their Impacts on Geoengineering Projects in High Geothermal Tunnels
Core Problem: High-geothermal tunnels face coupled thermal, stress, and stratigraphic controls that can govern excavation risk.
Key Innovation: The study analyzes coupling between geo-stress and stratigraphic structure and their impacts on high-geothermal tunnel engineering.
33. The new seismic catalog of the Gargano area (Southern Italy) after a decade of seismic monitoring by OTRIONS network
Core Problem: Seismic hazard assessment depends on dense, consistent regional catalogs, especially in structurally complex areas.
Key Innovation: A decade of OTRIONS monitoring is synthesized into a new seismic catalog for the Gargano area of southern Italy.
34. Residual wheat straw reinforcement for resilient cohesive soils against coupled cyclic wet-dry environment and seismic shaking
Core Problem: Cohesive soils degrade under cyclic wetting-drying and earthquake loading, creating resilience challenges for low-cost reinforcement.
Key Innovation: Residual wheat straw reinforcement is tested for improving cohesive-soil performance under coupled wet-dry environmental cycles and seismic shaking.
35. Multi-hazard fragility analysis of a single-span prefabricated concrete frame structure under sequential earthquake and blast loading
Core Problem: Structures may experience sequential hazards rather than single loading, but fragility functions often isolate earthquake effects.
Key Innovation: A single-span prefabricated concrete frame is assessed under sequential earthquake and blast loading to quantify multi-hazard fragility.
36. Shaking table test and numerical simulation of soil–corrugated steel utility tunnel interaction
Core Problem: Buried utility tunnels can alter and experience seismic ground response through soil-structure interaction.
Key Innovation: Shaking-table tests and validated 3D finite-element models quantify acceleration, strain, earth pressure, burial-depth, stiffness, and wave-incidence controls on tunnel deformation.
37. Precision, Detection Limits, and Uncertainty in Multi-Temporal Geomatic Glacier Monitoring: The Rutor Glacier Case Study
Core Problem: Glacier change products require transparent limits on precision, detection thresholds, and uncertainty across time.
Key Innovation: A Rutor Glacier case study evaluates precision, detection limits, and uncertainty in multi-temporal geomatic glacier monitoring.
38. Landsystems of the tropical high Peruvian Andes: Glaciers, lakes, wetlands and water resources in the Cordillera Vilcanota
Core Problem: Tropical Andean water security is changing as glaciers retreat, but the compensating roles of lakes, wetlands, moraines, and talus aquifers are unevenly resolved.
Key Innovation: Geomorphological mapping in the Cordillera Vilcanota shows that bofedal wetlands are often hydrologically independent of contemporary glaciers and may buffer water supply as ice storage declines.
39. An explainable physics-aware deep learning framework with improved spatiotemporal dependence matrices and signal decomposition for multi-station uncertainty daily runoff simulation
Core Problem: Basin-wide runoff simulation must represent nonstationarity, station interactions, reservoir regulation, and uncertainty at operational speed.
Key Innovation: GVMD-PA-STGMD combines signal decomposition, physics-aware node features, spatiotemporal TCN-GNN-attention learning, integrated gradients, and probabilistic intervals for Wei River runoff.
40. Climatic drying-wetting cycles induced dynamic coupling effect between 3D soil desiccation cracking and moisture redistribution
Core Problem: Drying-wetting cycles alter cracked clay hydraulics and mechanics, affecting slope and soil integrity in ways that surface inspection can miss.
Key Innovation: Field electrical-resistivity tomography and laboratory calibration reconstruct three-dimensional crack morphology and moisture redistribution during natural drying and simulated rainfall.
41. Simultaneous calibration of hydraulic head and deformation data for characterizing vertical aquitard heterogeneity via geostatistical inverse modeling
Core Problem: Aquitard heterogeneity controls drainage and land deformation, but hydraulic and deformation observations are usually calibrated separately.
Key Innovation: COMPAC simulations and quasi-linear geostatistical inversion jointly calibrate hydraulic head and deformation data to optimize observation design for Kv, Sske, and Sskp fields.
42. Impacts of Fire on Flow Magnitude and Variability in the Southeastern Amazon Basin
Core Problem: Amazon fire disturbances can alter catchment flow regimes, but their hydrological effect is less constrained than land-cover change.
Key Innovation: Water Resources Research quantifies how fire changes flow magnitude and variability in southeastern Amazon forest catchments.
43. Deformation characteristics and stability analysis of ecological geobag revetment in an inland arid river
Core Problem: Ecological geobag revetments must remain stable while accommodating deformation in inland arid rivers.
Key Innovation: The study analyzes deformation characteristics and stability of ecological geobag revetment, providing design evidence for riverbank protection.
44. Deformation mechanisms of salt-frost heave and dissolution-thaw collapse in sulfate saline soil
Core Problem: Sulfate saline soils deform through coupled water, vapor, heat, salt crystallization, and mechanical damage during freeze-thaw cycles.
Key Innovation: A coupled WVHSM model with microstructural observations explains a replenishment-crystallization-damage-dissolution feedback mechanism for heave and collapse.
45. Quantifying Decadal‐Scale Impacts of Flow and Sediment Regulation on an Alpine Gravel‐Bed River and Consequences for Fish Habitat: The Buëch River, France
Core Problem: Long-term dam operation and sediment management can drive channel incision and coarsening, affecting geomorphic stability and habitat.
Key Innovation: A one-dimensional morphodynamic model and habitat model quantify decadal effects of flow and sediment regulation on the Buëch River and evaluate repeated sediment reinjection.
46. Refinement of a Framework for Moving Aircraft River Velocimetry (MARV) and Application to Particle Tracking Along Alaskan Rivers
Core Problem: Flood and river monitoring need velocity measurements over large reaches without fixed instruments.
Key Innovation: The MARV framework is refined and applied to particle tracking along Alaskan rivers from moving aircraft imagery.
47. Nonlinear Responses Between Watershed Hydrological Functional Connectivity and Rainfall‐Runoff‐Sediment
Core Problem: Rainfall-runoff-sediment response can be nonlinear because watershed connectivity reorganizes during storms.
Key Innovation: The study quantifies nonlinear links between hydrological functional connectivity and rainfall-runoff-sediment behavior.
48. The economic implications of earthquake parameters for formulating financial risk management strategies: the Dead Sea Fault, Israel
Core Problem: Financial risk strategies for earthquakes depend on how rupture parameters translate into loss-relevant scenarios.
Key Innovation: The study evaluates economic implications of earthquake parameters along the Dead Sea Fault for financial risk management.
49. Perception-enhanced bidirectional attention networks for short-term time series forecasting of water level in cascade reservoir system
Core Problem: Cascade reservoir operation requires minute-scale water-level forecasts despite uncertainty and error propagation.
Key Innovation: PE-BANet integrates multi-scale atrous spatial pyramids, squeeze-excitation weighting, and ModernTCN to forecast one-hour water-level horizons with sub-0.15 m average error.
50. Transformer-enhanced extraction of operating rules for multi-objective reservoir groups: a min river basin case study
Core Problem: Basin-scale reservoir operation rules must preserve nonlinear feedbacks and water-level constraints.
Key Innovation: A triple-enhanced Transformer combines feature screening, multi-head attention, constraint-aware loss, and genetic hyperparameter optimization for Min River reservoir groups.
51. Diagnosing rainy-season heterogeneity in Huaihe Basin: a generalizable clustering and teleconnection-based analytical approach
Core Problem: Rainy-season variability affects flood and drought preparedness but can be spatially heterogeneous within a basin.
Key Innovation: A generalizable clustering and teleconnection-based framework diagnoses rainy-season heterogeneity across the Huaihe Basin.
52. Mechanical behavior of granite under cyclic loading after a seawater cyclic thermal shock
Core Problem: Coastal geothermal rock masses can experience thermal shock and cyclic loading that alter strength and failure mode.
Key Innovation: Triaxial cyclic loading, CT imaging, and seawater thermal-shock cycles quantify crack growth, strength loss, and shear failure evolution in granite.
53. Rainfall intensity and slope gradient interact to control runoff carbon–nitrogen losses and DOC–TN coupling
Core Problem: Storm runoff on sloping terrain transfers carbon and nitrogen in ways controlled by both rainfall intensity and slope gradient.
Key Innovation: The study tests how rainfall intensity and slope gradient interact to regulate runoff carbon-nitrogen losses and DOC-TN coupling.