TerraMosaic Daily Digest: August 2, 2026
Daily Summary
The day's landslide studies move beyond static terrain classification toward direct evidence of slope state and forcing history. InSAR deformation constrains susceptibility zones, mobile SLAM LiDAR tests road-corridor predictions, and repeat LiDAR-UAV surveys resolve lowland landslide evolution. Process models identify the controlling pathways: deep fissures can connect irrigation directly to groundwater and reduce a loess slope's factor of safety to 0.97, while groundwater cycling, wetting-drying and prolonged rainfall organize preferential flow and cumulative deformation in expansive-soil slopes.
The broader hazard record connects monitoring to decisions across collapse, flood, fire and compound climate extremes. Integrated geophysics delineates potential sinkholes; seismic observations and hydro-mechanical modelling reconstruct progressive mining collapse; and coastal profiles show how sandbars buffer extreme storm energy as tide level changes. Urban-flood surrogates, reach-specific routing, global wildfire attribution and non-stationary drought-heatwave analysis extend the same emphasis on process-resolving evidence across scales.
Method development is converging on transferable models that still respect how Earth observations are acquired and organized. Physics-guided change detection, geometry-aware SAR synthesis and self-supervised fire detection encode measurement structure, while TerraNova retains 1,024 physical and societal variables in their native spatial geometries with evidential uncertainty. The advance is not generic model scale alone, but the preservation of heterogeneity, dependence and uncertainty that hazard decisions require.
Key Trends
Five methodological shifts connect the day's work from slope-scale observations to regional warning and multi-hazard decisions.
- Deformation Becomes a Constraint, Not Just a Validation Layer: InSAR, SLAM LiDAR and repeat topography are used to adjust susceptibility classes, test spatial predictions and quantify evolving landslide geometry.
- Trigger Pathways Are Represented Explicitly: Fissure-fed infiltration, groundwater oscillation, wetting-drying, dynamic joint degradation and pulse-like shaking are resolved as causal pathways to instability.
- Early Warning Is Becoming Spatially Distributed: Highway-scale soil-water indices, low-coherence InSAR phase recovery and accelerated rainfall-response calibration target where and when failure develops, rather than issuing one basin-wide signal.
- Hazard Models Retain Timing, Heterogeneity and Dependence: Storm-tide alignment, reach-specific flood routing, physics-structured urban surrogates and non-stationary copulas preserve controls that homogeneous models suppress.
- Reusable Models Are Moving Closer to Earth-System Data: Geospatial foundation models, time-series experts, self-supervised fire detection and physics-guided change detection broaden transfer while exposing uncertainty and acquisition constraints.
Selected Papers
Observed deformation, hydrological triggers and operational warning performance define this issue. The leading studies connect repeat LiDAR, UAV and InSAR measurements to landslide evolution and susceptibility, then explain instability through fissure-fed seepage, groundwater cycling and structural-plane degradation. Companion analyses extend the same evidence chain to coastal storm buffering, sinkholes, seismic risk, floods, wildfire and compound climate extremes.
1. Evolution and Analysis of Landslides in Lowland Areas: The Case Study of Reuil in the Champagne Vineyard Region (Marne, France)
Core Problem: Low-relief vineyard landslides are poorly resolved by conventional inventories, leaving their volume, rate and climatic-land-use controls uncertain.
Key Innovation: Three LiDAR terrain models and two UAV surveys reconstruct movement and up to 900 cubic metres of displaced material, linking renewed activity to exceptional rainfall after drought and to vineyard land management.
2. Segment-Based Landslide Susceptibility Along Mountainous Road Corridors: Validating Random Forest Model with SLAM LiDAR in Northeastern Iraq
Core Problem: Road-corridor susceptibility models are commonly validated against inventories that lack the geometry and segment scale needed for maintenance decisions.
Key Innovation: A Random Forest model is evaluated along 20 km of mountain road using SLAM LiDAR; the test AUC is 0.725 and 88.9 percent of segments agree exactly or within one susceptibility class across scales.
3. Landslide Susceptibility Mapping Constrained by InSAR-Derived Deformation Using Multi-Source Data Integration
Core Problem: Static landslide susceptibility classes can be statistically accurate yet spatially inconsistent with deformation observed during the monitoring period.
Key Innovation: SBAS-InSAR from 31 Sentinel-1A images constrains an RF-BPNN susceptibility model; kernel-density and correlation-based class adjustment raises agreement with active deformation from 0.61 to 0.65.
4. A Novel Distributed Model for Predicting Runoff-Induced Multi-Instability Risk Along Highway Corridors Under Heavy Rainfall
Core Problem: A basin-averaged soil-water index cannot represent the spatially variable runoff and infiltration that trigger landslides along linear transport corridors.
Key Innovation: A distributed runoff model supplies spatial water depth to a tank model; for five Typhoon 10 landslides it predicts initiation within 1.5 hours and separates all failure sites from a non-landslide reference in the case study.
5. Early Identification of Subtle Deformations in Potential Debris Flow Source Areas Using Phase-Unwrapped Convolutional Neural Networks and Long-Time-Series InSAR Technology
Core Problem: Millimetre-scale creep in debris-flow source areas is obscured by low coherence and cross-pixel phase jumps, delaying recognition of accelerating deformation.
Key Innovation: A graph convolutional phase-unwrapping model propagates information across interferograms and combines temporal attention with acceleration criteria, identifying five source areas while keeping inversion error within plus or minus 1 mm in tested zones.
6. Utilizing High-Frequency Seismic Signals (HFSS) and other geophysical methods to locate potential sinkhole collapses: a case study from West of Alexandria, Egypt
Core Problem: Potential sinkhole collapse zones west of Alexandria cannot be located reliably from one geophysical observable in heterogeneous near-surface materials.
Key Innovation: High-frequency seismic signals are interpreted jointly with electrical resistivity, MASW, refraction, GPR and borehole evidence to cross-check void-prone zones.
7. State of Wildfires 2025–2026
Core Problem: Global wildfire assessment requires one consistent account of observed activity, attributable drivers, impacts and near-term risk across regions and seasons.
Key Innovation: The annual State of Wildfires framework integrates multiple global datasets to quantify the 2025-2026 fire year, attribute major regional extremes and evaluate seasonal outlooks.
8. Coupled groundwater fluctuation and wetting–drying effects on fissured expansive soil canal slopes: hydro-mechanical evolution from physical model testing
Core Problem: Canal slopes in fissured expansive soils accumulate damage through interacting groundwater fluctuation, wetting-drying and prolonged rainfall, but their coupled progression is insufficiently observed.
Key Innovation: A physical model tracks preferential flow, pore-pressure redistribution and cumulative deformation across repeated hydroclimatic cycles, resolving the transition to shallow sliding.
9. Sinkhole Formation over a Solution Mining Cavern Field in Maceió, Brazil: Analysis of the Geomechanical Processes — Part II
Core Problem: The progressive mechanism of the Maceio mining collapse remained uncertain despite surface deformation and seismic warning signals before sinkhole formation.
Key Innovation: Event chronology, seismic evidence and a three-dimensional hydro-mechanical model connect cavern-field solution mining to progressive roof failure and surface collapse.
10. Seepage and Stability Analysis of Loess Landslides Under the Coupled Effects of Long-Term Irrigation and Fissures
Core Problem: Long-term irrigation and deep loess fissures interact to raise groundwater, but their separate effects on seepage, deformation and slope stability are difficult to quantify.
Key Innovation: A saturated-unsaturated seepage-stress model shows that a 15 m fissure connects irrigation to groundwater, lowers the factor of safety to 0.97 and increases shoulder displacement by 54 percent relative to an intact slope.
11. Impact of an extreme storm event on a multiple intertidal barred coastal system
Core Problem: The capacity of multiple intertidal bars to protect a coast during an extreme storm depends on beach morphology and storm-tide timing, but those controls are rarely observed across an event.
Key Innovation: Profiles and wave measurements from Storm Barra separate erosional, accretional and transitional responses and show that subtidal and landward bars dissipated wave energy at successive tidal stages, preventing severe shoreline impact.
12. Morphodynamic adjustment in a watershed affected by megafires: Deadman River, British Columbia, Canada
Core Problem: Watershed-scale river adjustment after megafire is less well quantified than local post-fire mass wasting, limiting forecasts of channel mobility and sediment redistribution.
Key Innovation: Time-lapse photogrammetry and multispectral remote sensing show channel widening of up to 100 percent and meander-migration increases of up to 540 percent after the 2021 Sparks Lake Fire, with faster migration downstream of more severely burned terrain.
13. Ocean Surface Gravity Waves Excited by the 2022 Eruption of Hunga Tonga-Hunga Ha'apai Volcano
Core Problem: The force history of explosive eruptions is difficult to recover because atmospheric, oceanic and seismic signals overlap and propagate through different media.
Key Innovation: Ocean surface gravity waves at 15-40 mHz isolate the eruption source from the atmospheric Lamb wave, resolving a force of order 10 billion newtons, five hours of excitation and a later sub-event.
14. TerraNova: A Foundation Model for the Anthropocene
Core Problem: Earth-system foundation models usually force heterogeneous physical and societal variables into one raster representation and provide weak uncertainty information for unseen variables.
Key Innovation: TerraNova pre-trains across 1,024 variables in native gridded and administrative geometries and uses evidential uncertainty for sparse reconstruction and zero-shot variable prediction.
15. Multi-station tidal level forecasting based on a novel spatio-temporal graph convolutional neural network
Core Problem: Single-station tide forecasts do not represent how astronomical tides and storm-driven water levels propagate between coastal gauges.
Key Innovation: A multi-graph neural network learns spatial and temporal dependence jointly, generalizes to five additional stations and tests storm-tide behavior during Super Typhoon Jebi.
16. Evaluation of storm surge flooding response to climate warming and sea level rise: A case study of Metro Manila, Philippines
Core Problem: The separate contributions of climate warming and sea-level rise to future typhoon-driven flooding remain poorly quantified for Metro Manila.
Key Innovation: Coupled atmospheric-hydrodynamic experiments isolate warming, sea-level rise and their combined effect under two emissions pathways, identifying sea-level rise as the dominant control for a Nesat-like event.
17. Seismic displacement and failure characteristics of rock slopes considering dynamic degradation of rock structural planes under pulse-like ground motions
Core Problem: Rock-slope displacement estimates become unconservative when earthquake loading is separated from progressive degradation of structural planes.
Key Innovation: The analysis couples dynamic joint degradation with pulse-like ground motions, quantifying their combined effect on displacement and failure characteristics.
18. Mechanistic insights into lithology-normal stiffness coupling effects on stick-slip instability in fault-slip rockburst triggering
Core Problem: As underground engineering extends to depths of thousands of meters, fault-slip rockbursts induced by stick-slip instability pose a major safety threat.
Key Innovation: Controlled-normal-stiffness tests isolate how lithology and boundary stiffness govern stick-slip instability and rockburst triggering.
19. Hidden damage or hidden strength? Revealing the post-earthquake integrity of jet-grout columns
Core Problem: However, despite its increasing use, the post-earthquake integrity of jet-grout columns remains insufficiently understood, particularly regarding the potential for hidden discontinuities, internal cracking, or reductions in apparent column length.
Key Innovation: Field integrity tests and cores reveal hidden earthquake damage in jet-grout columns and its implications for liquefaction mitigation.
20. Integrated probabilistic seismic hazard and risk assessment of Hunza District, northern Pakistan
Core Problem: Mountain districts with sparse building and ground-motion data lack a coherent basis for comparing seismic hazard with expected structural consequences.
Key Innovation: Probabilistic hazard is linked to 8,470 exposed buildings in 15 typologies, producing an integrated risk baseline for Hunza rather than a hazard map alone.
21. Influence of reinforcement methods on the seismic response and energy evolution of high-fill embankment on sloping terrain
Core Problem: The relative seismic performance and failure modes of geogrids and anti-slide piles in high-fill embankments on sloping terrain are insufficiently constrained.
Key Innovation: Shaking-table tests show that geogrids reduce residual displacement but remain vulnerable to interface sliding, whereas composite reinforcement prevents overall sliding and reduces crack depth by 33 percent relative to geogrids alone.
22. Physics-guided urban flood modeling: spatial heterogeneity and diagnostic attribution via CMoE-HDAF
Core Problem: City-scale flood models are too slow for rapid scenario analysis, while uniform surrogates fail across strongly heterogeneous drainage regimes.
Key Innovation: A physics-structured conditional mixture of experts learns 25 hydrodynamic regimes from InfoWorks simulations, reaching Nash-Sutcliffe efficiency of 0.9761 across 243,944 nodes with 32-second inference.
23. Assessing compound drought-heatwave events in Oman using non-stationary copula models
Core Problem: Stationary dependence models mischaracterize compound drought-heatwave risk as climate baselines and marginal extremes change.
Key Innovation: Non-stationary copulas resolve evolving dependence and identify 2023-2024 as an exceptional compound event in Oman.
24. Seismic performance of colluvial slope reinforced by resilient anchor rod with negative stiffness damping
Core Problem: Conventional slope anchors have limited large-deformation adaptability and energy dissipation under strong earthquake loading.
Key Innovation: A negative-stiffness friction damper lowers crest acceleration and displacement by 34.1 and 33.3 percent relative to conventional anchorage, while shifting failure from global sliding to localized cracking.
25. Accelerating Landslide Temporal Forecasting Through a Re-engineered Software System: GASAKe 2.1
Core Problem: Genetic-algorithm calibration of rainfall-driven landslide activation models is computationally expensive when repeated across fitness functions and warning analyses.
Key Innovation: GASAKe 2.1 re-engineers the existing hydrological model for shared-memory execution, reproducing the Uncino calibration with speed-ups of 4.04-8.0 times.
26. Landslide Susceptibility and Risk Assessment in Kodagu, India, Using Machine Learning
Core Problem: Kodagu lacks a district-scale susceptibility assessment that integrates its terrain, rainfall, land-cover and geological heterogeneity.
Key Innovation: A GIS-based Random Forest-XGBoost ensemble maps five susceptibility classes and reaches an AUC of 0.84, exceeding either learner alone.
27. Code and processed data for land-cover-associated variation in recorded severe rainfall-triggered landslide susceptibility across Sumatra
Core Problem: Testing how land cover modifies recorded severe rainfall-triggered landslide susceptibility requires a reproducible regional data and analysis base.
Key Innovation: The release packages the processed Sumatra data and code needed to reproduce land-cover-associated susceptibility analyses.
28. Tsunami Simulations in Support of California’s Fifth Climate Change Assessment, in Tsunami Simulations in Support of California’s Fifth Climate Change Assessment
Core Problem: State-scale tsunami climate assessment requires a consistent and reusable archive of scenario simulations.
Key Innovation: The dataset releases the tsunami simulations prepared for California's Fifth Climate Change Assessment as a reproducible scenario resource.
29. Urban high-rise compact form amplifies humid heat risk from vegetation transpiration
Core Problem: How rapid urbanization and its associated three-dimensional (3D) urban form regulate vegetation’s coupled “cooling–humidifying” effect remains poorly understood.
Key Innovation: A 251-city analysis shows that compact high-rise form can turn vegetation transpiration into greater population-weighted humid-heat exposure.
30. Seismic Characteristics of the Gas Hydrate System in the Mackenzie Trough of the Canadian Beaufort Sea, Arctic Ocean
Core Problem: The authors present a comprehensive seismic characterization of the gas hydrate system in the outer continental shelf of the Mackenzie Trough, Canadian Beaufort Sea, based on recently acquired multichannel seismic (MCS) data, P‐wave velocity modeling, and 3D thermal modeling.
Key Innovation: Seismic velocity and thermal models map a large gas-hydrate system and its long-term sensitivity as a latent submarine geohazard.
31. Feature Interaction Modeling for Physics-Informed Neural Networks and Neural Operators
Core Problem: Physics-informed networks often model input features independently, limiting their ability to represent interactions within parameterized solution manifolds.
Key Innovation: Factorization-machine interaction modules are embedded in PINNs and neural operators to improve expressiveness on shock-dominated partial differential equations.
32. Locally Consistent Transductive Information Maximization for Few-Shot Remote Sensing Scene Classification
Core Problem: The regularizer enters as a single multiplicative factor in the closed-form q-update, adding negligible computational overhead.
Key Innovation: An open ten-dataset benchmark tests locally consistent transductive adaptation for few-shot remote-sensing scenes.
33. Training-Free Entity-Level Few-Shot Segmentation of Remote Sensing Images with Advection Refinement
Core Problem: Existing cross-domain few-shot segmentation approaches suffer from high training costs due to source-domain episodic training and pixel-wise dense prediction, while often producing fragmented and noisy predictions.
Key Innovation: Training-free entity segmentation combines foundation-model masks with advection-based refinement under sparse labels.
34. On four representations of the law of aftershock evolution
Core Problem: The relaxation of the earthquake source following the main shock has been investigated both theoretically and experimentally, based on data regarding the evolution of aftershocks.
Key Innovation: Alternative linear and discrete formulations describe aftershock decay through source deactivation and a source-specific time coordinate.
35. PhyUnfold-Net: Advancing Remote Sensing Change Detection with Physics-Guided Deep Unfolding
Core Problem: To stabilize this process, we introduce a staged Exploration-and-Constraint loss (S-SEC), which encourages component separation in early steps while constraining nuisance magnitude in later steps to avoid degenerate solutions.
Key Innovation: A physics-guided deep-unfolding architecture makes the stages of bi-temporal change detection explicit and testable.
36. TimeRFT: Stimulating Generalizable Time Series Forecasting for TSFMs via Reinforcement Finetuning
Core Problem: Time-series foundation models lose transfer skill when downstream records differ in distribution and training-data volume.
Key Innovation: Reinforcement fine-tuning rewards cross-dataset forecasting performance to improve generalization under temporal distribution shift.
37. EMINet-CD: An Edge-Enhanced Mixed Information Interaction Network for Remote Sensing Image Change Detection
Core Problem: Small targets, diffuse boundaries and illumination or shadow differences still limit bi-temporal remote-sensing change detection.
Key Innovation: EMINet-CD couples edge enhancement with mixed-information interaction to retain target boundaries while suppressing radiometric interference.
38. A Visibility-Aware Neural Gaussian Framework for SAR Image Generation and Data Augmentation
Core Problem: Many data-driven SAR approaches (e.g., GAN-based methods and SAR-NeRF) face challenges in optimization efficiency and physical interpretability.
Key Innovation: Visibility-aware neural Gaussians generate SAR observations while retaining acquisition-geometry constraints.
39. Forecast-based operation of repurposed small reservoirs for floods, farms, and (low) flows
Core Problem: The increased frequency and intensity of hydrological extremes, including drought, due to anthropogenic climate change will drive the need for enhanced water supply resilience, even in water-rich countries.
Key Innovation: Forecast-based control repurposes small reservoirs to balance flood attenuation, irrigation and low-flow objectives.
40. FIRE-BYOL: A Real-Time Grassland Active-Fire Detection Algorithm Fusing VIIRS Fire Products and Himawari-8/9 Data
Core Problem: Frequent grassland fires on the Mongolian Plateau require near-real-time active-fire detection, but current products trade spatial detail against revisit frequency and deep models depend on large labelled datasets.
Key Innovation: Self-supervised learning fuses VIIRS products with Himawari imagery for real-time grassland active-fire detection.
41. From Thermal Diagnosis to Spatial Allocation: A Remote-Sensing and Explainable Machine Learning Framework for Heat-Resilient Planning in Semi-Arid Grassland Towns
Core Problem: A constrained optimization stage generates alternative spatial allocations; the balanced scenario reduces mean land-surface temperature by 1.30 degrees Celsius while maintaining openness and improving economic benefit.
Key Innovation: Remote sensing and explainable machine learning connect thermal diagnosis to heat-resilient spatial planning in semi-arid towns.
42. Improving Convection-Allowing Ensemble Forecasts via Multi-Source Remote Sensing Data Assimilation Through Stepwise Cloud Analysis Initialization: A Remote Sensing Case Study
Core Problem: The “spin-up” problem, in which convection-permitting models require hours to develop realistic clouds from large-scale initial fields, critically limits short-term severe weather forecasting.
Key Innovation: Stepwise cloud analysis assimilates multi-source remote sensing into convection-allowing ensemble forecasts.
43. Disproportionate Soil Loss from Fragmented Sloping Cropland in Mountainous Northeastern Yunnan: Integrating Sentinel-2, CSLE, and Landscape Metrics
Core Problem: Soil erosion on sloping cropland is a major threat to agricultural sustainability and ecological security in mountainous regions, yet its spatial distribution and landscape-level structural characteristics remain insufficiently quantified.
Key Innovation: Sentinel-2, CSLE and landscape metrics quantify disproportionate soil loss from fragmented mountain cropland.
44. TimeHome: Heterogeneous Mixture-of-Experts for Time-Series Foundation Model
Core Problem: Time-series analysis is important for various scientific and industrial fields, such as remote sensing where observations may be disturbed by clouds, have irregular revisits and different sensors.
Key Innovation: A heterogeneous mixture-of-experts routes distinct temporal structures within a general time-series foundation model.
45. Climate-Adaptive Urban Planning: Quantitative Assessment of Drought Impact and Practical Strategies for Climate-Resilient Urban Green Spaces
Core Problem: Urban green spaces (UGSs) are vital for enhancing a city’s resilience and livability; however, their functionality is increasingly jeopardized by drought, particularly in water-scarce regions.
Key Innovation: Remote sensing quantifies drought impacts on urban green spaces and translates them into climate-adaptive planning strategies.
46. Advancing sustainable groundwater mapping and management in arid quaternary aquifers using machine learning and geospatial analytics integrating remote sensing and field hydrogeological data
Core Problem: Groundwater (GW) represents a critical resource for sustaining agriculture and rural communities across the arid regions of many developing countries.
Key Innovation: Machine learning and geospatial analytics fuse remote sensing with field hydrogeology for arid-aquifer characterization.
47. AI-based multimodal fusion of geophysical and geochemical data for subsurface characterization
Core Problem: Artificial intelligence (AI) increasingly supports subsurface characterization by integrating heterogeneous geophysical, geochemical, and spatial data.
Key Innovation: AI-based fusion combines geophysical and geochemical evidence for heterogeneous subsurface characterization.
48. Reconstruction and Prediction of Pre-peak Load–Displacement Curves in Pre-cracked Sandstone Fracture Using Multimodal Fusion
Core Problem: The pre-peak stage of three-point bending in pre-cracked sandstone reflects the coupled development of crack-tip localization and internal damage.
Key Innovation: Multimodal fusion reconstructs and predicts pre-peak load-displacement behaviour in pre-cracked sandstone.
49. Estimations of peak floor accelerations in multi-story self-centering concentrically braced frames
Core Problem: Self-centering concentrically braced frames (CBFs) belong to seismic resilient CBFs, owing to their excellent capacity of returning to original deformations after strong earthquakes.
Key Innovation: A flag-shaped hysteresis-aware formulation improves peak-floor-acceleration estimates for self-centering concentrically braced frames.
50. Plutonium isotopes (239+240Pu) in volcanic soils of Iceland – Depth patterns, behavior and applications for soil redistribution tracing
Core Problem: At the Blönduos, redistribution rates (from −3.08 to +3.03 t ha−1 yr) were likely overestimated due to the difficulty in finding a representative, undisturbed reference site.
Key Innovation: Plutonium-isotope depth profiles provide a tracer of soil redistribution in Icelandic volcanic terrain.
51. Differential response of floods to climate change and human activities in the Yalu River over the past two millennia: insights from hydrological modeling
Core Problem: Floods are severe natural disasters threatening the world, and their long-term evolution patterns are crucial for regional security and sustainable development.
Key Innovation: Two millennia of hydrological modelling separate climate and human controls on Yalu River floods across event magnitudes.
52. Effect of Desiccation Cracking on Moisture Flow and Heave-Subsidence response of Expansive Soils: Insights from Dual-Permeability Modeling
Core Problem: Conventional single-domain flow models do not represent how desiccation cracks alter moisture transport and coupled heave-subsidence in expansive soils.
Key Innovation: A dual-permeability formulation separates matrix and crack flow to quantify their contribution to moisture redistribution and ground deformation.
53. ismakulsoom/IPIM-T-Landslide-Susceptibility: IPIM-T v0.1
Core Problem: Open landslide-susceptibility implementations rarely expose how deformation constraints, physics regularization and multimodal fusion are assembled for reproducible reuse.
Key Innovation: IPIM-T v0.1 archives an InSAR-constrained, physics-informed multi-stream Transformer with training and evaluation code, while explicitly noting that the full study dataset and validation results are not included.
54. A Scalable Approach for Identification of Reach‐Scale Hydraulic Processes to Support Flood Prediction Models
Core Problem: Simplification of channel routing models results in errors in streamflow predictions and uncertain flood forecasts.
Key Innovation: Topographically defined reach classes identify where broad-scale flood models require more complex hydraulic routing.
55. A comprehensive multi-indicator urban flood-resilience evaluation system based on the PSR-CPM framework
Core Problem: Urban flood-resilience indices often aggregate pressure, state and response indicators without resolving nonlinear threshold behavior across rainfall return periods.
Key Innovation: A PSR-CPM evaluation coupled to SWMM-LISFLOOD-FP simulations identifies spatial resilience contrasts and a stepwise decline beyond the 20-year rainfall threshold in Jinan.
56. Assessment of urban flood resilience indicators: a review of two decades of research
Core Problem: Urban flood-resilience studies use heterogeneous indicators and assessment targets, making results difficult to compare across cities and governance settings.
Key Innovation: A bibliometric and systematic review organizes two decades of resilience, risk and susceptibility indicators and identifies persistent institutional, financial and microclimate gaps in Indian research.
57. Exploring Determinants of Community Flood Resilience in Southeast Asia: A Systematic Review
Core Problem: Flooding is the most devastating natural disaster in Southeast Asia, disproportionately affecting vulnerable communities and posing substantial public health challenges.
Key Innovation: Evidence from Southeast Asia is synthesized to identify the determinants that repeatedly shape community flood resilience.
58. Facing the Tide: How Can Woollahra’s Urban Planning Outsmart Flood Risks?
Core Problem: Established, affluent coastal suburbs combine severe flood exposure with extensive private land ownership, yet little is known about how multi-level planning systems perform when private property rights must be balanced against public flood resilience.
Key Innovation: A reproducible plan-evaluation framework exposes the gap between strategic flood intent and enforceable, current local practice.
59. Influence Mechanisms of Spatial Optimization of the “Terraced Fields‐Gully Land Reclamation Project ( GLRP )” Cascades on Erosion‐Transportation Processes in Watershed in the Loess Hilly‐Gully Region of China
Core Problem: Large terraced-field and gully-reclamation programs can alter runoff and sediment connectivity, but the effect of their spatial arrangement is poorly constrained at watershed scale.
Key Innovation: A three-dimensional physical watershed model compares cascade layouts and identifies how spatial optimization changes erosion, transport and sediment export.
60. Real-Time Earthquake Detection: A Study of CNN and SNN Approaches
Core Problem: Real-time earthquake classifiers must balance recognition accuracy against the latency and energy limits of edge deployment.
Key Innovation: CNN and spiking-neural-network models are compared on global earthquake attributes; the spiking model trades modest accuracy for substantially lower reported inference time and power use.
61. CorrelationFlow: A Training-Free Geometric Approach for LiDAR Scene Flow Estimation
Core Problem: LiDAR scene flow estimation has settled into a monoculture: nearly all recent methods share the same feed-forward architecture and the same family of self-supervised losses, inheriting each other's assumptions, and each other's blind spots.
Key Innovation: A training-free geometric method estimates LiDAR scene flow without learned correspondences, supporting moving-surface monitoring.
62. FieryGS: In-the-Wild Fire Synthesis with Physics-Integrated Gaussian Splatting
Core Problem: Recent scene modeling advances like 3D Gaussian Splatting (3DGS) enable high-fidelity real-world scene reconstruction, yet lack physical grounding for combustion.
Key Innovation: Physics-integrated Gaussian splatting synthesizes in-the-wild fire observations for scarce-data vision tasks.
63. A Model-Driven Approach for Developing Families of Reinforcement Learning Environments
Core Problem: Reinforcement-learning studies need families of controlled environment variants, yet building those variants manually is slow and error-prone.
Key Innovation: A model-driven generator derives related training environments from one specification and demonstrates the workflow in a wildfire-mitigation scenario.
64. Optimization and effectiveness evaluation of nonlinear energy sink (NES) in reducing dynamic responses of floating offshore wind turbine (FOWT) under multiple hazards
Core Problem: Narrow-band tuned mass dampers have limited effectiveness and excessive stroke for floating wind turbines exposed to combined earthquake, wind and wave loading.
Key Innovation: A nonlinear energy sink is optimized across single and compound hazards to broaden vibration suppression without relying on one tuned frequency.
65. DisentangledMamba: Decoupling Spectral-Spatial Dependencies for Hyperspectral Image Classification
Core Problem: Hyperspectral image classification faces a fundamental challenge in that existing deep learning methods process spectral and spatial information jointly, inevitably entangling two physically distinct types of dependencies and degrading the discriminative capacity of each dimension.
Key Innovation: Disentangled state-space branches separate spectral and spatial dependencies in hyperspectral imagery.
66. 3-D Spatial–Spectral Parameter Extraction of Complex Waveforms From Hyperspectral Lidar Using Multichannel B-Spline Decomposition
Core Problem: Weak and overlapping echoes in asymmetric hyperspectral-LiDAR waveforms are poorly represented by regular parametric signal models.
Key Innovation: Multichannel B-spline decomposition extracts three-dimensional spatial-spectral parameters while adapting to irregular waveform shape.
67. RFM-UNet: Hybrid Frequency–Mamba UNet for Remote-Sensing Road Extraction
Core Problem: Road-network extraction from very high-resolution (VHR) remote-sensing imagery remains a challenging task owing to the structural sparsity, topological complexity, and severe occlusions of road networks.
Key Innovation: Frequency and state-space features are combined for road extraction relevant to post-event access mapping.
68. Vegetation Optical Depth at Enhanced Spatial Resolution: Progress, Challenges, and Perspectives
Core Problem: Vegetation optical-depth products remain too coarse for many local ecosystem and disturbance applications, and available downscaling methods have uneven validation.
Key Innovation: The review organizes enhanced-resolution VOD methods, their data dependencies and validation limits, clarifying where finer products are scientifically defensible.
69. A Scalable Open Source Workflow for Riverbed Substrate Classification Using UAV Imagery
Core Problem: Accurate characterization of riverbed substrate from remote sensing imagery is essential for applications in fluvial geomorphology, habitat modeling, and river management.
Key Innovation: An open workflow turns UAV imagery into scalable riverbed-substrate classifications for fluvial monitoring.
70. Terrestrial water storage monitoring in Bulgaria using Global Navigation Satellite System (GNSS) displacements and comparison with GLDAS, GLWS and GRACE products
Core Problem: National terrestrial-water-storage estimates from satellite gravimetry are spatially coarse and need independent deformation-based constraints.
Key Innovation: Vertical displacements from 32 Bulgarian GNSS stations are converted to storage variations and compared with GRACE and land-surface products over 2004-2025.
71. Hydro-mechanical evolution of seepage-induced soil arching under geometric constraints: a CFD-DEM study
Core Problem: The interaction between seepage erosion and soil arching around existing tunnels poses significant challenges to the stability of underground infrastructure.
Key Innovation: CFD-DEM simulations resolve the hydro-mechanical evolution of soil arching under seepage and geometric confinement.
72. Macro- and microscopic mechanisms of pipeline uplift in granular soils under varying groundwater levels: a coupled SPH–DEM study
Core Problem: Buried pipelines may experience uplift buckling once the upward stress surpasses the overburden resistance, a risk that is not yet fully characterized in partially submerged to fully submerged soils.
Key Innovation: Coupled SPH-DEM modelling resolves pipeline uplift mechanisms as groundwater levels vary.
73. An Inverse Generative Framework for Evacuation Risk Mitigation in Multi-exit Buildings with Complex Layout
Core Problem: Existing evacuation studies predominantly follow a forward modeling paradigm, evaluating evacuation performance under predefined occupancy conditions while providing limited insight into the inverse relationship between occupancy organization and evacuation outcomes.
Key Innovation: An inverse generative framework optimizes multi-exit evacuation layouts under complex building geometry.
74. Reliability-oriented Digital Twin of a building-scale rainwater treatment and reuse system under uncertain demand and climate variability
Core Problem: Rainwater Harvesting, Treatment and Reuse Systems (RHTRSs) are a feasible option, especially for non-drinking purposes, but their effectiveness depends on maintaining service continuity and compliance with water quality standards under uncertain environmental and operating conditions.
Key Innovation: A reliability-oriented digital twin tests building-scale rainwater treatment under uncertain demand and climate forcing.
75. Mitigating terrain effects of GEDI data on regional accounting of aboveground biomass utilizing slope-adaptive waveform metrics
Core Problem: Terrain-induced waveform broadening biases GEDI biomass metrics in steep regions and propagates into regional carbon accounting.
Key Innovation: Slope-adaptive waveform metrics reduce terrain effects before regional aboveground-biomass estimation.
76. SuperDove satellite detection of matter-loaded windrows with a focus on extreme environmental events
Core Problem: Matter-loaded windrows are high-density submesoscale accumulation structures that form on the ocean surface and are often associated with extreme environmental events.
Key Innovation: SuperDove imagery and a U-Net detect matter-loaded windrows associated with extreme environmental events.
77. Two-stage instance segmentation method based on Tunnel-SPVNet for shield tunnel segments
Core Problem: In applications of tunnel digital modeling and structural analysis, accurately extracting segment structural units of shield tunnels from raw point clouds is a key prerequisite for digital reconstruction and subsequent structural analysis.
Key Innovation: A two-stage point-cloud instance-segmentation method identifies individual shield-tunnel segments for automated inspection.
78. Disentangling precipitation forcing and parameter uncertainty in a multi-objective surface–subsurface hydrologic modeling framework
Core Problem: Precipitation forcing uncertainty can substantially influence parameter estimation and predictive performance in coupled surface-subsurface hydrologic models, yet practical strategies for explicitly representing this uncertainty during calibration remain limited.
Key Innovation: A multi-objective surface-subsurface framework separates precipitation-forcing uncertainty from parameter uncertainty.
79. An attention-enhanced hybrid physics-informed neural network for multi-parameter inversion of heterogeneous reservoirs in CO2 geological storage
Core Problem: Geological parameter inversion in high-dimensional heterogeneous reservoirs under limited observations remains a strongly ill-posed problem.
Key Innovation: Attention-enhanced physics-informed learning jointly inverts multiple parameters in heterogeneous geological reservoirs.
80. High-resolution spatiotemporal monitoring of chlorophyll-a in small inland reservoirs using multi-source remote sensing data fusion
Core Problem: Small reservoirs are highly vulnerable to rapid cyanobacterial blooms, yet routine monitoring is constrained by the spatial-temporal trade-off of single-sensor satellite observations.
Key Innovation: Multi-source remote-sensing fusion reconstructs high-resolution chlorophyll-a dynamics in small reservoirs.
81. Oblique bearing capacity of strip footings near Hoek-Brown rock slopes using the modified stress characteristics method
Core Problem: The ultimate bearing capacity of strip footing near rock slope (SFNRS) under oblique loads remains a theoretical challenge, as the existing analytical methods predominantly rely on a priori assumed kinematic mechanisms or numerical bound optimizations.
Key Innovation: A modified stress-characteristics method estimates oblique footing capacity near Hoek-Brown rock slopes.
82. Mesoscopic Damage Evolution of Water-Bearing Mudstone Under Low-Strain-Rate Cyclic Dynamic Loading: A Particle-Flow Simulation Study
Core Problem: Groundwater-related weakening and repeated low-strain-rate disturbances can jointly affect the long-term stability of soft surrounding rock.
Key Innovation: Particle-flow simulations resolve mesoscopic damage evolution in water-bearing mudstone under repeated low-rate dynamic loading.
83. An AE-DIC-FEM prediction method for rock failure: uniaxial compression experiment case
Core Problem: Rock-failure warning requires both temporal precursors and spatial damage evidence, which acoustic-emission or deformation measurements alone do not fully provide.
Key Innovation: Acoustic emission, digital image correlation and finite-element inversion jointly track crack acceleration and weakening parameters, producing failure-time errors below 1.4 percent in granite and red-sandstone tests.
84. Uncertainty propagation in water systems decision-making: A multi-parameter robustness framework
Core Problem: Water systems are increasingly affected by hydrological variability, socio-economic change, and evolving management constraints, creating challenges for robust decision-making.
Key Innovation: A multi-parameter robustness framework traces uncertainty propagation through water-system decisions.
85. Human pressures exacerbate the impacts of hydroclimatic extremes on river greenhouse gas emissions
Core Problem: River greenhouse-gas inventories rarely represent how hydroclimatic extremes interact with human pressure across basins.
Key Innovation: Cross-basin observations quantify the amplification of greenhouse-gas responses when droughts and floods coincide with intensive human disturbance.