TerraMosaic Daily Digest: August 29, 2026
Daily Summary
The largest direct cohort centers on earthquake hazard and risk, where regionalization, site effects, and uncertainty quantification are being pushed beyond generic national curves toward basin-, station-, and orientation-specific estimates. Papers on the 2025 U.S. National Seismic Hazard Model for Puerto Rico and the U.S. Virgin Islands, updated probabilistic seismic hazard assessment for northeastern Brazil, directional PSHA, partially nonergodic variability, spatially varying ground-motion models, host-to-target adjustments, sediment-thickness corrections, and basin-geometry or velocity-based site response collectively show that seismic hazard is increasingly being resolved through structured regional heterogeneity. Complementary studies on aftershock sampling, GNSS-supported ShakeMaps, graph-based post-event shaking estimation, and expanded strong-motion databases strengthen the basis for rapid hazard characterization, while liquefaction-targeted design, regional liquefaction models, expected triggering-magnitude mapping, and Vancouver slope-displacement assessment extend the same probabilistic logic to ground failure.
Outside earthquakes, the direct geohazard papers resolve failure and mitigation through material state, geometry, and transport regime. Riverbank erosion mapping in Bangladesh shows that lightweight adaptation of a foundation segmentation model can recover erosion extent and area change from long image sequences, debris-flow flume experiments quantify how barrier loads decline across the deposition zone as effective momentum and deposition height decay, and slope-stability studies use microtremor-informed weak-layer characterization, equivalent-particle-size landslide simulations, and bounded-field reliability analysis to make spatial variability explicit. Flood and subsurface papers extend this state-based framing to human-disturbed and engineered systems: distributed hydrological modelling attributes basin response to changing land and groundwater conditions, urban low-impact-development layout is posed as a runoff-cost optimization problem, FDIP-constrained tunnel inflow prediction and DAS-based earthwork surveys convert subsurface structure into operational decisions, risk-calibrated settlement warning improves shield-tunnelling management, and thermodynamic frozen-soil, fracture-dissolution, and braided-river studies track evolving phase, conduit, and morphodynamic states.
A separate but important resilience and methods stream broadens how hazards are measured, interpreted, and operationalized. Earthquake-risk papers move from component damage toward functionality, accessibility, and consequence through fragility and exposure modelling, fatality and debris prediction, digital twins, healthcare accessibility, inspection-time inference, repair-quantity translation, functional-recovery design, and equity-oriented risk metrics. In parallel, transferable remote-sensing and scientific-machine-learning papers emphasize calibration, geometry, and physical structure rather than generic automation: standardized Earth-observation change-detection benchmarks, multimodal UAV alignment, memory-efficient multi-view stereo, coordinate-anchored large-scene reconstruction, appearance-robust Gaussian splatting, cross-modal image translation, distilled weather ensembles, structured neural operators, prequential Bayesian assimilation, diffusion-based inverse priors, and PDE-discovery agents all expand the monitoring and inference toolkit. These methods are clearly relevant as enabling capability, but the selected papers do not by themselves establish hazard-specific validation.
Key Trends
Five trajectories structure the August 29 literature: site-structured seismic hazard, probabilistic ground failure, function-centered resilience, state-aware geomorphic mechanics, and calibration-driven Earth observation and scientific inference.
- Regional seismic hazard models are becoming explicitly site-structured and nonergodic: The Puerto Rico and U.S. Virgin Islands hazard-model papers, northeastern Brazil PSHA, spatially varying ground-motion models, directional hazard formulations, host-to-target adjustments, coastal-plain path corrections, and basin-sensitive site-response studies all treat regional heterogeneity as a first-order quantity rather than residual noise. The trajectory is toward hazard surfaces that are conditional on local structure, orientation, and epistemic uncertainty.
- Ground-failure assessment is moving toward probabilistic performance targets: Liquefaction-targeted design basis, regional liquefaction model building, national triggering-magnitude mapping, Napier case-history testing, and probabilistic earthquake-induced slope-displacement mapping in Greater Vancouver show a common shift from binary susceptibility statements to risk-calibrated ground-failure metrics. This makes liquefaction and slope displacement more directly usable in engineering design and regional risk workflows.
- Risk analysis is expanding from damage states to functionality, accessibility, and equity: Across functional-recovery design requirements, repair and inspection models, healthcare accessibility, digital twins, fatality modelling, debris blockage prediction, exposure modelling, and disaggregated equity metrics, the earthquake-risk cohort increasingly evaluates whether systems remain usable and for whom. The scientific move is from loss estimation alone to operational consequence and recovery capacity.
- Direct geomorphic and geotechnical hazards are being parameterized by internal state and transport regime: Riverbank erosion segmentation, debris-flow barrier experiments, microtremor-informed slope reliability, equivalent-particle-size landslide simulations, bounded-field rock-slope analysis, tunnel inflow prediction, tunnelling settlement warning, frozen-soil phase-transition modelling, fracture dissolution, and braided-river state transitions all link hazard behaviour to evolving material, hydraulic, or morphodynamic state. The emphasis is on mechanism-resolving controls rather than trigger-only description.
- Transferable Earth-observation and scientific-ML methods now prioritize calibration, geometry, and physical priors: The methods cohort repeatedly addresses viewpoint distortion, multimodal misalignment, reconstruction scale, benchmark comparability, mass conservation, dynamical structure, and constrained inference. Standardized change-detection evaluation, geometry-aware UAV fusion, large-scene 3D reconstruction, cross-modal translation, distilled weather ensembles, operator learning, Bayesian assimilation, and diffusion-based inverse methods therefore read as a maturation of analytic infrastructure rather than as generic end-to-end AI claims.
Selected Papers
The selected papers interleave hazard-tested studies on earthquakes, liquefaction, landslides, flooding, erosion, tunnelling, frozen ground, and infrastructure recovery with a secondary methods cohort in remote sensing, 3D reconstruction, and scientific machine learning. Read the first group for substantive geohazard findings and the second for transferable analytical capability that still requires hazard-specific validation.
1. Riverbank Erosion Analysis in Bangladesh Using Spatiotemporal Segmentation
Core Problem: Manual riverbank erosion mapping from imagery is slow and inconsistent across large river networks.
Key Innovation: Fine-tunes only SAM's lightweight decoder to map erosion and area change accurately.
2. Effect of barrier location in the deposition zone on debris flow load reduction: a quantitative experimental study
Core Problem: The load-reduction effect of barrier placement within debris-flow deposition zones has not been quantified.
Key Innovation: Flume experiments separate how velocity loss, deposition height, and effective momentum control peak barrier loads.
3. A microtremor-informed system reliability analysis framework for weak interlayered rock slopes considering spatial variability
Core Problem: Assess reliability of weak interlayered rock slopes while accounting for spatial variability.
Key Innovation: Microtremor-informed reliability framework for spatially variable weak interlayered rock slopes.
4. Discrete element modeling of landslides: Equivalent particle-size dependence of mobility and mass entrainment on erodible slopes
Core Problem: Field-scale DEM landslide simulations need coarse equivalent particles, but the mobility bias this creates is poorly understood.
Key Innovation: Real-terrain entraining DEM experiments show strong equivalent-particle-size control on runout, entrainment, and energy conversion.
5. 2025 U.S. Geological Survey National Seismic Hazard Model for Puerto Rico and the U.S. Virgin Islands: Overview of Model and Hazard Results
Core Problem: Puerto Rico and the U.S. Virgin Islands needed a modernized hazard model incorporating two decades of new data and uncertainty treatment.
Key Innovation: Delivers the 2025 PRVI NSHM with subduction-interface inversion, updated ground-motion characterization, and gridwide epistemic uncertainty outputs.
6. Development of Liquefaction-Targeted Design Basis in U.S. Seismic Provisions
Core Problem: U.S. liquefaction screening criteria are not aligned with risk-targeted structural design objectives.
Key Innovation: Develops a probabilistic liquefaction-targeted design basis and proposed PGA metric calibrated to acceptable triggering risk.
7. Near-Source Ground Motion From Normal Faults: Insights From the Physics-Based Numerical Simulations of the 1980 Irpinia Earthquake, Southern Italy
Core Problem: Need physically grounded insight into near-source ground motion and deformation in normal-fault events.
Key Innovation: Uses 3D spectral-element simulation of the 1980 Irpinia rupture to resolve directivity, polarization, and permanent deformation patterns.
8. Evaluation of Earthquake-Induced Slope Displacements in Greater Vancouver Based on Probabilistic Framework and Using NBCC 2020 and EqDisp Code Platform
Core Problem: Need probabilistic estimates of earthquake-induced slope displacements that reflect local hazard sources and model uncertainty.
Key Innovation: Integrates PSHA with seismic-displacement prediction in the EqDisp platform to generate displacement hazard curves and regional landslide hazard maps.
9. Suite of Regional Seismic Velocity Models for the U.S. Atlantic and Gulf Coastal Plains (CPVMv2.0): Incorporating Epistemic Uncertainty in Velocity Models
Core Problem: Atlantic and Gulf Coastal Plain hazard studies lack regional velocity models with explicit epistemic uncertainty.
Key Innovation: Builds CPVMv2.0 from measured profiles and stratigraphic constraints, with alternative median profiles and quantified uncertainty.
10. Next-Generation Probabilistic Liquefaction Model Building at the Regional Scale
Core Problem: Regional liquefaction studies rely on manifestation models not built from modern large CPT case-history datasets.
Key Innovation: Builds severity-based probabilistic free-field liquefaction models with spatial correlation and probabilistic model-scoring procedures.
11. OpenAmp: An Open-Source Site Amplification Database to Unlock the Potential of Machine Learning
Core Problem: Near-surface amplification studies lack a consistently curated, metadata-rich site database.
Key Innovation: Releases 1,742 site amplification functions with new site descriptors and vector proxies tailored to geospatial and machine-learning use.
12. Probabilistic Hazard Assessment of Coseismic Elastic Surface Deformation for the Leech River Valley-Devil's Mountain Fault System Using Stochastic Source Modeling
Core Problem: Ignoring interaction between the Leech River Valley and Devil's Mountain faults may underestimate surface deformation hazard near Victoria.
Key Innovation: Builds a stochastic multisegment fault-slip deformation hazard model and quantifies the deformation increase from combined rupture scenarios.
13. Diffusion Distillation for Efficient Weather Ensembles
Core Problem: Diffusion weather ensembles are accurate but too expensive for fast operational sampling.
Key Innovation: Distills multi-step diffusion ensembles into a single-step student while preserving forecast and extreme-event skill.
14. GeBDA: Building Damage Assessment as Text-Based Sequence Prediction
Core Problem: Localize buildings and classify damage directly from bi-temporal satellite imagery using a general-purpose vision-language model.
Key Innovation: Recasts building damage assessment as autoregressive text sequence prediction of bounding boxes plus damage labels.
15. Influence of tropical cyclone activity on seismicity along the Pacific-Philippine plate boundary in the West North Pacific
Core Problem: Test whether variability in tropical cyclone activity is associated with earthquake occurrence in the West North Pacific.
Key Innovation: Shows depth-dependent correlations between cyclone activity, especially ACE, and shallow seismicity along the Pacific-Philippine boundary.
16. A comprehensive distributed hydrological framework for assessing flood responses to environmental change under multiple human disturbances
Core Problem: Assess flood responses to environmental change under multiple human disturbances within one distributed framework.
Key Innovation: Comprehensive distributed hydrological framework for scenario-based flood-response assessment.
17. Regional Site Response Model via Uncertainty Projected Mapping: Application to Kanto Region, Japan
Core Problem: Proxy-based regional site-response maps can misrepresent uncertainty and extrapolate poorly across space.
Key Innovation: Applies Uncertainty Projected Mapping to regional site response and compares it against proxy and kriging approaches.
18. Quantifying Multidimensional Effects in Linear Site Response Using 1D and 2D Simulations of Sedimentary Basins in Wellington, New Zealand
Core Problem: One-dimensional site-response assumptions can miss important topographic and stratigraphic amplification effects.
Key Innovation: Quantifies where 2D basin simulations materially change amplification relative to 1D models and to observed station behavior.
19. Spatial Cross-Correlation Models for Next-Generation Amplitude and Cumulative Intensity Measures
Core Problem: Next-generation intensity measures lacked spatial and cross-spatial correlation models for regional analysis.
Key Innovation: Develops PCA-based spatial cross-correlation models linking conventional and next-generation intensity measures.
20. Hazard-Consistent Aftershock Ground Motion Intensity Map and Arrival Time Sampling for Regional Seismic Risk Analysis
Core Problem: Full aftershock catalogs are too large for tractable regional seismic risk analysis immediately after major earthquakes.
Key Innovation: Samples aftershock scenarios, intensity maps, and arrival times while preserving the hazard implied by the full catalog.
21. The Influence of Fault Rupture Velocity on Ground Motions and Building Demands in the Near-Fault Region
Core Problem: The regional demand consequences of near-shear and supershear rupture propagation remain insufficiently quantified.
Key Innovation: Combines observed Turkey motions and fault-to-structure simulations to show strong demand amplification from faster ruptures.
22. Nonstationary Spatial Correlation of Earthquake Ground Motions in California
Core Problem: Stationary spatial-correlation assumptions hide regional differences in how earthquake residuals decay with distance.
Key Innovation: Builds a Bayesian nonstationary correlation model that captures geographically varying decay behavior and reduces residual variance.
23. Incorporating Ground Motion Directionality into Probabilistic Seismic Hazard Analysis
Core Problem: Conventional PSHA only gives central-tendency horizontal shaking and misses systematic directionality relevant to structures.
Key Innovation: Extends PSHA to specified horizontal orientations by adapting ground-motion model outputs to directional hazard.
24. Ground-Motion Aleatory-Variability Models for Puerto Rico and the U.S. Virgin Islands
Core Problem: Regional hazard studies in PRVI needed defensible aleatory-variability logic trees that blend global and local evidence.
Key Innovation: Constructs partially nonergodic variability models partitioned into event, site-to-site, and single-station components.
25. National-scale Mapping of the Expected Earthquake Magnitude for Soil Liquefaction Triggering Analyses in Italy
Core Problem: Liquefaction triggering analyses need a site-specific expected magnitude estimate that is not tied to one disputed shaking metric.
Key Innovation: Maps expected triggering magnitudes for Italy using historical seismicity and empirical distance-to-liquefaction relations.
26. Update of Mainshock Intensity Map Using Aftershock Records Collected From Temporary Seismic Stations: Case Study of 2017 M5.4 Pohang Earthquake
Core Problem: Mainshock intensity maps can remain biased when local site effects are unresolved at key locations.
Key Innovation: Uses aftershock-derived nonergodic site terms from temporary stations to materially improve mainshock PGA map accuracy.
27. Validation of ShakeMaps Created From Global Navigation Satellite System (GNSS) Velocities
Core Problem: ShakeMaps may underuse dense geodetic observations that could improve strong-motion interpolation.
Key Innovation: Validates inclusion of GNSS-derived velocities in ShakeMap generation across multiple real earthquakes.
28. An Extended and Updated Strong Ground-Motion Database for the Chilean Subduction Earthquakes
Core Problem: Chile lacked a larger modern strong-motion database with richer metadata, site terms, and distance calculations.
Key Innovation: Publishes a substantially expanded subduction-zone flatfile with event classification, site proxies, and multiple intensity measures.
29. Ground-Motion Characterization for the 2025 U.S. National Seismic Hazard Model for Puerto Rico and the U.S. Virgin Islands
Core Problem: Update regional ground-motion characterization and uncertainty treatment for the 2025 national seismic hazard model.
Key Innovation: Introduces scaled-backbone median models and independent aleatory-variability logic trees using regional data.
30. Advancing Probabilistic Seismic Hazard Assessment for Northeastern Brazil: Toward a Comprehensive Treatment of Epistemic Uncertainties
Core Problem: Need an updated probabilistic seismic hazard model for Northeastern Brazil with fuller uncertainty treatment.
Key Innovation: Builds an updated PSHA model with gridded hazard outputs and sensitivity analysis on key epistemic drivers.
31. Evaluating the U.S. Geological Survey's Earthquake Shaking Hazard Forecasts and Their Implications on Seismic Risk
Core Problem: Need to quantify how successive USGS hazard-model updates alter exposure and risk, not just hazard maps.
Key Innovation: Compares six NSHM update cycles and converts hazard changes into impacts on people, buildings, and economic exposure.
32. Suitability of Ground-Motion Models for Seismic Hazard Assessments in Puerto Rico and the US Virgin Islands
Core Problem: Available ground-motion models are biased for Puerto Rico and the U.S. Virgin Islands and may distort hazard estimates.
Key Innovation: Compiles a large regional dataset and derives region-specific site and constant-term adjustments for model selection in the 2025 NSHM.
33. Development of a Global Earthquake Impact Database (GEID) for the Verification and Calibration of Loss Models
Core Problem: Loss-model verification is limited by scarce, inconsistent global impact and shaking data.
Key Innovation: Builds a public global database linking damage, losses, shaking, and rupture information in OpenQuake-ready formats.
34. Influence of Site Amplification Approaches in PSRA: Insights From Five Different Cities
Core Problem: Proxy-based site amplification choices can materially bias regional seismic risk results.
Key Innovation: Compares 17 site-effects treatments across five cities to quantify their influence on probabilistic risk metrics.
35. Host-To-Target Site Response Adjustments for Site-Specific Probabilistic Seismic Hazard Analysis
Core Problem: Traditional buried-rock workflows complicate site-specific PSHA and may underrepresent uncertainty in site adjustments.
Key Innovation: Implements host-to-target site response adjustments with logic-tree uncertainty for a Level 3 PSHA case study.
36. Examination of Very High Peak Ground Acceleration and Short-Period Spectral Accelerations on Soft-Soil Sites from the New Zealand National Seismic Hazard Model
Core Problem: Very high short-period motions on soft-soil sites in the New Zealand hazard model may be overestimated by current nonlinear amplification treatment.
Key Innovation: Benchmarks hazard-model predictions against observations and proposes adjusted nonlinear PGA functions for soft-soil conditions.
37. Assessments of Liquefaction Triggering and Manifestation at Three Case History Sites in Napier Based on 1931 Hawke's Bay, New Zealand, Earthquake
Core Problem: Simplified triggering and manifestation models overpredict liquefaction severity at Napier case-history sites.
Key Innovation: Tests multiple susceptibility, fines-content, saturation, and thin-layer adjustments against three 1931 Hawke's Bay sites to diagnose persistent overprediction.
38. Dandelion: A Spherical Flower for Neural Simulation of Planetary Dynamics
Core Problem: Standard neural PDE architectures distort dynamics defined on the sphere.
Key Innovation: Introduces a warp-based spherical solver and benchmark suite spanning atmospheric and ocean dynamics.
39. Quanta Perception as Probabilistic Events
Core Problem: Photon-counting quanta sensors generate data too fast for practical real-time perception.
Key Innovation: Replaces frame reconstruction with recursive probabilistic events for kilohertz perception in extreme low-light conditions.
40. From Perspective to Fisheye Depth Estimation and Open-Vocabulary Segmentation
Core Problem: Perspective-trained depth and segmentation models fail on fisheye imagery because of radial-distortion shift.
Key Innovation: Learns distortion extenders that align fisheye latent embeddings with perspective-trained foundation models.
41. Beyond Pairwise Graphs in Science: Hypergraph Adaptive Wavelet Operators for Parametric PDEs
Core Problem: Pairwise graph operators miss group-wise couplings in PDE learning on complex geometries.
Key Innovation: Introduces a hypergraph wavelet neural operator with adaptive spectral coverage and stable multi-step rollouts.
42. GAAT: Geometry-Aware Alignment Transformer for Multimodal UAV Perception
Core Problem: Multimodal UAV perception suffers from local misalignment across RGB, IR, SAR, and depth sensors.
Key Innovation: Estimates local correspondence reliability before sparse cross-modal fusion and benchmarks acquisition-state effects.
43. ZipMVS: Multi-View Stereo with Compressed Cost Volumes
Core Problem: Standard MVS reconstruction is memory-heavy, limiting efficient high-quality 3D mapping.
Key Innovation: Compressed cost volumes and a new depth-hypothesis strategy greatly reduce GPU memory while preserving reconstruction quality.
44. Ex-Sim(3)-Reg: 2D-3D Correspondence Pruning via Extended Sim(3) Registration
Core Problem: 2D-3D correspondence pruning breaks down when monocular depth priors are noisy under low-inlier or distorted conditions.
Key Innovation: Recasts pruning as extended Sim(3) registration that explicitly models depth noise and improves registration recall.
45. WilLaGS: Latent-Conditional 3D Appearance Fields for Robust Gaussian Splatting In-the-Wild
Core Problem: In-the-wild scene reconstruction breaks when appearance varies continuously and transient objects violate multi-view consistency.
Key Innovation: Latent-conditional global appearance modeling, dynamic local illumination fields, and self-supervised perceptual masking for robust 3DGS.
46. A comprehensive and trustworthy benchmark of AI methods for change detection in Earth observation
Core Problem: EO change-detection research is hard to compare because protocols differ and efficiency is rarely evaluated alongside accuracy.
Key Innovation: A standardized open benchmark across ten datasets and ten architectures with matched training, pretraining, and efficiency reporting.
47. GeoFF3D: Coordinate-Anchored Feed-Forward Reconstruction for Large-Scale UAV Mapping
Core Problem: Feed-forward 3D reconstruction methods do not scale cleanly to large UAV surveys and Sim(3) aggregation can be unstable.
Key Innovation: Combines coordinate-anchored metric-frame reconstruction with hierarchical spatial chunking and shared-view prior propagation.
48. Learning the Target Priors Before Image Translation: A Decoupled Training Paradigm for Cross-Modal Image Translation in Remote Sensing
Core Problem: Translate between remote-sensing modalities while preserving source content and matching target-domain priors from scarce paired data.
Key Innovation: Decouples target-prior learning from conditional translation, leveraging large unpaired imagery and parameter-efficient adaptation.
49. Climate Physics Dynamic Matching
Core Problem: Forecast atmospheric dynamics while blending physical priors with data-driven generative modeling under modest compute.
Key Innovation: Introduces a variational, simulation-free dynamics-informed framework with advection-type physics priors for weather prediction.
50. Bayesian Experimental Design for Model Discrepancy Calibration: A Rivalry between Kullback--Leibler Divergence and Wasserstein Distance
Core Problem: BED utility choices behave differently when model discrepancy contaminates inference.
Key Innovation: Clarifies when KL or Wasserstein utilities are preferable for sequential source inversion.
51. Prequential posteriors
Core Problem: Deep generative forecasting models have intractable likelihoods that block standard Bayesian data assimilation.
Key Innovation: Defines prequential posteriors and couples them with scalable waste-free SMC for forecasting-model assimilation.
52. Bias-Aware Machine Learning Spatial Downscaling of GRACE Signals: Application to the Bug River Basin
Core Problem: GRACE terrestrial water-storage observations are too coarse and temporally gappy for sub-basin applications.
Key Innovation: Combines leakage-safe temporal imputation with bias-aware, mass-conserving ML downscaling of GRACE fields.
53. Non-probabilistic reliability analysis of rock structures using optimized super-ellipsoid convex modelling considering spatial variability
Core Problem: Reliability analysis of rock structures needs a way to represent spatial variability and bounded uncertainty without strong probabilistic assumptions.
Key Innovation: Combines optimized super-ellipsoid convex bounds, non-probabilistic series expansion, and FLAC-based slope analysis to derive stability bounds.
54. Component-To-Component Variability of Response Spectral Ordinates from Strike-Slip Earthquakes
Core Problem: Orientation-sensitive definitions can misestimate component-to-component variability in recorded shaking.
Key Innovation: Proposes orientation-independent equations for total variability of random horizontal spectral response.
55. Structural Implications of Near-Fault Directivity Representations for Bridges
Core Problem: New directivity-amplified spectra for bridge design lacked validation against nonlinear bridge performance.
Key Innovation: Compares hazard-consistent and code-factor directivity spectra using 3D bridge analyses across motion incidence angles.
56. Regional Earthquake Ground Motion Simulations for Southern California With EQSIM: Insights From the 2008 Chino Hills, 2024 Highland Park, and 2021 Carson Earthquakes
Core Problem: Practical confidence in deterministic 3D simulations for real urban earthquakes remains limited.
Key Innovation: Validates GPU-efficient EQSIM simulations against three Los Angeles events and demonstrates rapid ensemble-ready regional modeling.
57. Rapid City-Scale Earthquake Assessment by Combining Numerical Simulation and Sparse Sensing
Core Problem: City-scale seismic risk prediction is hard when dense sensor deployments and exhaustive scenario sets are infeasible.
Key Innovation: Combines scenario simulation, POD-based mode extraction, and optimized sensor placement for rapid city-scale assessment.
58. Site Response Models Based on Geometric Parameters for Southern California Sedimentary Basins
Core Problem: Standard one-dimensional basin proxies do not capture lateral and geometric controls on site amplification.
Key Innovation: Introduces multidimensional basin geometry descriptors that measurably reduce site-to-site residual variability in Southern California.
59. Bayesian Model Updating and Model Selection of Ground Motion Models for Crustal Earthquakes in Colombia
Core Problem: Colombia lacked a Bayesian comparison of local and global ground-motion models for national hazard use.
Key Innovation: Regionalizes candidate GMMs with Bayesian updating and compares predictive performance with fully Bayesian model selection.
60. Utilization of Spatially Distributed mHVSR to Assess the Effectiveness of Ground Response Analysis
Core Problem: Analysts lack an efficient field indicator for when one-dimensional ground response analysis is likely unreliable.
Key Innovation: Uses spatial variability in distributed mHVSR measurements to predict where transfer-function misfit and site complexity are likely.
61. Adaptive Attention Optimized Deep Learning With Vision Transformers for Fine Grained Earthquake Structural Damage Detection
Core Problem: Manual structural damage assessment after major earthquakes is slow, subjective, and hard to scale.
Key Innovation: Builds a vision-transformer damage classifier with strong accuracy and interpretable Grad-CAM localization on UAV imagery.
62. Observed Ground Motions That Exceeded Design Response Spectra in the Western United States
Core Problem: Debate persists over whether very high design spectra are physically realistic or overly conservative.
Key Innovation: Documents real recording exceedances of several contemporary U.S. design spectra and interprets them with hazard disaggregation.
63. Opportunities for the U.S. Geological Survey's National Seismic Hazard Model to Improve Seismic Risk Assessment of Critical Infrastructure
Core Problem: Current national hazard products omit extensions needed for realistic regional critical-infrastructure risk assessment.
Key Innovation: Demonstrates the added value of vertical hazard curves, stochastic event sets, and probabilistic ground-failure maps for infrastructure risk.
64. Partitioning Site Amplification and κ 0 in Earthquake Recordings
Core Problem: Site amplification and kappa are strongly coupled, making direct estimation from observed spectra unstable.
Key Innovation: Shows broadband Bayesian inversion is more robust than slope-based methods for separating amplification from spectral decay.
65. Assessment of Site Response and Building Damage Following the Mw 6.4 Copiapó Earthquake
Core Problem: A moderate 2025 Chile earthquake caused unusually strong motions and widespread nonstructural damage that needed explanation.
Key Innovation: Combines ambient-vibration site characterization and building dynamic-property surveys to interpret damage patterns after the Copiapo event.
66. Empirical Ground-Motion Model With Spatially Varying Coefficients for Improved Ground-Motion Prediction Accuracy at Seismic Observation Stations in Japan
Core Problem: Station-specific ground-motion prediction accuracy is limited by spatially uniform coefficients and imperfect site characterization.
Key Innovation: Introduces spatially varying coefficients plus empirical site terms to improve nonergodic peak-acceleration prediction in Japan.
67. Assessing Post-Earthquake Emergency Healthcare Accessibility Considering Damage to Hospital Buildings and Transportation Infrastructure
Core Problem: Post-earthquake healthcare access depends on coupled hospital damage and bridge-network disruption, which are rarely assessed together.
Key Innovation: Integrates hospital capacity loss and road-network damage in a regional accessibility model for Cascadia scenarios.
68. The 2025 Puerto Rico and Virgin Islands U.S. National Seismic Hazard Model Update: Ground Motion Model Selection and Comparison
Core Problem: PRVI hazard calculations required justified GMM selection, regional adjustments, and uncertainty treatment for modern conditions.
Key Innovation: Builds the PRVI GMM logic tree by combining global models, regional corrections, and explicit epistemic and aleatory uncertainty comparisons.
69. Conversion of Seismic Fragility and Vulnerability Models to Alternative Intensity Measures for Regional Risk Analysis
Core Problem: Regional risk analyses struggle when fragility and vulnerability models are defined on incompatible intensity measures.
Key Innovation: Systematically tests multiple conversion strategies across 2,260 cases and extends conversion logic to vulnerability models.
70. Seismic Fragility and Vulnerability of Partially Grouted Flexural-Controlled Reinforced Masonry Buildings in Colombia
Core Problem: Need representative fragility and vulnerability functions for Colombian partially grouted reinforced masonry buildings in national risk assessment.
Key Innovation: Builds census-informed archetypes and derives PEER-PBEE damage and loss curves from large nonlinear analyses.
71. From Warnings to Actions: Factors Influencing Response to Earthquake Early Warning-A Systematic Literature Review
Core Problem: Earthquake early warning effectiveness is limited by fragmented evidence on how people interpret and act on alerts.
Key Innovation: Systematically synthesizes 45 studies into interacting demographic, sociocultural, cognitive, and environmental drivers of EEW response.
72. A Performance-Based Earthquake Fatality Modeling Framework Using Collapse Volume Ratios
Core Problem: Binary collapse assumptions understate earthquake fatalities by ignoring severity of loss of survivable space.
Key Innovation: Elevates collapse volume ratio into a probabilistic fatality framework that can ingest observed or fragility-based collapse information.
73. Exposure Model of Urban Residential Buildings for Seismic Risk Assessment of Prioritized Municipalities in Colombia
Core Problem: National seismic risk assessment needs consistent municipal-scale residential exposure data.
Key Innovation: Builds standardized census-block exposure models for 83 Colombian municipalities with public taxonomies and protocols.
74. Post-Event Ground Motion Estimation using Graph Neural Networks
Core Problem: Estimate site-specific ground motion after earthquakes when recordings are missing.
Key Innovation: Uses graph neural networks to learn spatial dependence and combine multiple observation types for post-event intensity-measure estimation.
75. Validating V S 30 -Based Site Amplification Factors in Hybrid Broadband Ground-Motion Simulation of Small-Magnitude Earthquakes in New Zealand
Core Problem: Vs30-only amplification adjustments in hybrid broadband simulation need empirical validation and better application rules.
Key Innovation: Validates alternative site-factor implementations on New Zealand recordings and proposes basin-aware protocols that reduce residual bias.
76. Insights on Numerical Damping Formulations Gained From Calibrating 2D Ground Response Analyses at Downhole Array Sites
Core Problem: Standard small-strain damping choices make ground response analyses misfit observed transfer functions.
Key Innovation: Shows Rayleigh Mass damping with calibrated inflated Dmin better matches downhole-array transfer functions while remaining computationally efficient.
77. Adjustment for the Gulf Coast and Atlantic Coastal Plain Region Considering Path Terms and Site Terms With Respect to Rupture Distance and Sediment Thickness for the Median NGA-East Ground-Motion Characterization
Core Problem: NGA-East median predictions are biased in Gulf and Atlantic Coastal Plain settings because sediment effects are not captured well.
Key Innovation: Fits sediment-thickness- and rupture-distance-dependent adjustment factors from mixed-effects residual analysis.
78. Damage in Rio Dell From 2022 Ferndale M6.4 Earthquake
Core Problem: Need community-scale damage evidence tied to strong recorded shaking for future earthquake assessments.
Key Innovation: Combines field reconnaissance, state damage data, and an extreme in-city recording to derive empirical fragilities for common housing types.
79. Predicting Seismic Debris Distributions of Collapsed Unreinforced Masonry Structures Through Physics Engines and Machine Learning Algorithms
Core Problem: Debris footprints from unreinforced masonry collapse are hard to quantify but matter for emergency routing and recovery.
Key Innovation: Combines validated physics-engine collapse simulations with machine-learning heatmap prediction to estimate debris distributions and roadway blockage.
80. A Coupled LBM-FDEM Framework for Simulating Solute Transport and Mineral Dissolution Under H-M-C Conditions
Core Problem: CO₂-driven mineral dissolution in fractured reservoirs is hard to simulate under coupled flow-mechanics-chemistry.
Key Innovation: Couples LBM and FDEM with topology updating to resolve evolving fracture dissolution across Peclet regimes.
81. Revisiting Local Context for Long-Horizon Streaming 3D Reconstruction
Core Problem: Online 3D reconstruction from very long videos accumulates drift under bounded memory.
Key Innovation: Keeps context strictly local and predicts reference-frame-equivariant geometry and relative pose for stable long-horizon reconstruction.
82. VersaGauss: A Versatile Framework for Generating Multiphase Dynamics with 3D Gaussians
Core Problem: Existing Gaussian-based dynamic scene methods poorly handle realistic multiphase physical interactions.
Key Innovation: A unified 3D Gaussian framework with a coupled multiphase point method for simulating and rendering interacting materials.
83. Learning to Difference: Adaptive Reversible Differencing (AdaRDiff) for Time Series Forecasting
Core Problem: Deep forecasting models struggle with trends and seasonality when classical differencing choices are fixed by hand.
Key Innovation: Learns reversible weighted differencing and closed-form reconstruction as a plug-in module for long-horizon forecasting.
84. Euclidean Fourier Neural Operators
Core Problem: Make Fourier neural operators transfer consistently across periodic domains of different shapes and sizes.
Key Innovation: Parameterizes spectral kernels as continuous functions of physical wavevector to create domain-independent operator learning.
85. Video Generative Models as Geometry Learner
Core Problem: Learn monocular depth and surface normals efficiently without large geometry-specific labeled datasets.
Key Innovation: Repurposes pretrained video generative models as next-frame predictors for unified, data-efficient geometry estimation.
86. Hypothesize, Evaluate, Refine: A Scientific Agent for PDE Discovery with Unknown Spatial Coefficient Fields
Core Problem: Recover governing PDE structure and unknown spatial coefficient fields in heterogeneous media from noisy trajectories.
Key Innovation: Uses an agentic hypothesis-refinement loop with cross-excitation transfer scoring to jointly discover operators and coefficient fields.
87. Diffusion models as plug-and-play priors
Core Problem: Need a flexible way to combine learned priors with external differentiable constraints.
Key Innovation: Treats pretrained diffusion models as reusable priors for constrained inference.
88. Examining the robustness of Physics-Informed Neural Networks to noise for Inverse Problems
Core Problem: It is unclear whether PINNs are robust enough to solve noisy inverse problems competitively.
Key Innovation: Systematically compares PINNs against FEM-plus-optimization baselines and isolates noise-related failure modes.
89. Optimal layout of LID facilities at an urban basin scale in piedmont cities: Case study of Jinan City, China
Core Problem: Optimize basin-scale placement of low-impact development facilities in a piedmont city.
Key Innovation: Urban-basin optimization method for LID layout in flood-prone piedmont settings.
90. Nonstructural Design Requirements for Functional Recovery Performance
Core Problem: Prescriptive nonstructural design levels needed for functional recovery targets were not quantified.
Key Innovation: Calibrates component importance factors and equipment qualification needs across building systems using FEMA P-58 and ATC-138.
91. Quantification of Direct Benefits of Functional Recovery-Based Design for Seismic Resilience
Core Problem: The socioeconomic benefits and upfront costs of low-damage functional-recovery design remain weakly quantified.
Key Innovation: Estimates repair-cost, demolition-rate, and recovery-time reductions from combined low-damage structural and nonstructural measures.
92. A Comparison of Ground Motion Intensity Measures for Estimating Collapse
Core Problem: Common intensity measures vary in how efficiently and practically they predict collapse across building types.
Key Innovation: Shows velocity-based measures such as PGV and VSI are strong collapse predictors, especially for short-period structures.
93. Validation of Physics-Based Simulated Accelerograms for Seismic Fragility Analyses of Italian RC Existing Building Types
Core Problem: Fragility analysts need evidence that 3D physics-based simulated accelerograms can substitute for recorded motions.
Key Innovation: Validates physics-based simulated broadband motions by comparing resulting fragility curves against record-based benchmarks.
94. Achieving Functionality Targets for Seismically Isolated Steel Buildings Using Gaussian Process Modeling
Core Problem: Designing isolated buildings to meet repair-cost and downtime targets is computationally expensive and iterative.
Key Innovation: Uses Gaussian process surrogate models to map design variables directly to decision variables and optimize feasible designs.
95. A Bayesian Framework for Empirical Estimation of Post-Earthquake Building Inspection Time
Core Problem: Building-level inspection delays after earthquakes are poorly quantified despite their importance for recovery.
Key Innovation: Uses Bayesian inference on Canterbury inspection data to estimate stochastic distributions for rapid and detailed inspection timelines.
96. Assessing the Loss of Building Functionality Post-Earthquake Using Machine Learning
Core Problem: Post-earthquake building functionality depends on uncertain component interactions that are hard to model quickly.
Key Innovation: Applies random forests to identify key damage pathways and generate probabilistic functionality-state estimates from empirical data.
97. Operationalizing Equity: How Disaggregated Risk Metrics Support Equity in Emergency Management
Core Problem: Emergency managers lack practical risk metrics that operationalize equity for earthquake response decisions.
Key Innovation: Tests disaggregated risk metrics in practitioner workshops and shows they improve equity-focused task planning beyond standard loss summaries.
98. Notes on Spatial Correlation for Average Spectral Acceleration: Direct and Indirect Approaches
Core Problem: Practitioners lack clear guidance on modeling spatial correlation for average spectral acceleration without generating unrealistic fields.
Key Innovation: Compares direct and indirect correlation formulations and highlights major pitfalls in indirect ground-motion-field generation.
99. Analytical Solutions for Seismic Earth Pressure in Near-Fault Regions
Core Problem: Conventional methods poorly capture near-fault seismic earth pressure in backfill soils.
Key Innovation: Derives response-spectrum-based analytical active and passive earth-pressure solutions validated by centrifuge testing.
100. Digital Twins for Post-Earthquake Loss Assessment of Building Portfolios
Core Problem: Traditional portfolio damage assessments are too uncertain and do not fully use sensors and ML damage data.
Key Innovation: Proposes a building-portfolio digital twin that fuses detailed inventories, sensors, and ML outputs to reduce loss-assessment uncertainty.
101. Data-Driven Zonation for Seismic Exposure Modeling via Machine Learning
Core Problem: Expert-defined homogeneous zones for exposure models are slow, subjective, and hard to scale.
Key Innovation: Automates seismic exposure zonation from census, cadastral, and geospatial data using unsupervised ML and spatial clustering.
102. Beyond Deaths and Dollars: Modeling Requirements to Support Equity-Focused Disaster Risk Management
Core Problem: Standard loss metrics do not support equity-centered disaster management decisions.
Key Innovation: Derives practitioner-informed requirements for demographically disaggregated and actionable earthquake impact metrics.
103. Cross-Correlations Between Fourier Spectral Ordinates and Ground-Motion Intensity Measures: An Empirical Study Using the NGA-West2 Database
Core Problem: Engineering use of Fourier spectral ordinates is limited by missing empirical correlations with standard intensity measures.
Key Innovation: Maps total and partial correlations between Fourier amplitudes and multiple intensity measures across frequency and magnitude regimes.
104. Evaluating the Seismic Damage Dynamics of Buildings’ Vertical Growth in Lima, Peru
Core Problem: Incremental vertical growth of informal housing increases earthquake vulnerability but is rarely modeled dynamically.
Key Innovation: Combines Street View change comparison, field surveys, and Markov-chain growth modeling to project future collapse risk.
105. Collapse Safety of Reinforced Concrete Frames Under Building-Code Evolution
Core Problem: The aggregate collapse-safety effect of successive Colombian reinforced-concrete frame code eras was not quantified.
Key Innovation: Uses large nonlinear archetype sets to compare fragility and annual collapse risk across code eras and seismicity levels.
106. Morphological State Transitions and Sediment Transport Variability in Braided Rivers
Core Problem: Braided-river sediment transport variability is not well explained by local hydraulics alone.
Key Innovation: Models transport bursts as transitions among recurrent morphological states with a Markov process.
107. ShiftSplit-AD: Separating Domain Shift from Defects in Foundation-Feature Visual Anomaly Detection
Core Problem: Foundation-feature anomaly detectors confuse domain shift with actual defects.
Key Innovation: Decomposes nearest-normal residuals into low-rank and sparse parts to suppress shift-induced evidence.
108. ABCD: Alpha-Composited Block Coordinate Descent: Constant-VRAM Training for Large Radiance Fields
Core Problem: Large radiance fields exceed GPU memory during training.
Key Innovation: Uses alpha-composited block coordinate descent so only one spatial partition is active at a time.
109. uScenes: A Multimodal RGB and 3D Sonar Dataset for Underwater Robot Perception
Core Problem: 2D forward-looking sonar lacks elevation information for robust 3D underwater scene understanding.
Key Innovation: Releases synchronized RGB and 3D multibeam sonar data to support cross-modal 3D perception underwater.
110. Temporal Tree of Thought: Reasoning-Guided Visual Cue Search for Long-Video Understanding
Core Problem: Uniform frame sampling misses crucial evidence in long-video question answering.
Key Innovation: Builds a hierarchical temporal tree and reasons in a coarse-to-fine answer-retrieve-explore loop.
111. Training-Free Temporal Abstraction for General Video Understanding
Core Problem: Dense videos need reusable temporal structure without training task-specific models for each downstream use.
Key Innovation: Chunks videos into semantic units with frozen video-text embeddings and reuses them across boundary detection, retrieval, and frame selection.
112. 3D-USE: From Image-Level to Scene-Level Underwater Enhancement
Core Problem: Enhance degraded underwater multi-view scenes consistently at the 3D scene level rather than per image.
Key Innovation: Combines a medium-aware Gaussian scene representation with scene-global consensus targets transferred from 2D underwater enhancement.
113. Task-State Adaptation with Prototype Memory for Multi-Task Dense Prediction
Core Problem: A single shared visual backbone must support multiple dense prediction tasks with different image-dependent needs.
Key Innovation: Prototype-memory task states guide sparse expert routing and residual adaptation for multi-task dense prediction.
114. Dual-Stream Semantic Guidance with Prototype Anchor Calibration for Source-Fully-Free Adaptation of Vision-Language Models
Core Problem: Vision-language adaptation without source data suffers from semantic drift in both class anchors and generated captions.
Key Innovation: Dual-stream semantic guidance plus prototype-anchor calibration stabilizes source-free adaptation while reducing text-side cost.
115. WALDO: One-Shot Exemplar-Conditioned Object Detection in Cluttered Scenes
Core Problem: Find a specific object instance from a single exemplar and text cue, including correctly predicting absence.
Key Innovation: A lightweight exemplar-conditioned detection head over frozen world-model features with careful shortcut-resistant episode synthesis.
116. Non-Uniform Quantisation for 3DGS Compression
Core Problem: 3D Gaussian splatting models are too bitrate-heavy for practical storage, transmission, and standardization.
Key Innovation: Non-uniform importance-weighted quantization and merging tailored to 3DGS distributions for stronger compression.
117. Quantum SEDONet: Spectrally-Embedded Quantum Deep Operator Networks for Partial Differential Equations
Core Problem: Improve neural-operator approximation of PDE solutions by matching spectral embeddings to coordinate boundary conditions.
Key Innovation: Adds coordinate-wise Fourier or Chebyshev spectral embeddings to quantum DeepONets without extra qubits or circuit depth.
118. LitCurate: A Configuration-Driven AI-Assisted Framework for Scientific Database Construction with an Application to Lower-Mantle Equation-of-State Data
Core Problem: Turn dispersed scientific literature into structured, auditable databases that retain scientific context and provenance.
Key Innovation: Implements a stage-wise LLM curation workflow spanning discovery, screening, extraction, and provenance tracking, demonstrated on lower-mantle data.
119. On the Computational and Statistical Efficiency of the Empirical Maximum Entropy on the Mean Method
Core Problem: Establish stronger statistical and computational guarantees for empirical maximum-entropy-on-the-mean inverse methods.
Key Innovation: Proves faster convergence and recasts the dual MEM problem as expected-risk minimization for scalable stochastic optimization.
120. Localizing Global Discrepancies: Marginal Contributions and Contextual Anomaly Detection
Core Problem: Global discrepancy tests flag shifts without identifying responsible observations.
Key Innovation: Localizes global statistics through matched-context marginal contribution estimators.
121. Generalized Splines and Gaussian Processes
Core Problem: Infinite-dimensional Gaussian inverse problems lack a unified estimator view.
Key Innovation: Extends regularized least squares and GP equivalence to generalized splines on nuclear spaces.
122. Amortizing intractable inference in diffusion models for vision, language, and control
Core Problem: Posterior inference under diffusion priors is intractable for black-box constraints.
Key Innovation: Relative trajectory balance trains diffusion samplers for arbitrary posteriors.
123. Bringing SAM to new heights: Leveraging elevation data for tree crown segmentation from drone imagery
Core Problem: Tree-crown segmentation from drone imagery remains difficult under limited labels and varied forests.
Key Innovation: Integrates DSM elevation with tuned SAM for instance segmentation across forest types.
124. RegCL: Compact Continual SAM Adaptation for Visual Grounding in Multi-Sensorial Media
Core Problem: SAM adapters forget prior domains or grow too large under continual adaptation.
Key Innovation: Merges domain adapters into one replay-free SAM module using consistency and stored statistics.
125. Organisation and co-ordination of spontaneous volunteers in response to the 2023 Kahramanmaraş earthquake disaster in Türkiye
Core Problem: Understand how spontaneous volunteers were organized and coordinated after the 2023 Kahramanmaras earthquake.
Key Innovation: Case-study analysis of spontaneous-volunteer coordination during a major earthquake response.
126. Multi-task deep reinforcement learning for service-oriented post-flood recovery scheduling of large urban metro systems
Core Problem: Schedule service-oriented recovery of large urban metro systems after flooding.
Key Innovation: Applies multi-task deep reinforcement learning to post-flood metro recovery scheduling.
127. A quantitative prediction method for water inflow based on FDIP-derived hydraulic conductivity in tunnels
Core Problem: Predict tunnel water inflow from FDIP-derived hydraulic conductivity.
Key Innovation: Builds a quantitative tunnel inflow prediction method from FDIP-derived conductivity estimates.
128. A thermodynamic FEM-DEM algorithm for granular-scale phase transition and liquid-water flow in frozen soil
Core Problem: Capture coupled water-ice phase transition, liquid-water migration, and frost-heave evolution in frozen soils across continuum and granular scales.
Key Innovation: Couples continuum THM simulation with DEM-based ice-bond evolution and granular seepage forcing to resolve multiscale frost-heave processes.
129. Assessment of railway earthworks using DAS based active MASW surveys
Core Problem: Assess railway earthworks using DAS-based active MASW surveys.
Key Innovation: Applies distributed acoustic sensing to active MASW for large-scale railway earthwork characterization.
130. Surface settlement prediction-to-warning using machine learning with asymmetric-risk calibration for metro shield tunnelling
Core Problem: Point settlement predictions do not directly solve the asymmetric decision risk of under-warning versus over-warning.
Key Innovation: Risk-calibrated LightGBM converts settlement prediction into safer warnings while preserving interpretability with SHAP.
131. Toward Sustainable Seismic Retrofit: A Review of Performance-Based Earthquake Engineering, Sustainability, and Policy Frameworks
Core Problem: PBEE, sustainability assessment, and retrofit policy are rarely integrated into a unified decision framework.
Key Innovation: Synthesizes engineering, life-cycle, and policy mechanisms for retrofit strategy across project and governance scales.
132. User Acceptance and Perceptions of Earthquake Early Warning Systems as a Function of Information Type: The Case of Postearthquake Nepal
Core Problem: It is unclear how technical and limitation-focused information shapes public acceptance of EEWS in Nepal.
Key Innovation: Combines preregistered survey experiments and interviews to isolate how transparency about system limits changes perceived usefulness and interest.
133. Development of Full-Scale Experiment-Based Seismic Fragility Functions for Code-Compliant Fire Sprinkler Piping Systems
Core Problem: Existing fragility functions for sprinkler systems rely too heavily on judgment and may overstate minor damage.
Key Innovation: Derives full-scale experimental fragility functions and updated damage-state calibrations for sprinkler piping systems.
134. From Observed Damage to Executable Repair Quantities: The RELA Repair Module
Core Problem: Recovery workflows rarely translate observed building damage into executable repair quantities early enough for real decisions.
Key Innovation: Creates an observation-driven module that converts element-level damage into repair bills of quantities, time, and material needs.
135. Impact of Structural Configurations on Seismic Performance of Low- and Mid-Rise Thin Lightly Reinforced Concrete Wall Buildings
Core Problem: Identify which thin lightly reinforced concrete wall configurations drive unacceptable collapse risk in Colombian housing.
Key Innovation: Links hazard-consistent archetype fragilities to wall geometry and reinforcement patterns across 52 buildings.
136. Seismic Performance of Multidrum Free-Rocking Columns With Shear Key
Core Problem: Need to quantify how shear keys affect multidrum free-rocking column collapse behavior.
Key Innovation: Combines shaking-table tests, FE calibration, and failure-probability mapping, including tenon-detachment screening.
137. Perceived Enablers and Barriers of the Los Angeles Non-Ductile Concrete Seismic Retrofit Ordinance
Core Problem: Los Angeles retrofit uptake is slow and the barriers faced by owners are not well characterized.
Key Innovation: Uses stakeholder interviews to separate safety, financial, technical, relocation, and governance barriers to ordinance compliance.
138. Probabilistic Seismic Performance Assessment of Optimal Self-Centering Shape Memory Alloy-Based Steel Moment Frames Considering Nonstructural Elements
Core Problem: Optimal self-centering frames are usually evaluated without full structural and nonstructural loss consequences.
Key Innovation: Uses FEMA P-58 to compare repair cost, downtime, demolition, and carbon outcomes of conventional versus SMA self-centering frames.
139. Limiting Safety-Critical Structural Damage in Recovery-Based Seismic Design Provisions
Core Problem: Functional-recovery design needs risk-targeted limits that bound safety-critical structural damage.
Key Innovation: Defines a new Functional Recovery Earthquake check and derives allowable drift and response-modification combinations across systems.
140. Maximum Usable Period of Ground Motion Records for Engineering Applications
Core Problem: Signal-processing bandwidth limits may be too permissive for engineering response analyses, especially inelastic ones.
Key Innovation: Defines an engineering-based upper usable period threshold tied to tolerated response bias and analysis type.
141. Seismic Risk Assessment of Dry-Jointed Concrete Voussoir Arch in Highway Bridge
Core Problem: The seismic suitability of dry-jointed concrete voussoir arch bridges across Indian seismic zones is unknown.
Key Innovation: Uses discrete-element modeling, fragility analysis, and annual collapse metrics to benchmark the bridge against reinforced-concrete piers and code limits.