TerraMosaic Daily Digest: October 26, 2026
Daily Summary
Slope failure is examined through the interaction of water, mineralogy and geological structure. The Jinping investigation supports a role for salt dissolution and clay swelling in the deterioration of muddy sandstone, with preliminary mineral sampling limiting regional extrapolation. Rock-slope analysis shows how tensile failure of rock bridges can redirect a seismic failure path and amplify displacement under selected pulse-like motions. Along the G55 expressway, spatially independent validation strengthens a local debris-flow susceptibility comparison tied to shallow source failures.
The relationship between hazards matters as much as their individual intensity. A Napier framework conditions shaking, tsunami, liquefaction, lateral spreading and landslides on common earthquake realizations, preserving dependence across the modelling chain. It does not yet represent changes in vulnerability caused by earlier damage. Harvey and Ike hindcasts reveal another form of coupling: flood-driver dominance and the sign of nonlinear interactions change over the storm lifecycle, so a peak map can obscure how flooding develops.
Observation and evaluation studies test what can be inferred beyond a fitted model. SWOT comparisons quantify the effects of filtering and measurement scale on water-level agreement. GNSS pretraining improves displacement analysis and seismic-step localization. Flood-mapping experiments find a useful speed advantage for a compact pretrained model without consistent accuracy superiority over U-Net. Streamflow studies separately test outlet-state consistency, transfer to human-affected basins and failures in flood and drought regimes.
Material and infrastructure studies connect those regional questions to local response. Sand morphology influences undrained strength loss despite similar peak properties, while unequal confinement changes bolt anchorage behavior. Drilling-vibration measurements offer a field route to more spatially detailed rock-strength estimates. Glacial-lake development screening and scenario-based lifeline retrofit analysis then show how physical assessments can inform prioritization while retaining the distinction between a modelled benefit and a demonstrated reduction in risk.
Key Trends
The strongest connections run from material state and failure pathways to correlated hazards, observation conditions and tests of model transfer.
- Material state changes the failure pathway: Mineral dissolution, swelling, rock-bridge tension and particle morphology affect different stages of deterioration and rupture. The Jinping, jointed-slope and sand studies make those mechanisms explicit within their particular sampling and loading conditions.
- Hazard dependence has several forms: The Napier assessment preserves shared event conditions across earthquake perils. The hurricane comparison resolves changing flood drivers and interactions through time. These are complementary advances, and neither implies that sequential structural damage or every environmental control has been represented.
- Validation design can change the model ranking: Spatially independent debris-flow validation, excluded-event flood mapping and unseen-site bathymetry test different transfer questions. The compact flood model and neural-operator baseline studies also show why architecture claims need matched baselines and comparable tuning effort.
- Observations constrain specific quantities: Filtered SWOT elevations, daily GNSS displacement and current outlet discharge each provide a useful but different constraint. Accurate water level, seismic-step localization or issue-time flux consistency does not alone establish flood depth, earthquake prediction or unique physical storage.
- Prioritization retains scenario assumptions: The Poiqu lake assessment combines resource potential with development suitability, while lifeline-network analysis refines critical disruption scenarios to guide retrofit. Both help narrow decisions whose safety and benefit still depend on the adopted physical, exposure and performance assumptions.
Selected Papers
Water-driven weakening, rupture pathways and the dependence between hazards anchor this selection. The papers connect slope and material processes with earthquake and flood assessment, then examine how observations, transfer tests and engineering response constrain interpretation. Studies are ordered by relevance score in one continuous list.
1. Deformation and failure mechanism of Jinping landslide in Junlian County, Sichuan Province, China: insight into mineral disintegration characteristics
Core Problem: The Jinping failure raises a material question: how can intermittent infiltration weaken a muddy sandstone slope beyond the effects of immediate rainfall loading?
Key Innovation: Field observations, mineral analyses and wetting–drying tests support a role for glauberite dissolution and clay swelling in progressive deterioration. Preliminary XRD covers only two samples, so the evidence supports a plausible site mechanism rather than a universal mineral-content threshold for failure.
2. Permanent displacement limit analysis of discontinuous jointed rock slopes considering tensile-shear failure mode of rock bridges under near-fault pulse-like ground motions
Core Problem: Discontinuous joints can force a seismic failure surface through intact rock bridges, where tensile rupture changes both the path and permanent displacement.
Key Innovation: A graph-based path search couples tension-corrected limit analysis with Newmark displacement. Numerical comparisons show transitions toward stepped failure and strong pulse sensitivity. The reported 22–30-fold displacement contrast concerns selected equal-PGA records and model assumptions, rather than a general near-fault multiplier.
3. Debris flow susceptibility assessment along the G55 expressway in Guangdong Province, China
Core Problem: Four debris flows along the G55 expressway motivate corridor-scale screening that accounts for local source failures and uneven terrain conditions.
Key Innovation: Four classifiers are compared using three-fold spatially independent validation, with XGBoost performing best in the reported test. The study links the mapped corridor pattern to shallow collapses and landslides feeding debris flows; the suggested role of slippery weathering products remains an association.
4. SBAS-InSAR-ML Dynamic-static coupling for landslide susceptibility in Xinyuan’s ecologically sensitive mountains
Core Problem: Static conditioning factors can miss information contained in ongoing surface deformation, but adding a dynamic predictor does not by itself create a forecast of failure time.
Key Innovation: The Xinyuan case combines SBAS-InSAR time series with 13 static factors and six classifiers, reporting improved local susceptibility discrimination. Ensemble and single-model advantages differ by metric. Spatial overlap and SHAP interpretation support local analysis, while prospective warning skill requires a separate test.
5. An internally consistent framework for calculating cascading probabilistic earthquake risk and its application to a case study in New Zealand
Core Problem: Combining separately sampled hazards can lose the shared rupture and ground-motion conditions that connect their impacts on the same buildings.
Key Innovation: An event-conditioned Monte Carlo chain integrates shaking, tsunami, liquefaction, lateral spreading and landslides for Napier. Tsunami becomes more consequential in the rare-event tail. The framework preserves hazard dependence but does not model how prior shaking damage changes vulnerability to later hazards, and integrated observational validation remains limited.
6. Compound Flooding Evolves Across Hurricane Lifecycle: A Contrasting Analysis of Hurricanes Harvey and Ike in the Houston–Galveston Region.
Core Problem: The relative roles of rainfall, river flow and storm surge change during a hurricane, making a single peak-inundation map an incomplete description of compounding.
Key Innovation: Consistent SFINCS hindcasts of Harvey and Ike separate excess flooding over the dominant driver from departures from the sum of drivers. Interactions vary in sign, place and time, including damping over some marshes and floodplains. The two-storm comparison supports time-resolved analysis without implying that all compound interactions amplify flooding.
7. From Hazard to Resource: A Screening‐Level Assessment of Glacial Lake Hydropower Potential and Development Suitability in the Poiqu River Basin, Central Himalaya
Core Problem: Glacial-lake hydropower screening needs to consider changing inflow and storage together with environmental, technical and economic suitability.
Key Innovation: A 37-lake Poiqu framework combines climate scenarios, water budgets, hydraulic head and multi-criteria ranking. Modelled operations reduce maximum storage for selected high-risk lakes under the tested assumptions. This is regional screening, rather than a dam-safety assessment or demonstrated reduction in outburst probability.
8. Dependence of seismic hazard assessment on the observation time interval: insights from a synthetic earthquake catalogue in southeastern Spain
Core Problem: A millennium of recorded earthquakes can still sample only part of the behavior of slowly deforming faults.
Key Innovation: A million-year synthetic catalogue for southeastern Spain is divided into many observation-length windows, with selected windows used for hazard calculations. Variability grows at higher PGA levels in the tested setting. The exercise quantifies catalogue-window sensitivity conditional on the simulator and fault model, rather than identifying a uniquely correct return period.
9. Ground-motion simulation of multi-segment ruptures on the Alpine-Marlborough fault system
Core Problem: Alpine Fault scenarios need to consider rupture continuing onto Marlborough structures and the resulting shaking beyond the main fault corridor.
Key Innovation: Broadband simulations compare three multi-segment geometries and multiple kinematic realizations using a New Zealand crustal model. Northward directivity and basin response affect Wellington shaking. The scenarios retain the chosen southern hypocentre and limited realization set; they are not predictions of the next rupture or a complete sampling of source uncertainty.
10. Accuracy and efficiency of compact foundation models for surface-water mapping during floods
Core Problem: Compact pretrained models need matched comparisons with ordinary segmentation networks, especially when flood labels, computation and transfer conditions differ.
Key Innovation: Sen1Floods11 experiments compare Prithvi-Tiny, random initialization, U-Net and other baselines across label budgets and excluded events. Pretraining helps the tiny transformer, but U-Net remains slightly more accurate in the reported comparisons while Prithvi-Tiny is faster on the tested hardware. Surface-water accuracy is kept distinct from mapping newly inundated land.
11. Influence of Particle Roughness and Shape on the Monotonic and Cyclic Undrained Response of Terrigenous Sands
Core Problem: Similar gradation and peak friction can conceal different post-peak strength loss and cyclic pore-pressure response in sands.
Key Innovation: Two Colombian sands with contrasting roundness and roughness are compared in monotonic, cyclic and small-strain tests. The smoother, rounder sand loses more undrained strength and has lower cyclic resistance in the tested conditions. Cyclic tests cover one density, stress and fines combination, limiting generalization of the morphology contrast.
12. Experimental Study on Anchorage Performance and Pullout Failure Mechanisms of Rock Bolts Under Biaxial Unequal Confining Pressure
Core Problem: Radially uniform confinement may misrepresent bolt anchorage in the unequal principal stresses found in deep rock engineering.
Key Innovation: Biaxial pullout tests, strain and acoustic-emission monitoring distinguish interface debonding, rock fracture and progressive slip as confinement changes. A derived interface-stress distribution and empirical force relation complement the tests. The observed increase in pullout capacity remains conditional on the tested bolt, resin, rock and stress configurations.
13. Field Estimation of UCS for Open-Pit Mine Slopes from Blasting Borehole Drilling Vibrations
Core Problem: Slope-strength characterization is sparse when core tests provide only isolated measurements within heterogeneous rock masses.
Key Innovation: Drilling vibrations from 30 blast holes are combined with operating parameters and reference UCS tests, followed by five additional boreholes for validation. Performance varies by engineering-geological zone, including a larger relative error in the weakest zone. The method estimates borehole-scale strength; it does not directly establish slope stability or failure probability.
14. Landslide susceptibility and risk estimation across the Indian himalaya for informed decision-making
Core Problem: Regional susceptibility maps alone do not represent the distribution of exposed infrastructure, social vulnerability and response capacity across the Indian Himalaya.
Key Innovation: The study combines machine-learning susceptibility with infrastructure, socioeconomic and response indicators, then presents district-scale risk through a GEE application. High-risk areas differ from susceptibility alone. The resulting composite zoning supports prioritization and should not be read as a calibrated annual loss or landslide probability.
15. Hydrologically driven mass failure dominates sediment redistribution on reservoir bank slope: Insights from tracer experiments
Core Problem: Reservoir-bank sediment redistribution can involve both surface transport and hydrologically driven mass movement.
Key Innovation: The verified title identifies tracer experiments separating these processes on a reservoir-bank slope. This is a title-level scope summary: the matching record does not support numerical source fractions, operating-level thresholds or extrapolation to other reservoirs.
16. Effects of different infiltration conditions on the failure process of a layered slope in the Qinling Mountains: A material point method analysis
Core Problem: Layered slopes can respond differently to infiltration pathways and boundary conditions even under superficially similar wetting.
Key Innovation: The verified study uses material point analysis to examine infiltration-dependent failure in the Qinling Mountains. With no reliable matching abstract, the entry identifies the modelling question without assigning a specific failure sequence, infiltration threshold or validation result.
17. Mechanical response of a shield tunnel to normal fault displacement considering longitudinal bolt–segment interaction
Core Problem: Longitudinal bolt–segment interaction can influence how a shield tunnel accommodates normal-fault displacement.
Key Innovation: The verified title centers the analysis on that interaction rather than treating the lining as a homogeneous tube. Title-level evidence does not establish the governing failure location, allowable offset or a tested bolt-design recommendation.
18. Failure mechanisms and damage evolution of cross-fault tunnels under combined sliding and nonlinearly distributed fault displacement
Core Problem: Combined sliding and a non-uniform fault-displacement profile can produce tunnel damage patterns that a simple imposed offset may miss.
Key Innovation: The verified record examines damage evolution under these combined kinematics. A matching abstract is unavailable, so no comparative damage reduction, displacement tolerance or preferred lining configuration is inferred.
19. Damage characteristics of tunnels under different temporal sequences of reverse fault dislocation and intense seismic excitation
Core Problem: The order of fault dislocation and strong shaking may change the response of a tunnel crossing an active structure.
Key Innovation: The verified study explicitly compares temporal sequences of reverse-fault movement and seismic excitation. Available evidence identifies the sequence question but does not establish which order is most damaging or quantify its amplification.
20. Design implications of composite tunnel linings with varying stiffness and thickness in saturated sand: Insights from large-scale 1g shaking table tests
Core Problem: Composite lining stiffness and thickness affect tunnel response in saturated sand, where flexibility and load redistribution can compete.
Key Innovation: The title confirms large-scale 1g shaking-table tests of alternative lining configurations. Matching results are unavailable; a preferred stiffness ratio, thickness or field-scale performance benefit cannot be specified from this record.
21. A framework for evaluating soil liquefaction and weakening of dam foundation during earthquakes
Core Problem: Earthquake assessment of dams needs to represent foundation weakening as well as the potential for soil liquefaction.
Key Innovation: The verified record proposes a framework addressing both processes. This title-level entry does not infer a particular constitutive model, damage threshold or demonstrated dam-safety improvement.
22. Prediction of soil-pile-interaction in liquefiable ground conditions: from single element calibration to system response
Core Problem: A constitutive model calibrated on soil elements still needs testing against the coupled response of a pile and liquefying ground.
Key Innovation: The verified title explicitly connects single-element calibration to system response. Without a matching abstract, it is retained for this validation question rather than for an unverified prediction accuracy or design conclusion.
23. Climate change-driven impacts on the lifetime seismic performance of coastal bridges: accounting for coupled actions of carbonation, chloride corrosion and sea-level rise
Core Problem: Coastal bridge performance can evolve as corrosion and carbonation interact with changing marine exposure over the structure's lifetime.
Key Innovation: The verified study considers these processes together with sea-level rise in seismic assessment. Title-level evidence supports the coupled scope but provides no reliable lifetime loss estimate, dominant degradation mechanism or retrofit ranking.
24. Assessing extreme monsoon and storm surge height projections along the Singapore Strait in a warming climate
Core Problem: Singapore Strait coastal planning requires regional projections of extreme monsoon conditions and surge heights under warming.
Key Innovation: The citation and title are verified, but the supplied text describes a different, earlier-style projection study. Its sea-level values and conclusions are not transferred to this entry; only the new study's stated assessment scope is retained.
25. Regional non-stationarity of typhoon-wave return levels across the South China Sea and adjacent western North Pacific: Joint location–scale non-additivity
Core Problem: Typhoon-wave return levels may change through interacting location and scale parameters rather than a single uniform trend.
Key Innovation: The verified title addresses regional non-stationarity and joint parameter non-additivity across the South China Sea and nearby western North Pacific. Matching results are unavailable, so the direction, magnitude and coastal distribution of return-level changes are not asserted.
26. Visible Flood Progression Typology from CCTV Keyframe Sequences for Urban Flood Monitoring
Core Problem: CCTV sequences contain the progression of visible flooding, which a single selected frame cannot fully describe.
Key Innovation: The verified study develops a progression typology from keyframe sequences for urban monitoring. Title-level evidence identifies this temporal representation but does not demonstrate quantitative depth retrieval, warning lead time or transfer between cameras.
27. Comparative evaluation of two-dimensional maps and virtual reality for tsunami evacuation route learning: A field experiment in an urban environment
Core Problem: The medium used to learn an evacuation route may influence how people understand and remember a safe path through an urban environment.
Key Innovation: The verified title confirms a field comparison of two-dimensional maps and virtual reality. A matching abstract is unavailable, so no claim is made that either medium improves actual evacuation time, route recall or survival.
28. Wet-Year Precipitation Anomalies and Geohazard Economic Losses in China: A Province-Year Sensitivity and Climate-Exposure Assessment
Core Problem: Province-scale wet-year anomalies and exposure can covary with economic losses without establishing a single rainfall-to-loss response for every locality.
Key Innovation: The verified record addresses province-year sensitivity and climate exposure in China. This scope summary does not infer causal attribution, a monetary elasticity or a transferable local loss function.
29. Multi-resolution framework for efficient resilience assessment and retrofit of lifeline networks
Core Problem: Exhaustive disruption scenarios make network-level resilience assessment and retrofit prioritization expensive.
Key Innovation: A selective multi-resolution scheme refines critical regions and iteratively prioritizes components under prescribed resilience and probability thresholds. Demonstrations include hypothetical networks and a railway under an assumed earthquake scenario. Computational savings and the retrofit order depend on those scenario and performance assumptions.
30. STGAF-Net: A Multi-Source Spatiotemporal Gated Adaptive Fusion Network for Wind Speed Nowcasting
Core Problem: Multi-source wind nowcasting must retain useful structure and high-wind performance as forecast lead time increases.
Key Innovation: STGAF-Net fuses radar reflectivity and surface wind through gated spatial and temporal modules, predicting multiple future fields in parallel. Two-dataset comparisons report gains at high-wind thresholds within a two-hour horizon. The available evidence does not establish calibrated warning probabilities or universal transfer to other radar networks.
31. GNSS-DispFM: A Self-Supervised Foundation Model for Daily GNSS Displacement Time Series
Core Problem: GNSS archives offer abundant unlabeled displacement data, while task-specific models are constrained by labeled examples.
Key Innovation: GNSS-DispFM pretrains a displacement-and-increment encoder on more than 17,000 stations and fine-tunes it for 90-day displacement prediction and seismic-step localization. Improvements on those tasks support geodetic reuse, but forecasting displacement is not earthquake prediction and representations alone do not validate landslide warning.
32. Non-stationary low-flow frequency analysis with mixed Weibull components and Copula-based dependence framework
Core Problem: Annual low-flow extremes can combine changing summer and winter processes whose dependence also evolves over time.
Key Innovation: A mixed Weibull and copula framework introduces temporal change in seasonal location and inter-event dependence, tested across more than 700 reference catchments. It exposes sensitivity of higher-return-period estimates to non-stationarity; linear trends and the selected distribution family still delimit the inferred evolution.
33. Analytic outlet-state identification in a pathway-age neural transport operator enables cross-domain streamflow forecasting in human-affected basins
Core Problem: A neural runoff model can start from a storage state inconsistent with discharge observed at forecast issue time, weakening short-range transfer.
Key Innovation: OGPAT analytically scales a learned pathway-age state to current outlet discharge and uses a nonnegative transport ledger for 1–18-hour forecasts. Tests transfer from reference to human-affected US basins and report improved peak metrics against matched controls. Outlet consistency constrains one learned-state coordinate; it does not uniquely recover physical catchment storage.
34. EvoBench: An Evolving, Diagnostic Benchmark Framework for Environmental Modeling
Core Problem: A high average streamflow score can conceal failures at unmonitored basins, shifted climates and rare hydrological regimes.
Key Innovation: EvoBench applies diagnostic testing to a large US-basin collection, comparing physical, conceptual and data-driven models. The reported flood underprediction and changing rankings motivate regime-specific evaluation and targeted improvement. Its evolving benchmark is an evaluation framework, rather than evidence that one model reliably forecasts all extreme events.
35. Evaluating SWOT Satellite Data With Gage Information in the United States and Canada
Core Problem: Satellite water elevations require quality control and scale-aware comparison with point gauges before being used in hydraulic assessment.
Key Innovation: Version-D SWOT products are compared with 335,531 paired observations at 2,629 gauges in the US and Canada. Filtering markedly improves agreement, but coverage, waterbody eligibility and point-versus-area mismatch remain consequential. The reported filtered elevation errors are not flood-depth accuracy or a guarantee for individual sites.
36. Braided Rivers as Morphological States: A Data‐Driven Identification of Recurrent Planform Configurations
Core Problem: Scalar channel metrics can miss recurring braided-river configurations and their association with intermittent sediment transport.
Key Innovation: A month-long laboratory experiment combines image comparison, dimension reduction and clustering to identify recurrent planform states, residence times and preferred transitions. The approach provides a compact description of the tested river dynamics without demonstrating a field-scale avulsion or flood forecast.
37. Climate downscaling with likelihood functional flow matching
Core Problem: Generative downscaling is costly when uncertainty must be sampled on fine grids, and grid dependence constrains flexible output locations.
Key Innovation: CD-FFM uses a continuous-field flow-matching formulation and a fitted spatial-noise prior, reporting temperature downscaling gains over three test years in Fennoscandia and a precipitation extension. Temperature CRPS and speed results do not establish extreme-rainfall fidelity, hydrological impacts or performance in other climates.
38. Cross-View and Cross-Time Tracking of Crack Evolution in Concrete
Core Problem: Repeated crack imagery must preserve crack identity across camera changes before growth can be distinguished from projection differences.
Key Innovation: A geometry-assisted framework aligns views, assigns morphology-based instances and maps temporal extension in three annotated cracking scenarios. Loose correspondence is achievable, while strict overlap and growth localization remain limited by microcrack visibility and projection drift. It is a monitoring method, not a validated structural failure criterion.
39. Satellite-derived bathymetry of Arabian Gulf ports under leakage-free validation: a multi-site benchmark of empirical, machine-learning and deep-learning models against 285,000 multibeam soundings, with an outlook for UAE waters
Core Problem: Accuracy from randomly mixed depth pixels can overstate performance when optical bathymetry is applied to another part of a turbid coast.
Key Innovation: A six-port benchmark contrasts random, spatial-block and unseen-site validation against multibeam soundings. Error changes more with the split than with model choice, and a simple band ratio remains competitive in transfer. Depth, turbidity, local control and vertical-datum conditions bound use in coastal-model preparation.
40. Performance of Neural Operator Baselines on 'The Well' Dataset
Core Problem: A lightly tuned neural-operator baseline can become a misleading reference when later papers attribute all performance differences to architecture.
Key Innovation: A bounded hyperparameter search on The Well improves TFNO errors across most tested tasks while retaining the original wall-clock constraint. The result supports tuning-aware comparisons; it does not introduce a new hazard model or show that TFNO is optimal across physical systems.
41. From Sentinel-2 to 1-m Imagery: Real-World ×10 Cross-Sensor Super-Resolution via Visual Autoregressive Modeling
Core Problem: Cross-sensor reconstruction at tenfold linear magnification must contend with differences in acquisition, radiometry, spectra and geometry.
Key Innovation: CrossVAR combines hierarchical autoregression with residual refinement on geographically paired Sentinel-2 and Jilin-1 RGB scenes. Reported image metrics improve over the compared models. A one-metre output grid contains reconstructed detail and should not be treated as one-metre observed evidence of a small landslide or damaged structure.
42. Efficient probabilistic weather forecasting at quarter-degree resolution with two-sided transport attention
Core Problem: Probabilistic weather models face substantial training and ensemble-generation costs at quarter-degree resolution.
Key Innovation: Astra combines two-sided transport attention with a multiscale scoring objective and reports favorable ERA5-based benchmark skill and compute comparisons. These results concern the specified variables, leads and hardware accounting; hazard-specific extremes, reliability and downstream warning performance need their own assessment.
43. Constant-Volume Direct Simple Shear Behavior of Reconstituted Natural and Degraded Glauconite Sand
Core Problem: Glauconite sands with different geological origins can change very differently during degradation and specimen reconstitution.
Key Innovation: Constant-volume simple-shear tests distinguish clay-like contraction in degraded authigenic material from more sand-like behavior in the allogenic material. Preparation, degradation and stress state affect the response. Laboratory dispersion is a characterization procedure rather than proof of an equivalent in-situ weathering history.
44. CAS-Mamba: Change-Aware State-Space Modulation in the Encoder for Remote Sensing Change Detection
Core Problem: Bitemporal image differences can enter a change detector only after encoding, limiting how they shape the initial representations.
Key Innovation: CAS-Mamba conditions time-step and input-write pathways inside a shared state-space encoder, with comparisons on three change-detection datasets. It is a useful encoding modification. The supplied experiments support benchmark change localization, rather than demonstrated disaster attribution or a new general hazard foundation model.
45. Technical note: Machine learning metamodelling for global sensitivity analysis
Core Problem: Cheap model surrogates are useful for sensitivity analysis only when their importance measures have a defensible relation to the target indices.
Key Innovation: Tests with three hydrologic models and three metamodel families compare permutation importance and SHAP with Sobol total effects. Nonlinear surrogates reproduce rankings and relative magnitudes more reliably in the tested cases. The theoretical permutation link relies on input-independence assumptions, and SHAP ranking agreement does not make it identical to a Sobol index.
46. An Open Sentinel-1 Uncompressed SAR Dataset with FDBAQ Compression Performance Benchmarks
Core Problem: Lossy onboard compression complicates direct tests of its effects on Sentinel-1 radiometry and image quality.
Key Innovation: An openly released collection of 65 uncompressed Sentinel-1C/D Level-0 products enables controlled emulation of FDBAQ and quantitative comparison with nominal processing. This is a focused sensor-compression resource; it is not a landslide inventory, a global uncompressed archive or a demonstrated hazard-mapping improvement.
47. Improving satellite-derived soil moisture estimates in a data-scarce watershed using strategically designed sampling and machine learning
Core Problem: Local refinement of coarse satellite soil moisture is constrained by where scarce field observations are collected.
Key Innovation: Strategic cLHS sampling and RF/XGB fusion are evaluated in a Brazilian watershed, including leave-one-site-out tests. Site-dependent transfer complements the strong held-out-period scores. SHAP identifies model associations, while the results do not establish causal controls or a calibrated slope-failure moisture threshold.
48. Comparative Analysis of Groundwater Recharge Dynamics in Granular (Soil) and Fractured Aquifers Using Water Level Fluctuation Monitoring
Core Problem: Rapid shallow water-table rises and muted fractured-aquifer responses can yield very different apparent recharge estimates under the same rainfall sequence.
Key Innovation: Seven wells reveal contrasting response times while alternative rise calculations and specific-yield assumptions span wide recharge envelopes. Some apparent estimates exceed precipitation, underscoring structural and methodological uncertainty. These envelopes are not independently validated absolute recharge rates or interchangeable estimates across aquifer domains.
49. Quantitative Assessment of Grain Size Effects on Granite Fracture Characteristics: from Microcracking to Macroscopic Failure
Core Problem: Granite grain size can change tensile strength and microcracking in ways that one average grain-size descriptor does not fully capture.
Key Innovation: Two granite groups are compared using Brazilian splitting, acoustic emissions, image correlation and fracture-surface analysis. Coarser specimens show lower strength and more distributed cracking in the tests. The multi-technique contrasts inform material models but do not isolate grain size from every other microstructural difference or establish a field-scale warning signal.
50. Root-induced preferential flow in the critical zone: Observations, processes, mechanisms, and modeling
Core Problem: Roots can create preferential pathways as well as alter matrix structure, complicating the hydraulic representation of vegetated hillslopes.
Key Innovation: The verified review synthesizes observations, processes, mechanisms and modelling of root-induced preferential flow in the critical zone. The available record is title-level; it does not establish a universal direction or magnitude of the effect on slope stability.
51. Global rainfall erosivity: A comprehensive review and appropriate applications of empirical erosivity models
Core Problem: Empirical rainfall-erosivity models need application conditions that distinguish an erosive forcing index from actual sediment loss.
Key Innovation: The verified review addresses global erosivity and appropriate use of empirical models. With no matching abstract, specific best-performing formulas, geographic error estimates and implications for landslide triggering are left unassigned.
52. A unified density-based PF-GIMP framework from seepage-induced fracture to post-failure large deformation in saturated porous media
Core Problem: The transition from saturated seepage-driven fracture to post-failure motion is difficult to represent within one numerical formulation.
Key Innovation: The verified title proposes a density-based PF-GIMP framework spanning those stages. The supplied abstract is an unrelated general MPM review, so it is not used to claim benchmark accuracy, validation cases or a demonstrated breakthrough in hazard prediction.
53. One-dimensional compressive deformation behavior and underlying mechanisms of saturated coastal soft clay under vibratory loading
Core Problem: Cumulative compression of soft coastal clay depends jointly on vibration amplitude, frequency and initial consolidation state.
Key Innovation: Oedometer and thermogravimetric tests identify state-dependent strain trends and a shift in the most deforming tested frequency with consolidation pressure. Changes in bound water suggest a contributing mechanism. The laboratory amplitude transition and water-state interpretation require further evidence before becoming field settlement or seismic design rules.