TerraMosaic Daily Digest: October 24, 2026
Daily Summary
Earthquake observations and source models depend on different kinds of prior information. Coseis-Diff applies physics-based learned priors to recover deformation from a single interferogram without external auxiliary corrections at inference. Its selected real-event tests include comparisons with GNSS and a dislocation model. Reconstruction of the Kahramanmaraş doublet addresses the geological counterpart: segmentation, terminal fault complexity and deformation partitioning help explain the unusually broad rupture footprint.
At the material scale, several studies separate processes that can look similar in bulk measurements. Hydraulic gradients and cyclic loads jointly promote fine-particle migration in gravel–sand mixtures. Calcareous-sand tests show that stabilized breakage does not necessarily mean macroscopic shakedown, while effective-medium modelling links coordination and fabric to different cyclic moduli. Frozen and unfrozen compaction experiments further distinguish breakage-derived densification from the residual contribution attributed to rearrangement.
Engineering comparisons remain sensitive to site and evaluation conditions. A detailed soil–structure model of a tall building gives modest changes in most above-ground demands at a stiff site, alongside much larger local basement differences. Snow-emergency routing uses expert-weighted accessibility and redundancy criteria for preparedness. Environmental input studies add another constraint: rainfall-product rankings vary by scale and terrain, while a coastal suspended-solids reconstruction is evaluated mainly through satellite-derived references.
Key Trends
The selected work connects deformation recovery, source geometry and soil mechanics with the conditions used to evaluate engineering applications.
- Inference-time independence still uses learned information: Coseis-Diff reduces the need for external corrections during application, but relies on physics-based training priors. Its real-event checks help evaluate recovery in specific settings without supplying complete spatial ground truth.
- Fault complexity matters beyond the rupture trace: Bayesian slip modelling of the Kahramanmaraş doublet relates segmentation, afterslip and broad terminal damage zones to fault structure. The interpretations constrain the event's geometry and extent, rather than providing a prospective magnitude rule.
- Cyclic response needs several mechanical diagnostics: Fines migration, breakage, contractancy and cumulative deformation respond differently to loading and material state. The calcareous-sand tests and DEM-based moduli study show why a single degradation indicator may not identify the overall response class.
- Local structural demands can differ from global response: The tall-building SSI comparison retains similar above-ground behavior while changing some basement and foundation demands. The result is tied to the stiff site, illustrating the importance of checking the response location and supporting ground conditions.
- Environmental products require matched reference checks: The Helmand evaluation separates daily/monthly, seasonal and elevation effects. The China coastal TSS reconstruction uses satellite-derived reference information, so cross-product agreement remains distinct from independent in-situ concentration validation.
Selected Papers
The selection links earthquake-deformation recovery and source geometry to soil response and engineering protection. Experimental and numerical studies provide mechanism-level evidence, while the applied papers address infrastructure performance, emergency access and environmental inputs. All entries appear in one descending relevance-score sequence, with findings limited to the evidence available for each study.
1. Coseis‐Diff: A Data‐Independent Diffusion Transformer for Single Coseismic Interferogram Error Correction and Deformation Recovery
Core Problem: A single interferogram mixes deformation with atmospheric, orbital and other errors.
Key Innovation: Coseis-Diff uses physics-based learned priors and needs no external auxiliary corrections at inference. Synthetic and selected real-earthquake tests, including GNSS comparisons, support recovery in those settings. The inference-time independence still relies on physics-based training; an Okada comparison is a model constraint rather than complete spatial ground truth.
2. The Role of Fault Geometry in the Size of the 2023 Kahramanmaraş Earthquake Doublet
Core Problem: Fault complexity and parallel structures can affect rupture segmentation and spatial extent.
Key Innovation: Bayesian slip models and postseismic observations examine the 2023 Kahramanmaraş doublet. Straight and complex segments show different slip patterns; damage-related termination and strain partitioning are geological interpretations, not a demonstration that fault geometry alone predicts earthquake size.
3. Fine particle migration of gravel-sand mixtures under coupled hydraulic gradient and cyclic loading
Core Problem: Hydraulic gradients and repeated loading can jointly mobilize fines.
Key Innovation: Tests on two gravel–sand gradations show greater migration under cyclic loading and sensitivity to loading frequency and mean load. An empirical accumulation model is compared with study and literature data. The results do not define a universal piping threshold or a field dam-failure forecast.
4. Integrated High‐Fidelity Soil–Structure‐Interaction Analysis of a Tall Reinforced Concrete Core Wall Building
Core Problem: Soil compliance can change foundation and local structural demands differently from above-ground response.
Key Innovation: A nonlinear SSI model of a 25-storey core-wall building gives broadly similar above-ground demands to a fixed-base model but substantially different basement responses. The limited overall SSI influence reflects this relatively stiff site; the comparison should not be generalized to soft soils or all building heights.
5. Boulder-resolved MPM-DEM simulation for engineering-scale design of debris-flow baffle protection along railway corridors
Core Problem: Large boulders can affect the loads and performance of railway-corridor barriers.
Key Innovation: Title-level scope: engineering-scale MPM–DEM simulation with resolved boulders for baffle protection. Barrier layouts, loading gains, validation and field design limits remain unassessed without a matching abstract.
6. Erosion and deposition dynamics of debris flows: Reconstructing runoff-driven events in the Italian Alps
Core Problem: Runoff-driven events exchange material with the channel along their path.
Key Innovation: Title-level scope: reconstructing Italian Alpine events to assess erosion and deposition dynamics. Event observations, entrainment relations and model performance cannot be inferred from the citation alone.
7. Advances and challenges in monitoring, risk assessment, and early warning of ice avalanches in High Mountain Asia
Core Problem: Monitoring and warning of ice avalanches require connecting observations, movement and exposure.
Key Innovation: Title-level scope: advances and challenges in High Mountain Asia monitoring, risk assessment and early warning. The review's evidence selection, recommendations and validation limits remain unassessed without a matching abstract.
8. Energy-based machine learning models for predicting soil liquefaction potential
Core Problem: Model predictions depend on how seismic demand and soil resistance are represented.
Key Innovation: Verified title-level scope: energy-based machine learning for liquefaction potential. The supplied abstract instead describes SPT/stress-based ANN and SVM models and does not establish an exact match to the energy formulation; its dataset and performance claims are not used.
9. Shaking table test of high-speed railway bridges with novel variable friction pendulum bearings
Core Problem: Strong shaking can produce excessive displacement in friction-pendulum bearings.
Key Innovation: The available abstract introduces a variable-friction bearing and a 1/12-scale shaking-table setup. It does not provide comparative response results, so no displacement-reduction factor or full-scale performance claim is inferred.
10. Machine Learning‐Based Prediction of Drift Limit States of Concrete‐Filled Steel Tube Columns
Core Problem: Drift-limit estimates vary with column properties and the damage mechanism being considered.
Key Innovation: Experimental literature calibrates a numerical/ML workflow, followed by 10,000 sampled column configurations. These are model-generated cases, not 10,000 physical tests. Predicted damage limits and frame fragilities remain dependent on the calibration and numerical representation.
11. Pre-designation of emergency resource mobilization routes for heavy snowfall disaster response: a GIS-based multi-criteria decision framework integrating AHP and dijkstra’s algorithm
Core Problem: Shared road vulnerabilities can compromise several candidate emergency routes simultaneously.
Key Innovation: AHP weights from 60 experts and GIS/Dijkstra routing identify mobility, connectivity and redundancy priorities in Gangneung. The exercise supports predesignation; expert suitability weights are not measured route-failure probabilities or demonstrated real-event travel-time gains.
12. Assessing climate change impacts on coastal dynamics along the Pakistan coastline using remote sensing and machine learning
Core Problem: Coastal change varies across deltas, beaches and developed port settings.
Key Innovation: Landsat/RF mapping documents Pakistan coastline changes and supports conditional scenario projections. Relationships with sea-level anomaly and temperature do not uniquely attribute change to climate, and projected local rates depend on forcing and modelling assumptions.
13. Local scour around a submarine pipeline in a shell–sand mixed seabed
Core Problem: Shell content may change sediment behavior around exposed pipelines.
Key Innovation: Title-level scope: local scour in a shell–sand mixed seabed. Bed composition, scour evolution, measured effects and transferable protection criteria remain unassessed without an abstract.
14. Experimental and numerical study of wave attenuation and flow characteristics induced by emergent canopy vegetation
Core Problem: Canopy-induced changes to flow matter when evaluating wave protection.
Key Innovation: Title-level scope: experimental and numerical attenuation and flow characteristics for emergent vegetation. Attenuation values, plant configurations and model validation remain unassessed without a matching abstract.
15. A semi-analytical effective-medium model of small-strain stress-dilatancy during cyclic undrained shear: Mechanical links to coordination number and contact fabric
Core Problem: Contact coordination and fabric affect cyclic stiffness and contractancy.
Key Innovation: An effective-medium model with two empirical corrections reproduces small-strain moduli in DEM paths. It links microstructure to stress–dilatancy under the tested assumptions; agreement with DEM is not independent field validation of liquefaction resistance.
16. Stress-dependent cyclic behaviour and shakedown-based prediction model of calcareous sand for coral reef runway
Core Problem: Macroscopic shakedown and stabilization of particle breakage need not coincide.
Key Innovation: Drained cyclic tests of calcareous sand show response classes jointly dependent on stress ratio and confinement. An empirical model fits the calibration range, while no single stress ratio separates response classes across all pressures. The work concerns runway-material behavior rather than a direct liquefaction-warning criterion.
17. Particle-scale insights into sand deformation during open-ended pile installation using DIC and X-ray CT
Core Problem: Installation changes sand fabric and strain around an open-ended pile.
Key Innovation: Two model installations combine 2D DIC with post-installation X-ray CT, relating contact orientations to compressive strains. The measurements clarify local mechanisms but do not directly establish field pile capacity or long-term foundation behavior.
18. A modified triaxial test for characterising soil–structure interface behaviour
Core Problem: Interface shear measurements depend on apparatus stress state and boundary conditions.
Key Innovation: A modified triaxial configuration gives friction angles close to a modified direct-shear method, with simulations indicating relatively uniform stress mobilization. These comparisons support the apparatus assessment, without proving universal representativeness for all interfaces and stress paths.
19. Effects of xanthan gum biopolymer on elastic wave velocity and relative permittivity of repeatedly frozen–thawed soils
Core Problem: Elastic and electromagnetic response can help characterize treated soil under freezing cycles.
Key Innovation: Four-cycle tests examine xanthan-gum content at fixed density and saturation. More stable wave responses in some treatments are a proxy observation, not direct proof of long-term strength retention or field slope stabilization.
20. Separation of particle breakage and rearrangement contributions to compaction densification of sand at unfrozen and frozen conditions
Core Problem: Breakage and rearrangement make different contributions to densification.
Key Innovation: Tests at unfrozen and frozen temperatures partition a breakage-derived volume change from a residual attributed to rearrangement. This decomposition explains some expansive frozen responses but remains dependent on its measurement and attribution assumptions; fitted equations are not independently established across other soils.
21. Assessing the Reliability of Multi-Source Gridded Precipitation Products for Use as Inputs to Hydrological and Hydroclimatic Applications: A Case Study of the Helmand River Basin, Afghanistan
Core Problem: Sparse gauges and complex terrain complicate selection of rainfall inputs.
Key Innovation: Eight products are compared with 18 Helmand stations across scales, seasons and elevations. MSWEP and ERA5-Land are relatively balanced within this evaluation; ranking depends on resolution, extraction and reference data, and does not establish universal flood-forecast superiority.
22. A Decadal Total Suspended Solids Dataset for China’s Coastal Waters Reconstructed from Sentinel-2 Imagery Using a Spatiotemporal Graph Neural Network
Core Problem: Large-scale suspended-sediment histories are difficult to observe consistently.
Key Innovation: A 2016–2025 Sentinel-2/STGNN product is trained on satellite-derived reference TSS and compared with other satellite products. Spatial agreement supports reconstruction assessment, not independent in-situ concentration accuracy or demonstrated storm-event performance.
23. How many directional moments are enough? Sharp symmetry-aware wave-excitation certificates for offshore structures
Core Problem: Routine directional moments can leave several compatible loading predictions.
Key Innovation: Exact discrete-kernel identities bound the response spread consistent with retained moments, showing that a good reconstructed prediction can coexist with broad ambiguity. The certificate covers the specified discrete response model; buoy error and hydrodynamic-model uncertainty need separate treatment.
24. Denoising-guided SAR foundation model
Core Problem: SAR denoising and representation learning may offer complementary features.
Key Innovation: Title-level scope: a denoising-guided SAR foundation model. Pretraining data, transfer breadth, hazard-specific validation and availability remain unassessed; the foundation-model name alone does not establish a disruptive geohazard advance.
25. TSE-Net: Semi-supervised monocular height estimation from single remote sensing images
Core Problem: Single-image height inference needs evidence about scene geometry and training transfer.
Key Innovation: Title-level scope: semi-supervised monocular height estimation for remote-sensing images. Height errors, reference quality and transfer to unstable or damaged terrain remain unassessed without a matching abstract.
26. Quaternary tectonic activity and deformation mechanism of the Wenshu Shan in the northeastern Tibetan Plateau constrained by quantitative geomorphology
Core Problem: Long-term deformation provides geological context for regional tectonic structure.
Key Innovation: Title-level scope: quantitative geomorphology of Quaternary activity at Wenshu Shan. Deformation rates, chronological uncertainty and implications for present earthquake hazard remain unassessed.
27. Hydrologic responses to reforestation in tropical regions - a systematic review
Core Problem: Reforestation can alter several water-balance components with context-dependent effects.
Key Innovation: Title-level scope: systematic review of tropical hydrological responses to reforestation. The included evidence, runoff effects and implications for flood or landslide risk cannot be assessed from the available citation alone.